Qucs

Work Book

Chris Pitcher
Clemens Novak
Gunther Kraut
Mike Brinson
Thierry Scordilis
Stefan Jahn

Copyright
c 2005 Chris Pitcher <ozjp@chariot.net.au>
Copyright
c 2013 Clemens Novak <clemens@familie-novak.net>
Copyright
c 2006 Gunther Kraut <gn.kraut@online.de>
Copyright
c 2005 Thierry Scordilis <thierry.scordilis@free.fr>
Copyright
c 2006, 2007, 2014 Mike Brinson <mbrin72043@yahoo.co.uk>
Copyright
c 2005, 2006, 2007 Stefan Jahn <stefan@lkcc.org>

Permission is granted to copy, distribute and/or modify this document under the
terms of the GNU Free Documentation License, Version 1.1 or any later version
published by the Free Software Foundation. A copy of the license is included in
the section entitled ”GNU Free Documentation License”.
Contents
1 General Design Flow 12

2 Getting started with Qucs 14

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.2 Tool suite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.3 Setting up schematics . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.3.1 DC simulation - A voltage divider . . . . . . . . . . . . . . . 19
2.3.2 DC simulation - Characteristics of a transistor . . . . . . . . 35
2.3.3 AC simulation - Transit frequency of a bipolar transistor . . 43
2.3.4 AC simulation - A simple RC highpass . . . . . . . . . . . . 47
2.3.5 Transient simulation - Amplification of a bipolar transistor . 49
2.3.6 S-parameter simulation - Transit frequency of a BJT . . . . 51
2.3.7 S-parameter and AC simulation - A Bessel band-pass filter . 54

3 Understanding RF Data Sheet Parameters 60

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2 DC specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.3 Maximum ratings and thermal characteristics . . . . . . . . . . . . 61

4 DC Analysis, Parameter Sweep and Device Models 62

4.1 DC Static Circuits . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
4.2 When Things Vary . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.3 Models and Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 70

5 Getting Started with Digital Circuit Simulation 74

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
5.2 Simulating simple digital circuits . . . . . . . . . . . . . . . . . . . 75
5.2.1 Notes on drawing digital schematics . . . . . . . . . . . . . . 76
5.3 VHDL code generated by Qucs . . . . . . . . . . . . . . . . . . . . 77
5.4 Truth tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5.5 Digital subcircuits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
5.6 Building a digital component library . . . . . . . . . . . . . . . . . 86
5.6.1 Logic zero . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5.6.2 Logic one . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

2
 5.6.3 G2bit - 2 bit pattern generator . . . . . . . . . . . . . . . . 86
5.6.4 G4bit - 4 bit pattern generator . . . . . . . . . . . . . . . . 87
5.6.5 MUX2to1 - 2 input to 1 output multiplexer . . . . . . . . . 88
5.6.6 MUX4to1 - 4 input to 1 multiplexer . . . . . . . . . . . . . . 89
5.6.7 2 bit adder . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.7 Subcircuit VHDL code generated by Qucs . . . . . . . . . . . . . . 90
5.7.1 Gen2bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
5.7.2 2 bit adder . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.7.3 Notes on subcircuit VHDL generation . . . . . . . . . . . . . 92
5.8 Subcircuit nesting: A more complex design example . . . . . . . . . 93
5.8.1 4 bit RTL design . . . . . . . . . . . . . . . . . . . . . . . . 94
5.9 Update number one: May 2006 . . . . . . . . . . . . . . . . . . . . 98
5.9.1 Bugs, corrections and small changes to the Qucs digital sim-
ulation code . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.9.2 New digital simulation features . . . . . . . . . . . . . . . . 99
5.9.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5.9.4 Using the Qucs VHDL editor . . . . . . . . . . . . . . . . . 102
5.9.5 Linking VHDL entity-architecture models to Qucs schematic
device symbols . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.9.6 Generating VHDL code from Qucs schematic drawings . . . 116
5.10 Update number two: September 2006 . . . . . . . . . . . . . . . . . 125
5.10.1 Simulating VHDL code using Qucs and FreeHDL. . . . . . . 126
5.10.2 VHDL predefined packages and libraries. . . . . . . . . . . . 129
5.10.3 VHDL simulation code structures. . . . . . . . . . . . . . . . 129
5.10.4 VHDL data types. . . . . . . . . . . . . . . . . . . . . . . . 132
5.10.5 An example VHDL simulation employing integer signals. . . 133
5.10.6 Multivalued logic. . . . . . . . . . . . . . . . . . . . . . . . . 134
5.10.7 Run debugging of VHDL simulation code. . . . . . . . . . . 141
5.10.8 Testing digital systems using test vectors stored on disk. . . 149
5.11 End note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

6 Transient Domain Flip-Flop Models for Mixed-Mode Simulation 157

6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
6.2 Latches and flip-flops . . . . . . . . . . . . . . . . . . . . . . . . . . 158
6.3 The gated D latch . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158
6.4 Edge-triggered D type flip-flop . . . . . . . . . . . . . . . . . . . . . 161
6.5 The edge-triggered JK flip-flop . . . . . . . . . . . . . . . . . . . . . 163
6.6 The edge-triggered T flip-flop . . . . . . . . . . . . . . . . . . . . . 165
6.7 Two example digital circuits . . . . . . . . . . . . . . . . . . . . . . 167
6.8 VHDL code for the transient domain flip-flop models . . . . . . . . 171
6.9 Generating a library of mixed-mode digital components . . . . . . . 175

3
 6.10 Digital component propagation time delays and transient simulation
numerical stability . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
6.11 Mixed-mode example simulations . . . . . . . . . . . . . . . . . . . 177
6.12 End Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185

7 Modelling Operational Amplifiers 186

7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
7.2 The Qucs built-in operational amplifier model . . . . . . . . . . . . 186
7.3 Adding features to the Qucs OP AMP model . . . . . . . . . . . . . 193
7.4 Modular operational amplifier macromodels . . . . . . . . . . . . . 193
7.5 A basic AC OP AMP macromodel. . . . . . . . . . . . . . . . . . . 194
7.5.1 The input stage. . . . . . . . . . . . . . . . . . . . . . . . . 194
7.5.2 Voltage gain stage 1. . . . . . . . . . . . . . . . . . . . . . . 198
7.5.3 Derivation of voltage gain stage 1 transfer function . . . . . 198
7.5.4 Output stage. . . . . . . . . . . . . . . . . . . . . . . . . . . 200
7.5.5 A subcircuit model for the basic AC OP AMP macromodel . 200
7.6 A more accurate OP AMP AC macromodel . . . . . . . . . . . . . 204
7.6.1 Derivation of voltage gain stage 2 transfer function. . . . . . 204
7.6.2 Simulating OP AMP open loop differential gain . . . . . . . 205
7.7 Adding common mode effects to the OP AMP AC macromodel . . 207
7.7.1 Simulating OP AMP common-mode effects . . . . . . . . . . 208
7.8 Large signal transient domain OP AMP macromodels . . . . . . . . 212
7.8.1 Slew rate macromodel derivation . . . . . . . . . . . . . . . 212
7.8.2 Modelling OP AMP overdrive and output voltage limiting . 214
7.8.3 Modelling OP AMP output current limiting . . . . . . . . . 220
7.9 Obtaining OP AMP macromodel parameters from published device
data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
7.10 More complete design examples. . . . . . . . . . . . . . . . . . . . . 223
7.10.1 Example 1: State variable filter design and simulation . . . . 223
7.10.2 Example 2: Sinusoidal signal generation with the Wien bridge
oscillator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225
7.11 Update number one: March 2007 . . . . . . . . . . . . . . . . . . . 233
7.11.1 Building a library component for the modular OP AMP
macromodel . . . . . . . . . . . . . . . . . . . . . . . . . . . 233
7.11.2 Changing model parameters: use of the SPICEPP preprocessor235
7.11.3 The Boyle operational amplifier SPICE model . . . . . . . . 235
7.11.4 Model accuracy . . . . . . . . . . . . . . . . . . . . . . . . . 237
7.11.5 The PSpice modified Boyle model . . . . . . . . . . . . . . . 240
7.12 Constructing Qucs OPAMP libraries . . . . . . . . . . . . . . . . . 253
7.13 Extending existing OP AMP models . . . . . . . . . . . . . . . . . 255
7.14 End note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262

4
8 Modelling the 555 Timer 263
8.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 263
8.2 The Qucs 555 timer model . . . . . . . . . . . . . . . . . . . . . . . 264
8.2.1 The trigger comparator macromodel . . . . . . . . . . . . . 265
8.2.2 The threshold comparator macromodel . . . . . . . . . . . . 266
8.2.3 The digital logic macromodel . . . . . . . . . . . . . . . . . 267
8.2.4 The 555 timer output amplifier macromodel . . . . . . . . . 268
8.2.5 The discharge switch macromodel . . . . . . . . . . . . . . . 269
8.3 Published 555 timer test circuits . . . . . . . . . . . . . . . . . . . . 270
8.3.1 The 555 timer monostable pulse generator . . . . . . . . . . 271
8.3.2 The 555 timer astable pulse oscillator . . . . . . . . . . . . . 272
8.3.3 Pulse width modulation . . . . . . . . . . . . . . . . . . . . 275
8.3.4 Pulse position modulation . . . . . . . . . . . . . . . . . . . 278
8.4 Multiple 555 timer simulation examples . . . . . . . . . . . . . . . . 279
8.4.1 Sequential pulse train generation . . . . . . . . . . . . . . . 279
8.4.2 Frequency divider circuit . . . . . . . . . . . . . . . . . . . . 284
8.5 End note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287

9 Qucs Simulation of SPICE Netlists 288

9.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
9.2 The basic SPICE netlist format . . . . . . . . . . . . . . . . . . . . 289
9.3 Defining symbols for Qucs SPICE netlist components . . . . . . . . 294
9.4 Handling SPICE subcircuits . . . . . . . . . . . . . . . . . . . . . . 296
9.4.1 Subcircuit example 1: a multisection LC delay line . . . . . 296
9.4.2 Subcircuit example 2: a two section CMOS ring counter . . 296
9.5 Limitations when converting SPICE netlists . . . . . . . . . . . . . 303
9.6 Extending the SPICE netlist language . . . . . . . . . . . . . . . . 303
9.6.1 The SPICEPP preprocessor . . . . . . . . . . . . . . . . . . 304
9.7 Circuit template models . . . . . . . . . . . . . . . . . . . . . . . . 305
9.8 Building circuit design equations into netlists . . . . . . . . . . . . . 310
9.9 Global nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
9.10 End Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317

10 Biasing a BJT Transistor 318

10.1 Graphical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
10.1.1 Graphical approach shows trade-offs . . . . . . . . . . . . . 320
10.2 Simulation technics . . . . . . . . . . . . . . . . . . . . . . . . . . . 323

11 BJT Modeling and Verification 324

11.1 choice of transistor . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
11.2 library creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328

5
 11.3 device library verification . . . . . . . . . . . . . . . . . . . . . . . . 329
11.4 parasitic description of the package . . . . . . . . . . . . . . . . . . 333
11.5 small signal S parameter verification . . . . . . . . . . . . . . . . . 337

12 Power Amplifier Design 342

12.1 Field of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
12.2 System consideration . . . . . . . . . . . . . . . . . . . . . . . . . . 342
12.3 Biasing consideration . . . . . . . . . . . . . . . . . . . . . . . . . . 343
12.4 Why thermal design ? . . . . . . . . . . . . . . . . . . . . . . . . . 346
12.4.1 Thermal management . . . . . . . . . . . . . . . . . . . . . 346
12.5 DC Power dissipation . . . . . . . . . . . . . . . . . . . . . . . . . . 348
12.6 Small signal analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 349

13 Low Noise Amplifier Design 350

13.0.1 System consideration . . . . . . . . . . . . . . . . . . . . . . 350
13.0.2 Choice of transistor . . . . . . . . . . . . . . . . . . . . . . . 351
13.0.3 library creation . . . . . . . . . . . . . . . . . . . . . . . . . 352
13.0.4 DC study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
13.0.5 SP study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
13.0.6 Non linearities study . . . . . . . . . . . . . . . . . . . . . . 353
13.0.7 Possible improvement tips . . . . . . . . . . . . . . . . . . . 353

14 Microstrip Design 354

14.1 10dB Directional Coupler Design . . . . . . . . . . . . . . . . . . . 354
14.1.1 Some boring theory beforehand . . . . . . . . . . . . . . . . 354
14.1.2 Design equations . . . . . . . . . . . . . . . . . . . . . . . . 356
14.1.3 Applying the design equations . . . . . . . . . . . . . . . . . 356
14.1.4 What next? . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
14.1.5 Verification of the design . . . . . . . . . . . . . . . . . . . . 358
14.1.6 Suggested improvements . . . . . . . . . . . . . . . . . . . . 363
14.1.7 Remaining thinkabouts . . . . . . . . . . . . . . . . . . . . . 365

15 Measurement Expressions Reference Manual 366

15.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
15.2 Using Measurement Expressions . . . . . . . . . . . . . . . . . . . . 366
15.2.1 Entering Measurement Expressions . . . . . . . . . . . . . . 367
15.2.2 Changing Measurement Expressions . . . . . . . . . . . . . . 368
15.2.3 Syntax of Measurement Expressions . . . . . . . . . . . . . . 369
15.3 Functions Syntax and Overview . . . . . . . . . . . . . . . . . . . . 373
15.3.1 Functions Reference Format . . . . . . . . . . . . . . . . . . 373
15.3.2 Functions Listed by Category . . . . . . . . . . . . . . . . . 375

6
 15.4 Math Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381
15.4.1 Vectors and Matrices . . . . . . . . . . . . . . . . . . . . . . 381
15.4.2 Elementary Mathematical Functions . . . . . . . . . . . . . 391
15.4.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 454
15.5 Electronics Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 486
15.5.1 Unit Conversion . . . . . . . . . . . . . . . . . . . . . . . . . 486
15.5.2 Reflection Coefficients and VSWR . . . . . . . . . . . . . . . 490
15.5.3 N-Port Matrix Conversions . . . . . . . . . . . . . . . . . . . 495
15.5.4 Amplifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504

16 Component, compact device and circuit modelling using symbolic equa-

tions 521
16.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 521
16.2 Qucs electronic device and circuit modelling . . . . . . . . . . . . . 521
16.3 Extending circuit simulation capabilities with equations . . . . . . . 526
16.3.1 Low pass active filter design with embedded design equations 527
16.4 Introduction to Qucs subcircuit parameters . . . . . . . . . . . . . . 535
16.5 Building universal macromodels using subcircuits and parameters . 538
16.6 More complex nested subcircuit models . . . . . . . . . . . . . . . . 544
16.7 Introduction to equation defined devices (EDD) . . . . . . . . . . . 547
16.8 The Qucs EDD component . . . . . . . . . . . . . . . . . . . . . . . 549
16.9 Modelling nonlinear resistors . . . . . . . . . . . . . . . . . . . . . . 553
16.10Modelling nonlinear capacitors and inductors . . . . . . . . . . . . . 555
16.11Compact device modelling using EDD . . . . . . . . . . . . . . . . 558
16.12Constructing EDD compact device models and circuit macromodels 567
16.13End Note . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
16.14Appendix A: Qucs constants, operators and functions . . . . . . . . 568
16.15Appendix B: Constructing subcircuits with parameters . . . . . . . 571
16.15.1 Enter the series resonance circuit and add input and output
pins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
16.15.2 Change the component names to Ls, Cs and Rs . . . . . . . 572
16.15.3 Construct symbol for new subcircuit . . . . . . . . . . . . . 573
16.15.4 Add the names of the subcircuit parameters to the LCR symbol574
16.15.5 Test the LCR subcircuit . . . . . . . . . . . . . . . . . . . . 575

17 Subcircuit and Verilog-A RF Circuit Simulation Models for Axial and

Surface Mounted Resistors 576
17.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
17.2 RF Resistor Models . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
17.3 Analysis of the RF resistor model . . . . . . . . . . . . . . . . . . . 579

7
 17.4 Direct measurement of RF resistor impedance using a simulated
impedance meter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 581
17.5 Extraction of RF resistor data from measured S parameters . . . . 581
17.6 Extraction of RF resistor parameters from measured S data for a
nominal 1000 Ω axial resistor . . . . . . . . . . . . . . . . . . . . . 587
17.7 One more example: extraction of RF resistor parameters from mea-
sured S data for a nominal 100 Ω SMD resistor . . . . . . . . . . . 587
17.8 A Verilog-A RF resistor model . . . . . . . . . . . . . . . . . . . . . 589
17.9 Extraction of Verilog-A RF resistor model parameters from mea-
sured S data for a 100 Ω axial resistor . . . . . . . . . . . . . . . . . 592
17.10End Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 592

18 Using Qucs in Textmode 594

18.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
18.2 Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 594
18.3 Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
18.4 Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
18.5 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
18.6 AC Sweep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
18.6.1 Analysis with Octave . . . . . . . . . . . . . . . . . . . . . . 597
18.6.2 Analysis with Python . . . . . . . . . . . . . . . . . . . . . . 598
18.7 Transient Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . 601
18.8 Qucs Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
18.9 Passive Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
18.9.1 Resistor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 602
18.9.2 Capacitor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
18.9.3 Inductor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
18.10Nonlinear Components . . . . . . . . . . . . . . . . . . . . . . . . . 603
18.10.1 Diode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 603
18.10.2 Bipolar Junction Transistor with Substrate . . . . . . . . . . 604
18.10.3 Diac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 606
18.10.4 Silicon Controlled Rectifier (SCR) . . . . . . . . . . . . . . . 606
18.10.5 Triac (Bidirectional Thyristor) . . . . . . . . . . . . . . . . . 606
18.10.6 Resonance Tunnel Diode . . . . . . . . . . . . . . . . . . . . 607
18.10.7 Junction Field-effect Transistor . . . . . . . . . . . . . . . . 607
18.10.8 MOS field-effect transistor with substrate . . . . . . . . . . . 608
18.11Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 610
18.11.1 DC Voltage Source . . . . . . . . . . . . . . . . . . . . . . . 610
18.11.2 AC Voltage Source . . . . . . . . . . . . . . . . . . . . . . . 610
18.11.3 Voltage Controlled Voltage Source . . . . . . . . . . . . . . . 610
18.11.4 Voltage Controlled Current Source . . . . . . . . . . . . . . 610

8
 18.11.5 Current Controlled Voltage Source . . . . . . . . . . . . . . 611
18.11.6 Current Controlled Current Source . . . . . . . . . . . . . . 611
18.11.7 Voltage Pulse . . . . . . . . . . . . . . . . . . . . . . . . . . 611
18.11.8 Current Pulse . . . . . . . . . . . . . . . . . . . . . . . . . . 612
18.11.9 Rectangle Voltage . . . . . . . . . . . . . . . . . . . . . . . . 612
18.11.10Rectangle Current . . . . . . . . . . . . . . . . . . . . . . . 612
18.11.11Exponential Voltage Source . . . . . . . . . . . . . . . . . . 613
18.11.12Exponential Current Source . . . . . . . . . . . . . . . . . . 613
18.11.13AC Voltage Source with Ammplitude Modulator . . . . . . . 613
18.11.14AC Voltage Source with Ammplitude Modulator . . . . . . . 614
18.11.15AC Voltage Source with Phase Modulator . . . . . . . . . . 614
18.12Simulation Commands . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.13DC Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.14AC Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.15Parameter Sweep . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
18.16Transient Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . 616
18.17S Parameter Simulation . . . . . . . . . . . . . . . . . . . . . . . . 617
18.17.1 xxx . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
18.17.2 Silicon Controlled Rectifier (SCR) . . . . . . . . . . . . . . . 617
18.17.3 Triac (Bidirectional Thyristor) . . . . . . . . . . . . . . . . . 618
18.17.4 Resonance Tunnel Diode . . . . . . . . . . . . . . . . . . . . 618

19 Active Filters Design with Qucs and Qucsactivefilter 619

19.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 619
19.2 Interface description . . . . . . . . . . . . . . . . . . . . . . . . . . 619
19.3 Filter transfer function . . . . . . . . . . . . . . . . . . . . . . . . . 621
19.4 Filters parameters explanation . . . . . . . . . . . . . . . . . . . . . 621
19.5 Filter design example . . . . . . . . . . . . . . . . . . . . . . . . . . 623
19.6 Manual transfer function definition . . . . . . . . . . . . . . . . . . 626
19.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 627

9
Introduction
Important note and warning
You should take into account the fact that this document is written on the fly, so
some mistakes are still possible, and the author is not responsible for any damage
due to the use of this document.

This document is intended to be a work book for RF and microwave designers.

Our intention is not to provide an RF course, but some touchy RF topics. The
goal is to insist on some design rules and work flow for RF desings using CAD
programs. This work flow will be handled through different chapters on quite
different subjects.

Work book content

In this workbook, we will pass through some regular tasks. But there is a progres-
sion on the explanations, and due to the fact that we have to cover a huge amount
of information, some key point will be shown ony once, so it is recommanded to
read the chapters in order.

This work book will include:

Work flow: the regular process of project design is shown,

Understanding RF data sheets: a usual task, that could be hell, could turn a
project into a nightmare,

BJT Modeling: after having chosen a device, we always need to use in the CAD,
and usually this device does not exits in the CAD... how to create it and
verify

DC static: since all active devices have to be biased...

PA Desgin: the active component is found, and a small amplifier is designed

without to many constraints

10
LNA Design: a more constraint design using more rules, stability, noise etc.

oscillator design: a procedure that is typical from CAD issues, handling non
usual procedure,

vco design: a normal evolution from a oscillator,

detector: a design difficult to handle.

more will come . . .

11
1 General Design Flow
Knowing the fact that you are familiar with the regular design flow of RF, mi-
crowave circuits and or systems, we need to clarify how Qucs is intended to be
used for this type of circuits design.

As an RF research engineer, I’m still having some new graduate students. And I’m
always having some problems with the new methods that are teached. Usually they
arrive with some knowledge on CAD programs, but they do not really know how
to dimension their design. They use only the optimizer to replace their thinking.
What a pity! Of course not all of them are like this, but it is a common trend.
By since work book I want to show that there are some rules to follow, and that
a design can be calculated, and that it will not work due to a wizard!

For the experts, nothing very new herein, but only some particular use of Qucs,
since the design rules are the one that you could have on the workbench using a
paper and a pen.

The author.

Regular document organisation

We will try to have always the same kind of organization inside the different
chapters, that is to say:

a main topic: in order to say in which field of activity this design is intended to
be used

a block specification: in order to know what we have to do. This task will not
be explain at a first glance, since it is not the goal of this document (we’re
not dealing with system specification, it could be if the component present
in Qucs are increased . . . so why not in further version of this document.)

DC explanation: if the design includes a DC part, then we should provide the

DC study including thermal aspect if needed.

12
functional design: in order to explain how this functionality is designed either
in general or by the mean of Qucs. The second aspect should be always kept
in mind. Everything might not be straightforward on other CAD programs,
and therefore not considerated herein.

Hoping that these explanations clarifies the goal of this document.

13
2 Getting started with Qucs
2.1 Introduction
The following sections are meant to give an overview about what the Qucs software
can be used for and how it is used to achieve this.
Qucs is free software licensed under the General Public License (GPL). It can be
downloaded from http://qucs.sourceforge.net and comes with the complete
source code. Every user of the program is allowed and called upon (on a voluntary
basis of course) to modify it for their purposes as long as changes are made public.
Contact the authors to verify them and finally to incorporate it into the software.
The software is available for a variety of operating systems including

• GNU/Linux

• Windows

• FreeBSD

• MacOS

• NetBSD

• Solaris

On the homepage you’ll find the source code to build and install the software. Build
instructions are given. Also links for binary packages for certain distributions (e.g.
Debian, SuSE, Fedora) can be found.
Once the software has been successfully installed on your system you can start it
by issuing the

# qucs

command or by clicking the appropriate icon on your start menu or desktop. Qucs
is a multi-lingual program. So depending on your system’s language settings the
Qucs graphical user interface (GUI) appears in different languages.

14
 Figure 2.1: Qucs has been started

On the left hand side you find the Projects folder opened. Usually the projects
folder will be empty if you use Qucs for the first time. The large area on the
right hand side is the schematic area. Above you can find the menu bar and the
toolbars.
In the File → Application Settings menu the user can configure the language
and appearance of Qucs.

15
 Figure 2.2: Application setting dialog

To take effect of the language and font settings the application must be closed
either via the Ctrl + Q shortcut or the File → Exit menu entry. Then start
Qucs again.

2.2 Tool suite

Qucs consists of several standalone programs interacting with each other through
the GUI. There are

• the GUI itself,

The GUI is used to create schematics, setup simulations, display simulation
results, writing VHDL code, etc.

• the backend analogue simulator,

The analogue simulator is a command line program which is run by the GUI
in order to simulate the schematic which you previously setup. It takes a
netlist, checks it for errors, performs the required simulation actions and
finally produces a dataset.

• a simple text editor,

The text editor is used to display netlists and simulation logging informa-
tions, also to edit files included by certain components (e.g. SPICE netlists,
or Touchstone files).

• a filter synthesis application,

The program can be used to design various types of filters.

16
 • a transmission line calculator,
The transmission line calculator can be used to design and analyze different
types of transmission lines (e.g. microstrips, coaxial cables).

• a component library,
The component library manager holds models for real life devices (e.g. tran-
sistors, diodes, bridges, opamps). It can be extended by the user.

• an attenuator synthesis application,

The program can be used to design various types of passive attenuators.

• a command line conversion program

The conversion tool is used by the GUI to import and export datasets, netlists
and schematics from and to other CAD/EDA software. The supported file
formats as well as usage information can be found on the manpage of quc-
sconv.

Additionally the GUI steers other EDA tools. For digital simulations (via VHDL)
the program FreeHDL (see http://www.freehdl.seul.org) is used. And for
circuit optimizations ASCO (see http://asco.sourceforge.net) is configured
and run.

2.3 Setting up schematics

The following sections will enable the user to setup some simple schematics. For
this we first create a new project named “WorkBook”. Either press the New
button above the projects folder or use the menu entry Project → New Project
and enter the new project name.

Figure 2.3: New project dialog

Confirm the dialog by pressing the “Create” button. When done, the project is
opened and Qucs switches to the Content tab.

17
 Figure 2.4: New empty project has been created

In the Content tab you will find all data related to the project. It contains
your schematics, the VHDL files, data display pages, datasets as well as any other
data (e.g. datasheets). On the right hand side an “untitled” and empty schematic
window is displayed.
Now you can start to edit the schematic. The available components can be found
in the Components tab.

18
 Figure 2.5: Components tab

In fig. 2.5 is shown when clicking the Components tab. There are lumped com-
ponents (e.g. resistors, capacitors), sources (e.g. DC and AC sources), transmis-
sion lines (e.g. microstrip, coaxial cable, twisted pair), nonlinear components (e.g.
ideal opamp, transistors), digital components (e.g. flip-flops), file components (e.g.
Touchstone files, SPICE files), simulations (e.g. AC or DC analysis), diagrams (e.g.
cartesian or polar plot) and paintings (e.g. texts, arrows, circles).
Each of the components can placed on the schematic by clicking it once, then move
the mouse cursor onto the schematic and click again to put it on its final position.
During the mouse move you can right click in order to rotate the component into
its final position. The user can also drag-and-drop the components.

2.3.1 DC simulation - A voltage divider

The DC analysis is a steady state analysis. It computes the node voltage as well
as branch currents of the complete circuit. The given circuit in fig. 2.6 is going to
divide the voltage of a DC voltage source according to the resistor ratio.

19
 Figure 2.6: Components of the voltage divider place in the schematic area

Wiring components
Now you need to connect the components appropriately. This is done using the
wiring tool. You enable the wiring mode either by clicking the wire icon or by
pressing the Ctrl + E shortcut. Left clicking on the components’ ports (small
red circles) starts a wire, clicking on a second port finishes the wire. In order to
change the orientation of the wire right click it. You can leave the wiring mode by
the pressing Esc key.

20
 Figure 2.7: Components of the voltage divider appropriately wired

For any analogue simulation (including the DC simulation) there is a reference

potential required (for the nodal analysis). The ground symbol can be found in
the Components tab in the lumped components category. The user can also
choose the ground symbol icon or simply press the Ctrl + G shortcut. In the
given circuit in fig. 2.8 the ground symbol is placed at the negative terminal of the
DC voltage source.

Placing simulation blocks

The type of simulation which is performed must also be placed on the schematic.
You choose the “DC simulation” block which can be found in the Components
tab in the simulations category.

21
 Figure 2.8: Ground symbol as well as DC simulation in place

Labelling wires
If you want the voltage between the two resistors (the divided voltage) be output
in the dataset after simulation the user need to label the wire. This is done by
double clicking the wire and given an appropriate name. Wire labelling can also
be issued using the icon in the toolbar, by pressing the Ctrl + L shortcut or by
choosing the Insert → Wire Label menu entry.

Figure 2.9: Node label dialog

The dialog is ended by pressing the Enter key of pressing the “Ok” button.
Now the complete schematic for the voltage divider is ready and can be saved.
This can by achieved by choosing the File → Save menu entry, clicking the single

22
disk icon or by pressing the Ctrl + S shortcut.

Figure 2.10: File save dialog

23
 Figure 2.11: Final voltage divider schematic

The final DC voltage divider is shown in fig. 2.11.

Issuing a simulation
The schematic can now be simulated. This is started by choosing the Simulation
→ Simulate menu entry, clicking the simulation button (the gearwheel) or by
pressing the F2 shortcut.

24
 Figure 2.12: Empty data display after simulation finished

After the simulation has been finished the related data display is shown (see
fig.2.12). Also the Components tab has changed its category to “diagrams”.

Placing diagrams
Choose the tabular (list of values) diagram and place it on the data display page.
After dropping the tabular, the diagram dialog appears as shown in fig. 14.7.

25
 Figure 2.13: Diagram dialog

By double clicking the divided.V the graph (i.e. values in a tabular plot) is added
to the diagram. Beside the node voltage divided.V also the current through the
DC voltage source V1.I is available. Only items listed in the dataset list can be
put into the graph.

Available dataset items

Depending on the type of simulation the user performed you find the following
types of items in the dataset.

• node.V – DC voltage at node node

• name.I – DC current through component name

• node.v – AC voltage at node node

• name.i – AC current through component name

• node.vn – AC noise voltage at node node

• name.in – AC noise current through component name

• node.Vt – transient voltage at node node

• name.It – transient current through component name

26
 • S[1,1] – S-parameter value

Please note that all voltages and currents are peak values and all noise voltages
are RMS values at 1Hz bandwidth.

Figure 2.14: Diagram dialog with the node voltage added

Depending on the type of graph you have various options to choose for the graph.
For a tabular graph there is the the number precision as well as type of number
notation (important for complex values). Press the “Ok” button to close the dialog.

27
 Figure 2.15: Data display with tabular graph

In the tabular graph you see now the value of the node voltage divided.V which
is 0.5V. That was expected since the values of the resistors are equally sized and
the DC voltage source produces 1V.
Congratulations! You made your first successful simulation using Qucs.

Changing component properties

If you want to change the resistor ratio then switch back to your schematic either
by clicking on the divider.sch tab, by pressing the F4 shortcut or by choosing
the Simulation → View Data Display/Schematic menu entry. Afterwards
double click the R1 resistor. This opens the component property dialog shown in
fig. 2.16.

28
 Figure 2.16: Component property dialog for the R1 resistor

In the component property dialog all the properties of a given component can be
edited. A short description is given as well as there is a checkbox for each property
display in schematic which can be used to add the property name and value on
the schematic (or to hide it).

Allowed property values For component values either standard (1000), scientific
(1e-3) or an engineering (1k) number notation can be chosen. Some units are also
allowed. The units are

• Ohm – resistance / Ω

• s – time / Seconds

• S – conductance / Siemens

• K – temperature / Kelvin

• H – inductance / Henry

• F – capacitance / Farad

• Hz – frequency / Hertz

• V – voltage / Volt

• A – current / Ampere

• W – power / Watt

29
 • m – length / Meter (not usable standalone, see paragraph below)

The available engineering suffixes are

• dBm – 10 · log (x/0.001)

• dB – 10 · log (x)

• T – 1012

• G – 109

• M – 106

• k – 103

• m – 10−3

• u – 10−6

• n – 10−9

• p – 10−12

• f – 10−15

• a – 10−18

Please note that all units and engineering suffixes are case sensitive and also note
the conflict in m. When specifying one millimeter you can use 1mm. One meter
(1m) cannot be specified and will always be interpreted as one milli (engineering
notation).
Now you can change the resistor value to 1Ω, see fig. 2.17.

30
 Figure 2.17: Component property dialog for the R1 resistor

Press the “OK” button to close the dialog. You will get the following schematic.

31
 Figure 2.18: Value of resistor R1 changed

In order to change the value of the resistor R2 you can just click on the 50 Ohm
value directly on the schematic and edit the value.

Figure 2.19: Change value of resistor R2 directly on schematic

Change the value to “3” which will give a resistor ratio of 3/(1 + 3) = 0.75. Now
you have the following schematic.

32
 Figure 2.20: Value of resistor R2 changed

Diagrams are not limited to be placed on the data display, they can also reside on
the schematic directly. Thus you can place again now a tabular diagram on the
schematic and add the divided.V value. The diagram will show the result from
the previous simulation.

Changing document properties

If you do not want Qucs to change automatically to the associated data display
you can change the behaviour in the document setting dialog. You can go to the
document settings dialog by right clicking on free space on the schematic area and
choose the Document Settings menu item in the context menu which pops up
or by choosing the File → Document Settings menu entry.

33
 Figure 2.21: Document settings dialog

In the dialog you uncheck the open data display after simulation item. Press
the “OK” button to apply the change. If you now resimulate the schematic by
pressing the F2 shortcut the “Qucs Simulation Messages” dialog window opens
and can be left by pressing Esc . The tabular diagram now show the new value
for divided.V.

34
 Figure 2.22: Divider schematic after new simulation

2.3.2 DC simulation - Characteristics of a transistor

We are now going ahead and will setup schematics for some characteristic curves
of a bipolar transistor using DC simulation and the parameter sweep.

35
 N
2
4
0
1
Q
_
1 P
a
r
a
m
e
w
s
W
1t
e
e
pr
I=
1 V
0
=
b UT S
i
D
m
=
l
y
p
e
a
r
m
P
1
t
t
S
o
pC
1
o
g
I
b
=
0
n
m
i
s
=
P
n1
c
d
s
C
D
1i
l
m
a
ui
t
o
n
Figure 2.23: Swept DC simulation setup

In the schematic in fig. 2.23 there is a bipolar transistor placed in a common emitter
configuration. Additionally a parameter sweep has been placed. Please note the
Sim property of the parameter sweep. It contains the instance name of the DC
simulation DC1 which is going to be swept. The parameter which is swept is Ib
(the base current) and is put into the Param property of the parameter sweep.
The parameter Ib is also put into the I property of the DC current source I1.

Using the component library

The bipolar transistor has been taken from the component library. You can start
the program by choosing the Tools → Component Library menu entry or by
pressing the Ctrl + 4 shortcut.

36
 Figure 2.24: Component library tool

When choosing the “Transistor” category with the combobox you find the “2N4401”
transistor. By clicking the “Copy to clipboard” button the component is available
in the clipboard and can be inserted in the schematic using the Ctrl + V shortcut
or by choosing the Edit → Paste menu entry. The component can also by dragged
onto the schematic by clicking on the symbol in the library tool.
So what do we want to simulate actually? It is the current transfer curve of the
bipolar transistor. The input current (at the base) is given by the swept parameter
Ib. The output current (at the collector) flows through the DC voltage source V1.
The current transfer curve is:

βDC = f (IC ) = IC /IB

The current through the voltage source V1 is the collector current flowing out of
the transistor.

Placing equations on the schematic

In order to compute the necessary values for the transfer curve we need to place
some equations on the schematic. This is done by clicking the equation icon or by
choosing the Insert → Insert Equation menu entry. When double clicking the
equation component you can edit the equations to be computed.

37
 Figure 2.25: Equation dialog

In the upper edit box you enter the name of the equation and in the lower one the
computation formula. The resulting schematic is shown in fig. 2.26.

38
 N
2
4
Q01
_
1 P
a
r
a
m
e
w
s
W
1t
e
e
pr
I=
1
b UT V
0
= S
i
D
m
=
l
y
p
e
a
r
m
P
1
t
t
S
o
pC
1
o
g
I
b
=
0
n
m
i
s
=
P
n1
c
d
s
C
Di
l
m
a
u
1Bi
t
o
n Eq
E
q
n
Ic
=
ei
t
a
o
n
u
1
I
:
.
V
/
b
ta
c
=
v
s
c
_
_lo
=
P (
tV
s
e
B,Ic
ta )
Figure 2.26: Swept DC simulation setup with equations

Note that three equations have been added. The first one Ic=-V1.I is the collec-
tor current flowing into the transistor (current though voltage sources flow from
the positive terminal to the negative terminal). The equation Beta=Ic/Ib com-
putes the current gain and finally Beta vs Ic=PlotVs(Beta,Ic) changes the
data dependency of the current gain to be the collector current. The original data
dependency is the swept parameter Ib.

The internal help system

The full list of available functions in the equation solver can be seen in the internal
help system. It is started by pressing the F1 shortcut or by choosing the Help
→ Help Index menu entry. In the sidebar choose the “Short Description of
mathematical Functions” entry.

39
 3 0
Figure 2.27: Internal help system

ctsvae_2101e81e71e61e51e41e30.10.11
IB
The help can be closed using the Ctrl + Q shortcut.

Configuring cartesian diagrams

In fig. 2.28 the final simulation result is shown. In the diagram dialog the Beta vs Ic

Ic
dataset entry was chosen.

Figure 2.28: Simulation result

Additionally the x-axis has been chosen to be logarithmic. The x-axis label is Ic.

40
 3 0 t1eras_7uv.01Ice126:406.31e951ec41e30.10.11
ctsvae_2101e8V
IB e
B
Figure 2.29: Editing diagram properties

Working with markers in diagrams

The current gain curve in diagram in fig. 2.28 shows a maximum value. If you

I
want to know the appropriate values it is possible to use markers for this purpose.

Figure 2.30: Cartesian diagram with marker

This is achieved by pressing the Ctrl + B shortcut, clicking the marker icon or
choosing the Insert → Set Marker on Graph menu entry. Then click on the
diagrams curve you want to have the marker at. If the marker is selected you can
move it by pressing the arrow keys ← , → and ↑ or ↓ for multi-dimensional
graphs.

41
 Figure 2.31: Marker dialog

Double clicking the marker opens the marker dialog. There you can configure the
precision as well as the number notation of the displayed values.

A multi-dimensional sweep
Now we are going to create a schematic for the output characteristics of the bipolar
transistor. The characteristic curve is defined as follows:

P
a
r
a
m
e
wt
e
e
prPa
r
a
m
e
wt
e
e
pr
IC = f (IB , VCE )

N
2
4
Q0
1
_
1 s
W
1 s
W2
Thus it is necessary to modify the schematic from the previous sections a bit.

I=
1
bU V
=c
e
VTS
i
D
m
=
l
y
p
e
a
r
m
P
0
t
t
S
4
o
pC
1
i
n
c
e
=
V S
y
T
P
t
Si
m
=
S
l
p
e
a
r
m
0
t
o
pW
1
i
n
I
b
=
1
m
.
9
i
n
s8
1
=P
P i
n
s
=5
c
d
si
l
m
a
u
1E
C
D i
t
o
n t
q
a
u
E
1
q
n
Ici
o
n
.I
=
:
V
Figure 2.32: Sweep setup for the output characteristics

A second parameter sweep has been added. The first order sweep is Vce specified
in the parameter sweep SW1. The Sim parameter points to the instance name of
the DC simulation DC1. The second order sweep is Ib specified in the parameter
sweep SW2. The Sim parameter of this second sweep points to the instance
name of the first sweep SW1. The first order sweep variable Vce is put into the
U property of the DC voltage source V1.

42
 0
.Ic015.2
Ibce0:.13485V
501V c2e34
Figure 2.33: Output characteristics of a NPN bipolar transistor

2.3.3 AC simulation - Transit frequency of a bipolar transistor

Pa
r
a
mt
e
er
In the next section we are going to determine the transit frequency of the bipolar

N
2
4
Q0
1
_ se
wp
transistor used in the previous DC sections. First a bias point is chosen. In fig. 2.34

1 W
1
the DC setup was a bit modified.

I=
1
b T V
=
U0 S
P
Si
D
m
=
l
y
p
e
a
r
m
1
t
t
o
pC
1
o
g
I
b
=
0
n
m
Pi
s
=
n1
c
d
s
C
Di
l
m
a
u
1Bi
t
o
n Ei
t
q
a
o
u
E
1
q
n
Ic
I
=
:
.
V
/
t
e
a
c
=n
b q
u
E
2
q
E
n
t
e
a
B
_i
t
a
o
d
0
=n
i
f
(
I
c
,b
)
v
s
_
_l
c
o
=
P(
t
s
e
V
BI
)
t
a
c
,
Figure 2.34: DC setup for determining a bias point for AC simulation

There is now an additional equation computing the RF current gain for zero fre-
quency which is Beta 0=diff(Ic,Ib). The equation denotes
∂Ic
βRF (f = 0) =
∂Ib

43
 3 0 b
:
I
t
e
B5
a.
2
_ e
:  05
7
tae210001e801e72143581e61eIb5B
0_B Ibe:t0a.214e614381e30.1
In fig. 2.35 the DC current gain from fig. 2.30 is plotted versus the base current
Ib choosing Beta in the diagram dialog instead of Beta vs Ic. The appropriate
base current shown in the marker is 140µA.

Figure 2.35: DC current gain vs. base current

It can be seen that the maximum AC current gain (257 @ 53µA) differs from the
maximum DC gain. Also the AC current gain almostly equals the DC current

I2
gain at the base current for the maximum DC current gain. For maximum RF

A2
1
u
=
performance the base current with the maximum AC current gain could be chosen.

a
r
at
m
ee
r a
c
si
l
m
ui
t
a
on
But there may be other consideration, e.g. DC power dissipation, so we choose

N
40 1 P
s
e
wp
the bias point with the maximum DC current gain – arbitrarily.

I=
1
b TQ _
V
U1
= 0W
1
S
A
m
=
i
y
p
eC
1
l
t
s
i
I
b C
A
1
y
p
e
T
t
t
a
r
S
o
n
P
il
o
g
=
k
H
1
z
G
0
1
s
=
a
r
m
P
l
u
V=
[
5
3
u1
4
0
;]
5
0
u
;
u
d
c
s
Ci
l
m
a
u
1b
D i
t
o
n E
cq
E
q
=
ien
(a
u
1 tio
n
.i/b
IV
taIc
=
2 Figure 2.36: Bias dependent AC simulation setup

In fig. 2.36 is a DC bias dependent AC simulation setup shown. The DC base

44
current Ib is swept for 53µA, 140µA and 500µA. Additionally the AC simulation
block has been placed on the schematic.
The Sim parameter of the SW1 parameter sweep is set to the instance name of the
AC simulation AC1. Qucs automatically “knows” that the DC simulation has to
be run before each AC simulation since it is required to determine the appropriate

3
bias points.

0
5
The AC current current source I2 is in parallel to the DC current source and has

2
an AC amplitude of 1µA. During the AC simulation the DC current source I1 is

0
an ideal open and the DC voltage source V1 is an ideal short.

a1
5
bte0501e3aIbc:ftr0e.q2u15en14/c4y0:5.24e°
In the equations V1.i (mark the small i letter) denotes the AC current through the

+
03
DC voltage source V1. The AC base current ib is taken from the input parameter
I2.I denoting the value of the property I of the AC current source I2 (1µA).
After pressing F2 – to start the simulation – the following cartesian diagram can

1e5acfrq1ue6ncy1e71e81e9
be placed on the data display page, see fig. 2.37.

Figure 2.37: AC current gain of the bipolar transistor

The marker clearly shows for the low frequency range (f → 0) the DC current
gain of 246 (for IB = 140µA) which was already determined in fig. 2.35.
In the next AC simulation setup shown in fig. 2.38 the parameter sweep is dropped
to concentrate on the determination of the transit frequency. The transit frequency
of a bipolar transistor denotes the frequency where the AC current gain drops to
1 (0 dB).
fT ← |h21 |2 = 1
Expressed in h-parameters of a general two-port the AC current gain is:

i2 
βRF = h21 = 
i1 v2 =0

45
 I2
A2
1
u
= a
ci
s
m
u l
ai
t
on
whereas port 1 is the base and port 2 the collector. The side condition (v2 = 0) is

N4
0 1
given in our setup since the DC voltage source is an ideal AC short.

I=
1
4
0A
u PQ _
V
U1
0
= C
A
1
y
p
T
t
a
r
S
o
nl
e
o
g
=
k
H
1
t
G
0
s
i=z
1
d
c
s
Ci
l
m
a
u
1b
D i
t
o
n q
E
q
iE
c
=n
'
ta
e a
u
1
Vi
t
o
n
.
i
/ q
a
u
E
2
q
n
E
d
t
e
a
b
_i
t
o
n
B
d
B
t(
=
lu
( /d
ic1
e
'
B)
,6
0
Ic
=
I2 x
v
bf=e
e
ba_

B
d 40
e_2201e3a
tba n0cy1:6e25.a58cfer+
bcftr1e_eqd4uB 08 n
u
1b
m
e
rt
f
8
4
e
.
20
8
+
Figure 2.38: AC setup for determining the transit frequency

There are two more equations in the setup. One calculates the AC current gain

e
16 7
e
18
e
1
19
e
in dB (which is 20 · log (beta) and the other one is ft=xvalue(beta dB,0). The

quncy
equation searches for the nearest given x-value (in this case the frequency) where
beta dB approaches 0.

Figure 2.39: Bode plot of the current transfer function

In fig. 2.39 the Bode plot (double logarithmic plot) of the current transfer func-
tion of the bipolar transistor is shown. The current gain is constant up to the
corner frequency and then drops by 20dB/decade. The marker finally denotes
where the gain is finally 0dB. The equation for ft worked correctly as seen in the
beside tabular. The transit frequency of the bipolar transistor in this bias point is
approximately 288MHz.

46
2.3.4 AC simulation - A simple RC highpass
Simple circuit AC analysis (circuit frequency response analysis) can be carried out
easily by using the AC Simulation block.

For instance, a simple high pass RC filter can be analyzed by constructing first the
schematic displayed on figure 2.40 which corresponds to a high pass RC network.

Figure 2.40: simple RC high-pass filter schematic

Performing the actual AC analysis is as easy as dragging and dropping an AC

Simulation block available under the Simulations tab as can be seen in figure 2.41.

47
 Figure 2.41: AC simulation block placed

Once this is done one must configure the ranges of the simulation analysis by
clicking twice on the AC Simulation box as can be seen in figure 2.42.

Figure 2.42: AC simulation block configuration dialog

Finally by pressing F2 the simulation takes places and a graphic report can be
generated by selecting the right plot as seen in the previous sections. The final view
of the network with its respective frequency analysis can be seen on figure 2.43.

48
 Figure 2.43: AC simulation results

2.3.5 Transient simulation - Amplification of a bipolar

transistor
Based on the schematic in fig. 2.38 we are now going to simulate the bipolar
transistor in the time domain.

49
 I2
1
f=A
u
k
H
z t
N
2
4
0
Q1
_
1 i
r
a
n
s
i
l
s
m
a
u
R
1e
n
t
i
on
I=
1
4AUS
0
u V
=0 T
l
i
y
p
e
n
=
0
t
t
a
r
5
o
i
2
n
s
=
Pm
s
0
1
q
E
q
n
E
I
c
=<
Vt
a
u
1i
o
n
I
t
.m t
q
a
u
E
2
q
n
E
C
t
e
a
B
Di
o
n
1
=
<
VI
/I
1
.. i
c
s
m
d
C
1
Dl
a
ui
o
n
t
ta
e
B
H R
T
(H
c
a
=
(t/)
Ic
a
x I2
<.Iin
m (
Ic
)
/2
Figure 2.44: Transient simulation setup

As shown in fig. 2.44 the transient simulation block was placed on the schematic.
Also the frequency f of the AC current source I2 was set to 1kHz. The start time
of the transient simulation is set to 0 and the stop time to 5ms which will include
5 periods of the input signal.
The additional DC simulation block is not necessary for the transient simulation
but left there for some result comparison.
The collector current in the equations is denoted by the transient current -V1.It.
The peak value if the collector current is determined by the equation for IcHat.
The current gain during transient simulation is calculated using BetaTR=IcHat/I2.I
whereas I2.I denotes the component property I of the the current source I2 (which
is 1µA peak). The current gain BetaDC is computed for convenience.
The equation blocks imply that the order of appearance of assignments does not
matter (e.g. IcHat is used before computed). The equation solver will take care
of such dependencies.

50
.Ic000..33344426 n1um ber0Ic.H 4e5taT
at2452B R2B C
4e6taD
00.2tim
e0.4 Figure 2.45: Transient results

Fig. 2.45 shows the results of the transient as well as DC simulation. The time
dependent collector current oscillates around its bias point. The current gain of
the transient signal corresponds perfectly with the DC value. That is because a
rather small frequency of 1kHz was chosen.

2.3.6 S-parameter simulation - Transit frequency of a BJT

In the following section the S-parameter simulation is introduced. The S-parameter
simulation is – similar to the AC simulation – a small signal analysis in the fre-
quency domain.

51
 PZN1u5m 1 3
I
1
40 A
Culation=uSsip1aurlm
0=dDCc1sim Q 2N4
0
1
_1
V
0
=
U X 1 2
P
N
u
m
=
5
0
Z
iTSytPoapinre=sGgk0H1zoentrbHEe=qnw2u_adptB[iorn1(]S,b'etaH)
Figure 2.46: S-parameter simulation setup for the bipolar transistor

Similar to the AC setup in fig. 2.38 the S-parameter setup in fig. 2.46 uses the
same biasing. The setup will be used to determine the transit frequency of the
bipolar transistor.
The two AC power sources P1 and P2 are required for a two-port S-parameter
simulation. They can be found in the Components tab in the sources category.
Depending on the number of these kind of sources one-port, two-port and multi-
port simulations are performed. The Num property of the sources determines
the location of the matrix entries in the resulting S-parameter matrix. The Z
properties define the reference impedance of the S-parameters.
The additional DC block C1 at the base node and the bias tee X1 on the collector
is used to decouple the signal path of the biasing DC sources from the internal
impedance of the AC power sources. Also the bias tee ensures that the AC signal
from the P2 source is not shorted by the DC source V1. The same functionality
is achieved by the DC current source I3 at the base. It represents an ideal AC
open.
The S-parameter simulation itself is selected by placing the S-parameter block
SP1 on the schematic. The same frequency range is chosen as in the previous AC
simulations.
The equations contain a two-port conversion function which convert the resulting
S-parameter S into the appropriate H-parameters H. Again the AC current gain
h21 is calculated and converted in dB.

52
 ,[ 20460810[],S
]S
1 21][, 0.50.10.15[],S
12
frequncy frequncy
B 40
Figure 2.47: S-parameters of the bipolar transistor

d
tb _
a
e 2 c1ye0:52f.r68q5eu1+
fj:c 201e31fbree4tqau_dnB 0 8
In fig. 2.47 the four complex S-parameters are displayed in two Polar-Smith
Combi diagrams. They represent what can be expected from a typical bipolar
transistor.

en6cy1e71e81e9
Using the computed H-parameters we can now compare the S-parameter simulation
results with those of the AC simulation. Fig. 2.48 shows that the curves beta dB
of both simulation setups cover perfectly each other. Again the transit frequency
is approximately 288MHz.

Figure 2.48: Comparison between S-parameter and AC result

The diagram implies that you can compare data curves from different setups. This
is indicated by the bjtacft: prefix. The appropriate dataset file bjtacft.dat can
be selected in the diagram dialog as shown in fig. 2.49.

53
 Figure 2.49: Choosing graphs from different datasets

The current S-parameter setup is called bjtsp and the setup shown in fig. 2.38
was called bjtacft. Please note that only datasets from the same project can be
compared with each other.

2.3.7 S-parameter and AC simulation - A Bessel band-pass

filter
The interested reader may have noticed that there seems to be a relationship
between AC analysis and the S-parameter simulation. In the next section we are
going to explain this relationship using a simple filter design.

54
 Figure 2.50: Filter synthesis application

In fig. 2.50 the filter synthesis program coming with Qucs is shown. You can start
it by the Ctrl + 2 shortcut or by choosing the Tools → Filter synthesis
menu entry. The user can choose between different types of filters and the filter
class (lowpass, highpass, bandpass or bandstop). Also the appropriate corner
frequencies and the order must be configured. When setup correctly you press the
Calculate and put into Clipboard button. The program will indicate if it was
possible to create the appropriate filter schematic. If so, the application passes the
schematic to the system wide clipboard.
Back in the schematic editor you can paste the filter design into the schematic
using the Ctrl + V shortcut or by choosing the Edit → Paste menu entry.

55
 P1 m L
12
C
3
1
=
.
2
8L
F
8
n
H2
4
0
=
.H
3
6
u4
C
=
L
3
4
9F
1
3
2
.
n
HL
4
8
=H
1
.
uL
51
7
6H
22
P
m
N
Zu 5 0=1
C
5
=F
4
9
.
pui
tC3
2
5
9
=
.
nFu 5
C
7
1
8
=
.=
F
n.u=
N
u
5
0
Z
rs
S
iT
lS p
iytoP
m 1 a
u m
ls2o0gM
apinre= t
e
r
o
n
kH
z q
u
E
n
S
B
da
o
n
1
S
B
(
2
=
d[
]
2
1
,)B
1
i
ml
e
s
b
H
z
.
M
d
pd
a
n
J
H
2
z
M
c
e
mp
a
s
f
I
y
J
,
P
h
t
c
ni
l
t
e
r
p
,
0
g
5h
Om
Figure 2.51: Schematic for 5th order Bessel band-pass filter

0
The schematic shown in fig. 2.51 was automatically created by the filter synthesis

d211050e5fr1eq6uncy1e72e7S
S ][1,frequncyS
][2,frequncy
program and can be simulated as is. It contains the LC-ladder network form-

B
ing the actual filter, the two S-parameter ports (the AC power sources) as well
the S-parameter simulation block with the appropriate frequencies pre-configured.
Additionally there is an equation computing the transmission and reflection of the
filter network in dB.

Figure 2.52: S-parameters of the band-pass filter

The results of the S-parameter simulation are depicted in fig. 2.52. In the logarith-
mic cartesian diagram the transmission of the filter clearly shows the band-pass
behaviour between the selected frequencies 1MHz and 2MHz. Additionally the

56
 P1 L 2
=4.03
6uH L
4
8
=.1uH 2
P
input- and output reflections can be seen in the two Smith charts.

R
1 C2
3
1
=
.8nF C 4
1
=F
3
2
.
n R
2
5
0
Now two AC setups will be created to calculate the same S-parameters as found

5
=0 L
1 L3 L5 =
in the previous simulation. In fig. 2.53 the LC-ladder network is unchanged but

2H
8
u 4 9 H
u 1
7
6
=
.H
2
u
the S-parameter ports are replaced by a 50Ω resistor and an AC voltage source in

1
V 2
V
0
series. Also there is now an AC simulation block with the same frequency sweep

=
Uacsim =
U
chosen as in the previous S-parameter simulation.

lA
lP
yS
T C p1e= ou
g1
C
5
4
9
=
.
p
i
t
o
n
aFa
S E
n
=q
1
(a
u
.
P
bt
v
/
ai
+o
Z
1C
n
0 3
2
=
*
B
V.5
1
.i9
)n
/
(F
2*s
q rt(
Z)
0 C5
7
= F
1
8
.
n
q
E
2
n
B
d
Si
t
a
o
u
B
1
=
dn
(
)
1
S
2
kH
toainrs20MzZ 2
0 R2
b1=(P
.vBZ0*V
.0i)/(*sqrt(Z
2 0)
1acfrequncyS
S 1
BdB1050e5acf1re6
2
B quncy1e72e7
Figure 2.53: S-parameters at port 1 of the band-pass filter using AC analysis

At this point some theory must be stressed.

57
S-parameters are defined by ingoing (a) and outgoing (b) power waves:

V + Z0 · I
a= √
2 · Z0
V − Z0 · I
b= √
2 · Z0
whereas Z0 denotes the reference impedance the S-parameters will be normalized
to. With this definition the two-port S-parameters can be written as:
   
b1  b2  b2  b1 
S11 =  S21 =  S22 =  S12 = 
a1 b2 =0 a1 b2 =0 a2 b1 =0 a2 b1 =0

Back at the schematic in fig. 2.53. The amplitude of the AC voltage source V1 is
set to 1V (but can be any other value different from zero) and the side condition
b2 = 0 is fulfilled by setting the amplitude of the AC voltage source V2 to 0V.
The additional equations just calculate the S-parameters as they are defined from
the AC simulation values.
Please note the current directions through the AC voltages sources V1.i and V2.i.
They must be considered by the unary minus in the equations.
The results of this simulation again show the filter transmission function as we
already know it from the S-parameter simulation. Also the reflections at port 1
look identical.
In the second schematic shown in fig. 2.54 the second port is handled. The am-
plitude of the AC voltage source V2 is set to 1V and the side condition b1 = 0
considered by a zero AC voltage source V1. Again the appropriate equations are
used to compute the two remaining S-parameters.
The below simulation results again verified that we can perform a partial S-
parameter analysis using the AC simulation block and some additional equations.
The diagrams in fig. 2.54 and fig. 2.52 are identical.

58
1
R
5
0
= P1L12
C
3
1
8
=
.
H
2
8F
nL 2=40
.3
6H
u L
3
4C4
=
91 .3
H 2F
nL
4
8
=
.H
1
uL
5
1
7
6H
2 2
P
2
R
5
0
=
1
V
0
=
Uacsim
lA 1
C
F
5
4
9
=
.
p
i u C3
2
=5
.
i
t F
9
n u C5
7
1
8
=
.
n=
.
Fu
i
t 2
V
1
=
U
C
lT
yS 1
roine=
tP
a p ot
o
n
u
a
g
H
0
z
kZ
s2M a
S2 E
n
=(q
u
1
P
b
b201=(R
Pa
v
.
/
2
v
.o
Z
+
a
Z
@ 0*
0
@
V
*i
2
.
1)/
( *s
q rt(
Z
0
i21@500)/(*sqrt(Z
V2
. )
0) q
u
E
2
n
B
d
S
1a
o
n
B
(
2
=
d
S
1)
2acfrequncyS
S B
d@01e5ac1fre6quncy1e72e7
Figure 2.54: S-parameters at port 2 of the band-pass filter using AC analysis

Recapitulating we learned from this example that a S-parameter simulation is

a number of AC simulations with some additional calculation formulas. This is
true though the actual simulation algorithms implemented in Qucs are completely
different.

59
3 Understanding RF Data Sheet
Parameters
. . . prepared by Norman E.Dye from Motorola RF Division : AN 11071 . Since this
AN is essential to our topics, it is good to make a small reference to it. All AN
from Motorola are a reference is this field. This chapter is only an extract, but the
main points are hilighted herein. . . .
The author.

3.1 Introduction
Data sheets are often the sole source of information about the capability and
characteristics of a product. This is particularly true of unique RF semiconductor
devices that are used by equipment designers all over the world. Because the
circuit designer often cannot talk directly with the factory, he relies on the data
sheet for his device information. And for RF devices, many of the specifications
are unique in themselves. Thus it is important that the user and the manufacturer
of RF products speak a common language, what the semiconductor manufacturer
says about his RF device is understood fully by the circuit designer.
This paper reviews RF transistor and amplifier module parameters from maximum
ratings to functional characteristics. It is divided into five basic sections:
1. DC specifications,
2. power transistors,
3. low power transistor,
4. power modules,
5. linear modules.
Comments are made about critical specifications about how values are determined
and what are their significance.
1
This note could be found on old application notes databook from Motorola, if you have one
keep them, it is a real treasure.

60
3.2 DC specifications
Basically, RF transistors are characterized by two types of parameters: DC and
functional. The ”DC” specs consist of breakdown voltage, leakage current, hF E (
DC β ) and capacitances, while the functional specs cover gain, ruggedness, noise
figure, Zin and Zout , S parameters, distortion, etc . . . . Thermal characteristics do
not fall cleanly into either category since thermal resistance and power dissipation
can be either DC or AC. Thus we will treat the spec of thermal resistance as a
special specification and give it its own heading called ”thermal characteristics”.

3.3 Maximum ratings and thermal characteristics

61
4 DC Analysis, Parameter Sweep
and Device Models
4.1 DC Static Circuits
A favourite question in electronics courses used to be:

You have twelve one ohm resistors; you connect them together so
that each resistor lies along the edge of a cube. What is the resistance
between opposite corners of the cube?

The intention may have been to teach soldering, as more than one student solved
it by making just such a cube! These days we can do that without touching the
soldering iron; we simulate the circuit.

Here is my attempt to make a cube in Qucs; anyone is welcome to try and improve
it.

62
 Figure 4.1: resistor cube schematic

All I did was select resistance in the left hand component window and paste them
down, rotating as necessary, until I had twelve on the schematic. Then I wired two
sets of four into squares, then connected the remaining four between the corners
of the squares. Which I’m sure is topologically the same as a cube.

Which all might seem trivial, but is a good reminder right at the beginning that
we are creating a virtual representation of a physical circuit. Sometimes we have
to bend and squeeze things to get it into a format that our simulator will accept,
which leaves us wondering whether we are working with an accurate representation.

The Rule is: if we can correlate the junctions of our components with those of
the real circuit, we are accurately representing the physical circuit. And, I might
add, it is ALWAYS worth checking that we have done it right; simulate the wrong
circuit and it will tell you lies.

With my cube of resistors accurately drawn, I only have to hit the simulation
button and the tabulated results will show me the voltage at the corner node. As I
am forcing a constant current through the cube from one corner to another, Ohm’s

63
Law tells me that the voltage between those corners will give me the resistance.
If I use a current of one amp, the output voltage will be equal to the resistance in
ohms.1

Those with good attention to detail will be complaining about now that I haven’t
really solved the problem, as the question mentioned one ohm resistors while I
have used fifty ohms. Well, yes, I cheated. Which I often do in simulations.

To set all the resistances to the correct value I would have had to open the Prop-
erties Editor window twelve times; here is how it looks...

Figure 4.2: component property dialog

and the highlighted value is inviting me to type in an alternative. I could have

done this, but natural laziness got the better of me. I reasoned that fifty ohms is
fifty times too high, but if I reduced the current source from one amp to twenty
milliamps, the output voltage would be the same. You will find such laziness (or
acute perception, depending on is telling the story!) can save much time and effort.

4.2 When Things Vary

All of which is interesting, but not nearly as interesting as when we start changing
things like the supply voltage and see the effects. For linear devices with a DC
1
I could tell you the value my simulation gave, but why should I spoil your fun.... go ahead
and run it yourself. If you really want to be thorough you could then also build the circuit
and measure the result.....

64
supply, the answer would be: not much. It’s when we introduce non-linear elements
that things start to happen.

The simplest non-linear element is the diode, and the question we ask most often
about a diode is: how does the diode forward voltage vary with current? So back
to Qucs and draw this circuit...

This circuit looks deceptively simple, but it introduces a few more features of Qucs,
so let’s go through them in order.

The components were again selected from the left hand window and wired together.
Then the two boxes were selected from the simulations window.

The DC simulation box can be pretty much left as is for now, but take note of the
name of the simulation: DC1.

The Parameter sweep box properties dialog looks like this when opened...

65
The first two items to take note of are the Simulation entry (here DC1, corre-
sponding to the name of the simulation box) and the Sweep Parameter entry,
here entered as Id1. If you look at the current source driving our diode you will
see that it just happens to be labeled Idrive. So the result of all this is that the
component property value Id1 of the current source’s property I will be swept
through a range of values as determined by our parameter sweep function named
SW1.2

The rest of the entries set the type of sweep (here logarithmic) and the range of
values over which to sweep. You can try different values in any of these to see the
effect; one of the advantages of a simulator over a physical prototype is that you
can’t blow up your diode by feeding too much current through it!

So I hit the simulation button and it passed me over the results page, and I created
a couple of graphs of the output. This is how my screen looked...

2
You can change this name if you wish, in the Properties menu of the Edit properties window.

66
In each case I have a plot of diode forward voltage (Y-axis) against forward current
(X-axis). The left hand graph has a logarithmic scale for forward current, while
the right hand graph uses a linear current scale. How did I do that? Well, you
should know by now that all things are easy with Qucs!

When you select a graph type from the left hand window and drag it into the
viewing area, it creates a graph and opens a dialog which looks like this

67
The left hand window shows the available variables and whether they are depen-
dent or independent. Here the current Id1 is the independent variable, and the
forward voltage Vdf.V is the dependent. Double-click on the entry for Vdf.V and
it is transferred to the right hand side; hit OK and the graph will be drawn.

That should give you something like the right hand graph in my screenshot above.
Do it all again, but this time before clicking OK open the Properties window,
which looks like this.

68
Here I’ve selected a logarithmic X Axis, which gave me the graph on the left hand
side. I’ve also moved them around and re-sized them to pretty them up; you can
do all kinds of fancy things if you want.

Now I’ve sneaked in another test to see if you are really following this. Those of you
who did run this simulation are probably wondering about now why your graphs
look rather different to mine. In particular, at high currents on the logarithmic
scale your curve is a straight line, while mine curves upwards alarmingly. What is
happening ?

What I did was open the Properties dialog for the diode and set some parameters.
This is what the dialog box looks like...

69
and each of these entries sets one parameter of the virtual component we are using
to model the diode.

So, what are these parameters? Time to explore one of the delights of computer
circuit simulation, device modeling...

4.3 Models and Parameters

When the computer creates that small piece of virtual reality which represents
your physical circuit, it uses sets of equations which describe the operation of each
device you insert. The equation which relates the diode DC forward voltage as a
function of current is
 V 
d
Id = Is · e n · Vt − 1

where Vt is the forward voltage drop at 25 degrees C of an ideal junction, also

given by
kB · T
Vt =
q

70
where

kB = Boltzmann’s constant
T = temperature in degrees Kelvin
q = charge of the electron

most of these are constants that the program already knows about. The ones we
need to supply are the ones listed in the properties editor window. For the DC
characteristics, most of the time, the only ones we need to worry about are Is, the
saturation current, and T, the temperature. If we are going to push relatively high
currents through the diode we can also include an estimate for the series resistance
Rs; if we are worried about low current behaviour then we need to add the reverse
current parameter Isr.

How do we know what values to insert? Much could be written about device
modeling; much indeed has been written about device modeling. As always, we
really have two choices: use a value from someone else, or find our own values,
usually by trial and error.

There are a great many models available for various simulation programs. Probably
the most freely available are those for spice, many of which can be downloaded
from the semiconductor companies. Here, for example, is a typical spice model for
a 1N4148 diode:3

.model 1N4148 D(Is=0.1p Rs=16 CJO=2p Tt=12n Bv=100 Ibv=0.1p)

85-??-?? Original library

Any values not supplied are assumed to be the defaults.

The other way is to create your own device parameters, which is a bit like catching
worms before you can go fishing. Insert values, plot the resulting characteristics,
see how they compare with the published data sheet values, go back and adjust
the values; continue until satisfied or exhausted.

Here, for example, is a circuit for quickly comparing the forward characteristics of
diodes with different parameter values.

3
I don’t know where this came from, so I can’t acknowledge the author. Most libraries are
copyright, even if freely available.

71
 Vi4

Idrive
I=Id1 D1
Is=1e-15 A
Equation N=1
Eqn1 Cj0=10 fF
M=0.5 Vi3
Vd2=Vi2.V-Vd1.V
Vd3=Vi3.V-Vi2.V Vj=0.7 V
Vd4=Vi4.V-Vi3.V D2
Export=yes Is=1e-14 A
N=1
Cj0=10 fF
Vi2
Parameter M=0.5
sweep Vj=0.7 V

SW1 D3
Sim=DC1 Is=3e-18A
Type=log N=1
Vd1
Param=Id1 Cj0=10 fF
Start=1e-6 M=0.5
Stop=1 Vj=0.7 V
Points=1000
D4
dc simulation Is=1e-9A
N=1.025
DC1 Cj0=10 fF
M=0.5
Vj=0.35 V

And here is the plotted output...

0.9

0.8

0.7

0.6
Vd1.V
Vd4
Vd3
Vd2

0.5

0.4

0.3

0.2

0.1

0
1.0e-6 1.0e-5 1.0e-4 1.0e-3 0.01 0.1 1
Id1
Id1
Id1
Id1
Figure 4.3: Diode Forward Voltage

The green and purple curves are typical of 1N4148 and 1N4448 devices; the others
are medium and low-barrier Schottky devices. I have done a first pass compare
with the data sheets, but I can’t guarantee that these curves are any more than
my best estimates.4
4
I’m assuming you are sick of screenshots by now, so I’ve just printed the schematic and display
files from Qucs; you’ll find the print item in the file menu, and if you ask it nicely it will print
a postscript file.

72
If you want to know more details of what each parameter does, there has been a
great deal written over the years, particularly for spice, on the subject; a google
search will quickly reveal most of it. Qucs comes with a document which lists the
details of its models, and, being open source, there is always the code itself.

Most of us end up taking a great deal on trust, and matching curves to data
sheets as best we can. This is yet another instance of one of the fundamentals of
engineering, the Duck Principle5 : If you can’t detect any difference between the
behaviour of your model and the physical device, then they are, for engineering
purposes, the same. Put it another way, when the difference between the model
and the real device drops below the usual level of measurement uncertainty, it does
matter any more.

In any case, component spreads in the real world tend to make the fine details
of model inaccuracies somewhat academic, as we shall see when we model more
complex devices.

5
Usually expressed as: If it looks like a duck, walks like a duck, quacks like a duck and tastes
like a duck, then, for all practical purposes, it is a duck.

73
5 Getting Started with Digital
Circuit Simulation

5.1 Introduction
On 21 January 2006 Qucs 0.0.8 was released by the Qucs development team. This
is the first version of the package to include digital circuit simulation based on
VHDL. FreeHDL1 being chosen as the VHDL engine. In the period following the
release of Qucs 0.0.8 there has been considerable activity centred around finding
and correcting a number of bugs in the Qucs digital simulation code. Many of these
fixes are now included in the latest CVS code and will eventually form part of the
next Qucs release. This tutorial note is an attempt on my part to communicate
to other Qucs users a number of background ideas concerning the capabilities and
limitations of the current state of Qucs VHDL simulation. Much of the information
reported here was assembled by the author while assisting Michael Margraf to test
and debug the VHDL code generated by Qucs. In the future, if there is enough
interest in these notes, or indeed in Qucs VHDL simulation in general, I will update
them as the Qucs digital simulation features are improved.

Qucs digital simulation follows a complex set of steps that are mostly transparent
to the software user. In step one, a schematic representing a digital circuit under
test is drawn. This schematic consists of an interconnected group of Qucs digital
components, one or more user defined digital subcircuits (if required), and a copy
of the digital simulation icon with the timing or truth table parameters set. In
step two, the information recorded on a circuit schematic is converted into a text
file containing VHDL statements. These describe the circuit components, their
connection, and a testbench for simulating circuit performance. Next, FreeHDL is
launched by Qucs to convert the VHDL code file into a C++ source program. This
is compiled to form an executable machine code simulation of the original circuit.
Finally, Qucs runs this program, collects signal data as digital signal events take
place and displays signal waveforms as a function of time or digital data in a truth
table format.
1
The FreeHDL Project, http://www.freehdl.seul.org/.

74
The VHDL code generated by Qucs 0.0.8 is limited in its scope by the following
factors:

• Digital gates are described by data flow concurrent statements.

• Flip-flops and the digital signal generator are described by process state-
ments.

• Component connection wires (signals) can only be of type bit as defined in

the standard VHDL library2 .

• Digital bus structures are not allowed in this release of the Qucs package.

• Digital subcircuits can be drawn as schematics and associated with a symbol

in a similar fashion to analogue subcircuits.

• Digital subcircuit pins can have type in, out, inout or analog. Qucs treats
pins of type analog the same as VHDL pin type inout.

• Once defined digital subcircuits may be placed and connected to other com-
ponents on schematics.

• Multiple copies of the same digital subcircuit are allowed on a single schematic.

• Digital subcircuits may also be nested; nesting has been tested to a depth of
four.

5.2 Simulating simple digital circuits

The most basic form of digital circuit that can be simulated is one consisting
entirely of Qucs predefined digital components drawn on a schematic having only
one level of design hierarchy. The truth table for a simple combinational circuit of
this type is shown in Table 8.1.

Output F can be expressed in sum of products Boolean form as

F = A.B.C + A.B.C + A.B.C + A.B.C

On minimisation, using Boolean algebra or a Karnaugh map, output F becomes
2
Signal type bit only defines logic signals ’0’ and ’1’. Care must be taken to ensure that
signal contention does not occur during simulation because the resulting logic state cannot
be modelled with type bit. Signal contention can happen when two or more digital devices
attempt to drive the same wire with logic ’0’ and logic ’1’ signals at the same time. Moreover,
it is not possible to simulate the performance of tristate devices using VHDL signal type bit.

75
 A B C F
0 0 0 0
0 0 1 1
0 1 0 1
0 1 1 0
1 0 0 0
1 0 1 1
1 1 0 1
1 1 1 0

Table 5.1: Truth table for a logic circuit with inputs A, B, C and output F.

F = A.C + B.C
The schematic for example 1 is illustrated in Fig. 5.1. This diagram was con-
structed using the same techniques employed for drawing analogue schematics.

5.2.1 Notes on drawing digital schematics

• The only predefined Qucs components that can be used to draw a digital
circuit schematic are (1) the digital components listed in the digital com-
ponents icon window, (2) the ground symbol, and (3) the digital simulation
icon.

• A useful tip when drawing digital schematics is to adopt the matrix approach
shown in Fig. 5.1. Input signals flow from top to bottom of the schematic
and output signals are positioned on the right-hand side of a horizontal line.
This makes checking the circuit schematic for errors much easier than the
case where diagrams have wires connecting components in an unstructured
way.

• Input and output wires (signals) should be given names consistant with the
circuit being simulated, A, B, C and F in Fig. 5.1. If the signal wires are
not named by the user, Qucs will allocate them different arbitrary names.
This can make identification and selection of signals for display on an output
waveform graph, and indeed checking for errors in a large circuit, much more
difficult than it need be.

• Notice in Fig. 5.1 the international symbols for the logic gates are shown on
the schematic.

76
 A B C

digital
simulation
Y1

1
Digi1
A B
Type=TimeList
C
time=200ns

&
1 F
CB Y2

Y4
&

Y3

Figure 5.1: Qucs schematic for minimised logic function F.

5.3 VHDL code generated by Qucs

Clicking the Qucs Simulate menu button (or pressing key F2) starts the simulation
process. At an early phase in this process Qucs writes a text file to disk that
contains the VHDL code for the circuit being simulated. This file can be displayed
by clicking on the show last netlist drop down menu or by pressing key F6. The
VHDL code produced by Qucs for the circuit shown in Fig. 5.1 is presented in
Table 6.1.

Signals identified by nnnet0 and nnnet1 in Table 6.1 have been allocated these
names by Qucs; nnnet0 and nnnet1 are internal signal nets that are not named
on the circuit schematic shown in Fig. 5.1. Fig. 5.2 illustrates the starting section
of a typical Qucs digital functional waveform plot. This style of plot illustrates
signal events without component delays. If required, signal delays can be specified
for individual gates and other components (from the component edit properties
menu). The VHDL code generated for components with delays will then reflect
such changes, for example adding a 10 ns delay to signal CB in Table 6.1 generates
VHDL code
CB <= not C a f t e r 10 ns ;
Readers will probably have observed that the Qucs version number referred to in
Table 6.1 VHDL listing is 0.0.9. This is the current CVS development version
number. Qucs 0.0.9 includes a number of important bug fixes. The remainder of

77
−− Qucs 0 . 0 . 9 t u t 1 e x 1 . s c h
entity TestBench i s
end entity ;
use work . a l l ;

architecture Arch TestBench of TestBench i s

s i g n a l CB, A, B, F , C,
nnnet0 ,
nnnet1 : b i t ;
begin
nnnet0 <= C and A;
nnnet1 <= CB and B ;
CB <= not C ;

A: process
begin
A <= ’ 0 ’ ; wait f o r 40 ns ;
A <= ’ 1 ’ ; wait f o r 40 ns ;
end process ;

B : process
begin
B <= ’ 0 ’ ; wait f o r 20 ns ;
B <= ’ 1 ’ ; wait f o r 20 ns ;
end process ;

F <= nnnet1 or nnnet0 ;

C: process
begin
C <= ’ 0 ’ ; wait f o r 10 ns ;
C <= ’ 1 ’ ; wait f o r 10 ns ;
end process ;

end architecture ;

Table 5.2: VHDL code for the circuit shown in Fig. 5.1.

78
these notes assume readers have downloaded, and recompiled, the latest CVS code
from Sourceforge.net3 .
dtime 0 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n 110n 120n 130n 140n 150n 160n 170n 180n 190n 200n
a.X
b.X
c.X
f.X

Figure 5.2: Digital functional waveforms for the circuit shown in Fig. 5.1.

5.4 Truth tables

Truth tables are one of the most fundamental and convenient ways of displaying
digital circuit data. Qucs has a built-in facility that allows a truth table to be
generated from a schematic drawing. This feature is particularly useful when
checking minimised logic designs for errors. Lets consider a simple but instructive
example: A logic circuit has four binary inputs A, B, C, and D, and one output
P. Output P is logic ’1’ when inputs ABCD are numbers in the decimal sequence
3, 5, 7, 11 and 13. In Boolean sum of product form
P = A.B.C.D + A.B.C.D + A.B.C.D + A.B.C.D + A.B.C.D
This simplifies to
P = D.(A.B + B ⊕ C)
The schematic for the sum of products equation for P is shown in Fig. 5.3(a).
Similarly Fig. 5.3(b) presents the schematic for a minimised P equation. Setting
the digital simulation type to TruthTable, rather than TimeList, causes Qucs on
pressing key F2, to generate a truth table based on the information provided
on a circuit schematic. The number of truth table inputs, and indeed outputs,
correspond to the number of input generators and the number of named outputs.
Truth tables for both schematics are given in Table 5.3(a) and 5.3(b). Comparing
these two tables clearly indicates that they are not identical and moreover confirms
that the minimised solution is not correct. Reworking the minimisation procedure
points to the error being a missing signal inversion. The correct Boolean equation
for P is
P = D.(A.B + B ⊕ C)
3
Please note, Qucs Linux release 0.0.8 will normally simulate single hierarchy digital circuits
without error. However, Qucs 0.0.8 does fail at the VHDL to C++ conversion phase if a
schematic includes more than one ground symbol.

79
 A B C D
digital
A Y1 B
Y2 C Y3 D simulation

1
Digi1
Type=TruthTable

&

Y6
&
1

P
Y7
& Y10

Y8
&

Y9

&

Y11

5.3(a): Schematic diagram for sum of products equation P

A B C D digital
D
A B C simulation
Y3
Digi1
1

Type=TruthTable

&
1
& P
Y7
Y8
=1 Y6

Y4

5.3(b): Schematic diagram for minimised equation P

80
 5.3(a): Truth table for sum of products
equation P
a.X b.X c.X d.X p.X
00000 0 0 0 0 0
00001 0 0 0 1 0
00010 0 0 1 0 0
00011 0 0 1 1 1
00100 0 1 0 0 0
00101 0 1 0 1 1
00110 0 1 1 0 0
00111 0 1 1 1 1
01000 1 0 0 0 0
01001 1 0 0 1 0
01010 1 0 1 0 0
01011 1 0 1 1 1
01100 1 1 0 0 0
01101 1 1 0 1 1
01110 1 1 1 0 0
01111 1 1 1 1 0

5.3(b): Truth table for minimised equation

P
a.X b.X c.X d.X p.X
00000 0 0 0 0 0
00001 0 0 0 1 0
00010 0 0 1 0 0
00011 0 0 1 1 1
00100 0 1 0 0 0
00101 0 1 0 1 1
00110 0 1 1 0 0
00111 0 1 1 1 0
01000 1 0 0 0 0
01001 1 0 0 1 0
01010 1 0 1 0 0
01011 1 0 1 1 1
01100 1 1 0 0 0
01101 1 1 0 1 1
01110 1 1 1 0 0
01111 1 1 1 1 1

81
5.5 Digital subcircuits
Although it is possible to draw complex schematic diagrams using only the pre-
defined digital components supplied with Qucs, this technique can be extremely
tedious, and is of course, prone to error. When drawing large schematics we re-
quire a design procedure that naturally subdivides groups of digital components
into self contained units. These units can then be treated in the same way as
basic digital components when placing and connecting them on a schematic draw-
ing. In the world of analogue and digital circuit design such units are often called
subcircuits.4 A subcircuit is defined by three major attributes plus a number of
other properties. The major attributes are, firstly a digital circuit that defines
circuit function, secondly a circuit symbol that depicts a circuit in a higher level
of a design hierarchy, and thirdly the subcircuit input/output pins shown on the
subcircuit symbol. Other properties include for example, signal path delays. The
process for generating digital subcircuits is identical to that used for analogue sub-
circuits. It is best demonstrated by considering an example. Figure 5.4 shows the
schematic for a four input combinational circuit.

After drawing a subcircuit schematic, input and output5 pins are attached to signal
ports. Input port pins of type in are shown on circuit diagrams as a green symbol,
signals W, X, Y, and Z, in Fig. 5.4. Ouput port pins of type out are coloured red,
signal G in Fig. 5.4. Signal flow through a port is indicated by the direction of the
port symbol arrow head. Input/output signals, and any other signals that need
to be easily identified, are also named. Once the subcircuit schematic is complete,
pressing key F3 causes Qucs to generate a subcircuit symbol. The drawing tools
listed as icons in the Qucs paintings window can be used to edit Qucs generated
subcircuit symbols. The input/output port pins on a subcircuit symbol have the
same type and name as those on the original subcircuit schematic. Fig. 5.5 shows
the finished symbol for subcircuit COMB1. In these notes, symbol outlines are
shown drawn in accordance with the international code for logic symbols6 . To test
our new subcircuit we place it’s symbol on a blank drawing sheet and apply test
signals to the input pins and observe the signals at the output pin. Fig. 5.6 shows a
typical test circuit. Subcircuit Gen4bit generates a 4 bit test pattern synchronised
4
The circuit simulator SPICE is a well known example of a widely used CAD program that
makes extensive use of subcircuits in circuit design.
5
Qucs 0.0.8 has a bug which causes a VHDL compile error when subcircuit pins are specified
as type out. A work around for this bug is to specify subcircuit output pins as type analog.
The Qucs routines that generate the circuit VHDL code convert pin type analog into VHDL
type inout. FreeHDL is then able to compile the generated VHDL code without error. This
bug has been corrected in Qucs 0.0.9.
6
Ian, Kampel, A practical introduction to the new logic symbols, Butterworths, 1985, ISBN
0-408-01461-X.

82
 Z Y X W
Y
Y3 Y2 Y1

1
Z XB
YB
& IN1

Y4
X WB

IN2
&
1 G

Y5
G
Y8
& IN3

Y6

& In4

Y7

Figure 5.4: Combinational logic circuit with inputs W, X, Y, Z, and output G.

to the input of a digital clock. The specification for Gen4bit is given in the next
section of these notes7 . The test pattern waveform and output signal G are shown
plotted as a function of time in Fig. 5.7.

7
Subcircuit Gen4bit includes other nested subcircuits. Qucs 0.0.8 has a bug that causes VHDL
compile errors with some configurations of nested subcircuits. This has been fixed in version
0.0.9.

COMB1
W W

X X
G G
Y Y

Z Z

SUB
File=name

Figure 5.5: Qucs symbol for a logic circuit with inputs W, X, Y, Z, and output G.

83
 R B0 COMB1
R B0 W
R B1 B1
times=5ns; 1sec B2 X
G
B3 B2 G
CLOCK Y
CLOCK
times=10ns; 10ns Gen4bit B3
Z
SUB2
File=gen4bit.sch

SUB1
digital File=dtut1_ex2.sch
simulation

COMB1
Type=TimeList
time=1000 ns

Figure 5.6: Test schematic for a logic circuit with inputs W, X, Y, Z, and output
G.

84
 dtime 5n 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n 110n
r.X
clock.X
b0.X
b1.X
b2.X
b3.X
g.X

Figure 5.7: Digital functional waveforms for a logic circuit with inputs W, X, Y,
Z, and output G.

85
5.6 Building a digital component library
The Qucs graphical user interface includes good project handling features. Com-
bining these features with the Qucs subcircuit capabilities provides all the tools
required for the development of a library of common digital components. Such
a library can be stored in a master project and the individual component files
imported into other projects when required. Here are a few components that I
developed during a recent series of tests aimed at detecting bugs in the VHDL
code generated by Qucs.

5.6.1 Logic zero

L0
L0 0
SUB
File=name

5.6.2 Logic one

L1 1 1 L1
L1

SUB
File=name Y1

5.6.3 G2bit - 2 bit pattern generator

RES R B0 B0

CLK B1 B1

Gen2bit

SUB
File=name

86
 B0
JK B0
1
SUB1
0 B1
SUB2 B1
J S Q J SQ
CLK
B1b
CLK K R Q K R Q
B0b

R FF0 FF1
RES

5.6.4 G4bit - 4 bit pattern generator

R R B0 B0
B1 B1
B2 B2
CLK B3 B3

Gen4bit

SUB
File=name

B0 B1 B2 B3
B0 B1 B2 B3
1
SUB1
CLK
J S Q 0 J S Q J S Q J S Q
SUB2
CLK K R Q K R Q K R Q K R Q
Q0B Q1B Q2B Q3B
FF0 FF1 FF2 FF3
R
JK

87
5.6.5 MUX2to1 - 2 input to 1 output multiplexer
EN A Y
1 X L
0 0 D0
0 1 D1

MUX
ENB EN

A 0}G 0
Y
1
D0 0

D1 1

SUB
File=name

A 1
&
A Y
Y1 1 Y
Y2
D0
& Y4
D0
D1
Y3

D1

88
5.6.6 MUX4to1 - 4 input to 1 multiplexer

B A EN Y
X X 1 0
0 0 0 D0
0 1 0 D1
1 0 0 D2
1 1 0 D3

MUX
ENB EN

A 0
B
}G 0
1 3
Y

D0 0
D1 1
D2 2
D3 3

SUB
File=name

D3 D2 D1 D0 B A ENB

ENB

Y8 Y7 Y6
1

EN
AB &
B BB A

Y1

D2 D1 D0
D3 &

1 Y
Y
Y2

& Y5

Y3

&

Y4

89
 Σ
Σ
5.6.7 2 bit adder

A1 0
B1 1
}A
0 S1
{
1 S2
A2 0
}B
B2 1

CI CI CO C0

SUB
File=name

CI B2 A2 B1 A1
=1 S1
S1

Y1
&

CI B2 A2 B1 A1 1
Y4
& Y7
=1
Y5

Y10
=1 S2
S2

Y2

&
=1
C0
1 CO
Y8
Y6
& Y3

Y9

5.7 Subcircuit VHDL code generated by Qucs

Qucs generates a separate entity-architecture model for each subcircuit. These
component definitions are compiled into the work library by FreeHDL. Here is the
VHDL code from two of the previous examples.

5.7.1 Gen2bit

90
entity S u b g e n 2 b i t i s
port (CLK: in b i t ;
R: in b i t ;
nnout B0 : out b i t ;
nnout B1 : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b g e n 2 b i t of S u b g e n 2 b i t i s
s i g n a l B0b ,
B1b ,
JK ,
nnnet0 ,
B0 ,
B1 : b i t ;
begin
FF0 : process ( nnnet0 , R, CLK)
begin
i f (R= ’1 ’) then B0 <= ’ 0 ’ ;
e l s i f ( nnnet0 = ’1 ’) then B0 <= ’ 1 ’ ;
e l s i f (CLK= ’1 ’ and CLK’ e v e n t ) then
B0 <= (JK and not B0 ) or ( not JK and B0 ) ;
end i f ;
end process ;
B0b <= not B0 ;

FF1 : process ( nnnet0 , R, B0b )

begin
i f (R= ’1 ’) then B1 <= ’ 0 ’ ;
e l s i f ( nnnet0 = ’1 ’) then B1 <= ’ 1 ’ ;
e l s i f ( B0b= ’1 ’ and B0b ’ e v e n t ) then
B1 <= (JK and not B1 ) or ( not JK and B1 ) ;
end i f ;
end process ;
B1b <= not B1 ;

SUB2 : entity S u b l o g i c z e r o port map ( nnnet0 ) ;

nnout B0 <= B0 or ’ 0 ’ ;
nnout B1 <= B1 or ’ 0 ’ ;
SUB1 : entity S u b L o g i c o n e port map (JK ) ;
end architecture ;

5.7.2 2 bit adder

entity S u b f a d d 2 b i t i s

91
 port (A1 : in b i t ;
B1 : in b i t ;
A2 : in b i t ;
B2 : in b i t ;
CI : in b i t ;
nnout S1 : out b i t ;
nnout S2 : out b i t ;
nnout CO : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b f a d d 2 b i t of S u b f a d d 2 b i t i s
s i g n a l nnnet0 ,
nnnet1 ,
nnnet2 ,
nnnet3 ,
nnnet4 ,
nnnet5 ,
nnnet6 ,
S2 ,
CO,
S1 : b i t ;
begin
S1 <= CI xor B1 xor A1 ;
nnnet0 <= B2 xor A2 ;
nnnet1 <= nnnet0 and nnnet2 ;
nnnet3 <= B2 and A2 ;
nnnet2 <= nnnet4 or nnnet5 ;
nnnet4 <= nnnet6 and CI ;
nnnet5 <= B1 and A1 ;
S2 <= B2 xor A2 xor nnnet2 ;
CO <= nnnet3 or nnnet1 ;
nnnet6 <= B1 xor A1 ;
nnout S2 <= S2 or ’ 0 ’ ;
nnout CO <= CO or ’ 0 ’ ;
nnout S1 <= S1 or ’ 0 ’ ;
end architecture ;

5.7.3 Notes on subcircuit VHDL generation

• Qucs predefined digital components generate concurrent data flow signal
statements or process statements.

• Previously defined subcircuit symbols generate VHDL port map statements.

92
 • Type out entity port signals are prevented from being read as input signals
by masking each output signal using the logic function signal-name OR
’0’.8

• A VHDL
use work . a l l ;

statement is included before each subcircuit architecture definition to ensure

that FreeHDL can find any nested subcircuits 9 .

• The complete VHDL code file for a digital design is composed from an outer
test bench entity-architecture model plus entity-architecture models for each
subcircuit specified in the design,

5.8 Subcircuit nesting: A more complex design

example
In theory there is no limit to the depth of subcircuit nesting allowed by Qucs. In
practice most digital circuit schematics can be constructed with a maximum of
four or five levels of design hierarchy. Figure 5.8 shows an example that was used
to test Qucs subcircuit nesting performance. The design is a simple RTL function
that uses a multiplexer to transfer data from one of two input registers to a single
output register. The next section of these notes outlines in detail the specification
of the subcircuits needed to build the RTL design. A set of sample simulation
waveforms showing the register transfer operation are illustrated in Fig. 5.9.

8
Attempting to read entity port signals of type out results in a VHDL compile error.
9
Strictly speaking it should not be necessary to specifically state the use of the work library
as this library is normally visible at all times when compiling entity-architecture models.
However, at this stage in the development of FreeHDL it does appear that it is necessary
when using the default FreeHDL VHDL library mapping.

93
5.8.1 4 bit RTL design

CONTROL1

REG 4bit CONTROL1

1 AND_LOAD
R1D R1Q0 CONTROL2
0
D0 Q0
R1Q1 CONTROL2
SUB5 Y1
D1 Q1
R1Q2 Num=2
D2 Q2
R1Q3 0
D3 Q3 0} G
1
REG 4bit
MUX R3D0 R3Q0
0 D0 Q0
1 LOAD R3D1
LOAD 1 R3Q1
D1 Q1
CLOCK 0 R3D2 R3Q2
D2 Q2
SUB13 1
CLOCK R3D3 R3Q3
D3 Q3
SUB3 0
File=reg_4bit.sch
1

0 LOAD

REG 4bit 1
R2Q0
1 R2D SUB1
D0 Q0
R2Q1 File=quad_mux2to1.sch SUB2
SUB14 D1 Q1 File=reg_4bit.sch
R2Q2
D2 Q2 R2Q3
D3 Q3

VHDL Test circuit 2

digital Multi-hierarchy RTL circuit
LOAD simulation
CONTROL2:R(SUB2)<- R(SUB4)
not(CONTROL2) and CONTROL1 : R(SUB2)<-R(SUB3)

SUB4 Digi1
File=reg_4bit.sch Type=TimeList
time=1000 ns

Figure 5.8: Top level schematic.

Reg4bit

REG 4bit

D0 D0 Q0 Q0

D1 D1 Q1 Q1

D2 D2 Q2 Q2

D3 D3 Q3 Q3

LOAD LOAD

CLOCK

SUB
File=name

94
 D0
D
Q0
D0 Q0
Num=1 EN

SUB1
File=d_flip_flop_l.sch
D1
D
Q1
D1 Q1
Num=2 EN

SUB2
File=d_flip_flop_l.sch

D2
D
Q2
D2 Q2
Num=3 EN

SUB3
File=d_flip_flop_l.sch

D3
D
Q3
D3 Q3
Num=4 EN
LOAD

LOAD SUB4
Num=5 File=d_flip_flop_l.sch
CLOCK

CLOCK
Num=6

95
D flip-flop with load enable

D D

EN EN Q
CLOCK

SUB
File=name

EN Mux2to1
0 G 0
1 Q Q
EN
0 D Q
D R
1
D SUB1 Y1
File=mux2to1.sch

0
CLOCK

CLOCK
SUB2
File=logic_zero.sch

Mux2to1
Mux2to1
A 0 G 0
1
D0 0 Y

D1 1
SUB
File=name

A
1
& Y
A Y
Num=4
Num=1 Y1 1
D0 Y2

& Y4
D0
Num=2 D1
Y3

D1
Num=3

96
QuadMux
0
SEL 0} G
1

A0 MUX
0 Y0
B0
1
A1 0 Y1
B1 1

A2 0
Y2
B2 1

A3 0
Y3
B3 1

SUB
File=name

Mux2to1
Y0
0 G 0
1 Num=10
SEL
Num=1 0
A0
Num=2 1
B0 SUB1
Num=6 File=mux2to1.sch

Mux2to1
Y1
0 G 0
1 Num=11
0
A1
Num=3 1
B1 SUB2
Num=7 File=mux2to1.sch

Mux2to1
Y2
0 G 0
1 Num=12
0
A2
Num=4 1
B2 SUB3
Num=8 File=mux2to1.sch

Mux2to1
Y3
0 G 0
1 Num=13
0
A3
Num=5 1
B3 SUB4
Num=9 File=mux2to1.sch

97
 dtime 370n 380n 390n 400n 410n 420n 430n 440n 450n 460n 470n 480n 490n 500n 510n 520n 530n 540n 550n
clock.X
control1.X
control2.X
load.X
r1q0.X
r1q1.X
r1q2.X
r1q3.X
r2q0.X
r2q1.X
r2q2.X
r2q3.X
r3d0.X
r3d1.X
r3d2.X
r3d3.X
r3q0.X
r3q1.X
r3q2.X
r3q3.X

Figure 5.9: Sample simulation waveforms for RTL design.

5.9 Update number one: May 2006

Although it is only a short time since the first version of these digital tutorial notes
was posted on the Qucs Sourceforge Web site, much has happened in the world of
Qucs digital simulation. Bugs in the Qucs code have been found, and fixed, and
a range of new features added to the software. These expand the power of Qucs
digital simulation and give users a glimpse of how the package will evolve in the
future. The purpose of these notes is firstly to update readers as to the changes to
Qucs digital simulation and secondly to explain how to use the new Qucs features.
Please note however, they are not intended to teach readers how to program using
VHDL.10

5.9.1 Bugs, corrections and small changes to the Qucs digital

simulation code
All the bugs reported in the first version of these notes have been corrected in
the latest Qucs CVS code. These corrections are, of course, also included in Qucs
release 0.0.9. During testing a number of other annoying, but significant, bugs
have also been found and eliminated, these include
10
A good introduction to the VHDL language and it’s application in digital system design can be
found in Digital System Design using VHDL by Charles H. Roth, Jr, PWS Publishing
Company, 1997, ISBN 0-534-95099-X.

98
 • Multiple input gates (three or more inputs) of types nand and nor failed at
the FreeHDL compile stage due to an error in the VHDL code generated by
Qucs.

• Signals names and, for example, component names constructed from a single
letter that was an abbreviation for a physical unit failed to compile.

• Changing digital component time delays caused component connections on

a schematic to be removed.

• GUI problems caused by errors in the symbol rotation and mirror code.

• Qucsconv code conversion errors caused the Qucs digital simulation cycle to
fail before plotting TimeList waveforms.

A number of changes to either the VHDL code generated by Qucs or the schematic
capture GUI have been introduced, these include

• The VHDL code generated by Qucs for the ground symbol has been changed
from
gnd <= gnd and ’ 0 ’ ;

to
gnd <= ’ 0 ’ ;

• The symbol for digital inout ports has been changed from the analogue pin
symbol to one that consists of the digital in and out pins drawn back-to-back.
This reflects the bidirectional status of an inout port.

A more complete list of all the bug corrections and other program modifications
can be found in the Qucs change log files.

5.9.2 New digital simulation features

The flow diagram illustrated in Fig. 5.10 shows a number of different simulation
routes for a digital circuit under test. The Qucs digital simulation facilities have
been improved to include direct simulation of VHDL testbench code and the sim-
ulation of circuit schematics that include digital components specified by VHDL
entity-architecture models. The various combinations that users can adopt for
Qucs digital circuit entry are as follows:

99
 1. Schematic circuit entry using predefined digital component symbols, subcir-
cuits generated using the same symbols and a copy of the digital simulation
icon; this is the approach described in the first version of these tutorial notes.

2. Circuit entry identical to 1 plus symbols for digital components specified by

VHDL entity-architecture models.

3. Circuit entry using the Qucs VHDL code editor. The text entered describes
both the circuit under test and the test vectors needed to drive the circuit
inputs during simulation.

Once the circuit under test has been entered into Qucs, clicking the Simulate menu
button, or pressing key F2, starts the Qucs digital simulation process.

5.9.3 Limitations
Before describing the new digital simulation features it is important that readers
understand the limitations that are inherent in the various digital simulation routes
described in the last section and illustrated in the flow diagram shown in Fig. 5.10.
Qucs schematic capture allows users to draw circuits consisting of predefined com-
ponent symbols and subcircuit symbols. At this stage in the development of the
GUI digital signals must be of type bit (as defined in the VHDL standard library
- library STD in the FreeHDL package) where individual signals flow through a
single wire. Qucs schematic drawing bus structures of VHDL type bit-vector, for
example, have not been implemented yet. This implies that the device symbol port
pins must represent single signals. Similarly the nets connecting pins on more than
one device can only be single signal nets and not bus structures. It is anticipated
that this will change in a future Qucs release.

Although the current release of FreeHDL is 0.0.1 the package implements a sub-
stantial subset of the entire VHDL language11 . The major features not supported
by release 0.0.1 are:

• Shared variables.

• The following attributes; transaction, quiet, stable and delayed.

• User defined attributes.

• Groups.
11
A complete description of the 1987 and 1993 specifications of the VHDL language can be
found in The Designer’s Guide to VHDL by Peter J Ashenden, second edition 2002, Morgan
Kaufmann Publishers, ISBN 1-55860-674-2.

100
VHDL Generate Predefined Qucs
entity/architecture VHDL digital component
code model symbol symbols

Qucs GUI
Digital
VHDL
subcircuit
STD VHDL testbench Circuit drawing symbols
library code
elements Entered using Qucs Generated
Entered using schematic capture using Qucs
Qucs VHDL schematic capture
editor

Digital circuit under test

SIMULATE

VHDL Testbench code

Compile VHDL code

and generate machine
FreeHDL code simulation program
for circuit under test

Machine code simulation

Run of circuit under test

Qucs TimeList plot

Simulation
Output
data
Qucs Truth Table

View

Figure 5.10: Flow diagram of Qucs digital simulation routes.

101
 • Guarded signal assignments.
• Currently drivers cannot be switched off.
The Qucs TimeList plotting program uses signal data output by the machine code
simulation program generated by the FreeHDL package12 . A current limitation of
the TimeList plotting program is that it can only display signals of type bit. Bus
signal waveforms cannot be displayed.

Given the above limitations it is therefore possible to write VHDL code that can
be compiled by FreeHDL but will cause problems at either the schematic drawing
or output waveform plotting stages in the Qucs simulation cycle. As Qucs develops
it is expected that these limitations will be removed. On the subject of limitations
one final point to note: FreeHDL can simulate circuits described by the data types
and other features found in the
IEEE.std_logic_1164
library and other predefined libraries. However, at this stage in the development
of the Qucs software only the VHDL standard library may be used, implying that
data type bit must be used to represent logic signals.

5.9.4 Using the Qucs VHDL editor

Qucs release 0.0.9 includes a VHDL text editor13 that has all the usual edit features
plus colour coding of the various VHDL language statements. One unusual feature
of this editor is a zoom control that allows the text size to be increased or decreased
in a similar way to the schematic drawing zoom. The VHDL editor is included
in the Qucs package for two primary purposes, firstly for purely text file VHDL
simulation14 and secondly for the development of VHDL entity-architecture models
that can be linked to schematic capture symbols. The latter increases significantly
the capabilities of the Qucs software in that it allows libraries of hand-crafted
device models to be constructed. These new library devices will, given support by
the general Qucs user community, greatly expand the potential use of the Qucs
package. In this section the use of the VHDL text editor is demonstrated through a
series of digital circuit simulation examples. The included VHDL listings indicate
typical Qucs use of a number of the basic VHDL data types. The text also outlines
any limitations imposed by Qucs.
12
The machine code simulation program outputs signal data in VCD format. This is then
converted to the Qucs TimeList data format by the qucsconv utility program.
13
To launch the new VHDL editor click on the second icon from the left on the Qucs toolbar. It
can also be activated using the key sequence Ctrl+Shift+V.
14
This is still the preferred method amongst many experienced users of VHDL. However, the
circuit schematic drawing approach does seem to be growing in popularity.

102
 • Example 1: A sum of products (SOP) combinational digital circuit.
The Boolean equation15 for a SOP combinational circuit is:

f = W .X.Y .Z + W .X.Y .Z + W.Y .Z + W.X.Y.Z

The VHDL code for a structural model of this combinational circuit and its
associated testbench is given in the following listing.
−− Qucs VHDL e d i t o r example 1
−−
entity t e s t v e c t o r i s −− Test v e c t o r g e n e r a t o r .
port ( z , y , x , w : out b i t
);
end entity t e s t v e c t o r ;
−−
architecture b e h a v i o u r a l of t e s t v e c t o r i s
begin
pz : process i s
begin
z <= ’ 0 ’ ; wait f o r 20 ns ;
z <= ’ 1 ’ ; wait f o r 20 ns ;
end process pz ;
py : process i s
begin
y <= ’ 0 ’ ; wait f o r 40 ns ;
y <= ’ 1 ’ ; wait f o r 40 ns ;
end process py ;
px : process i s
begin
x <= ’ 0 ’ ; wait f o r 80 ns ;
x <= ’ 1 ’ ; wait f o r 80 ns ;
end process px ;
pw : process i s
begin
w <= ’ 0 ’ ; wait f o r 160 ns ;
w <= ’ 1 ’ ; wait f o r 160 ns ;
end process pw ;
end architecture b e h a v i o u r a l ;
−−
entity and4 i s −− 4 i n p u t and g a t e .
15
The Boolean equation for function f has not been minimised. It is in a form derived directly
from a truth table and is introduced purely as an example to demonstrate the use of the Qucs
VHDL editor.

103
 port ( in1 , in2 , in3 , i n 4 : in b i t ;
out1 : out b i t
);
end entity and4 ;
−−
architecture d a t a f l o w of and4 i s
begin
out1 <= i n 1 and i n 2 and i n 3 and i n 4 ;
end architecture d a t a f l o w ;
−−
entity and3 i s −− 3 i n p u t and g a t e .
port ( in1 , in2 , i n 3 : in b i t ;
out1 : out b i t
);
end entity and3 ;
−−
architecture d a t a f l o w of and3 i s
begin
out1 <= i n 1 and i n 2 and i n 3 ;
end architecture d a t a f l o w ;
−−
entity o r 4 i s −− 4 i n p u t or g a t e .
port ( in1 , in2 , in3 , i n 4 : in b i t ;
out1 : out b i t
);
end entity o r 4 ;
−−
architecture d a t a f l o w of o r 4 i s
begin
out1 <= i n 1 or i n 2 or i n 3 or i n 4 ;
end architecture d a t a f l o w ;

entity i n v i s −− I n v e r t e r .
port ( i n 1 : in b i t ;
out1 : out b i t
);
end entity i n v ;
−−
architecture d a t a f l o w of i n v i s
begin
out1 <= not i n 1 ;
end architecture d a t a f l o w ;
−−

104
 entity t e s t b e n c h i s −− Test bench o u t e r e n t i t y wrapper .
end entity t e s t b e n c h ;
−−
l i b r a r y work ;
use work . a l l ;
−−
architecture s t r u c t u r a l of t e s t b e n c h i s −− T e s t b e n c h a r c h i t e c t u r e .
s i g n a l b0 , b1 , b2 , b3 , zb , yb , xb , wb , a , b , c , d , f : b i t ;
begin
d1 : entity t e s t v e c t o r port map( b0 , b1 , b2 , b3 ) ;
d2 : entity i n v port map( b0 , wb ) ;
d3 : entity i n v port map( b1 , xb ) ;
d4 : entity i n v port map( b2 , yb ) ;
d5 : entity i n v port map( b3 , zb ) ;
d6 : entity and4 port map( zb , yb , b1 , wb , a ) ;
d7 : entity and4 port map( zb , yb , xb , wb , b ) ;
d8 : entity and3 port map( zb , yb , b0 , c ) ;
d9 : entity and4 port map( b0 , b1 , b2 , b3 , d ) ;
d10 : entity o r 4 port map( a , b , c , d , f ) ;
end architecture s t r u c t u r a l ;

On entry of this code into the Qucs VHDL text editor the text is colour
coded. Unfortunately, the colour coding is lost when printed, or pasted
into a word processor, or a layout package like LaTeX. The structure of the
VHDL listing follows the normal convention for text based VHDL simulation.
All component entity-architecture models must be defined before they are
referenced in other component models. The simulation test bench must be
the last entity-architecture model in the VHDL listing. During the VHDL
compile phase FreeHDL compiles the component entity-architecture models
to the work library16 . These compiled models are then made available to the
simulation test bench through the use of the VHDL use statement inserted
in the listing prior to the testbench architecture statement. Once the VHDL
listing for the simulation has been typed into the Qucs VHDL code editor,
pressing key F2 starts the simulation process. The simulation duration can be
set using the Document Settings in the File dropdown menu (or by pressing
the Ctrl+. keys). Any VHDL syntax errors, or indeed typos, are written to
file and can be viewed by pressing key F5. Obviously if errors are reported
these need to be corrected using the VHDL text editor and the simulation
cycle restarted. A typical TimeList output for editor example 1 is shown in
16
In most VHDL implementations library work is always visible and there is no requirement to
make it visible by using the library and use statements. However, FreeHDL appears to need
these statements at the linking phase otherwise the VHDL compiler fails.

105
 Fig. 5.11.
dtime 0 20n 40n 60n 80n 100n 120n 140n 160n 180n 200n 220n 240n 260n 280n 300n 320n
b0.X
b1.X
b2.X
b3.X
f.X

Figure 5.11: Sample simulation waveforms for VHDL editor example 1 design.

• Example 2: VHDL editor example 1 modelled using dataflow VHDL state-

ments.
The VHDL code for the second example is given in the next listing. The
VHDL style chosen to model the circuit is based on VHDL dataflow con-
current signal assignments. The input text vectors are generated using a
simple state machine rather than separate process statements. The test vec-
tor generator port specification uses entirely single signal bit types and can
be easily interfaced, without problems, to other components connected on a
Qucs schematic diagram. The procedure for generating schematic capture
component symbols from entity - architecture models is introduced in a later
section of these notes. The use of bit vector bus constructions is also il-
lustrated in this example. Qucs allows the use of bit vectors as signals or
variables in VHDL models provided all signals in the port statement of en-
tity declaration are of type bit only.17 A typical TimeList output for editor
example 2 is shown in Fig. 5.12.
−− Qucs VHDL e d i t o r example 2
−−
entity t e s t v e c t o r a i s
port ( RESET, CLOCK : in b i t ;
B0 , B1 , B2 , B3 : out b i t
);
end entity t e s t v e c t o r a ;
−−
architecture b e h a v i o u r a l of t e s t v e c t o r a i s
s i g n a l p r e s e n t s t a t e , n e x t s t a t e : b i t v e c t o r ( 3 downto 0):= ”1111 ” ;
begin
17
This is a restriction of Qucs 0.0.9 and will be removed in a later release of the package. Also
note signals of type bit vector that are declared in architecture definitions are listed in the
TimeList plot signal dialogue. However, a text message saying no data results if an attempt
is made to display them. Again this limitation will be removed in a later release of Qucs.

106
−−
p1 : process (CLOCK ) i s
begin
i f (CLOCK’ e v e n t and CLOCK= ’1 ’) then
p r e s e n t s t a t e <= n e x t s t a t e ;
end i f ;
end process p1 ;
−−
p2 : process (RESET, present state ) is
begin
i f (RESET = ’ 1 ’ ) then n e x t s t a t e <= ”1111 ” ;
end i f ;
case p r e s e n t s t a t e i s
when ”0000 ” => n e x t s t a t e <= ”0001 ” ;
when ”0001 ” => n e x t s t a t e <= ”0010 ” ;
when ”0010 ” => n e x t s t a t e <= ”0011 ” ;
when ”0011 ” => n e x t s t a t e <= ”0100 ” ;
when ”0100 ” => n e x t s t a t e <= ”0101 ” ;
when ”0101 ” => n e x t s t a t e <= ”0110 ” ;
when ”0110 ” => n e x t s t a t e <= ”0111 ” ;
when ”0111 ” => n e x t s t a t e <= ”1000 ” ;
when ”1000 ” => n e x t s t a t e <= ”1001 ” ;
when ”1001 ” => n e x t s t a t e <= ”1010 ” ;
when ”1010 ” => n e x t s t a t e <= ”1011 ” ;
when ”1011 ” => n e x t s t a t e <= ”1100 ” ;
when ”1100 ” => n e x t s t a t e <= ”1101 ” ;
when ”1101 ” => n e x t s t a t e <= ”1110 ” ;
when ”1110 ” => n e x t s t a t e <= ”1111 ” ;
when ”1111 ” => n e x t s t a t e <= ”0000 ” ;
end case ;
B3 <= n e x t s t a t e ( 3 ) ; B2 <= n e x t s t a t e ( 2 ) ;
B1 <= n e x t s t a t e ( 1 ) ; B0 <= n e x t s t a t e ( 0 ) ;
end process p2 ;
end architecture b e h a v i o u r a l ;
−−
l i b r a r y work ;
use work . a l l ;
−−
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture d a t a f l o w of t e s t b e n c h i s
s i g n a l r e s e t , c l k , b0 , b1 , b2 , b3 , zb : b i t ;

107
 s i g n a l yb , xb , wb , a , b , c , d , f : b i t ;
begin
p1 : process i s
begin
c l k <= ’ 0 ’ ; wait f o r 10 ns ;
c l k <= ’ 1 ’ ; wait f o r 10 ns ;
end process p1 ;
−−
p2 : process i s
begin
r e s e t <= ’ 1 ’ ; wait f o r 10 ns ;
r e s e t <= ’ 0 ’ ; wait f o r 2000 ns ;
end process p2 ;
−−
d1 : entity t e s t v e c t o r a port map( r e s e t , c l k , b0 , b1 , b2 , b3 ) ;
−−
−− Data f l o w model o f c o m b i n a t i o n a l c i r c u i t
wb <= not b0 ; xb <= not b1 ; yb <= not b2 ; zb <= not b3 ;
a <= (wb and b1 ) and ( yb and zb ) ;
b <= (wb and xb ) and ( yb and zb ) ;
c <= b0 and ( yb and zb ) ;
d <= ( b0 and b1 ) and ( b2 and b3 ) ;
f <= a or b or c or d ;
end architecture d a t a f l o w ;

dtime 0 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n 110n 120n 130n 140n 150n 160n 170n
reset.X
b0.X
b1.X
b2.X
b3.X
f.X

Figure 5.12: Sample simulation waveforms for VHDL editor example 2 design.

• Example 3: VHDL editor example 1 modelled using VHDL process state-

ments and variables.
The VHDL code for the third example is given in the listing at the end of
this paragraph. In this example the use of VHDL variables is illustrated.
The VHDL code for the vector generator is a little unusual in that rather
than using the traditional two process design employing signals, a single
process statement employing variables undertakes both the calculation of

108
the next state data and the transfer of the next state information to the
present state. This approach is necessary because FreeHDL does not allowed
shared variables. Once again in this example only single bit data is passed
via the entity statement to the device under test. The device under test is
represented by a truth table encoded in a process statement. This is not the
most elegant code but it does serve the purpose of demonstrating the use
of different VHDL constructions and data types in Qucs digital simulation.
A typical TimeList plot for VHDL editor example 3 is shown in Fig. 5.13.
Comparison of the three output plots for the VHDL editor examples indicates
that all the simulation results are very similar with some slight differences in
the start up phase following the RESET pulse changing from logic ’1’ to logic
’0’. This is probably an effect due to the different initialisation sequences for
each of the test vector models.

−− Qucs VHDL e d i t o r example 3

−−
entity t e s t v e c t o r b i s
port ( RESET, CLOCK : in b i t ;
B0 , B1 , B2 , B3 : out b i t
);
end entity t e s t v e c t o r b ;
−−
architecture b e h a v i o u r a l of t e s t v e c t o r b i s
begin
p1 : process (RESET, CLOCK) i s
variable p r e s e n t s t a t e , n e x t s t a t e :
b i t v e c t o r ( 3 downto 0):= ”0000 ” ;
begin
i f (RESET = ’ 1 ’ ) then n e x t s t a t e := ”0000 ” ;
e l s i f (CLOCK’ e v e n t and CLOCK= ’1 ’) then
p r e s e n t s t a t e := n e x t s t a t e ;
case p r e s e n t s t a t e i s
when ”0000 ” => n e x t s t a t e := ”0001 ” ;
when ”0001 ” => n e x t s t a t e := ”0010 ” ;
when ”0010 ” => n e x t s t a t e := ”0011 ” ;
when ”0011 ” => n e x t s t a t e := ”0100 ” ;
when ”0100 ” => n e x t s t a t e := ”0101 ” ;
when ”0101 ” => n e x t s t a t e := ”0110 ” ;
when ”0110 ” => n e x t s t a t e := ”0111 ” ;
when ”0111 ” => n e x t s t a t e := ”1000 ” ;
when ”1000 ” => n e x t s t a t e := ”1001 ” ;
when ”1001 ” => n e x t s t a t e := ”1010 ” ;
when ”1010 ” => n e x t s t a t e := ”1011 ” ;

109
 when ”1011 ” => n e x t s t a t e := ”1100 ” ;
when ”1100 ” => n e x t s t a t e := ”1101 ” ;
when ”1101 ” => n e x t s t a t e := ”1110 ” ;
when ”1110 ” => n e x t s t a t e := ”1111 ” ;
when ”1111 ” => n e x t s t a t e := ”0000 ” ;
end case ;
end i f ;
B3 <= n e x t s t a t e ( 3 ) ; B2 <= n e x t s t a t e ( 2 ) ;
B1 <= n e x t s t a t e ( 1 ) ; B0 <= n e x t s t a t e ( 0 ) ;
end process p1 ;
end architecture b e h a v i o u r a l ;
−−
l i b r a r y work ;
use work . a l l ;
−−
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture d a t a f l o w of t e s t b e n c h i s
s i g n a l r e s e t , c l k , b0 , b1 , b2 , b3 , f : bit ;
begin
p1 : process i s
begin
c l k <= ’ 0 ’ ; wait f o r 10 ns ;
c l k <= ’ 1 ’ ; wait f o r 10 ns ;
end process p1 ;
−−
p2 : process i s
begin
r e s e t <= ’ 1 ’ ; wait f o r 10 ns ;
r e s e t <= ’ 0 ’ ; wait f o r 2000 ns ;
end process p2 ;
−−
d1 : entity t e s t v e c t o r b port map( r e s e t , c l k , b0 , b1 , b2 , b3 ) ;
−−
−− B e h a v i o u r a l model o f c o m b i n a t i o n a l c i r c u i t
p3 : process ( b3 , b2 , b1 , b0 ) i s
variable SEL : b i t v e c t o r ( 3 downto 0 ) ;
begin
SEL := b3&b2&b1&b0 ;
i f (SEL = ”0010 ”) then f <= ’ 1 ’ ;
e l s i f (SEL = ”0000 ”) then f <= ’ 1 ’ ;
e l s i f (SEL = ”1111 ”) then f <= ’ 1 ’ ;

110
 e l s i f (SEL = ”0001 ”) then f <= ’ 1 ’ ;
e l s i f (SEL = ”0011 ”) then f <= ’ 1 ’ ;
e l s e f <= ’ 0 ’ ;
end i f ;
end process p3 ;
end architecture d a t a f l o w ;

dtime 0 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n 110n 120n 130n 140n 150n 160n 170n
reset.X
b0.X
b1.X
b2.X
b3.X
f.X

Figure 5.13: Sample simulation waveforms for VHDL editor example 3 design.

5.9.5 Linking VHDL entity-architecture models to Qucs

schematic device symbols
VHDL was originally developed as a hardware description language for specify-
ing digital systems. Indeed many engineers still prefer to describe digital systems
entirely in VHDL statements rather than use schematic drawings. Once writ-
ten VHDL code is saved as a text file and becomes the input data for a VHDL
compiler/simulation package. Through popular demand a number of digital syn-
thesis/simulator CAD tools18 have started to include a facility that links VHDL
model code to a schematic capture symbol. It is then, of course, possible to use
a schematic diagram as the main entry media19 when designing and simulating a
digital design. Qucs release 0.0.9 has such a facility, allowing VHDL code models
to be linked to schematic symbols. When drawing digital design schematics, these
user defined symbols may be mixed with the Qucs predefined digital symbols and
other user defined subcircuit symbols. The process for linking VHDL code to Qucs
schematic drawing symbols is straightforward and will be illustrated in these notes
through two examples.

• Example 4: A 4 bit test vector pattern generator.

18
See for example the XILINX, WebPACK software at http//www.xilinx.com/ise/logic_
design_prod/webpack.htm.
19
Please note that at the start of the VHDL simulation process schematic drawings are converted
into a VHDL text file.

111
 Shown in Table 5.4 is the VHDL entity-architecture model listing for a 4
bit binary pattern generator. The VHDL code is identical to the test vector
code introduced in the third VHDL editor example. After entering the VHDL
entity-architecture model code using the Qucs VHDL editor the finished text
is saved in a file with a suitable name and file extension vhdl. Qucs then
lists the model under the VHDL project category. Simply clicking on a model
name in the VHDL category, with the left hand mouse button, then moving
the mouse pointer to a suitable position on a schematic, causes Qucs to
move a symbol that represents the model onto the schematic drawing sheet.
Placement of the symbol at the position located by the mouse pointer is
achieved by clicking the left hand mouse button. The procedure is identical
to that used to select and place the Qucs predefined symbols on a schematic
drawing. Qucs automatically generates a rectangular symbol with a name
called VHDL that has the same number of pins as the port statement listed
in the VHDL model entity statement. Each of the pins is given a name
that corresponds to a name in the entity statement. Qucs fixes the order
of the pins on the generated symbol. It appears that it is not possible to
edit this symbol. However, subcircuit in, out or inout port symbols can be
attached to symbol VHDL and a user edited symbol generated. Fig. 5.14
shows the Qucs generated VHDL symbol with attached ports for the model
listed in Table 5.4. The edited symbol for the 4 bit binary pattern generator
is illustrated in Fig. 5.15. Notice that in Fig. 5.15 the order of the pins has
been changed to reflect the natural order for a device with it’s input pins
on the left and output pins on the right. VHDL model symbols can also
be generated by placing the VHDL file component, this is located in the
digital components viewlist, on a schematic. On editing the VHDL file name
property of this device to the name of a VHDL entity-architecture model file,
Qucs automatically generates a VHDL symbol. Defining your own symbol
then proceeds in a similar fashion to the way described above.

• Example 5: A 4 bit full adder.

VHDL model symbols may be combined with either the Qucs predefined
digital component symbols or other subcircuit symbols. In this example a
VHDL model for a simple one bit full adder is connected four times in a
serial fashion to form a 4 bit full adder. The VHDL model code for a simple
one bit full adder is given in Table 5.5. The associated symbol diagrams for
the one bit full adder are illustrated in Fig. 5.16 and Fig. 5.17.

112
entity p a t g e n 4 b i t i s
port ( RESET, CLOCK : in b i t ;
B0 , B1 , B2 , B3 : out b i t
);
end entity p a t g e n 4 b i t ;
−−
architecture b e h a v i o u r a l of p a t g e n 4 b i t i s
begin
p1 : process (RESET, CLOCK) i s
variable p r e s e n t s t a t e , n e x t s t a t e :
b i t v e c t o r ( 3 downto 0):= ”0000 ” ;
begin
i f (RESET = ’ 1 ’ ) then n e x t s t a t e := ”0000 ” ;
e l s i f (CLOCK’ e v e n t and CLOCK= ’1 ’) then
p r e s e n t s t a t e := n e x t s t a t e ;
case p r e s e n t s t a t e i s
when ”0000 ” => n e x t s t a t e := ”0001 ” ;
when ”0001 ” => n e x t s t a t e := ”0010 ” ;
when ”0010 ” => n e x t s t a t e := ”0011 ” ;
when ”0011 ” => n e x t s t a t e := ”0100 ” ;
when ”0100 ” => n e x t s t a t e := ”0101 ” ;
when ”0101 ” => n e x t s t a t e := ”0110 ” ;
when ”0110 ” => n e x t s t a t e := ”0111 ” ;
when ”0111 ” => n e x t s t a t e := ”1000 ” ;
when ”1000 ” => n e x t s t a t e := ”1001 ” ;
when ”1001 ” => n e x t s t a t e := ”1010 ” ;
when ”1010 ” => n e x t s t a t e := ”1011 ” ;
when ”1011 ” => n e x t s t a t e := ”1100 ” ;
when ”1100 ” => n e x t s t a t e := ”1101 ” ;
when ”1101 ” => n e x t s t a t e := ”1110 ” ;
when ”1110 ” => n e x t s t a t e := ”1111 ” ;
when ”1111 ” => n e x t s t a t e := ”0000 ” ;
end case ;
end i f ;
B3 <= n e x t s t a t e ( 3 ) ; B2 <= n e x t s t a t e ( 2 ) ;
B1 <= n e x t s t a t e ( 1 ) ; B0 <= n e x t s t a t e ( 0 ) ;
end process p1 ;
end architecture b e h a v i o u r a l ;

Table 5.4: VHDL code for a 4 bit pattern generator.

113
 CLOCK
Num=2
RESET CLOCK

RESET B1
Num=1 Num=4
B0 B1
vhdl
B0 B3
Num=3 Num=6
B2 B3

B2
Num=5 X1

Figure 5.14: Qucs generated VHDL symbol with subcircuit ports for test pattern
generator.

SUB
patgen File=name
4bit
RESET R B0 B0
B1 B1
CLOCK B2 B2
B3 B3

Figure 5.15: User defined 4 bit pattern generator symbol.

114
−− F u l l adder − 1 b i t
entity f u l l a d d e r i s
port ( a , b , c i n : in b i t ;
sum , c o u t : out b i t
);
end entity f u l l a d d e r ;
−−
architecture d a t a f l o w of f u l l a d d e r i s
begin
sum <= ( a xor b ) xor c i n ;
c o u t <= ( a and b ) or ( a and c i n ) or ( b and c i n ) ;
end architecture d a t a f l o w ;

Table 5.5: VHDL code for a 1 bit full adder.

b
a b

a
sum
cin sum
vhdl
cin

cout

cout
X1
Figure 5.16: Qucs generated VHDL symbol with subcircuit ports for one bit full
adder.

115
 SUB
File=name
a sum
b
CO cout
cin CI

Figure 5.17: User defined one bit full symbol.

Figure 5.18 shows the schematic for a simple 4 bit ripple adder. The corre-
sponding user defined symbol for the 4 bit full adder is given in Fig. 5.19.
SUB1
sum0

a0
CO
b0 CI

cin
SUB2
sum1

a1
CO
b1 CI

SUB3
sum2

a3
CO
b2 CI

SUB4
sum3

a4
CO
b3 CI cout

Figure 5.18: 4 bit full adder schematic.

5.9.6 Generating VHDL code from Qucs schematic drawings

Pressing key F2 causes Qucs to simulate the design entered by the Qucs user.
The input data for a simulation is either a VHDL text file, saved from the VHDL
text editor, or a VHDL code file generated by Qucs using the information encoded
on a schematic drawing. In this section of these tutorial notes a larger design
is introduced and the resulting VHDL code and simulation results are discussed.

116
 a0 0 sum0
a1
a3 } a { sum1
sum2
a4 3 sum3

b0 0
b1 cout
b2 }b CO

b3 3

cin CI

SUB
File=name

Figure 5.19: User defined 4 bit full adder symbol.

b3 b2 b1 b0
a3 a2 a1 a0
a0b3 a0b2 a0b1 a0b0
a1b3 a1b2 a1b1 a1b0
a2b3 a2b2 a2b1 a2b0
a3b3 a3b2 a3b1 a3b0
r7 r6 r5 r4 r3 r2 r1 r0

Table 5.6: Product table for a 4 bit by 4 bit combinational multiplier.

The example chosen for this purpose is a 4 bit by 4 bit combinational digital
multiplier. Both the 4 bit pattern generator and the 4 bit full adder outlined in
the last section form part of the central core of the 4 bit multiplier design and
it’s associated testbench. Table 5.6 shows the multiplication product table for a 4
bit by 4 bit combinational binary multiplier. Inputs to the device are binary bits
a3 a2 a1 a0 and b3 b2 b1 b0. The 4 by 4 multiplier device requires 16 and gates
(to generate the multiplier product terms), three four bit full adders (to sum the
output r terms) and two 4 bit pattern generators to test the 256 possible input
states. The multiplier output is represented in Table 5.6 by r7 r6 r5 r4 r3 r2 r1
and r0. The circuit schematic for the 4 bit by 4 bit multiplier and test bench are
given in Fig. 5.20.

The VHDL code for this example is presented in the following listing. This listing

117
was generated by Qucs20 . A small section of the TimeList waveform plot for the
digital multiplier is shown in Fig. 5.21. At 1.74 micro seconds input a is ”0101”,
input b is ”0111”and the output r is ”00100011”which is 35 in decimal. Taking a few
random checks of the simulation results indicates that the 4 bit by 4 bit multiplier
design works correctly. Notice that the VHDL code generated by Qucs for the 4
bit multiplier does not contain any propagation delay timing data. This could be
added to the and gates, if required. However, at this stage in the development of
Qucs digital simulation passing timing data, and other parameters, from device
symbols generated from VHDL models has not been implemented yet. The use
of VHDL generics is an obvious way this could be done. Generics are allowed, of
course, in text based VHDL simulations.

20
Some readers will have noticed that the naming scheme for internal signal nets is different in
the multiplier VHDL listing when compared to the VHDL listings in the first version of these
notes. Towards the end of the 0.0.9 development phase the naming convention employed by
Qucs was changed to give a more flexible structure.

118
 Y1
CLOCK
digital
1 simulation
R CLOCK
Digi1
SUB3 SUB1
Type=TimeList
R time=5000 ns
R

R
patgen

patgen
4bit

4bit

SUB5
0
B3
B2
B1
B0

B3
B2
B1
B0

A3 A2 A1 A0 B3 B2 B1 B0
R0
&

Y2 & R1

0
Y3 &
}a {
Y4 &
3

Y5 & 0

Y6 & }b CO

3
Y7 &
CI
R2
Y8 &
SUB4 0

Y9 & }a {
3
&
Y10
0
&
Y11
}b CO R3
R4
0
&
Y12 3

}a { R5
R6
CI
&
Y13
3

& SUB6 0 R7
Y14

&
Y15 }b CO

3
&
Y16
CI

Y17
SUB7
File=full_adder_4 bit.sch

Figure 5.20: A 4 bit by 4 bit combinational digital multiplier.

119
−− Qucs 0 . 0 . 9
−− /mnt/ hda2 / v h d l c o m p l i b p r j / m u l t i p l i e r 4 b x 4 b i t . s c h

entity p a t g e n 4 b i t i s
port ( RESET, CLOCK : in b i t ;
B0 , B1 , B2 , B3 : out b i t
);
end entity p a t g e n 4 b i t ;
−−
architecture b e h a v i o u r a l of p a t g e n 4 b i t i s
begin
p1 : process (RESET, CLOCK) i s
variable p r e s e n t s t a t e , n e x t s t a t e :
b i t v e c t o r ( 3 downto 0 ) := ”0000 ” ;
begin
i f (RESET = ’ 1 ’ ) then n e x t s t a t e := ”0000 ” ;
e l s i f (CLOCK’ e v e n t and CLOCK= ’1 ’) then
p r e s e n t s t a t e := n e x t s t a t e ;
case p r e s e n t s t a t e i s
when ”0000 ” => n e x t s t a t e := ”0001 ” ;
when ”0001 ” => n e x t s t a t e := ”0010 ” ;
when ”0010 ” => n e x t s t a t e := ”0011 ” ;
when ”0011 ” => n e x t s t a t e := ”0100 ” ;
when ”0100 ” => n e x t s t a t e := ”0101 ” ;
when ”0101 ” => n e x t s t a t e := ”0110 ” ;
when ”0110 ” => n e x t s t a t e := ”0111 ” ;
when ”0111 ” => n e x t s t a t e := ”1000 ” ;
when ”1000 ” => n e x t s t a t e := ”1001 ” ;
when ”1001 ” => n e x t s t a t e := ”1010 ” ;
when ”1010 ” => n e x t s t a t e := ”1011 ” ;
when ”1011 ” => n e x t s t a t e := ”1100 ” ;
when ”1100 ” => n e x t s t a t e := ”1101 ” ;
when ”1101 ” => n e x t s t a t e := ”1110 ” ;
when ”1110 ” => n e x t s t a t e := ”1111 ” ;
when ”1111 ” => n e x t s t a t e := ”0000 ” ;
end case ;
end i f ;
B3 <= n e x t s t a t e ( 3 ) ; B2 <= n e x t s t a t e ( 2 ) ;
B1 <= n e x t s t a t e ( 1 ) ; B0 <= n e x t s t a t e ( 0 ) ;
end process p1 ;
end architecture b e h a v i o u r a l ;

entity S u b p a t g e n 4 b i t i s

120
 port ( n e t n e t 0 : in b i t ;
n e t n e t 5 : in b i t ;
n e t o u t n e t n e t 1 : out b i t ;
n e t o u t n e t n e t 3 : out b i t ;
n e t o u t n e t n e t 2 : out b i t ;
n e t o u t n e t n e t 4 : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b p a t g e n 4 b i t of S u b p a t g e n 4 b i t i s
signal net net1 ,
net net2 ,
net net3 ,
net net4 : bit ;
begin
n e t o u t n e t n e t 1 <= n e t n e t 1 or ’ 0 ’ ;
n e t o u t n e t n e t 2 <= n e t n e t 2 or ’ 0 ’ ;
n e t o u t n e t n e t 3 <= n e t n e t 3 or ’ 0 ’ ;
n e t o u t n e t n e t 4 <= n e t n e t 4 or ’ 0 ’ ;
X1 : entity p a t g e n 4 b i t port map ( n e t n e t 0 , n e t n e t 5 ,
net net1 , net net3 , net net2 , net net4 ) ;
end architecture ;

−− l o g i c z e r o . v h d l
entity l o g i c z e r o i s
port ( Y : out b i t
);
end entity l o g i c z e r o ;
−−
architecture d a t a f l o w of l o g i c z e r o i s
begin
Y <= ’ 0 ’ ;
end architecture d a t a f l o w ;

entity S u b l o g i c z e r o i s
port ( n e t o u t n e t Y : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b l o g i c z e r o of S u b l o g i c z e r o i s
s i g n a l netY : b i t ;
begin
X1 : entity l o g i c z e r o port map ( netY ) ;

121
 n e t o u t n e t Y <= netY or ’ 0 ’ ;
end architecture ;

−− F u l l adder − 1 b i t
entity f u l l a d d e r i s
port ( a , b , c i n : in b i t ;
sum , c o u t : out b i t
);
end entity f u l l a d d e r ;
−−
architecture d a t a f l o w of f u l l a d d e r i s
begin
sum <= ( a xor b ) xor c i n ;
c o u t <= ( a and b ) or ( a and c i n ) or ( b and c i n ) ;
end architecture d a t a f l o w ;

entity S u b f u l l a d d e r 1 b i t i s
port ( n e t n e t 0 : in b i t ;
n e t n e t 1 : in b i t ;
n e t n e t 2 : in b i t ;
n e t o u t n e t n e t 3 : out b i t ;
n e t o u t n e t n e t 4 : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b f u l l a d d e r 1 b i t of S u b f u l l a d d e r 1 b i t i s
signal net net3 ,
net net4 : bit ;
begin
X1 : entity f u l l a d d e r port map ( n e t n e t 0 , n e t n e t 1 ,
net net2 , net net3 , net net4 ) ;
n e t o u t n e t n e t 3 <= n e t n e t 3 or ’ 0 ’ ;
n e t o u t n e t n e t 4 <= n e t n e t 4 or ’ 0 ’ ;
end architecture ;

entity S u b full adder 4bit is

port ( n e t n e t 0 : in b i t ;
net n e t 1 : in b i t ;
net n e t 2 : in b i t ;
net n e t 3 : in b i t ;
net n e t 4 : in b i t ;

122
 n e t n e t 5 : in b i t ;
n e t n e t 6 : in b i t ;
n e t n e t 1 3 : in b i t ;
n e t n e t 7 : in b i t ;
n e t o u t n e t n e t 8 : out b i t ;
n e t o u t n e t n e t 9 : out b i t ;
n e t o u t n e t n e t 1 0 : out b i t ;
n e t o u t n e t n e t 1 1 : out b i t ;
n e t o u t n e t n e t 1 2 : out b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b f u l l a d d e r 4 b i t of S u b f u l l a d d e r 4 b i t i s
signal net net14 ,
net net15 ,
net net16 ,
net net8 ,
net net9 ,
net net10 ,
net net11 ,
net net12 : bit ;
begin
n e t o u t n e t n e t 8 <= n e t n e t 8 or ’ 0 ’ ;
n e t o u t n e t n e t 9 <= n e t n e t 9 or ’ 0 ’ ;
n e t o u t n e t n e t 1 0 <= n e t n e t 1 0 or ’ 0 ’ ;
n e t o u t n e t n e t 1 1 <= n e t n e t 1 1 or ’ 0 ’ ;
n e t o u t n e t n e t 1 2 <= n e t n e t 1 2 or ’ 0 ’ ;
SUB4 : entity S u b f u l l a d d e r 1 b i t port map ( n e t n e t 3 , n e t n e t 1 3 ,
net net14 , net net11 , net net12 ) ;
SUB3 : entity S u b f u l l a d d e r 1 b i t port map ( n e t n e t 2 , n e t n e t 6 ,
net net15 , net net10 , net net14 ) ;
SUB2 : entity S u b f u l l a d d e r 1 b i t port map ( n e t n e t 1 , n e t n e t 5 ,
net net16 , net net9 , net net15 ) ;
SUB1 : entity S u b f u l l a d d e r 1 b i t port map ( n e t n e t 0 , n e t n e t 4 ,
net net7 , net net8 , net net16 ) ;
end architecture ;

entity TestBench i s
end entity ;
use work . a l l ;

architecture Arch TestBench of TestBench i s

s i g n a l netA0 , netA1 , netA2 , netA3 , netR , netB0 ,
netB1 , netB2 , netB3 , netR0 , netR1 , netR2 ,

123
 netR3 , netR4 , netR5 , netR6 , netR7 , netCLOCK ,
net net0 , net net1 , net net2 , net net3 , net net4 ,
net net5 , net net6 , net net7 , net net8 , net net9 ,
net net10 , net net11 , net net12 , net net13 , net net14 ,
net net15 , net net16 , net net17 , net net18 , net net19 ,
net net20 , net net21 , net net22 , net net23 ,
net net24 : bit ;
begin
SUB3 : entity S u b p a t g e n 4 b i t port map ( netR , n e t n e t 0 ,
netA0 , netA1 , netA2 , netA3 ) ;
SUB1 : entity S u b p a t g e n 4 b i t port map ( netR , netCLOCK ,
netB0 , netB1 , netB2 , netB3 ) ;

R: process
begin
netR <= ’ 1 ’ ; wait f o r 10 ns ;
netR <= ’ 0 ’ ; wait f o r 2000 ns ;
end process ;

CLOCK: process
begin
netCLOCK <= ’ 0 ’ ; wait f o r 10 ns ;
netCLOCK <= ’ 1 ’ ; wait f o r 10 ns ;
end process ;

n e t n e t 0 <= not netB3 ;

netR0 <= netA0 and netB0 ;
n e t n e t 1 <= netA0 and netB1 ;
n e t n e t 2 <= netA0 and netB2 ;
n e t n e t 3 <= netA0 and netB3 ;
SUB5 : entity S u b l o g i c z e r o port map ( n e t n e t 4 ) ;
n e t n e t 5 <= netA1 and netB0 ;
n e t n e t 6 <= netA1 and netB1 ;
n e t n e t 7 <= netA1 and netB2 ;
n e t n e t 8 <= netA1 and netB3 ;
n e t n e t 9 <= netA2 and netB0 ;
n e t n e t 1 0 <= netA2 and netB1 ;
n e t n e t 1 1 <= netA2 and netB2 ;
n e t n e t 1 2 <= netA2 and netB3 ;
SUB4 : entity S u b f u l l a d d e r 4 b i t port map ( n e t n e t 1 , n e t n e t 2 ,
net net3 , net net4 , net net5 , net net6 , net net7 ,
net net8 , net net4 , netR1 , n e t n e t 1 3 , n e t n e t 1 4 ,

124
 net net15 , net net16 ) ;
SUB6 : entity S u b f u l l a d d e r 4 b i t port map ( n e t n e t 1 3 , n e t n e t 1 4 ,
net net15 , net net16 , net net9 , net net10 , net net11 ,
n e t n e t 1 2 , n e t n e t 4 , netR2 , n e t n e t 1 7 , n e t n e t 1 8 ,
net net19 , net net20 ) ;
n e t n e t 2 1 <= netA3 and netB0 ;
n e t n e t 2 2 <= netA3 and netB1 ;
n e t n e t 2 3 <= netA3 and netB2 ;
n e t n e t 2 4 <= netA3 and netB3 ;
SUB7 : entity S u b f u l l a d d e r 4 b i t port map ( n e t n e t 1 7 , n e t n e t 1 8 ,
net net19 , net net20 , net net21 , net net22 ,
n e t n e t 2 3 , n e t n e t 2 4 , n e t n e t 4 , netR3 , netR4 ,
netR5 , netR6 , netR7 ) ;
end architecture ;

dtime 1.67u 1.68u 1.69u 1.7u 1.71u 1.72u 1.73u 1.74u 1.75u 1.76u 1.77u 1.78u 1.79u
clock.X
a0.X
a1.X
a2.X
a3.X
b0.X
b1.X
b2.X
b3.X
r0.X
r1.X
r2.X
r3.X
r4.X
r5.X
r6.X
r7.X

Figure 5.21: A section of the 4 bit by 4 bit combinational digital multiplier

TimeList output waveforms.

5.10 Update number two: September 2006

Update number two in this tutorial series reports on the major changes that have
taken place to Qucs digital simulation since the first update was posted on the Qucs

125
Web site roughly three months ago. During this period a number of significant, and
very critical, extensions have been implemented. Previous releases concentrated
on establishing a fundamental base for digital circuit simulation using the VHDL
language. The primary vehicle for representing circuit signals being the VHDL bit
and bit-vector signal types. The next release of Qucs (version 0.0.10) and FreeHDL
(version 0.0.3) extends the allowed signal types to include IEEE std_logic_1164
nine level logic, integers, and reals. Readers will appreciate that these changes are
the result of a great deal of work by the Qucs team and must be considered as very
much work in progress because not all the features offered by the FreeHDL imple-
mentation of the VHDL language are currently available via the Qucs schematic
capture and VHDL text file simulation routes. Although a significant amount of
testing has taken place it is likely that software bugs will come to light as more
Qucs users try the new features - if you find a bug please report it by posting a
note on the Qucs Web site. Adding new signal types to Qucs digital simulation
affects all sections of the simulation route from schematic capture to plotting and
tabulating input and output signals. Hence, although it may seem the wrong way
round, the place to first implement the necessary changes to accommodate the
new signal types is at the simulation results reporting stages of the Qucs package.
In release 0.0.10 no attempt has been made to add the new signal types to the
schematic capture part of the Qucs package.21 Recent work on the digital sections
of the Qucs package has concentrated on (1) improvements to VHDL language
entry using the Qucs colour coded VHDL text editor22 , (2) modifications to Free-
HDL which allow a cleaner interface between Qucs and FreeHDL, (3) upgrades to
the data conversion of simulation results from the FreeHDL value change dump
format to the native Qucs format, and (4) major changes to the results reporting
routines that are accessed from the Qucs diagrams icon dialogue. A detailed list of
the software changes and bug fixes can be found in the Qucs and FreeHDL change
log files.

5.10.1 Simulating VHDL code using Qucs and FreeHDL.

The flow diagram drawn in Fig. 5.10 shows the relationship between Qucs and
FreeHDL, and the sequence that takes place during digital circuit simulation. This
flow diagram does not however, outline the details of the stages that are performed
when converting (1) VHDL circuit code into a machine code simulation program,
and (2) simulation output results into a format that can be plotted and tabulated
by Qucs. These are illustrated in the flow diagram presented in Fig. 5.22. The shell
script qucsdigi controls each of the stages in this sequence. A basic understanding
21
Adding new signal types to Qucs schematic capture is on the to-do list.
22
A number of editor bugs have been fixed and it is now possible for users to define their own
colour scheme for the various classes of VHDL reserved words and data types.

126
of the process employed by Qucs and FreeHDL is needed if users of the software are
to be able to write meaningful VHDL code and simulate it using the two packages.
VHDL code is either generated from a schematic diagram automatically by Qucs
or entered using the Qucs VHDL text editor. The use of the schematic entry
route was described in update one of these tutorial notes. However, a number
of readers will probably have spotted that included in the VHDL code generated
by Qucs are references to VHDL libraries. The VHDL language uses libraries to
provide features that are not specified in the basic language definition but are
commonly used by all language processing systems; two such libraries are STD
and IEEE. When simulating digital circuits a basic knowledge of the structure of
a simulation task and how these employ VHDL libraries is essential. This implies
that users of the Qucs/FreeHDL software must appreciate how the system compiles
and simulates a VHDL circuit simulation task. Once the VHDL simulation code
has been entered via the VHDL text editor clicking the Qucs simulation button
runs shell script qucsdigi performing the sequence shown in Fig. 5.2223 . Program
freeehdl-v2cc converts VHDL code into C++ functions. These are then compiled
along with a main C++ function. The next stage in the sequence links the compiled
object code with the object code from any references to items in the predefined
VHDL libraries to produce an executable digital simulation program. This is then
run by Qucs outputting a set of simulation results in value change dump (VCD)
format24 . Finally a program called qucsconv converts the VCD simulation results
into the Qucs native data format ready for post processing as graphical or tabular
diagrams by Qucs.

23
For the FreeHDL package to operate correctly the directory where the software is installed
must be included in the shell PATH from which Qucs is launched.
24
The value change dump language was originally designed as a simulation waveform interchange
format for Verilog HDL. The specification of the VCD format can be found at http://www-
ee.eng.hawaii.edu/ msmith/ASICs/HTML/Verilog/LRM/HTML/15/ch15.2.htm

127
 VHDL source text

file_name.vhdl OR
file_name.vhd

FreeHDL V2CC
Running C++ VHDL -> C++
conversion.... conversion

}
file_name.cc

Compiling functions... FreeHDL Compile

C++ functions

file_name.o

FreeHDL Compile
Compiling main... C++ main function

file_name_main_.o
Linking....

VHDL std library

standard.o
FreeHDL Linking
IEEE library
log.txt log.txt
std_logic_1164.o file_name

RUN file_name log.txt

Simulating.... simulation program
ERROR and other
results data

file_name.vcd

QUCS TimeList plot

QUCSCONV
Running VCD VCD -> QUCS
conversion... data conversion QUCS Truth table
QUCS Tabulation

Figure 5.22: Detailed flow diagram showing VHDL code compilation and simula-
tion results processing.

128
5.10.2 VHDL predefined packages and libraries.
All VHDL language processing systems provide a predefined VHDL package called
standard. This package defines many of the fundamental VHDL data types, for
example bit, character, integer and real. The predefined types, subtypes and
other functions in the package standard are stored in a library called STD. The
FreeHDL version of library STD includes an additional VHDL package called textio
which is used to input and output signal data from and to files. A second library
called IEEE defines (1) multivalued logic signals defined by nine different encoding
values, making it possible to model digital circuits that are composed from different
technology components, (2) logic signal subtypes and (3) an extensive range of
useful functions, procedures and overloaded operators. The FreeHDL version of
the IEEE library consists of the following packages:
1. std_logic_1164
2. numeric_bit
3. math_real
4. numeric_std
5. std_logic_arith
6. std_logic_unsigned
7. vital_timing
One other library is always defined by VHDL code processing systems namely the
work library. This library holds user compiled VHDL entity/architecture design
units.

5.10.3 VHDL simulation code structures.

In its most basic form VHDL circuit simulation code is structured as an entity-
architecture test bench which includes input signal test information.25 An example
outline of the basic format is
entity t e s t b e n c h i s
−− e n t i t y body s t a t e m e n t s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
−− a r c h i t e c t u r e body s t a t e m e n t s
end architecture b e h a v i o u r a l ;
25
Test signals are often called test vectors.

129
VHDL data types, functions and operators in package standard are always visible
to VHDL test bench code and reference to their use need not be added explicitly.
However, if the test bench entity-architecture uses data types or other items de-
fined in other libraries, for example the std_logic type in the IEEE library, then
reference to them needs to be added before each entity-architecture pair where they
are used. Libraries are referenced using the VHDL library and use statements. An
example showing how these statements are employed is outlined in the following
VHDL code segment:
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity t e s t b e n c h i s
−− e n t i t y body s t a t e m e n t s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
−− a r c h i t e c t u r e body s t a t e m e n t s
end architecture b e h a v i o u r a l ;

Here the VHDL code word all signifies that all items in a specific library are to be
made available for use in the following entity/architecture pair; testbench in the
above example. If more than one library is to be used then a library/use statement
is needed for each library reference. Most complete VHDL circuit simulation pro-
grams consist of more than one entity/architecture pair. In such cases the circuit
test bench, with its signal test vectors, must be the last entry in the program. An
example of a more complex VHDL program structure is
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity comp1 i s
−− e n t i t y body s t a t e m e n t s
end entity comp1 ;
−−
architecture b e h a v i o u r a l of comp1 i s
−− a r c h i t e c t u r e body s t a t e m e n t s
end architecture b e h a v i o u r a l ;
−−
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity comp2 i s
−− e n t i t y body s t a t e m e n t s

130
end entity comp2 ;
−−
architecture b e h a v i o u r a l of comp2 i s
−− a r c h i t e c t u r e body s t a t e m e n t s
end architecture b e h a v i o u r a l ;

−−
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
use work . a l l ;
−−
entity t e s t b e n c h i s
−− e n t i t y body s t a t e m e n t s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
−− a r c h i t e c t u r e body s t a t e m e n t s
end architecture b e h a v i o u r a l ;
During the conversion of VHDL code to a machine code simulation program each
entity/architecture pair, prior to the final test bench entry, is compiled as a sep-
arate design unit and stored in the work library26 . Compiled design units held in
the work library can be referenced in other entity/architecture models provided
the VHDL statement use work.all; 27 is inserted in the VHDL simulation code prior
to each entity/architecture statement where they are referenced.

26
The testbench entity/architecture pair is also, of course, compiled but this design unit is the
one that is run as the executable simulation program.
27
References to individual items are also allowed by inserting, for example, use.work.comb1;
use.work.comb2; in the VHDL code.

131
5.10.4 VHDL data types.

VHDL data
types

Scalar File Access Composite

Integer Real Enumerated Physical Array Record

Figure 5.23: VHDL data types

The chart shown in Fig. 5.23 indicates the different data types that are available
in the VHDL language. FreeHDL implements all these data types. In practical
circuit simulation the different VHDL data types are normally used to specify
(1) signals, (2) variables and (3) constants28 . During simulation Qucs/FreeHDL
automatically stores the values of integer, real and enumerated bit signals as simu-
lation time progresses. Furthermore, bit_vector and IEEE signal types including
std_logic_vector are also stored. Signals of these types are then available for
plotting and tabulation using the Timing, Truth table, Tabular and Cartesian out-
put diagrams. Selected elements in user defined composite signals, those that are
stored in arrays for example29 , can be assigned to the basic signal types then dis-
played.30 . An example of how this is done is given in later sections of these update
tutorial notes. Note - the values of variables and constants are not recorded during
simulation.
28
Type file is of course different in that it is used to store either test vectors, component data
such as ROM contents and output simulation results.
29
Please note that signal types based on the composite type record will probably cause the Qucs
simulation cycle to fail - work on this data type has been added to the to-do list.
30
Qucs/FreeHDL also automatically collects waveform data for composite signals based on arrays
of bit and IEEE signal types. However, in the case of large arrays care is needed when plotting
or tabulating these directly because the entire contents of an array is output each time a signal
is displayed.

132
5.10.5 An example VHDL simulation employing integer signals.
The following VHDL code demonstrates how the integer data type can be used
to represent signals. In this example signals A, B change state on the rising edge
of clock clk. The code tests the addition of integer signals and constants using
arithmetic operators defined in library STD.31 The results from this simulation
are shown in Fig. 5.24.
−− A v e r y b a s i c t e s t o f d a t a t y p e i n t e g e r .
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
s i g n a l A, B, C : i n t e g e r := 0 ;
signal c l k : b i t ;
begin
p0 : process i s −− Generate c l o c k s i g n a l .
begin
c l k <= ’ 0 ’ ; wait f o r 10 ns ;
c l k <= ’ 1 ’ ; wait f o r 10 ns ;
end process p0 ;
−−
p1 : process ( c l k ) i s
begin
i f ( c l k ’ e v e n t and c l k = ’1 ’) then
A <= A + 1 ;
B <= B + 2 ;
end i f ;
end process p1 ;
C <= A + B ;
end architecture b e h a v i o u r a l ;

dtime 0 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n
clk.X
a.R 0 1 1 2 2 3 3 4 4 5
b.R 0 2 2 4 4 6 6 8 8 10
c.R 0 3 3 6 6 9 9 12 12 15

Figure 5.24: Output results for a simple test bench example employing integer
signals.

31
The specification for the FreeHDL library STD can be found in text file freehdl-
0.0.3/std/standard.vhdl.

133
5.10.6 Multivalued logic.
Although signal types bit and bit-vector are widely employed when simulating dig-
ital systems one of their great weaknesses is the fact that it is difficult to represent
signal bus systems simply using only logic ’0’ and logic ’1’ signal encoding. More-
over, circuits where bus signal contention occurs often result in simulation failure.
The IEEE std_logic_1164 package overcomes this limitation through the intro-
duction of a multivalued logic system which defines nine different logic values to
represent signal types and signal strengths. Not only is the bus contention problem
solved through logic resolving functions but the multivalued logic system allows
devices constructed from different manufacturing technologies to be simulated at
the same time, ensuring that the simulation process mirrors real circuit design
practices. The next two simulation examples introduce the nine value logic sys-
tem and demonstrate it’s use in the design of digital bus systems. Signals of type
real are also introduced to show their representation by Qucs. Listed below is the
VHDL code for a basic simulation which generates a set of IEEE std_logic, inte-
ger and real signals. Figure 5.25 illustrates how the Qucs Timing diagram displays
different signal types. A section of tabulated results are also given in Fig. 5.26.

library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
signal c l k : b i t ;
s i g n a l bv1 : b i t v e c t o r ( 8 downto 0 ) ;
s i g n a l s t d l 1 : s t d l o g i c v e c t o r ( 8 downto 0 ) ;
s i g n a l INT1 : i n t e g e r := 0 ;
s i g n a l INT2 : i n t e g e r := 9 9 ;
s i g n a l R1 : r e a l := 0 . 3 3 ;
s i g n a l R2 : r e a l := 9 9 . 0 ;
s i g n a l R3 : r e a l := 0 . 0 ;
s i g n a l R4 : r e a l := 0 . 0 ;
begin
p0 : process i s
begin
c l k <= ’ 0 ’ ; wait f o r 10 ns ;
c l k <= ’ 1 ’ ; wait f o r 10 ns ;
end process p0 ;
−−

134
p1 : process ( c l k ) i s
variable v1 : i n t e g e r := 0 ;
begin
i f ( c l k ’ e v e n t and c l k = ’ 1 ’ ) then
v1 := v1 +1;
case v1 i s
when 1 => bv1 <= ”000000000 ” ; stdl1 <= ”000000000 ” ;
when 2 => bv1 <= ”000000001 ” ; stdl1 <= ”000000001 ” ;
when 3 => bv1 <= ”000000011 ” ; stdl1 <= ”00000001X” ;
when 4 => bv1 <= ”000000111 ” ; stdl1 <= ”0000001XZ” ;
when 5 => bv1 <= ”000001111 ” ; stdl1 <= ”000001XZU” ;
when 6 => bv1 <= ”000011111 ” ; stdl1 <= ”00001XZUW” ;
when 7 => bv1 <= ”000111111 ” ; stdl1 <= ”0001XZUWL” ;
when 8 => bv1 <= ”001111111 ” ; stdl1 <= ”001XZUWLH” ;
when 9 => bv1 <= ”111111111 ” ; stdl1 <= ”01XZUWLH−” ;
when others => v1 := 0 ;
end case ;
end i f ;
end process p1 ;
p3 : process ( c l k ) i s
begin
i f ( c l k ’ e v e n t and c l k = ’1 ’) then
INT1 <= INT1 + 1 ;
INT2 <= INT2 −20;
end i f ;
−−
i f ( INT1 >= 9 ) then
INT1 <= 0 ;
INT2 <= 9 9 ;
end i f ;
end process p3 ;
−−
p4 : process ( c l k ) i s
Variable V2 : r e a l ;
begin
i f ( c l k ’ e v e n t and c l k = ’1 ’) then
R1 <= R1 + 1 . 0 ;
R2 <= R2 −20.0;
R3 <= R1∗R2 ;
R4 <= R2/ (R1 + 0 . 0 0 0 1 ) ;
end i f ;
−−
i f (R1 >= 2 0 . 0 ) then

135
 R1 <= 0 . 0 ;
R2 <= 9 9 . 0 ;
end i f ;
end process p4 ;
end architecture b e h a v i o u r a l ;

dtime 0 10n 20n 30n 40n 50n 60n 70n

clk.X
r1.R 0.33 1.33 1.33 2.33 2.33 3.33 3.33 4.33
r2.R 99 79 79 59 59 39 39 19
r3.R 0 32.67 32.67 105.07 105.07 137.47 137.47 129.87
r4.R 0 299.909 299.909 59.394 59.394 25.3208 25.3208 11.7114
stdl1.X XXXXXXXXX 000000000 000000000 000000001 000000001 00000001X 00000001X 0000001XZ
int1.R 0 1 1 2 2 3 3 4
int2.R 99 79 79 59 59 39 39 19
bv1.X 000000000 000000000 000000000 000000001 000000001 000000011 000000011 000000111

dtime 70n 80n 90n 100n 110n 120n 130n 140n

clk.X
r1.R 4.33 4.33 5.33 5.33 6.33 6.33 7.33 7.33
r2.R 19 19 1 1 21 21 41 41
r3.R 129.87 129.87 82.27 82.27 5.33 5.33 132.93 132.93
r4.R 11.7114 11.7114 4.38789 4.38789 0.187614 0.187614 3.31748 3.31748
stdl1.X 0000001XZ 0000001XZ 000001XZX 000001XZX 00001XZX0 00001XZX0 0001XZX00 0001XZX00
int1.R 4 4 5 5 6 6 7 7
int2.R 19 19 1 1 21 21 41 41
bv1.X 000000111 000000111 000001111 000001111 000011111 000011111 000111111 000111111

Figure 5.25: Output results illustrating the TimeList representation of signals.

The VCD waveform interchange standard encodes digital signals as four different
logic levels. These are ’0’, ’1’, ’Z’ (high impedance) and ’X’ (unknown). Table 5.7
lists how the nine ieee.std_logic signal levels are represented using the VCD
format. Until the VCD standard is revised the Qucs/FreeHDL package is restricted
to displaying simulation output data using the basic ’0’, ’1’, ’Z’ and ’X’ signal en-
coding. The next example shows how the IEEE std_logic signal type can be used
to simulate bus logic. The demonstration has been kept simple in order to keep

136
 VHDL signal levels VCD
’0’ Forcing logic 0 ’0’
’1’ Forcing logic 1 ’1’
’X’ Forcing unknown ’X’
’Z’ High impedance ’Z’
’U’ Uninitialised ’X’
’W’ Weak unknown ’0’
’L’ Weak logic 0 ’0’
’H” Weak logic 1 ’1’
’-’ Don’t care ’X’

Table 5.7: IEEE multivalue logic and VCD representation.

the VHDL code short. The code fragment simulates two tri-state buffers which
pass their outputs to bus drivers who’s outputs connect on a common signal bus.
The bus drivers ensure that the outputs from the tri-state buffers are kept sepa-
rate before combining onto the common bus line. This allows the output signals
from the tri-state buffers and the combined signal to be plotted separately. The
resulting waveforms clearly show the std_logic resolution function in operation,
see Fig. 5.27 . Note the effect of the 7 ns delay on the plotted waveforms and the
use of the VHDL generic statement to set the invert device delay value.

−− Demonstration o f a s i m p l e bus s t r u c t u r e u s i n g
−− t h e IEEE s t d l o g i c d a t a t y p e .
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity buf i s
generic ( d e l a y : time := 0 ns ) ;
port ( in1 , c o n t r o l : in s t d l o g i c ;
out1 : out s t d l o g i c
);
end entity buf ;
architecture b e h a v i o u r a l of buf i s
begin
p0 : process ( in1 , c o n t r o l ) i s
begin
i f ( c o n t r o l = ’ 1 ’ ) then out1 <= i n 1 a f t e r d e l a y ;
e l s e out1 <= ’ Z ’ ;
end i f ;

137
 end process p0 ;
end architecture b e h a v i o u r a l ;
−−
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
entity i n v e r t i s
generic ( d e l a y : time := 0 ns ) ;
port ( i n 1 : in s t d l o g i c ;
out1 : out s t d l o g i c
);
end entity i n v e r t ;
−−
architecture b e h a v i o u r a l of i n v e r t i s
begin
out1 <= not i n 1 a f t e r d e l a y ;
end architecture b e h a v i o u r a l ;
−−
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−

entity buf2 i s
port ( i n 1 : in s t d l o g i c ;
out1 : out s t d l o g i c
);
end entity buf2 ;
−−
architecture d a t a f l o w of buf2 i s
begin
out1 <= i n 1 ;
end architecture d a t a f l o w ;
−−
library i e e e ;
use i e e e . s t d l o g i c 1 1 6 4 . a l l ;
−−
use work . a l l ;
−−
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture s t r u c t u r a l of t e s t b e n c h i s
signal data in 1 , data in 2 : s t d l o g i c ;

138
signal data out 1 , data out 2 : s t d l o g i c ;
signal data control , control buf1 : std logic ;
signal r e s u l t : s t d l o g i c ;
−−
begin
p0 : process i s
begin
d a t a i n 1 <= ’ 0 ’ ; wait f o r 5 ns ;
d a t a i n 1 <= ’ 1 ’ ; wait f o r 5 ns ;
end process p0 ;
−−
d a t a i n 2 <= not d a t a i n 1 ;
−−
p1 : process i s
begin
d a t a c o n t r o l <= ’ 1 ’ ; wait f o r 40 ns ;
d a t a c o n t r o l <= ’ 0 ’ ; wait f o r 40 ns ;
end process p1 ;
−−
c1g1 : entity buf port map( i n 1 => d a t a i n 1 , c o n t r o l => d a t a c o n t r o l ,
out1 => d a t a o u t 1 ) ;
c1g2 : entity i n v e r t generic map ( d e l a y => 7 ns )
port map( i n 1 => d a t a c o n t r o l , out1 => c o n t r o l b u f 1 ) ;
c1g3 : entity buf port map( i n 1 => d a t a i n 2 , c o n t r o l => c o n t r o l b u f 1 ,
out1 => d a t a o u t 2 ) ;
c1g4 : entity buf2 port map( i n 1 => d a t a o u t 1 , out1 => r e s u l t ) ;
c1g5 : entity buf2 port map( i n 1 => d a t a o u t 2 , out1 => r e s u l t ) ;
−−
end architecture s t r u c t u r a l ;

139
 dtime clk.X int1.R int2.R r1.R r2.R r3.R r4.R bv1.X stdl1.X
0 0 0 99 0.33 99 0 0 000000000 XXXXXXXXX
1e-8 1 1 79 1.33 79 32.7 300 000000000 000000000
2e-8 0 1 79 1.33 79 32.7 300 000000000 000000000
3e-8 1 2 59 2.33 59 105 59.4 000000001 000000001
4e-8 0 2 59 2.33 59 105 59.4 000000001 000000001
5e-8 1 3 39 3.33 39 137 25.3 000000011 00000001X
6e-8 0 3 39 3.33 39 137 25.3 000000011 00000001X
7e-8 1 4 19 4.33 19 130 11.7 000000111 0000001XZ
8e-8 0 4 19 4.33 19 130 11.7 000000111 0000001XZ
9e-8 1 5 -1 5.33 -1 82.3 4.39 000001111 000001XZX
1e-7 0 5 -1 5.33 -1 82.3 4.39 000001111 000001XZX
1.1e-7 1 6 -21 6.33 -21 -5.33 -0.188 000011111 00001XZX0
1.2e-7 0 6 -21 6.33 -21 -5.33 -0.188 000011111 00001XZX0
1.3e-7 1 7 -41 7.33 -41 -133 -3.32 000111111 0001XZX00
1.4e-7 0 7 -41 7.33 -41 -133 -3.32 000111111 0001XZX00
1.5e-7 1 8 -61 8.33 -61 -301 -5.59 001111111 001XZX001
1.6e-7 0 8 -61 8.33 -61 -301 -5.59 001111111 001XZX001
1.7e-7 1 9 -81 9.33 -81 -508 -7.32 111111111 01XZX001X
1.8e-7 0 0 99 9.33 -81 -508 -7.32 111111111 01XZX001X
1.9e-7 1 1 79 10.3 -101 -756 -8.68 111111111 01XZX001X

clk.X int1.R int2.R r1.R r2.R r3.R r4.R bv1.X stdl1.X

00000 0 0 99 0.33 99 0 0 000 0 0 0 0 0 0 XXXXXXXXX
00001 1 1 79 1.33 79 32.67 299.909 000 0 0 0 0 0 0 000000000
00010 0 1 79 1.33 79 32.67 299.909 000 0 0 0 0 0 0 000000000
00011 1 2 59 2.33 59 105.07 59.394 000 0 0 0 0 0 1 000000001
00100 0 2 59 2.33 59 105.07 59.394 000 0 0 0 0 0 1 000000001
00101 1 3 39 3.33 39 137.47 25.3208 000 0 0 0 0 1 1 00000001X
00110 0 3 39 3.33 39 137.47 25.3208 000 0 0 0 0 1 1 00000001X
00111 1 4 19 4.33 19 129.87 11.7114 000 0 0 0 1 1 1 0 0 0 0 0 0 1 XZ
01000 0 4 19 4.33 19 129.87 11.7114 000 0 0 0 1 1 1 0 0 0 0 0 0 1 XZ
01001 1 5 1 5.33 1 82.27 4.38789 000 0 0 1 1 1 1 0 0 0 0 0 1 XZX
01010 0 5 1 5.33 1 82.27 4.38789 000 0 0 1 1 1 1 0 0 0 0 0 1 XZX
01011 1 6 21 6.33 21 5.33 0.187614 000 0 1 1 1 1 1 0 0 0 0 1 XZX0
01100 0 6 21 6.33 21 5.33 0.187614 000 0 1 1 1 1 1 0 0 0 0 1 XZX0
01101 1 7 41 7.33 41 132.93 3.31748 000 1 1 1 1 1 1 0 0 0 1 XZ X0 0
01110 0 7 41 7.33 41 132.93 3.31748 000 1 1 1 1 1 1 0 0 0 1 XZ X0 0
01111 1 8 61 8.33 61 300.53 5.59338 001 1 1 1 1 1 1 0 0 1 XZX0 0 1
10000 0 8 61 8.33 61 300.53 5.59338 001 1 1 1 1 1 1 0 0 1 XZX0 0 1
10001 1 9 81 9.33 81 508.13 7.32284 111 1 1 1 1 1 1 0 1 XZ X0 0 1 X
10010 0 0 99 9.33 81 508.13 7.32284 111 1 1 1 1 1 1 0 1 XZ X0 0 1 X
10011 1 1 79 10.33 101 755.73 8.68158 111 1 1 1 1 1 1 0 1 XZ X0 0 1 X
10100 0 1 79 10.33 101 755.73 8.68158 111 1 1 1 1 1 1 0 1 XZ X0 0 1 X

Figure 5.26: Output results illustrating tabular representation of signals.

140
dtime 0 5n 7n 10n 15n 20n 25n 30n 35n 40n 45n 47n 50n 55n 60n 65n 70n 75n 80n

data_in_1.X
data_in_2.X
data_out_1.X Z Z Z Z Z Z Z Z Z
data_out_2.X Z Z Z Z Z Z Z Z Z Z Z
data_control.X
control_buf1.X X X
result.X Z Z

dtime 80n 85n 87n 90n 95n 100n 105n 110n 115n 120n 125n 127n 130n 135n 140n 145n 150n 155n

data_in_1.X
data_in_2.X
data_out_1.X Z Z Z Z Z Z Z Z Z
data_out_2.X Z Z Z Z Z Z Z Z Z
data_control.X
control_buf1.X
result.X Z Z Z

Figure 5.27: Signal waveforms for the simple bus example.

5.10.7 Run debugging of VHDL simulation code.

The VHDL language has a number of built in features that allow the debugging of
VHDL code at simulation time. In this section the VHDL reserved words assert,
report and severity are introduced and their use as code debugging aids explained
by way of a more detailed design example. In the previous digital tutorial update
a structural design of a 4 bit digital multiplier was introduced as an example that
employed the Qucs schematic capture digital simulation route. The next example
extends the previous multiplier design to 16 bits. However, at a structural level
the larger multiplier becomes very detailed and it’s design can be prone to error.
To demonstrate the power of VHDL the 16 bit multiplier has been redesigned at a
functional level. A block diagram of the multiplier simulation test bench is given in
Fig. 5.28: firstly a clock strobes a data generator unit which generates a sequence of
integer numbers. These are converted to 16 bit_vectors and applied to the 16 bit
multiplier unit as inputs x and y; secondly the 16-bit multiplier on sensing a change
in inputs x or y converts these signals from 16 bit_vectors to integers, multiples
them and finally converts the integer result to 32 bit_vector output Res_bit.
Although standard library STD defines arithmetic operations for integers it does
not provide functions for the conversion of integers to bit_vectors or the reverse
operation. The following VHDL listing gives the complete simulation test bench

141
program for the 16 bit multiplier including the required data conversion functions.
VHDL debug or message reporting code using the reserved words assert, report and
severity have been added to the data_generator and functional_multiplier
architecture code. During simulation these text strings, and the simulation time
when they were actioned, are written to the Qucs log.txt file, giving a trace record
of the simulation activity. In cases where an error occurs at severity level failure
the simulation will terminate. FreeHDL allows VHDL report statements without
an accompanying assert statement.32 A typical Timing diagram plot for this design
is shown in Fig. 5.29

X Res_bit
Data 16 bit
16
functional
generator Y 32
multiplier
16

CLK
CLOCK
X => bit_vector(15 downto 0)

Y => bit_vector(15 downto 0)

Res_bit => bit_vector(31 downto 0)

Figure 5.28: Block diagram of a 16 bit functional multiplier.

−− 16 b i t d i g i t a l m u l t i p l i e r example .
−− S i m u l a t i o n t r a c e u s i n g a s s e r t , r e p o r t and s e v e r i t y s t a t e m e n t s .
−−
entity c l o c k i s
port ( c l k : out b i t ) ;
end entity c l o c k ;
−−
architecture b e h a v i o u r a l of c l o c k i s
begin
p0 : process i s
32
One of the changes at the 1993 revision of the IEEE VHDL 1076-1987 standard was to al-
low report statements without the previous mandatory assert clause. FreeHDL attempts to
comply with the 1993 revision.

142
 begin
c l k <= ’ 0 ’ ; wait f o r 10 ns ;
c l k <= ’ 1 ’ ; wait f o r 10 ns ;
end process p0 ;
end architecture b e h a v i o u r a l ;
−−
entity d a t a g e n e r a t o r i s
port ( c l k : in b i t ;
x , y : out b i t v e c t o r ( 1 5 downto 0 )
);
end entity d a t a g e n e r a t o r ;
−−
architecture b e h a v i o u r a l of d a t a g e n e r a t o r i s
type mem array 16 i s array ( 1 to 8 ) of i n t e g e r ;
s i g n a l count : i n t e g e r := 0 ;
−−
function i n t e g e r t o v e c t o r 1 6 ( i n t n o : i n t e g e r ) return b i t v e c t o r
is
variable n i : i n t e g e r ;
variable r e t u r n v a l u e : b i t v e c t o r ( 1 5 downto 0 ) ;
begin
assert ( n i < 0 )
report ”Function i n t e g e r t o v e c t o r 3 2 : i n t e g e r number must be >= 0 ”
severity f a i l u r e ;
n i := i n t n o ;
for i in r e t u r n v a l u e ’ Reverse Range loop
i f ( ( n i mod 2 ) =1 ) then r e t u r n v a l u e ( i ) := ’ 1 ’ ;
e l s e r e t u r n v a l u e ( i ) := ’ 0 ’ ;
end i f ;
n i := n i / 2 ;
end loop ;
return r e t u r n v a l u e ;
end i n t e g e r t o v e c t o r 1 6 ;
−−
begin
p1 : process ( c l k ) i s
variable x i : mem array 16 := ( 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 ) ;
variable y i : mem array 16 := ( 2 , 4 , 6 , 8 , 1 0 , 1 2 , 1 4 , 1 6 ) ;
variable xh , yh : i n t e g e r ;
variable c o u n t i : i n t e g e r ;
begin

c o u n t i := count +1;

143
 i f ( c o u n t i > 8 ) then
c o u n t i := 1 ;
end i f ;
xh := x i ( c o u n t i ) ;
yh := y i ( c o u n t i ) ;
x <= i n t e g e r t o v e c t o r 1 6 ( xh ) ;
y <= i n t e g e r t o v e c t o r 1 6 ( yh ) ;
count <= c o u n t i ;
report ”In p r o c e s s p1 . d a t a g e n e r a t o r . ” ;
end process p1 ;
end architecture b e h a v i o u r a l ;
−−
−−
entity f u n c t i o n a l m u l t i p l i e r i s
port ( x , y : in b i t v e c t o r ( 1 5 downto 0 ) ;
r e s b i t : out b i t v e c t o r ( 3 1 downto 0 )
);
end entity f u n c t i o n a l m u l t i p l i e r ;
−−
−−
architecture b e h a v i o u r a l of f u n c t i o n a l m u l t i p l i e r i s
−−
function v e c t o r t o i n t e g e r ( v1 : b i t v e c t o r ) return i n t e g e r i s
variable r e t u r n v a l u e : i n t e g e r : = 0 ;
a l i a s v2 : b i t v e c t o r ( v1 ’ l e n g t h −1 downto 0 ) i s v1 ;
begin
f o r i in v2 ’ h i g h downto 1 loop
i f ( v2 ( i ) = ’ 1 ’ ) then
r e t u r n v a l u e := ( r e t u r n v a l u e +1)∗2;
else
r e t u r n v a l u e := r e t u r n v a l u e ∗ 2 ;
end i f ;
end loop ;
i f v2 ( 0 ) = ’ 1 ’ then r e t u r n v a l u e := r e t u r n v a l u e +1;
end i f ;
return r e t u r n v a l u e ;
end v e c t o r t o i n t e g e r ;
−−
function i n t e g e r t o v e c t o r 3 2 ( i n t n o : i n t e g e r ) return b i t v e c t o r
is
variable n i : i n t e g e r ;
variable v a l u e : b i t v e c t o r ( 3 1 downto 0 ) ;
begin

144
 assert ( n i < 0 )
report ”Function i n t e g e r t o v e c t o r 3 2 : i n t e g e r number must be >= 0 ”
severity f a i l u r e ;
n i := i n t n o ;
for i in 0 to 31 loop
i f ( ( n i mod 2 ) =1 ) then v a l u e ( i ) := ’ 1 ’ ;
e l s e v a l u e ( i ) := ’ 0 ’ ;
end i f ;
i f n i > 0 then n i := n i / 2 ;
e l s e n i := ( ni −1)/2;
end i f ;
end loop ;
return v a l u e ;
end i n t e g e r t o v e c t o r 3 2 ;
−−
begin
p0 : process ( x , y ) i s
variable xi , yi , prod mult : i n t e g e r ;
begin
x i := v e c t o r t o i n t e g e r ( x ) ;
y i := v e c t o r t o i n t e g e r ( y ) ;
prod mult := x i ∗ y i ;
r e s b i t <= i n t e g e r t o v e c t o r 3 2 ( prod mult ) ;
report ”In p r o c e s s p1 . f u n c t i o n a l m u l t i p l i e r ” ;
end process p0 ;
end architecture b e h a v i o u r a l ;
−−
entity t e s t 2 v h d l 1 i s
end entity t e s t 2 v h d l 1 ;
−−
architecture b e h a v i o u r a l of t e s t 2 v h d l 1 i s
signal c l k : b i t ;
s i g n a l x , y : b i t v e c t o r ( 1 5 downto 0 ) ;
s i g n a l r e s b i t : b i t v e c t o r ( 3 1 downto 0 ) ;
−−
begin
d1 : entity work . c l o c k port map ( c l k ) ;
d2 : entity work . d a t a g e n e r a t o r port map( c l k , x , y ) ;
d3 : entity work . f u n c t i o n a l m u l t i p l i e r port map ( x , y , r e s b i t ) ;

end architecture b e h a v i o u r a l ;

145
 dtime 10n 20n 30n
clk.X
res_bit.X 00000000000000000000000000001000 00000000000000000000000000010010
x.X 0000000000000010 0000000000000011
y.X 0000000000000100 0000000000000110

dtime 20n 30n 40n

clk.X
res_bit.X 00000000000000000000000000010010 00000000000000000000000000100000
x.X 0000000000000011 0000000000000100
y.X 0000000000000110 0000000000001000

dtime 40n 50n 60n

clk.X
res_bit.X 00000000000000000000000000110010 00000000000000000000000001001000
x.X 0000000000000101 0000000000000110
y.X 0000000000001010 0000000000001100

dtime 60n 70n 80n

clk.X
res_bit.X 00000000000000000000000001100010 00000000000000000000000010000000
x.X 0000000000000111 0000000000001000
y.X 0000000000001110 0000000000010000

Figure 5.29: Typical timing diagram for the 16 bit functional multiplier.

More advanced output debug messages, and results tables, can be written to Qucs
message file log.txt by using the predefined data handling routines in STD library
package textio33 . This package contains functions for reading and writing STD
data types from and to files34 . The next segment of VHDL code illustrates how a
simple table of results can be written to file log.txt. The results table is shown in
Table 5.8.
−− Test t e x t i o p a c k a g e .
−−
l i b r a r y STD ;
use STD . t e x t i o . a l l ;
−−
entity Q u c s w r i t e t e s t i s
33
The specification for the FreeHDL package textio can be found in text file freehdl-
0.0.3/std/textio.vhdl.
34
VHDL allows data to be read from and written to the standard input and output streams as
well as user defined files. At this time only writing data to file log.txt and reading data from
user defined data files has been tested. Please note that the use of the textio package is very
much a cutting edge feature of the Qucs/FreeHDL software and is probably not bug free.

146
end entity Q u c s w r i t e t e s t ;
−−
architecture b e h a v i o u r a l of Q u c s w r i t e t e s t i s
begin
w r i t e t e s t : process i s
variable i n p u t l i n e , o u t p u t l i n e : l i n e ;
variable i n t 1 : i n t e g e r := 1 0 ;
begin
write ( output line , string ’ ( ” ” ) ) ;
w r i t e l i n e ( output , o u t p u t l i n e ) ;
w r i t e ( o u t p u t l i n e , s t r i n g ’ ( ” S t r i n g −> l o g . t x t ” ) ) ;
w r i t e l i n e ( output , o u t p u t l i n e ) ;
−−
test L1 : f o r i c in 1 to 5 loop
i n t 1 := i n t 1 + 1 ;
write ( output line , string ’ ( ”int1 = ” ) );
write ( output line , int1 ) ;
write ( output line , string ’ ( ” i n t 1 ˆ2 = ” )
);
write ( output line , int1 ∗ int1 ) ;
w r i t e l i n e ( output , o u t p u t l i n e ) ;
end loop t e s t L 1 ;
report ” F i n i s h e d t e s t f o r l o o p . ” ;
end process w r i t e t e s t ;
end architecture b e h a v i o u r a l ;

147
Output:
----------
Starting new simulation on Thu 24. Aug 2006 at 13:10:56
running C++ conversion... done.
compiling functions... done.
compiling main... done.
linking... done.
simulating...
Output to STD output -> log.txt
int1 = 11 int1^2 = 121
int1 = 12 int1^2 = 144
int1 = 13 int1^2 = 169
int1 = 14 int1^2 = 196
int1 = 15 int1^2 = 225
0 fs + 0d: NOTE: Finished test for loop.
running VCD conversion... done.
Simulation ended on Thu 24. Aug 2006 at 13:10:57
Ready.
Errors:
--------

Table 5.8: Log.txt file showing tabular output results.

148
5.10.8 Testing digital systems using test vectors stored on disk.
In an attempt on my part to review all the new features introduced in the previous
sections of this update the final example demonstrates how test vectors stored
on disk, as a text file, can be read by the simulation program at the start of a
simulation, then applied to the inputs of the digital system under test. The code
for this example is given in the following listing:
−− T e s t i n g d i g i t a l c i r c u i t s u s i n g t e s t v e c t o r s
−− s t o r e d as a t e x t f i l e on d i s k .
−−
entity comb1 i s
port ( a , b , c , d : in b i t ;
y : out b i t
);
end entity comb1 ;
−−
architecture d a t a f l o w of comb1 i s
begin
y <= ( a nand b ) or ( c and d ) ;
end architecture d a t a f l o w ;
−−
l i b r a r y STD ;
use STD . t e x t i o . a l l ;
−−
entity t e s t b e n c h i s
end entity t e s t b e n c h ;
−−
architecture b e h a v i o u r a l of t e s t b e n c h i s
signal clock : b i t ;
s i g n a l v1 , v2 , v3 , v4 , y o u t : b i t ;
type a r r a y l i s t i s array ( 1 to 2 0 ) of b i t ;
s i g n a l v1sd , v2sd , v3sd , v4sd : a r r a y l i s t ;
−−
Procedure s t o r e d a t a ( variable number : out i n t e g e r ) i s
variable d1 , d2 , d3 , d4 : b i t ;
variable i n l i n e , o u t l i n e : l i n e ;
variable i : integer ;
variable m y s t r i n g : s t r i n g ( 1 to 2 0 ) := c r & ”C o n s t r a i n e d s t r i n g ” & c r ;
f i l e i n f i l e : t e x t open read mode i s ”/mnt/ hda2 / qucs − 0 . 0 . 1 0 f / t e s t 1 d a t a ” ;
begin
report m y s t r i n g ;
i := 1 ;
while not ( e n d f i l e ( i n f i l e ) ) loop

149
 readline ( infile , in line );
r e a d ( i n l i n e , d4 ) ;
r e a d ( i n l i n e , d3 ) ;
r e a d ( i n l i n e , d2 ) ;
r e a d ( i n l i n e , d1 ) ;
v1sd ( i ) <= d1 ;
v2sd ( i ) <= d2 ;
v3sd ( i ) <= d3 ;
v4sd ( i ) <= d4 ;
report ”In f i l e r e a d l o o p . ” ;
i := i +1;
i f ( i > 2 0 ) then e x it ;
end i f ;
number:= i ;
end loop ;
end procedure s t o r e d a t a ;
−−
begin
p0 : process i s −− Generate a c l o c k s i g n a l .
begin
c l o c k <= ’ 1 ’ ; wait f o r 10 ns ;
c l o c k <= ’ 0 ’ ; wait f o r 10 ns ;
end process p0 ;
−−
g0 : entity work . comb1 port map ( v1 , v2 , v3 , v4 , y o u t ) ;
−−
p1 : process i s −− Read t e s t v e c t o r s from d i s k and
−− apply data to c i r c u i t inputs .
variable n o r e a d s : i n t e g e r ;
variable i n l i n e , o u t l i n e : l i n e ;
begin
store data ( no reads ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ”count = ”) ) ;
w r i t e ( o u t l i n e , n o r e a d s −1);
w r i t e l i n e ( output , o u t l i n e ) ;
−−
f o r k in 1 to n o r e a d s −1 loop −− Count up .
wait u n t i l ( c l o c k ’ e v e n t and c l o c k = ’ 1 ’ ) ;
v1 <= v1sd ( k ) ;
v2 <= v2sd ( k ) ;
v3 <= v3sd ( k ) ;
v4 <= v4sd ( k ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ”Time = ” ) , l e f t , 8 ) ;

150
 w r i t e ( o u t l i n e , now , r i g h t , 1 0 ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ” Test v e c t o r s −> ” ) , r i g h t , 20 ) ;
w r i t e ( o u t l i n e , v4 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v3 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v2 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v1 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ”k = ” ) , r i g h t , 10 ) ;
write ( out line , k ) ;
w r i t e l i n e ( output , o u t l i n e ) ;
wait u n t i l ( c l o c k ’ e v e n t and c l o c k = ’ 0 ’ ) ;
end loop ;
−−
for k in n o r e a d s −1 downto 1 loop −− Count down .
wait u n t i l ( c l o c k ’ e v e n t and c l o c k = ’ 1 ’ ) ;
v1 <= v1sd ( k ) ;
v2 <= v2sd ( k ) ;
v3 <= v3sd ( k ) ;
v4 <= v4sd ( k ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ”Time = ” ) , l e f t , 8 ) ;
w r i t e ( o u t l i n e , now , r i g h t , 1 0 ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ” Test v e c t o r s −> ” ) , r i g h t , 20 ) ;
w r i t e ( o u t l i n e , v4 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v3 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v2 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , v1 , l e f t , 2 ) ;
w r i t e ( o u t l i n e , s t r i n g ’ ( ”k = ” ) , r i g h t , 10 ) ;
write ( out line , k ) ;
w r i t e l i n e ( output , o u t l i n e ) ;
wait u n t i l ( c l o c k ’ e v e n t and c l o c k = ’ 0 ’ ) ;
end loop ;
wait ;
end process p1 ;
end architecture b e h a v i o u r a l ;
Although the listing above is relatively short, careful study of it’s contents should
allow readers to identify many of the new Qucs/FreeHDL features introduced ear-
lier. Moreover in some sections, the code illustrates extra features which will be
familiar to those Qucs/FreeHDL users who have a more advanced knowledge of
the VHDL language. These are listed below with a number of general points:
• The VHDL code simulates the performance of a simple combinational logic
circuit called comb1: this has four inputs (a, b, c, d) of type bit and one
output (y) of type bit35 .
35
Type bit was chosen for this example rather than one of the IEEE signal types because package

151
 • The testbench being simulated consists of two processes: process p0 generates
a clock signal with a period of 20 ns; process p1 inputs test data held in file
test1_data 36 and stores it in four signal arrays (v1sd, v2sd, v3sd and v4sd),
applying this data to the inputs of the circuit under test at the leading edges
of the clock pulse. Note process p1 only executes once due to the wait
statement at its end.

• An instantiation of the comb1 component is included in the testbench archi-

tecture. Note the use of the VHDL entity work.comb1 construction, this is
an alternative for use work.all ;

• The test vector data held in file test_data is read by procedure store_data
which returns the number of lines of data read in variable number. File
handling, including reading data from disk, is undertaken with predefined
routines in package textio.

• The first report statement in procedure store_data writes string my_string

to file log.txt. My_string is an example of the VHDL constrained string type,
consisting of non-printable control characters37 concatenated with printable
characters.

• Two loops are employed in process p1 to apply signal test vectors to the
input of comb1: the first loop counts up from one and the second loop
counts down from the number of lines of test vectors read by procedure
store_data, effectively generating test vectors in a way similar to using an
up-down pattern generator counter. Note that the signal data is applied to
the circuit under test on the rising edge of the clock signal and that the
applied signal vector sequence is really up to the imagination of the VHDL
programmer.

• The write statements in the process p1 for loops demonstrate the formatted
version of the textio write statement. This greatly assists in setting up
tables of results. Table 5.9 gives a typical log.txt content for the comb1 test
simulation.

• In process p1 signals v1, v2, v3 and v4 are assigned an indexed value from
(type array_list) v1sd, v2sd, v3sd and v4sd signals. During simulation
textio does not handle the IEEE multivalue logic types.
36
I use the Knoppix version of the Linux/GNU operating system for all work on the Qucs project.
The absolute location of the test data file will depend on where Qucs and FreeHDL have been
installed and the location where work files are kept.
37
Type character in package standard lists the two letter codes used by VHDL to represent
non-printable control characters.

152
 Qucs/FreeHDL stores signal values as a simulation progresses. Hence, it
is theoretically possible to display both the standard and composite signal
types. A typical waveform plot for signals v1, v2, v3, v4 and y_out is given in
Fig. 5.30. Fig. 5.31 illustrates a waveform plot of the composite signals v1sd,
v2sd, v3sd and v4sd. In Fig. 5.31 each group is plotted at a clock edge change
yielding identical groups of values; each vertical set of bits represents the bit
values for a single line in file test1_data. Compare the displayed values in
Fig. 5.31 with the contents of the test1_data file shown in Fig. 5.32. As
mentioned before some care is needed when plotting, or tabulating, composite
signals, particularly when the array sizes are large; array dimensions above
roughly 50 become difficult to plot on a normal resolution screen. In such
cases it is better to slice part of an array and assign the required values to a
signal that can be easily displayed.

dtime 0 10n 20n 30n 40n 50n 60n 70n 80n 90n 100n 110n 120n 130n 140n 150n 160n 170n 180n

v1.X
v2.X
v3.X
v4.X
y_out.X

dtime 180n 190n 200n 210n 220n 230n 240n 250n 260n 270n 280n 290n 300n 310n 320n 330n 340n 350n 360n

v1.X
v2.X
v3.X
v4.X
y_out.X

dtime 220n 230n 240n 250n 260n 270n 280n 290n 300n 310n 320n 330n 340n 350n 360n 370n 380n 390n 400n
v1.X
v2.X
v3.X
v4.X
y_out.X

dtime 400n 410n 420n 430n 440n 450n 460n 470n 480n 490n 500n 510n 520n 530n 540n 550n 560n 570n 580n
v1.X
v2.X
v3.X
v4.X
y_out.X

Figure 5.30: Typical timing diagram for comb1 simulation.

153
Output :
−−−−−−−−−−
S t a r t i n g new s i m u l a t i o n on F r i 25 . Aug 2006 a t 14 : 35 : 48
r u n n i n g C++ c o n v e r s i o n . . . done .
c o m p i l i n g f u n c t i o n s . . . done .
c o m p i l i n g main . . . done .
l i n k i n g . . . done .
simulating . . .
0 f s + 0d : NOTE:
Constrained s t r i n g
0 f s + 0d : NOTE: In f i l e r e a d l o o p .
.
0 f s + 0d : NOTE: In f i l e r e a d l o o p .
count = 16
Time = 0 ns Test v e c t o r s −> 0 0 0 0 k = 1
Time = 20 ns Test v e c t o r s −> 0 0 0 0 k = 2
Time = 40 ns Test v e c t o r s −> 0 0 0 1 k = 3
Time = 60 ns Test v e c t o r s −> 0 0 1 0 k = 4
.
Time = 200 ns Test v e c t o r s −> 1 0 0 1 k = 11
Time = 220 ns Test v e c t o r s −> 1 0 1 0 k = 12
Time = 240 ns Test v e c t o r s −> 1 0 1 1 k = 13
Time = 260 ns Test v e c t o r s −> 1 1 0 0 k = 14
Time = 280 ns Test v e c t o r s −> 1 1 0 1 k = 15
Time = 300 ns Test v e c t o r s −> 1 1 1 0 k = 16
Time = 320 ns Test v e c t o r s −> 1 1 1 1 k = 16
Time = 340 ns Test v e c t o r s −> 1 1 1 1 k = 15
Time = 360 ns Test v e c t o r s −> 1 1 1 0 k = 14
Time = 380 ns Test v e c t o r s −> 1 1 0 1 k = 13
Time = 400 ns Test v e c t o r s −> 1 1 0 0 k = 12
.
Time = 560 ns Test v e c t o r s −> 0 1 0 0 k = 4
Time = 580 ns Test v e c t o r s −> 0 0 1 1 k = 3
r u n n i n g VCD c o n v e r s i o n . . . done .
S i m u l a t i o n ended on F r i 25 . Aug 2006 a t 14 : 35 : 50
Ready .
Errors :

Table 5.9: An edited version of the formatted tabular output results written to file
log.txt.

154
dtime 40n 50n 60n 70n 80n
v1sd.X 01010101010101010000 01010101010101010000 01010101010101010000 01010101010101010000
v2sd.X 00110011001100110000 00110011001100110000 00110011001100110000 00110011001100110000
v3sd.X 00001111000011110000 00001111000011110000 00001111000011110000 00001111000011110000
v4sd.X 00000000111111110000 00000000111111110000 00000000111111110000 00000000111111110000

Figure 5.31: Typical timing diagram for composite signals v1sd, v2sd, v3sd and
v4sd.

0 0 0 0
0 0 0 1
0 0 1 0
0 0 1 1
0 1 0 0
0 1 0 1
0 1 1 0
0 1 1 1
1 0 0 0
1 0 0 1
1 0 1 0
1 0 1 1
1 1 0 0
1 1 0 1
1 1 1 0
1 1 1 1

Figure 5.32: Comb1 simulation test vectors.

5.11 End note

Qucs 0.0.8 added digital simulation to the impressive list of features already avail-
able in the Qucs package. The 0.0.8 release represented a significant step forward
in the development of the Qucs project. The fact that there were bugs in the first
version of the digital simulator was not surprising given the complexity of the soft-
ware. Release 0.0.9 goes a long way to correcting the most annoying of these bugs.
It also adds a number of new features, the most notable being the new VHDL edi-

155
tor and the automatic generation of component symbols from hand crafted VHDL
model code. Qucs 0.0.10 and FreeHDL 0.0.3 adds a range of new features to the
software, particularly important are the use of the IEEE std_logic_1164 package
and the file handling routines found in the textio package. My thanks to Michael
Margraf and Stefan Jahn for all their encouragement during the period that I have
been testing the Qucs VHDL digital simulation and the subsequent writing of these
notes.

156
6 Transient Domain Flip-Flop
Models for Mixed-Mode
Simulation

6.1 Introduction

One of the primary aims of the Qucs project is the development of a universal
circuit simulator that allows circuit performance to be investigated from DC to
microwave frequencies. Adding performance analysis in the digital domain makes
Qucs a truly universal simulator. Qucs 0.0.8 was the first release to include digi-
tal simulation. Qucs digital simulation centres around VHDL using the FreeHDL
VHDL compiler to generate a machine code simulation of a circuit under test. Re-
lease 0.0.8 includes built-in models for the basic digital gates and a number of the
common sequential flip-flops. The Qucs gate models can be used in both digital
and transient simulation. Unfortunately, the flip-flop models are only available
in digital simulation. The current version of Qucs models flip-flops using VHDL
and does not provide time domain models for transient simulation. This is an
important omission which limits the Qucs simulator mixed-mode simulation ca-
pabilities. Mixed-mode simulation is a term commonly employed to describe the
simulation of circuits that contain both analogue and digital components. In the
real world circuits are, of course, not subdivided into neat boxes labelled analogue,
S-parameter, digital or any other physical domain. So it is of some importance
that Qucs device modelling be developed to allow circuits consisting of a range of
different analogue and digital components, to be simulated at the same time. Nor-
mally such systems are simulated in the time domain using large signal transient
simulation. Performance data being both analogue and digital expressed in tabu-
lar or graphical form. This tutorial note presents a number of transient simulation
models for flip-flops based on structural digital circuits, describes their use, and
outlines a number of example simulations derived from practical circuits.

157
6.2 Latches and flip-flops
Sequential digital devices generically known as flip-flops (SR, D, JK and T types)
are commonly classified into three major groups.

• Latches: basic or gated

• Pulse triggered flip-flops: master slave devices with or without data-lockout

• Edge-triggered flip-flops: leading or trailing edge triggered.

As the speed of electronic systems has increased so has the popularity of the
single edge-triggered flip-flops over the slower master slave devices. Today most
IC designs are based on D type edge-triggered devices rather than the earlier JK
master slave devices. Our concern here is the development of a consistant set of
models that allow the common flip-flops to be modelled accurately, and reliably,
in the transient time domain. In order to keep these models simple the D gated
and edge-triggered devices have been chosen as the fundamental building blocks
for the transient domain Qucs models. Using basic Boolean logic concepts it is
straightforward to show that JK and T edge-triggered flip-flop models can be
derived from the D flip-flop models.

6.3 The gated D latch

The circuit diagram for a gated D latch constructed from two input nand gates is
shown in Fig. 6.11 . Outputs Q and not Q (QB in Fig. 6.1) are derived from the two
cross coupled nand gates connected as a basic SR nand latch. Fig. 6.2 shows the
performance characteristics for this circuit. These were obtained using the simple
test configuration shown in Fig. 6.3. Logic one digital signals are represented by
1V and logic 0 signals by 0V in the transient analysis domain. Propagation delays
through the various circuit gates can be set by changing the delay time for each
gate. Cross coupled gates are often a cause of simulation failure due to the fact
that DC analysis fails to converge to a stable solution at the start of a transient
simulation. One approach that helps to force a stable DC solution is to set Q and
QB to known values, say logic 0 and logic 1, at the start of a simulation. In circuits
like the basic gated D latch shown in Fig. 6.1, where asynchronous set and reset
inputs are not included, this is not possible. However, flip-flops with asynchronous
set and reset inputs do allow the state of a flip-flop to be determined at a given
time in a simulation. In the examples that follow, whenever possible, the state of
the latch or flip-flop devices is set at the start of a simulation. In the majority
1
Richard S. Sandige, Modern Digital Design, 1990, McGraw-Hill International Editions.

158
of the example circuits, device delays have also been set to zero. It therefore
follows that most waveform plots show functional data rather than accurate timing
characteristics. In many mixed-mode simulations the digital elements present in a
design are often modelled as functional devices whose primary task is to generate
the signals needed for the overall circuit to function. A more detailed discussion
of the effects on transient simulation caused by including device timing delays is
presented in a later section of these notes.
D

& Q
&
D
times=20ns; 20ns Y3
C Y1
QB
&
C &
times=5ns; 5ns &
Y2
Y4
Y5

transient
simulation

TR1
Type=lin
Start=0
Stop=100 ns

Figure 6.1: Gated D latch with digital signal generators D and C

159
 1
D.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time
1
C.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

1
Q.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

1
QB.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

Figure 6.2: Gated D latch simulation waveforms

D Q

D Q transient
simulation
D
C QB
TR1
C Q Type=lin
C SUB1 Start=0
Num=2 File=gated_d_latch.sch Stop=200 ns

Figure 6.3: Gated D latch test circuit

160
6.4 Edge-triggered D type flip-flop
The schematic for a positive edge-triggered D flip-flop is shown in Fig. 6.42 . Asyn-
chronous set (SET) and reset (RESET) control inputs allow the flip-flop outputs
Q and not Q (QB in Fig. 6.4) to be set to known values at the start of a simu-
lation. The nand gates forming each of the cross coupled SR latches have their
delay times set at 0 ns. The edge-triggered D device is a building block for both
the JK and T types of flip-flop. A typical set of transient simulation test results
for the D flip-flop model are illustrated in Fig. 6.5. These where obtained using
the basic test configuration shown in Fig. 6.6.

SET

&
SET

Y7
I0

RESET & Q Q
&
I1
RESET Y10
CLOCK Y8

QB
& I2 &
CLOCK
Num=1 QB

Y11 Y2

I3

&
DIN

DIN Y9
Num=2

Figure 6.4: Positive edge-triggered D flip-flop circuit

2
David A. Hodges and Horace G. Jackson, Analysis and Design of Digital Integrated Circuits,
1998, Second edition, McGraw-Hill Book Company.

161
 1
R.Vt

0
0 5e-8 1e-7 1.5e-7 2e-7 2.5e-7 3e-7 3.5e-7 4e-7 4.5e-7 5e-7
time

1
DIN.Vt

0
0 5e-8 1e-7 1.5e-7 2e-7 2.5e-7 3e-7 3.5e-7 4e-7 4.5e-7 5e-7
time
CLOCK.Vt

0
0 5e-8 1e-7 1.5e-7 2e-7 2.5e-7 3e-7 3.5e-7 4e-7 4.5e-7 5e-7
time

1
Q.Vt

0 5e-8 1e-7 1.5e-7 2e-7 2.5e-7 3e-7 3.5e-7 4e-7 4.5e-7 5e-7
time

1
QB.Vt

0 5e-8 1e-7 1.5e-7 2e-7 2.5e-7 3e-7 3.5e-7 4e-7 4.5e-7 5e-7
time

Figure 6.5: Transient waveforms for the circuit shown in Fig. 6.6

1
DIN SUB2
S Q
D Q transient
S4 simulation
times=40ns; 40ns
CLOCK QB TR1
Q Type=lin
R Start=0
S2
times=5ns; 5ns Stop=500 ns
SUB1
R

S3
times=20ns; 1000ns

Figure 6.6: D flip-flop test circuit

162
6.5 The edge-triggered JK flip-flop
A leading edge-triggered JK flip-flop can be constructed using a positive edge-
triggered D flip-flop and external logic3 . The external logic generates the required
JK flip-flop characteristic equation given by

Q+ = J.Q + K.Q
Were Q, Q, J and K are the current state values of the device signals and Q+ is
the next state value of Q following the rising edge of the device clock pulse. The
schematic diagram for the edge triggered flip flop is shown in Fig. 6.7 and a typical
set of test waveforms in Fig. 6.8. These were obtained using the test circuit shown
in Fig. 6.9.

SET Q
&
1 S
D Q
J Y1 QB
Y2

CLOCK
1 Q
&
R
K
Y4
Y3 SUB1
File=dff_sr.sch
RESET

Figure 6.7: Positive edge-triggered JK flip-flop circuit

3
M. Morris Mano and Charles R Kime, Logic and Computer Design Fundamentals, 2004, Third
edition, Pearson Education International, Prentice Hall

163
 1

RESET.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7
time

1
CLOCK.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7
time

1
Q.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7
time

1
QB.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7
time

Figure 6.8: Transient waveforms for the circuit shown in Fig. 6.9

1
SUB2

Q
S
J Q transient
CLOCK simulation

TR1
QB
CLOCK Type=lin
Num=1 K Q Start=0
R Stop=100 ns

RESET SUB1

RESET

Figure 6.9: JK flip-flop test circuit showing JK operating in toggle mode

164
6.6 The edge-triggered T flip-flop
The characteristic equation for a leading edge-triggered flip-flop is4

Q+ = T ⊕ Q
where the symbols have the same meaning as the JK flip-flop. The circuit dia-
gram, test waveforms and test circuit for the edge-triggered flip-flop are given in
Figures 6.10 to 6.12.

SET

SET
TFF QQ
=1 S
D Q
TFF Y1
CLOCK QB
QB

Q
CLOCK
R

R SUB1
File=dff_sr.sch

Figure 6.10: Positive edge-triggered T flip-flop circuit

4
See footnote 2.

165
 1
SET.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time
1
CLOCK.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time
1
TFF.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

1
Q.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

1
QB.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7
time

Figure 6.11: Transient waveforms for the circuit shown in Fig. 6.12

SET

transient
SET
simulation
times=10 ns; 1000 ns
TFF Q TR1
S
T Q Type=lin
Start=0
T1 QB Stop=200ns
times=30 ns; 60 ns IntegrationMethod=Trapezoidal
Q Order=2
CLOCK
R

CLOCK 1
times=5 ns; 5 ns
SUB2
SUB1 File=Logic_one.sch
File=tff.sch

Figure 6.12: T flip-flop test circuit

166
6.7 Two example digital circuits
• A synchronous BCD up-counter: Figure 6.13 shows a synchronous BCD
up-counter constructed from four edge-triggered JK flip flops connected as
toggle flip-flops. The input signal waveforms and the corresponding counter
outputs Q0, Q1, Q2 and Q3 are illustrated in Fig. 6.14. These simulation
results were obtained using the default trapezoidal integration method with
order 2.
SUB5
& & &
1

File=Logic_one.sch

Y1 Y2 Y3
Q0 Q1 Q2 Q3
S S S S
J Q J Q J Q J Q

CLOCK

K Q K Q K Q K Q
R R R R
CLOCK
times=5 ns; 5ns

COUNT &
SUB1 SUB2 SUB3 SUB4
File=jkff.sch File=jkff.sch File=jkff.sch File=jkff.sch
COUNT Y4
times=5 ns; 1000ns transient
simulation
CLEAR 1
TR1
CLEAR Type=lin
Num=3 Y5 Start=0
times=10 ns; 1000ns Stop=120ns
IntegrationMethod=Trapezoidal
Order=2

Figure 6.13: Synchronous BCD up-counter circuit

At the start of simulation signal CLEAR is set to logic 1 which in turn causes
the counter to be reset to 0000. Similarly signal COUNT has to be set to 1
for counting to take place. Notice that the counter counts from 0 to 9 and
then resets to 0.

• A simple state machine: Figure 6.15 shows a simple sequential state

machine with input X and outputs Y1 and Y2. The outputs are synchronised
to the input clock. The state equations for this example are

J = X, K = 1, Y 1 = Q0.X, Y 2 = Q0

167
CLEAR.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
CLOCK.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
COUNT.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

1
Q0.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q1.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q2.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q3.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

Figure 6.14: Transient waveforms for the circuit shown in Fig. 6.13

168
 1
SUB4
X File=Logic_one.sch
Y2
1 S S
J Q D Q transient
X simulation
Y2 20ns
times=100ns;
TR1
Type=lin
K Q Q
R Start=0
CLOCK R Stop=350 ns
IntegrationMethod=Trapezoidal
SUB2
Order=2
CLOCK File=dff_sr.sch
times=5ns; 5ns
SUB1
File=jkff.sch

Y1
& S
D Q digital
simulation
Y1
Digi1
Type=TimeList
Q time=350 ns
R
RESET
SUB3
File=dff_sr.sch
RESET
times=15ns; 1000ns

Figure 6.15: A simple state machine

169
 1
CLOCK.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7 2.2e-7 2.4e-7 2.6e-7 2.8e-7 3e-7 3.2e-7 3.4e-7
time
1
RESET.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7 2.2e-7 2.4e-7 2.6e-7 2.8e-7 3e-7 3.2e-7 3.4e-7
time

1
X.Vt

0
0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7 2.2e-7 2.4e-7 2.6e-7 2.8e-7 3e-7 3.2e-7 3.4e-7
time

1
Y1.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7 2.2e-7 2.4e-7 2.6e-7 2.8e-7 3e-7 3.2e-7 3.4e-7
time

1
Y2.Vt

0 2e-8 4e-8 6e-8 8e-8 1e-7 1.2e-7 1.4e-7 1.6e-7 1.8e-7 2e-7 2.2e-7 2.4e-7 2.6e-7 2.8e-7 3e-7 3.2e-7 3.4e-7
time

Figure 6.16: Transient waveforms for the circuit shown in Fig. 6.15

170
6.8 VHDL code for the transient domain flip-flop
models
Although the primary purpose for developing the transient domain flip-flop mod-
els is the simulation of mixed-mode circuits, it is worth noting that because the
models have been constructed from Qucs gate primitives using a bottom-up de-
sign approach, Qucs can also use the models for digital simulation. Moreover,
provided the circuit being simulated does not contain any purely analogue com-
ponents Qucs will generate a VHDL model testbench that describes the function
and test sequence for the circuit being simulated. Shown in Fig. 6.17 is a digital
timelist waveform plot for the synchronous BCD up-counter introduced in the pre-
vious section of these notes. Listing 6.1 lists the VHDL code generated by Qucs
for the synchronous BCD up-counter example.
dtime 0 5n 10n 15n 20n 25n 30n 35n 40n 45n 50n 55n 60n 65n 70n 75n 80n 85n 90n 95n
clear.X
count.X
clock.X
q0.X
q1.X
q2.X
q3.X

Figure 6.17: Digital TimeList waveforms for the circuit shown in Fig. 6.13

Listing 6.1: VHDL testbench code for the circuit shown in Fig. 6.13
−− Qucs 0 . 0 . 9
−− /mnt/ hda2 / D i g i t a l S u b c i r c u i t s p r j / Sync BCD counter . s c h

entity S u b L o g i c o n e i s
port ( nnout L1 : out b i t ) ;
end entity ;
use work . a l l ;
architecture Arch Sub Logic one of S u b L o g i c o n e i s
s i g n a l gnd ,
L1 : b i t ;
begin
gnd <= ’ 0 ’ ;
L1 <= not gnd ;
nnout L1 <= L1 or ’ 0 ’ ;

171
end architecture ;

entity S u b d f f s r i s
port (CLOCK: in b i t ;
DIN : in b i t ;
nnout Q : out b i t ;
nnout QB : out b i t ;
RESET: in b i t ;
SET : in b i t ) ;
end entity ;
use work . a l l ;
architecture A r c h S u b d f f s r of S u b d f f s r i s
s i g n a l I0 ,
I2 ,
I1 ,
I3 ,
QB,
Q : bit ;
begin
nnout QB <= QB or ’ 0 ’ ;
nnout Q <= Q or ’ 0 ’ ;
I 1 <= not (CLOCK and RESET and I 0 ) ;
I 3 <= not (DIN and I 2 and RESET ) ;
QB <= not (RESET and I 2 and Q) ;
Q <= not ( I 1 and QB and SET ) ;
I 0 <= not ( I 3 and I 1 and SET ) ;
I 2 <= not (CLOCK and I 3 and I 1 ) ;
end architecture ;

entity S u b j k f f i s
port ( nnnet6 : in b i t ;
nnnet1 : in b i t ;
nnnet8 : in b i t ;
nnout nnnet3 : out b i t ;
nnout nnnet7 : out b i t ;
nnnet9 : in b i t ;
nnnet10 : in b i t ) ;
end entity ;

172
use work . a l l ;
architecture A r c h S u b j k f f of S u b j k f f i s
s i g n a l nnnet0 ,
nnnet2 ,
nnnet4 ,
nnnet5 ,
nnnet7 ,
nnnet3 : b i t ;
begin
nnnet0 <= not nnnet1 ;
nnnet2 <= nnnet3 and nnnet0 ;
nnnet4 <= nnnet2 or nnnet5 ;
nnnet5 <= nnnet6 and nnnet7 ;
nnout nnnet7 <= nnnet7 or ’ 0 ’ ;
nnout nnnet3 <= nnnet3 or ’ 0 ’ ;
SUB1 : entity S u b d f f s r port map ( nnnet8 , nnnet4 , nnnet3 ,
nnnet7 , nnnet10 , nnnet9 ) ;
end architecture ;

entity TestBench i s
end entity ;
use work . a l l ;

architecture Arch TestBench of TestBench i s

s i g n a l CLEAR,
COUNT,
CLOCK,
Q3 ,
Q0 ,
Q1 ,
Q2 ,
nnnet0 ,
nnnet1 ,
nnnet2 ,
nnnet3 ,
nnnet4 ,
nnnet5 ,
nnnet6 ,
nnnet7 ,
nnnet8 ,

173
 nnnet9 : b i t ;
begin
SUB5 : entity S u b L o g i c o n e port map ( nnnet0 ) ;
nnnet1 <= Q0 and nnnet2 ;
nnnet3 <= Q1 and nnnet1 ;
nnnet4 <= Q2 and nnnet3 ;
SUB2 : entity S u b j k f f port map ( nnnet1 , nnnet1 , nnnet5 ,
Q1 , nnnet6 , nnnet0 , nnnet7 ) ;

SUB3 : entity S u b j k f f port map ( nnnet3 , nnnet3 , nnnet5 ,

Q2 , nnnet8 , nnnet0 , nnnet7 ) ;
SUB1 : entity S u b j k f f port map ( nnnet0 , nnnet0 , nnnet5 ,
Q0 , nnnet9 , nnnet0 , nnnet7 ) ;
nnnet5 <= COUNT and CLOCK;
nnnet7 <= not CLEAR;

CLEAR: process
begin
CLEAR <= ’ 1 ’ ; wait for 10 ns ;
CLEAR <= ’ 0 ’ ; wait for 1000 ns ;
end process ;

COUNT: process
begin
COUNT <= ’ 0 ’ ; wait for 5 ns ;
COUNT <= ’ 1 ’ ; wait for 1000 ns ;
end process ;

CLOCK: process
begin
CLOCK <= ’ 0 ’ ; wait for 5 ns ;
CLOCK <= ’ 1 ’ ; wait for 5 ns ;
end process ;

SUB4 : entity S u b j k f f port map ( nnnet4 , Q0 , nnnet5 ,

Q3 , nnnet2 , nnnet0 , nnnet7 ) ;
end architecture ;

174
6.9 Generating a library of mixed-mode digital
components
The Qucs project facilities offer users a simple and convenient approach to devel-
oping libraries of components that are linked by a common theme; in these notes
this is digital component models for transient simulation. To form a library create
a new folder, at a point on a disk file system that users have read/write access,
giving it a suitable name, for example
f l i p f l o p models t r a n sim p r j .
Next move into the new library folder a copy of each of the schematic capture files
for the flip-flop models introduced in these notes. These are:
d f f s r . sch , j k f f . sch , t f f . sch , and gated d l a t c h . sch .
A copy of the schematic for setting nodes to logic one is also required
( l o g i c one . sch ) .
These models are then freely available for use in any projects which users are
working on. They can be copied into such projects using the ”Add files to Project...”
menu button found under the Qucs Project drop-down menu. Similarly any new
models developed as part of a project can be added to the library and used again
in the future.

6.10 Digital component propagation time delays and

transient simulation numerical stability
Transient simulation is in general much slower than digital simulation using VHDL
generated machine code. The large signal transient simulation models of flip-flops
and other sequential digital devices are intended for use in mixed-mode circuit
simulation rather than being used for pure digital circuit simulation. An interest-
ing, and indeed very important question, relates to the efficiency, and accuracy,
of the numerical analysis algorithms employed in the integration routines that are
central to transient circuit simulation. Qucs allows users to select the algorithm
they wish to employ for transient simulation. The available algorithms are Trape-
zoidal, Euler, Gear and Adams Moulton; in each case the algorithm order can be
set from 1 to 6. The second order Trapezoidal integration algorithm is used by
Qucs as the default for transient simulation. To test which of these algorithms
offers the most time efficient solution to the transient simulation of digital circuits,
that include flip-flops, the BCD counter shown in Fig. 6.13 was used as a test case
and repeatedly simulated using different integration routines and algorithm orders.

175
 Order Trapezoidal Euler Gear Adams Moulton
1 1 1.62 1.65 1.62
2 1 1.62 0.44 1
4 1 1.62 1.28 0.39
6 1 1.62 0.28 0.18

Table 6.1: Relative simulation times for the circuit shown in Fig. 6.13

Order Number or rejections Average time step

1 1470 5.17737e-12
2 1750 9.4585e-12
4 1454 2.866e-11
6 61 5.76646e-11

Table 6.2: Number of rejections and average time step data for the Adams Moulton
algorithm

The test results are shown in Table 6.1. Very little difference was found between
circuits where the cross coupled gates both had zero propagation delays and the
case where one gate had 0.5ns delay and the other zero delay.

One obvious fact emerges from the data given in Table 6.1; namely that the Adams
Moulton higher order integration routines appear to be faster than the default
trapezoidal algorithm. This is corroborated by the average time step and number
of rejection data points output by Qucs at the end of a simulation. Table 6.2 lists
this data for the Adams Moulton algorithm tabulated in Table 6.1.

Table 6.2 points to the increase in average time step and the dramatic reduction
in the number of simulation solution rejections as the probable reason for the
reduction in transient simulation time when using the higher order Adams Moulton
integration routines. However, other factors may influence the choice of integration
routine. Often speed is not the only criteria that is of importance when simulating
large complex circuits. Consider the following case (the circuit shown in Fig. 6.13
with order 6 Adams Moulton transient analysis integration); setting one of the
gate delays to 1ns, and the other to 0ns, in each of the RS latches in the edge-
triggered D flip-flop yields the signal waveforms illustrated in Fig. 6.18. Clearly
here the solution is incorrect pointing to probable numerical instability caused by
the choice of integration routine.

176
 CLEAR.Vt
1

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
CLOCK.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
COUNT.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q0.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q1.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q2.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
1
Q3.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

Figure 6.18: Digital TimeList waveforms for the circuit shown in Fig. 6.13

6.11 Mixed-mode example simulations

Mixed-mode simulation involves the simulation of circuits that contain electronic
devices and circuits from different physical domains; the most obvious being cir-
cuits with a mixture of analogue and digital components. Qucs has developed to
a point where it can handle this type of circuit given device models that can span
across the different physical domains. In the future such circuits are likely to incor-
porate components from other domains, including for example, digital signal pro-
cessing components (DSP) and possibly nano mechanical devices. Multi-domain
simulation adds additional complexity to the simulation process not normally found
in single domain simulation. Each domain usually represents signal data in a spe-
cific way attributed to a given domain; voltage and current for analogue quantities,
boolean ’1’ and ’0’ for digital signals and floating point numbers for DSP. Hence,
signals passing from one domain to another have to be converted from one format
to another. These conversion elements are often called node-bridges and form an
essential part of the mixed-mode simulation process. The three examples that are
introduced in this section of these notes have been chosen to illustrate a number
of the basic ideas concerned with mixed-mode simulation of circuits containing
analogue and digital components, and to show how Qucs deals with this type of
simulation. In the last section the importance of correct selection of integration
routine when simulating circuits in the time domain was stressed. Mixed-mode

177
circuits often include a wide diversity of components that exhibit widely differing
time constants. This makes the problem of numerical stability versus simulation
run time even more critical. With the explicit numerical integration routines, like
the trapezoidal routine, numerical instability results if the simulation time step be-
comes much larger than the smallest time constant in a circuit. Hence, to achieve
successful completion of a simulation the integration time step must be reduced
which in turn makes the overall simulation time increase significantly. The implicit
Gear algorithm5 does not suffer from this problem and is the natural choice for
circuits with components that have widely differing time constants.

• Example 1: Analogue waveform driven digital devices with output node-

bridge.
The circuit in Fig. 6.19 shows an analogue voltage source driving a digital
inverter with a node-bridge element processing the inverter output signal.
The input signal is a sinusoidal voltage of amplitude 1V peak. The inverter
output signal, V1 in Fig. 6.19, has an nonsymmetrical mark to space ra-
tio because the threshold point for the inverter is set at 0.5V; the halfway
point for the two logic levels. The node-bridge element is basically a voltage
controlled voltage source where the device gain and time delay can be pro-
grammed. In this first example the gain has been set to 5 and the time delay
to 0.5ns. Figure 6.20 illustrates the simulation TimeList waveforms for this
example mixed-mode circuit. The node-bridge shown in Fig. 6.19 is a very
basic device. Moreover, by adding additional features, parameters like fall
and rise time can set to specific values. The next example demonstrates the
use of an active node-bridge.

5
The Gear integration algorithm is a powerful method for solving stiff systems of differential
equations, see Donald A. Calahan, Computer Aided Network Design, Revised edition, 1972,
McGraw-Hill.

178
 Vin V1 V5D
1 VINP VOUTP
V1
D to A
U=1 V Node
Y1
Bridge
transient
simulation VINN VOUTN

SUB1
TR1 File=a_node_bridge.sch
Type=lin
Start=0
Stop=20us
IntegrationMethod=Gear
Order=6

Figure 6.19: Analogue waveform driven digital device with output node-bridge
Vin.Vt

0 2e-6 4e-6 6e-6 8e-6 1e-5 1.2e-5 1.4e-5 1.6e-5 1.8e-5 2e-5
time

1
V1.Vt

0 2e-6 4e-6 6e-6 8e-6 1e-5 1.2e-5 1.4e-5 1.6e-5 1.8e-5 2e-5
time
5
V5D.Vt

0
0 2e-6 4e-6 6e-6 8e-6 1e-5 1.2e-5 1.4e-5 1.6e-5 1.8e-5 2e-5
time

Figure 6.20: Digital TimeList waveforms for the circuit shown in Fig. 6.19

179
 • Example 2: Pulse driven digital inverter with an active node bridge.
Illustrated in Fig. 6.21 is a similar circuit to the previous example. In
Fig. 6.21 a pulse generator drives a digital inverter. The inverter output
signal is processed by an active node-bridge derived from a basic BJT switch-
ing amplifier. The output waveforms for this circuit are shown in Fig. 6.22.
Notice that the pulse rise and fall times are determined by the node-bridge
amplifier and that the resulting analogue signal amplitude is set to 5V.

transient
simulation
V2
TR1
U=5 V
Type=lin
Start=0 R2
Stop=30ns R=4.7 k Ohm
IntegrationMethod=Gear
VC
Order=6

VPIN V1 VB T1 C1
1 Type=npn C=0.1 pF
V3 Is=1e-16
U1=0 V R1 Nf=1
Y1 R=10k Ohm
U2=1 V Vaf=0
T1=5ns Bf=100
T2=20ns

Figure 6.21: Pulse driven digital inverter with active node-bridge

180
 1
VPIN.Vt

0
0 2e-9 4e-9 6e-9 8e-9 1e-8 1.2e-8 1.4e-8 1.6e-8 1.8e-8 2e-8 2.2e-8 2.4e-8 2.6e-8 2.8e-8 3e-8
time
1
V1.Vt

0 2e-9 4e-9 6e-9 8e-9 1e-8 1.2e-8 1.4e-8 1.6e-8 1.8e-8 2e-8 2.2e-8 2.4e-8 2.6e-8 2.8e-8 3e-8
time
1
VB.Vt

0 2e-9 4e-9 6e-9 8e-9 1e-8 1.2e-8 1.4e-8 1.6e-8 1.8e-8 2e-8 2.2e-8 2.4e-8 2.6e-8 2.8e-8 3e-8
time

4
VC.Vt

0
0 2e-9 4e-9 6e-9 8e-9 1e-8 1.2e-8 1.4e-8 1.6e-8 1.8e-8 2e-8 2.2e-8 2.4e-8 2.6e-8 2.8e-8 3e-8
time

Figure 6.22: Digital TimeList waveforms for the circuit shown in Fig. 6.21

• Example 3: A more complex mixed-mode simulation example.

The circuit shown in Fig. 6.23 brings together a number of the ideas outlined
in these tutorial notes. A 4-bit digital signal is generated from a simple asyn-
chronous binary counter operated from a digital clock signal. The counter
output is transformed to the analogue domain using a simple node-bridge,
of the type introduced in mixed-mode example 1. A 4-bit binary weighted
DAC converts the transformed node-bridge signals into the final analogue
output signal. The DAC operational amplifier is modelled as a gain block
with a single pole frequency response and DC voltage output limiting. The
output waveforms for this example are shown in Fig. 6.24 and the details of
the operational amplifier model in Fig. 6.25.

181
 CLOCK
SUB5
S1 1
Num=1
SUB9
T

File=tff.sch
R

File=Logic_one.sch
Q

B0
VINP VOUTP

D to A R1
Node
SUB6 Bridge R=10k Ohm A_VOUT
T

File=tff.sch
R

VINN VOUTN
Q

SUB10 R2
File=a_node_bridge.sch R=10k Ohm
B1 SUB14
File=spole_op_amp.sch
VINP VOUTP

D to A R10 V-
Node V1
SUB7 Bridge R=5k Ohm
U=18 V
T

File=tff.sch
R

VINN VOUTN
Q

SUB11
File=a_node_bridge.sch +
B2 V+
V2
VINP VOUTP U=18 V

SUB8
D to A R4
Node
R=2.5k
T

File=tff.sch Bridge transient

R

simulation
Q

VINN VOUTN
RESET
SUB12
TR1
File=a_node_bridge.sch
B3 Type=lin
S2 VINP VOUTP Start=0
Num=2 Stop=40 m
D to A R5
Node IntegrationMethod=Gear
Bridge R=1.25k Ohm Order=6
VINN VOUTN

SUB13
File=a_node_bridge.sch

Figure 6.23: A more complex analogue-digital mixed-mode simulation example

182
 RESET.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time
CLOCK.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time
1
B0.Vt

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time
1
B1.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time

1
B2.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time

1
B3.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time
A_VOUT.Vt

-20
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 0.024 0.026 0.028 0.03 0.032 0.034 0.036 0.038 0.04
time

Figure 6.24: Digital TimeList waveforms for the circuit shown in Fig. 6.23

183
 VDCP
Num=4

D1
Is=1e-15 A
N=1
Cj0=10 fF
M=0.5
Vj=0.7 V R5
R=5 Ohm
VOUT
Num=3

R1 C1
VINP R4 R3
R=200k Ohm C=3.2uF R6
Num=1 R=10k Ohm R=50 Ohm
R=5 Ohm

SRC2 SRC1
VINN
G=200k G=1
Num=2
T=0 T=0 D2
Is=1e-15 A
N=1 VDCN
Cj0=10 fF Num=5
M=0.5
Vj=0.7 V

Figure 6.25: Operational amplifier model with Rin = 200k Ω, pole frequency =
5Hz, DC differential gain = 200k and Rout = 50 Ω

184
6.12 End Note
The examples described in these notes were all simulated using the latest CVS
code version of Qucs. Since release of version 0.0.8, Qucs has matured enough
to allow it to be used for mixed-mode simulation and many of the known bugs
in Qucs 0.0.8 will be corrected with the release of Qucs 0.0.9 some time in the
future. Release 0.0.9 will represent another important step in the development of
a truly universal simulator. However, much more work needs to be done on the
development of models for use across the different physical domains. My thanks
to Michael Margraf and Stefan Jahn for all their hard work in correcting the bugs
which surfaced while the examples presented in this tutorial note where being
tested.

185
7 Modelling Operational Amplifiers
7.1 Introduction
Operation amplifiers (OP AMP) are a fundamental building block of linear elec-
tronics. They have been widely employed in linear circuit design since they were
first introduced over thirty years ago. The use of operational amplifier models for
circuit simulation using SPICE and other popular circuit simulators is widespread,
and many manufacturers provide models for their devices. In most cases, these
models do not attempt to simulate the internal circuitry at device level, but use
macromodelling to represent amplifier behaviour as observed at the terminals of a
device. The purpose of this tutorial note is to explain how macromodels can be
used to simulate a range of the operational amplifier properties and to show how
macromodel parameters can be obtained from manufacturers data sheets. This
tutorial concentrates on models that can be simulated using Qucs release 0.0.9.

7.2 The Qucs built-in operational amplifier model

Qucs includes a model for an ideal operational amplifier. It’s symbol can be found
in the nonlinear components list. This model represents an operational amplifier
as an ideal device with differential gain and output voltage limiting. The model is
intended for use as a simple gain block and should not be used in circuit simulations
where operational amplifier properties are crucial to overall circuit performance.
Fig. 7.1 shows a basic inverting amplifier with a gain of ten, based on the Qucs OP
AMP model. The simulated AC performance of this circuit is shown in Fig. 7.2.
From Fig. 7.2 it is observed that the circuit gain and phase shift are constant and
do not change as the frequency of the input signal is increased. This, of course,
is an ideal situation which practical operational amplifiers do not reproduce. Let
us compare the performance of the same circuit with the operational amplifier
represented by a device level circuit. Shown in Fig. 7.3 is a transistor circuit dia-
gram for the well known UA741 operational amplifier1 . The gain and phase results
for the circuit shown in Fig. 7.1, where the OP AMP is modelled by the UA741
1
The UA741 operational amplifier is one of the most studied devices. It is almost unique in that
a transistor level model has been constructed for the device. Details of the circuit operation
and modelling of this device can be found in (1) Paul R. Grey et. al., Analysis and Design of

186
transistor level model, are given in Fig. 7.4. The curves in this figure clearly illus-
trate the differences between the two simulation models. When simulating circuits
that include operational amplifiers the quality of the OP AMP model can often
be a limiting factor in the accuracy of the overall simulation results. Accurate OP
AMP models normally include a range of the following device characteristics: (1)
DC and AC differential gain, (2) input bias current, (3) input current and voltage
offsets, (4) input impedance, (5) common mode effects, (6) slew rate effects, (7)
output impedance, (8) power supply rejection effects, (9) noise, (10) output volt-
age limiting, (11) output current limiting and (12) signal overload recovery effects.
The exact mix of selected properties largely depends on the purpose for which
the model is being used; for example, if a model is only required for small signal
AC transfer function simulation then including the output voltage limiting section
of an OP AMP model is not necessary or indeed may be considered inappropri-
ate. In the following sections of this tutorial article macromodels for a number of
the OP AMP parameters listed above are developed and in each case the neces-
sary techniques are outlined showing how to derive macromodel parameters from
manufacturers data sheets.

Analog Integrated Circuits, Fourth Edition, 2001, John Wiley and Sons INC., ISBN 0-471-
32168-0, and (2) Andrei Vladimirescu, The SPICE book, 1994, John Wiley and Sons, ISBN
0-471-60926-9.

187
 Equation

R3 Eqn1
R=47 k d1=dB(Vout.v)
OP1 d2=phase(Vout.v)
G=1e6
Vin
Vout
V1
U=1 V R4
R=4.7 k

dc simulation ac simulation

DC1 AC1
Type=log
Start=1 Hz
Stop=100 MHz
Points=801

Figure 7.1: Qucs schematic for a basic OP AMP inverting amplifier:Qucs OP AMP
has G=1e6 and Umax=15V.

188
 0
Vout.v

-10

-20
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz
40
dB(Vout.v)

20

0
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz
400
phase(Vout.v) Degrees

200

0
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz

Figure 7.2: Gain and phase curves for a basic OP AMP inverting amplifier.

189
 T6 T7 T14 T15
P_VCC T24

T27

R10
R=50

R5
R=39k T18
T3 T4 VIN_N
R11
C1 R=25
VIN_P T21 VOUT
C=30pF

T26
T2 T5 R9
R=40k
T17 T23

T13
T8 T10 T19
T11 T12

T16
R1
R=3k R8
R6 R7 R=50k T20
R4 R3 R2 R=50k R=50
R=1k R=50k R=1k

P_VEE

Figure 7.3: Transistor level circuit for the UA741 operational amplifier.

190
 10
Vout.v

1 10 100 1e3 1e4 1e5 1e6 1e7

Frequency Hz

20
dB(Vout.v)

-20

1 10 100 1e3 1e4 1e5 1e6 1e7

Frequency Hz
200
phas(Vout.v)

100

0
1 10 100 1e3 1e4 1e5 1e6 1e7
Frequency Hz

Figure 7.4: Gain and phase curves for a times 10 inverting amplifier with the OP
AMP represented by a transistor level UA741 model.

R2
R=47k Ohm
Vin
Vout
V1 R1 C1
U=1 V R3 R=200k Ohm C=159.15nF
R=4.7k Ohm

SRC1 OP1
G=1 S G=1
T=0

Equation
dc simulation ac simulation
Eqn1
DC1 AC1 d2=phase(Vout.v)
Type=lin d1=dB(Vout.v)
Start=1 Hz
Stop=10 MHz
Points=1800

Figure 7.5: Modified Qucs OP AMP model to include single pole frequency
response.

191
 10
Vout.v

0
1 10 100 1e3 1e4 1e5 1e6 1e7
Frequency Hz

20
dB(Vout.v)

-20

1 10 100 1e3 1e4 1e5 1e6 1e7

Frequency Hz
200
phase(Vout.v) Degrees

150

100

1 10 100 1e3 1e4 1e5 1e6 1e7

Frequency Hz

Figure 7.6: Gain and phase curves for the circuit shown in Fig. 7.5.

192
7.3 Adding features to the Qucs OP AMP model
In the previous section it was shown that the Qucs OP AMP model had a frequency
response that is independent of frequency. By adding external components to the
Qucs OP AMP model the functionality of the model can be improved. The UA741
differential open loop gain has a pole at roughly 5Hz and a frequency response
that decreases at 20 dB per frequency decade from the first pole frequency up to
a second pole frequency at roughly 3 MHz. The circuit shown in Fig. 7.5 models
the differential frequency characteristics of a UA741 from DC to around 1 MHz.
Figure 7.6 illustrates the closed loop frequency response for the modified Qucs
OP AMP model.

7.4 Modular operational amplifier macromodels

Macromodelling is a term given to the process of modelling an electronic device
as a ”black box” where individual device characteristics are specified in terms of
the signals, and other properties, observed at the input and output terminals
of the black box. Such models operate at a functional level rather than at the
more detailed transistor circuit level, offering considerable gain in computational
efficiency.2 Macromodels are normally derived directly from manufacturers data
sheets. For the majority of operational amplifiers, transistor level models are not
normally provided by manufacturers. One notable exception being the UA741
operational amplifier shown in Fig. 7.3.
A block diagram of a modular3 general purpose OP AMP macromodel is illustrated
in Fig. 7.7. In this diagram the blocks represent specific amplifier characteristics
modelled by electrical networks composed of components found in all the popular
circuit simulators4 . Each block consists of one or more components which model a
single amplifier parameter or a group of related parameters such as the input offset
current and voltage. This ensures that changes to one particular parameter do not
indirectly change other parameters. Local nodes and scaling are also employed
in the macromodel blocks. Furthermore, because each block operates separately,
2
Computational efficiency is increased mainly due to the fact that operational amplifier macro-
models have, on average, about one sixth of the number of nodes and branches when compared
to a transistor level model. Furthermore, the number of non-linear p-n junctions included in
a macromodel is often less than ten which compares favorable with the forty to fifty needed
to model an amplifier at transistor level.
3
Brinson M. E. and Faulkner D. J., Modular SPICE macromodel for operational amplifiers,
IEE Proc.-Circuits Devices Syst., Vol. 141, No. 5, October 1994, pp. 417-420.
4
Models employing non-linear controlled sources, for example the SPICE B voltage and current
sources, are not allowed in Qucs release 0.0.9. Non-linear controlled sources are one of the
features on the Qucs to-do list.

193
scaled voltages do not propagate outside individual blocks. Each block can be
modelled with a Qucs subcircuit that has the required specification and buffering
from other blocks. Moreover, all subcircuits are self contained entities where the
internal circuit details are hidden from other blocks. Such an approach is similar to
structured high-level computer programming where the internal details of functions
are hidden from users. Since the device characteristics specified by each block are
separate from all other device characteristics only those amplifier characteristics
which are needed are included in a given macromodel. This approach leads to a
genuinely structured macromodel. The following sections present the detail and
derivation of the electrical networks forming the blocks drawn in Fig. 7.7. To
illustrate the operation of the modular OP AMP macromodel the values of the
block parameters are calculated for the UA741 OP AMP and used in a series of
example simulations. Towards the end of this tutorial note data are presented for
a number of other popular general purpose operational amplifiers.

7.5 A basic AC OP AMP macromodel.

A minimum set of blocks is required for the modular macromodel to function as
an amplifier: an input stage, a gain stage and an output stage. These form the
core modules of all macromodels.

7.5.1 The input stage.

The input stage includes amplifier offset voltage, bias and offset currents, and the
differential input impedance components. The circuit for the input stage is shown
in Fig. 7.8, where

1. R1 = R2 = Half of the amplifier differential input resistance (RD).

2. Cin = The amplifier differential input capacitance (CD).

3. Ib1 = Ib2 = The amplifier input bias current (IB ).

4. Ioff = Half the amplifier input offset current (IOFF ).

5. Voff1 = Voff2 = Half the input offset voltage ( VOFF ).

Typical values for the UA741 OP AMP are:

1. RD = 2 MΩ and R1 = R2 = 1MΩ

2. CD = Cin1 = 1.4 pF.

194
3. IB = Ib1 = Ib2 = 80 nA.

4. IOFF = 20 nA and Ioff1 = 10 nA.

5. VOFF = 0.7 mV and Voff1 = Voff2 = 0.35 mV.

195
 In+ In-

Common Input Stage

mode stage

Signal
adder

Slew rate
limiting stage

Voltage gain
stage 1

Overdrive limiting
Vcc stage
In+
Out Voltage gain
stage 2

In-
Vee

Output stage
RPD

Vee Vcc

Current limiting
Vcc Vee stage

Voltage limiting
stage

Figure 7.7: Block diagram of an operational amplifier macromodel.

196
The differential output signal (VD) is given by V D− P 1−V D− N 1 and the common
mode output signal (VCM ) by (V D− P 1 + V D− N 1)/2.

Voff1
U=0.35mV

Input
Ib1 R1
IN_N1 VD_N1 stage
I=80nA R=1M Ohm
In- Vd-
Cin1
Ioff1 VCM1
C=1.4 pF
I=10nA Vcm

Ib2 R2
Voff2 In+ Vd+
I=80nA R=1M Ohm
U=0.35mV

SUB1
IN_P1 VD_P1
File=input_stage.sch

Figure 7.8: Modular OP AMP input stage block.

197
7.5.2 Voltage gain stage 1.
The circuit for voltage gain stage 1 is shown in Fig. 7.9, where
1. RD1 = 100 MΩ = A dummy input resistor - added to ensure nodes IN− P 1
and IN− N 1 are connected by a DC path.
2. GMP1 = 1 S = Unity gain voltage controlled current generator.
3. RADO = The DC open loop differential gain ( AOL(DC) ) of the OP AMP.
4. CP1 = 1/(2*π*GBP), where GBP = the OP AMP gain bandwidth product.
Typical values for the UA741 OP AMP are:
1. RADO = 200kΩ. (AOL(DC) = 106 dB)
2. CP1 = 159.15 nF (The typical value for UA741 GBP is 1 MHz).

7.5.3 Derivation of voltage gain stage 1 transfer function

Most general purpose operational amplifiers have an open loop differential voltage
gain which has (1) a very high value at DC (2) a dominant pole (fp1 ) at a low
frequency - typically below 100 Hz, and (3) a gain response characteristic that
rolls-off at 20 dB per decade up to a unity gain frequency which is often in the
MHz region. This form of response has a constant gain bandwidth product (GBP )
over the frequency range from fp1 to GBP. A typical OP AMP differential open
loop response is shown in Fig. 7.10. The voltage gain transfer function for this
type of characteristic can be modelled with the electrical network given in Fig. 7.9,
where the the AC voltage transfer function is

GM P 1 ∗ (V (IN− P 1) − V (IN− N 1)) ∗ RADO

vout(P OLE− 1− OU T 1) = (7.1)
1 + j(ω ∗ RADO ∗ CP 1)

POLE1
IN+
RD1 RADC1 CP1
IN_P1 OUT
R=100M R=200k Ohm C=159.15 nF
POLE_1_OUT1 IN-

GMP1
IN_N1 SUB1
G=1 S
T=0 File=pole1.sch

Figure 7.9: Modular OP AMP first voltage gain stage.

198
 Aol

Aol(DC)

fp1 GBP f Hz

Figure 7.10: OP AMP open loop differential voltage gain as a function of frequency.

Where
1
fP 1 = (7.2)
2π ∗ RADO ∗ CP 1
Let RADC = Aol(DC) and GMP1 = 1 S. Then, because fp1*AOL(DC) = GBP,
1
CP 1 = (7.3)
2π ∗ GBP

199
 OUTSTG_OUT1
Output
stage
RD1 In+
IN_P1 ROS1
R=100M
R=75 Ohm Out

EOS1 In-
IN_N1
G=1
T=0
SUB1
File=out_stage.sch

Figure 7.11: Modular macromodel output stage.

7.5.4 Output stage.

The electrical network representing a basic output stage is given in Fig. 7.11, where

1. RD1 = 100 MΩ = A dummy input resistor - added to ensure nodes IN− P 1

and IN− N 1 are connected by a DC path.

2. EOS1 G = 1 = Unity gain voltage controlled voltage generator.

3. ROS1 = OP AMP output resistance.

A typical value for the UA741 OP AMP output resistance is ROS1 = 75Ω.

7.5.5 A subcircuit model for the basic AC OP AMP

macromodel
The model for the basic AC OP AMP macromodel is shown in Fig. 7.12. The
input stage common mode voltage (V cm) is not used in this macromodel and has
been left floating. To test the performance of the AC macromodel it’s operation
was compared to the transistor level UA741 model. Figure 7.13 shows a schematic
circuit for two inverting amplifiers, each with a gain of ten, driven from a common
AC source. One of the amplifiers uses the simple AC macromodel and the other
the transistor level UA741 model. Figure 7.14 illustrates the output gain and
phase curves for both amplifiers. In general the plotted curves are very similar.
However, at frequencies above the GBP frequency the basic AC macromodel does
not correctly model actual OP AMP performance. This is to be expected because
the simple AC macromodel does not include any high frequency modelling com-
ponents. Notice also that the DC output voltages for vout and vout3 are very
similar, see the DC tabular results given in Fig. 7.13.

200
 SUB2
File=input_stage.sch

In+ Vd+ POLE1

IN_P1 IN+ Output
stage In-
Vcm OUT In+
OUT1
OP AMP
IN- Out IP1O
In- Vd-
IN_N1 Input In-
SUB4
stage In+
File=pole1.sch
SUB3
File=out_stage.sch SUB5
File=op_amp_ac_IP1O.sch

Figure 7.12: Simple AC OP AMP macromodel.

201
 dc simulation ac simulation Equation

Eqn1
DC1 AC1
yp=phase(vout.v)
Type=lin
yp3=phase(vout3.v)
number vout.V vout3.V Start=1 Hz
ydb=dB(vout.v)
1 0.0068 0.0069 Stop=10 MHz
ydb3=dB(vout3.v)
Points=1801

R1
R=10k Ohm
vin

In-
V1 R2 vout
U=1 V R=1k Ohm OP AMP
IP1O

SUB5
In+

R3
R=10k Ohm

SUB6

- V2
vout3 U=15 V
R8 VEE
R=1k Ohm
UA741_tran
VCC

+ V3
U=15 V

Figure 7.13: Test circuit for an inverting amplifier. Output signals: (1) vout for
AC macromodel, (2) vout3 for UA741 transistor model.

202
 200
20

phase(vout.v) in degrees
dB(vout.v)

150
0

100
-20

1 10 100 1e3 1e4 1e5 1e6 1e7 1 10 100 1e3 1e4 1e5 1e6 1e7
Frequency Hz Frequency Hz

200
20
phase(vout3.v) in degrees
db(vout3.v)

0
100

-20

0
1 10 100 1e3 1e4 1e5 1e6 1e7 1 10 100 1e3 1e4 1e5 1e6 1e7
Frequency Hz Frequency Hz

Figure 7.14: Simulation test results for the circuit shown in Fig. 7.13.

203
7.6 A more accurate OP AMP AC macromodel
Most general purpose OP AMPs have a high frequency pole in their differential
open loop gain characteristics. By adding a second gain stage to the simple AC
macromodel the discrepancy in the high frequency response can be corrected. The
model for the second gain stage is shown in Fig. 7.15. This additional gain stage
has a structure similar to the first gain stage, where
1. RD2 = 100 MΩ = A dummy input resistor - added to ensure nodes IN_P2
and IN_N2 are connected by a DC path.
2. GMP2 = 1 S = Unity gain voltage controlled current generator.
3. RP2 = 1Ω.
4. CP2 = 1/(2π*fp2), where f p2 = the second pole frequency in Hz.
A typical value for the UA741 OP AMP high frequency pole is fp2 = 3M Hz

7.6.1 Derivation of voltage gain stage 2 transfer function.

The differential voltage gain transfer function for voltage gain stage 2 is given by

GM P 2 ∗ (V (IN− P 2) − V (IN− N 2)) ∗ RP 2

vout(P OLE− 2− OU T 1) = (7.4)
1 + j(ω ∗ RP 2 ∗ CP 2)
Let RP2 = 1Ω and GMP2 = 1 S. Then

V (IN− P 2) − V (IN− N 2)
vout(P OLE− 2− OU T 1) = (7.5)
1 + j(ω ∗ CP 2)
and
1
CP 2 = (7.6)
2π ∗ f p2

POLE2
IN+
RD2 RP2 CP2
IN_P2 OUT
R=100M R=1 Ohm C=53.05nF
POLE_2_OUT1 IN-

GMP2
IN_N2 SUB1
G=1 S
T=0 File=pole2.sch

Figure 7.15: Modular OP AMP second voltage gain stage.

204
 ac simulation

AC1
R1 Type=log
R=10M Ohm Start=1 Hz
Stop=100MHz
In- Points=241
C1
vout
C=100mF Equation
OP AMP
IP1P2O Eqn1
vin
In+ y4=rad2deg(unwrap(angle(vout3.v)))
y=dB(vout.v)
V1 y3=dB(vout3.v)
U=1 V SUB1
File=op_amp_ac_IP1P2O.sch y2=phase(vout.v)

dc simulation
R2
R=10M Ohm DC1
SUB2
V2
U=15 V

C2
-
vout3
C=100 mF VEE
UA741_tran
VCC
V3
+ U=15 V

Figure 7.16: Test circuit for simulating OP AMP open loop differential gain.

7.6.2 Simulating OP AMP open loop differential gain

The circuit shown in Fig. 7.16 allows the open loop differential gain (Aol(f )) to
be simulated. This circuit employes a feedback resistor to ensure DC stability.
Fig. 7.16 illustrates two test circuits driven from a common AC source. This
allows the performance of the AC macromodel and the UA741 transistor level
model to be compared. The AC voltage transfer function for the test circuit is

Aol(f )
vout(f ) = vin(f ) (7.7)
Aol(f )
1+
1 + jω ∗ R ∗ C
vout(f )
where vout(f ) = (V + − V − ) ∗ Aol(f ), V + = vin(f ), and V − =
1 + jω ∗ R ∗ C
Aol(f )
Provided << 1, equation (7) becomes vout(f ) ⇒ Aol(f ) ∗ vin(f ).
ω∗R∗C
Hence, for those frequencies where this condition applies vout(f ) = Aol(f ) when
vin(f ) = 1 V. Figure 17 shows plots of the open loop simulation data. Clearly with
the test circuit time constant set at 1e6 seconds the data is accurate for frequencies
down to 1 Hz.

205
 1e5
1e5
1e4 1e4
1e3
1e3
100

vout3.v
vout.v

10 100
1
10
0.1
0.01 1
1e-3 0.1
1e-4 0.03
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz Frequency Hz

100 100
dB(vout3.v)
dB(vout.v)

-100
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz Frequency Hz

0
phase(vout3.v) degrees
phase(vout.v) degrees

-100
-200

-200
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz Frequency Hz

Figure 7.17: Simulation test results for the circuit shown in Fig. 7.16.

206
7.7 Adding common mode effects to the OP AMP
AC macromodel
The open-loop differential gain AD (f ) for most general purpose operational am-
plifiers can be approximated by
1
AD (f ) = AD(0) (7.8)
f
1+j
fP D
Similarly, the common-mode gain ACM (f ) can be represented by the same single-
pole response and a single zero response given by
f
1+j
fCM Z
ACM (f ) = ACM (0) (7.9)
f
1+j
fP D
Defining the common-mode rejection ratio CM RR(f ) of an OP AMP as
AD (f )
CM RR(f ) = (7.10)
ACM (f )
gives
1
CM RR(f ) = CM RR(0) (7.11)
f
1+j
fCM Z
where
AD (0)
CM RR(0) = (7.12)
ACM (0)
Common-mode effects can be added to OP AMP macromodels by including a
stage in the modular macromodel that introduces a zero in the amplifier frequency
response. Output VCM from the macromodel input stage senses an amplifier com-
mon mode signal. This signal, when passed through a CR network generates the
required common mode zero. Figure 18 gives the model of the zero generating
network, where.

1. RDCMZ = 650 MΩ = common-mode input resistance/2.

2. RCM1 = 1 MΩ
RCM 1
3. ECM1 G = 31.623 = RCM 2 . (NOTE: RCM1/RCM2 is a scaling factor.)
CM RR(0)

207
 RCM1 CMV_OUT1
R=1M
RDCMZ RCM2
IN_P1 CMZERO
R=650M R=1
IN+

CCM1 OUT
ECM1
IN_N1 C=795.8 pF
G=31.623
T=0 IN-

SUB1
File=cmzero.sch

Figure 7.18: Common-mode zero macromodel

1
4. CCM1 = 795.8 pF = .
2π ∗ RCM 1 ∗ fCM Z
5. RCM2 = 1 Ω
Typical values for the UA741 OP AMP are:
1. Common-mode input resistance = 1300 MΩ.
2. CMRR(0) = 90 dB
3. fCM Z = 200 Hz.
The AC voltage transfer function for the common-mode zero transfer function is

 
RCM 2 1 + jω ∗ RCM 1 ∗ CCM 1
V out(CMV_OUT1) = G(ECM 1) [V (IN_P1) − V (IN_N1)]
RCM 1 1 + jω ∗ RCM 2 ∗ CCM 1
(7.13)
RCM 2
As << 1, the pole introduced by the common-mode RC network is at a
RCM 1
very high frequency and can be neglected. Combining the common-mode zero with
the previously defined stage models yields the macromodel shown in Fig. 7.19. In
this model the differential and common-mode signals are combined using a simple
analogue adder based on voltage conrolled current generators.

7.7.1 Simulating OP AMP common-mode effects

OP AMP common-mode effects can be simulated using the circuit shown in Fig. 7.20.5
The resulting output voltages (vout.v and vout3.v) for a test circuit with matched
5
Brinson M.E. and Faulkner D.J., New approaches to measurement of operational amplifier
common-mode rejection ratio in the frequency domain, IEE Proc-Circuits Devices Sys., Vol
142, NO. 4, August 1995, pp 247-253.

208
 SUB6

VSUM
In+ Vd+ IN1+

IN_N1
IN1- POLE1
Vcm OUT IN+ POLE2
IN2+
OUT IN+ Output
In- Vd- stage
Input IN2- IN- OUT In+
IN_P1 OUT1
stage
SUB5 IN- Out
SUB4
CMZERO
IN+ SUB3 In-

OUT
In- SUB2
IN- OP_AMP
ICMZP1P2O
SUB1
File=cmzero.sch In+

SUB7
File=op_amp_ac_ICMZP1P2O.sch

WHERE IN1_P1
RSUM1
R=1
SUM_OUT1
VSUM
IN1+
GMSUM1
IN1_N1
IN1- G=1 S
T=0
OUT

IN2+

IN2- IN2_P1

SUB8
File=VSUM.sch GMSUM2
IN2_N1
G=1 S
T=0

Figure 7.19: AC macromodel including common-mode zero.

209
 R1
R=10k
dc simulation
vin In-
DC1
R2 vout
OP_AMP
V1 R=10k ICMZP1P2O
ac simulation U=1 V
In+
AC1 R3
Type=log R=10k SUB1
Start=1 Hz
Stop=10 kHz R4File=op_amp_ac_ICMZP1P2O.sch
Points=401 R=10k

R6
SUB2 R=10k
File=ua742_tran.sch

- V2
R5 vout3 U=15 V
VEE
R=10k UA741_tran
VCC
V3
+ U=15 V

R7
R=10k R8
R=10k

Figure 7.20: Simulation of OP AMP common-mode performance.

vout(0) 1
resistors are shown plotted in Fig. 7.21, where = . Clearly the
vin CM RR(0)
test results for the macromodel and the UA741 transistor model are very similar.
In the case of the macromodel typical device parameters were used to calculate the
macromodel component values. However, in the transistor level model the exact
values of the component parameters are unknown.6

6
The UA741 transistor level model is based on an estimate of the process parameters that
determine the UA741 transistor characteristics. Hence, the device level model is unlikely to
be absolutely identical to the model derived from typical parameters values found on OP
AMP data sheets. From the simulation results the CMRR(0) values are approximately (1)
macromodel 90 dB, (2) UA741 transistor model 101 dB. Similarly, the common-mode zero
frequencies are approximately (1) macromodel 200 Hz, (2) UA741 transistor model 500 Hz.

210
 1e-3
vout.v

1e-4

3e-5
1 10 100 1e3 1e4
Frequency Hz
1e-4
vout3.v

1e-5

3e-6
1 10 100 1e3 1e4
Frequency Hz

Figure 7.21: Simulation test results for the circuit shown in Fig. 7.20.

211
7.8 Large signal transient domain OP AMP
macromodels
The modular macromodel introduced in the previous sections concentrated on
modelling OP AMP performance in the small signal AC domain. Large signal
models need to take into account the passage of signals through an OP AMP in
the time domain and limit the excursion of voltage and current swings to the prac-
tical values found in actual amplifiers. Starting with the AC domain macromodel
introduced in the previous sections, adding a slew rate limiting stage and a over-
drive stage will more correctly model OP AMP high speed large signal limitations.
Furthermore, by adding output voltage and current limiting stages the OP AMP
macromodel will correctly model large signal effects when signal levels approach
circuit power supply voltages or the OP AMP output current limits.

7.8.1 Slew rate macromodel derivation

The slew rate of an OP AMP can be modelled by limiting the current charging
CP 1 in the first voltage gain stage POLE1. From Fig. 7.9

V (P OLE− 1− OU T 1) dV (P OLE− 1− OU T 1)
GM P 1 (V (IN− P 1) − V (IN− N 1)) = +CP 1∗
RADO dt
(7.14)
Hence, provided RADO is large7

dV (P OLE− 1− OU T 1)
GM P 1 (V (IN− P 1) − V (IN− N 1)) ' CP 1 ∗ (7.15)
dt
1
But CP 1 =
2π ∗ GBP
Yielding

1 dV (P OLE− 1− OU T 1)
GM P 1 (V (IN− P 1) − V (IN− N 1)) ' ∗ (7.16)
2π ∗ GBP dt
dV (P OLE− 1− OU T 1)
Moreover, if is set equal to the OP-AMP slew rate then the
dt
current

7
This condition is normally true because RADO is set to the DC open loop differential gain in
macromodule POLE1.

212
 GMSRT1
D1
G=0.01 S
SLEWRT
T=0
IN+
RSCALE1 VSR1 RSRT1
IN_P1 SLEWRT_OUT1
R=100 Ohm U=7.26 V R=1 OUT

IN-
SRC1
IN_N1
G=1 S SUB1
T=0 File=slewrt.sch

Figure 7.22: OP AMP slew rate macromodel.

charging CP1 will be limited to the maximum allowed. In Fig. 7.9 GM P 1 is 1 S.

Therefore, voltage difference V (IN− P 1) − V (IN− N 1)

1 dV (P OLE− 1− OU T 1)
must be set to ∗ .
2π ∗ GBP dt
This is done by the network SLEWRT shown in Fig. 7.22, where

1. RSCALE1 = 100 Ω = Scaling resistance (Scale factor x 100).

2. SRC1 G = 1 S.

3. VSR1 = V1.

4. GMSRT1 G = 0.01 S. (Scale factor = 1/100).

5. RSRT1 = 1 Ω

And,

100 ∗ P ositive− slew− rate

1. V 1 = − 0.7V
2π ∗ GBP
100 ∗ N egative− slew− rate
2. V 2 = − 0.7V
2π ∗ GBP
3. The diode parameters are IS=1e-12 IBV=20mA BV=V1+V2, others default.

Typical values for the UA741 OP AMP are:

1. P ositive− slew− rate = N egative− slew− rate = 0.5V/µS.

2. V 1 = V 2 = 7.25V.

213
Scaling is used in the slew rate model to allow the use of higher voltages in
the clamping circuit. Increased voltages reduce errors due to the forward biased
junction voltage. Current limiting results by clamping the voltage across resistor
RSCALE1 with a diode. This diode acts as a zener diode and saves one nonlinear
junction when compared to conventional clamping circuits. The output section of
the SLEWRT circuit removes the internal scaling yielding an overall gain of unity
for the module.

The circuit in Fig. 7.23 demonstrates the effect of slew rate limiting on OP AMP
transient performance. Three identical OP AMP inverter circuits are driven from
a common input 10 kHz AC signal source. Voltage controlled voltage sources are
used to amplify the input signal to the second and third circuits. The three input
signals are (1) 5 V peak, (2) 10 V peak and (3) 15 V peak respectively. The input
and output waveforms for this circuit are illustrated in Fig. 7.24. The effect of
slew rate limiting on large signal transient performance is clearly demonstrated by
these curves. In the case of the 15 V peak input signal the output signal (vout3.Vt)
has a slope that is roughly 0.5 V per µS.

7.8.2 Modelling OP AMP overdrive and output voltage limiting

Large transient signals can overdrive an OP AMP causing it’s output voltage to
saturate. On removal of the overdrive signal an OP AMP takes a finite time to
recover8 and return to normal linear circuit behaviour. When saturated the output
voltage is clamped at a voltage close to the plus or minus power rail voltage. The
overdrive and voltage clamping properties of an OP AMP are related and macro-
models for both effects need to be added to an OP AMP model when simulating
OP AMP overdrive characteristics. However, in many circuit simulations the over-
drive macromodel can be left out without loss of functionality or accuracy.

The effect of overdrive signals can be modelled by a voltage clamping circuit which
takes account of OP AMP recovery time from voltage overdrive. This extra element
clamps the output of the POLE1 module at a level above the OP AMP DC supply
voltages. The overall effect of the overdrive circuit is to delay the restoration of
linear circuit behaviour when an overload signal is removed. In contrast to the
overdrive module the output voltage limiting module clamps the output voltage to
a voltage close to the power rail voltages, clipping any output voltage excursions
above the power rail voltage levels. Figure 7.25 illustrates the macromodels for
the overdrive and output voltage limiting models, where
8
Overload recovery time of an OP AMP is the time required for the output voltage to recover
to a rated output voltage from a saturated condition. Typical values are in the µ S region.

214
 transient
simulation
In- TR1
OP_AMP vout1
Type=lin
ICMZ Start=0
vin SLEWRT Stop=200us
P1P2O
In+
V1
U=5 V
SUB1
File=op_amp_ac_ICMZP1P2O.sch

In-
OP_AMP vout2
ICMZ
SLEWRT
P1P2O
In+

SUB2
File=op_amp_ac_ICMZP1P2O.sch
SRC1
G=2
T=0

In-
OP_AMP vout3
ICMZ
SLEWRT
P1P2O
In+

SUB3
File=op_amp_ac_ICMZP1P2O.sch
SRC2
G=3
T=0

Figure 7.23: OP AMP slew rate test circuit.

215
 5
vin.Vt

-5

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4
time

5
vout1.Vt

-5

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4
time

10
vout2.Vt

-10

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4
time
20

10
vout3.Vt

-10

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4
time

Figure 7.24: OP AMP slew rate simulation waveforms for the circuit shown in
Fig. 7.23.

216
 P_VCC1
P_VCC2
VOVDRV1
U=2.5 V VLIM1
U=2 V

DOVRV1
Is=8e-16 A VCC
DVL1
OVDRV SUB1 Is=8e-16 A
File=OVDRV.sch VCC
IN SUB2
IN VLIMIT File=vlimit.sch
DVL2
P_IN1 DOVDRV1 VEE P_IN2
Is=8e-16 A
Is=8e-16 A VEE

VOVDRV2 VLIM2
U=2.5 V U=2 V

P_VEE2
P_VEE1

Figure 7.25: OP AMP overdrive and output voltage limiting macromodels.

1. VOVDR1 = 2.5 V = (Positive slew rate)*(Amplifier recovery time).

2. VOVDR2 = 2.5 V = (Negative slew rate)*(Amplifier recovery time).

3. VLIM1 = 2.0 V = (+ supply voltage) - (Maximum positive output voltage)

+ 1 V.

4. VLIM2 = 2.0 V = (- supply voltage) - (Maximum negative output voltage)

+ 1 V.

5. The diode parameters are Is = 8e-16 A, others default.

Typical values for the UA741 OP AMP are:

1. Amplifier recovery time 5 µS.

2. + supply voltage = 15 V.

3. - supply voltage = -15 V.

4. Maximum positive output voltage = 14 V.

5. Maximum negative output voltage = -14 V.

The test circuit given in Fig. 7.26 illustrates the effects of signal overdrive and
output voltage clamping on a unity gain buffer circuit. The test input signal is a 1
kHz signal with the following drive voltages (1) vin1 = 10 V peak, (2) vin2 = 18 V
peak, and (3) vin3 = 22 V peak. The corresponding output waveforms are shown
in Fig. 7.27. These indicate that increasing overdrive signals results in longer OP
AMP recovery times before the amplifier returns to linear behaviour.

217
 V2
U=15 V
In-
VCC
OP_AMP vout1
ICMZ
vin1 SLEWRT OVDRV
P1P2O
V1 In+ VEE
U=10 V
SUB1 V3
U=15 V
transient
simulation

TR1
Type=lin
Start=0
Stop=1.20 ms

In-
VCC
vout2 dc simulation
OP_AMP
vin2 ICMZ
DC1
SLEWRT OVDRV
P1P2O
In+ VEE

SUB2

SRC1
G=1.8
T=0

In-
VCC
OP_AMP vout3
vin3 ICMZ
SLEWRT OVDRV
P1P2O
In+ VEE

SUB3

SRC2
G=2.2
T=0

Figure 7.26: OP AMP overdrive and output voltage limiting test circuit.

218
 10
vout1.Vt
vin1.Vt

-10

0 1e-4 2e-4 3e-4 4e-4 5e-4 6e-4 7e-4 8e-4 9e-4 1e-3 0.0011 0.0012
Time

20
vout2.Vt
vin2.Vt

-20

0 1e-4 2e-4 3e-4 4e-4 5e-4 6e-4 7e-4 8e-4 9e-4 1e-3 0.0011 0.0012
Time

20
vout3.Vt
vin3.Vt

-20

0 1e-4 2e-4 3e-4 4e-4 5e-4 6e-4 7e-4 8e-4 9e-4 1e-3 0.0011 0.0012
Time

Figure 7.27: OP AMP overdrive and output voltage limiting waveforms for the
circuit shown in Fig. 7.26.

219
 D2 D1
Is=1e-15 A Is=1e-15 A

P_OUT1 CLIMIT

IN OUT
P_IN1
RDCL1 HCL1
R=100M G=36 SUB1
T=0 File=CLIMIT.sch

ECL1
G=1
T=0

Figure 7.28: OP AMP output current limiter macromodel.

7.8.3 Modelling OP AMP output current limiting

Most general purpose OP AMPs have a network at the circuit output to protect
the device from high load currents generated by shorting the output terminal to
ground or some other situation where a high current flows through the OP AMP
output stage. The electrical network shown in Fig. 7.28 acts as a current limiter:
current flowing between pins P_IN1 and P_OUT1 is sensed by current controlled
voltage generator HCL1. The voltage output from generator HCL1 is in series
with voltage controlled generator ECL1. The connection of these generators is in
opposite polarity. Hence, when the load current reaches the maximum allowed by
the OP AMP either diode DCL1 or DCL2 turns on clamping the OP AMP output
voltage preventing the output current from increasing. The parameters for the
current limiter macromodel are given by

1. RDCL1 = 100 MΩ = Dummy resistor.

2. ECL1 G = 1.

3. HCL1 G = 36Ω = 0.9 V/(Maximum output current A).

4. The diode parameters are Is = 1e-15 A, others default.

A typical value for the UA741 OP AMP short circuit current is 34 mA at 25o C.

Figures 7.29 and 7.30 show a simple current limiter test circuit and the resulting
test waveforms. In this test circuit time controlled switches decrease the load
resistors at 1 mS intervals. When the load current reaches roughly 34 mA the
output voltage is clamped preventing further increases in load current.

220
 V2
U=15 V

In-
VCC
OP_AMP vout
ICMZ
vin SLEWRT OVDRV
P1P2O CLIMIT
In+ VEE

V1 SUB1
U=10 V
S1 S2 S3 S4 S5
V3
U=15 V

R1 R2 R3 R4 R5 R6
R=1k R=1k R=2k R=2k R=2k R=2k
dc simulation transient
simulation
DC1
TR1
Type=lin
Start=0
Stop=8 ms

Figure 7.29: OP AMP output current limiter test circuit.

10
vin.Vt

-10

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005 0.0055 0.006 0.0065 0.007 0.0075 0.008
Time

10
vout.Vt

-10

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005 0.0055 0.006 0.0065 0.007 0.0075 0.008
Time

Figure 7.30: Simulation waveforms for current limiter test circuit shown in
Fig. 7.29.

221
 Parameter UA741 OP27 OP42 OPA134 AD746 AD826
Offset voltage (V) 7e-4 30e-6 4e-4 5e-4 3e-4 5e-4
Bias current (A) 80e-9 15e-9 130e-12 5e-12 110e-12 3-3e-6
Offset current (A) 20e-9 12e-9 6e-12 2e-12 45e-12 25e-9
Differential input res. (ohm) 2e6 4e6 1e12 1e13 2e11 300e3
Differential input cap. (F) 1.4e-12 6e-12 2e-12 5.5e-12 1.5e-12
Avd(0) dB 106 125 120 120 109 75
fp1 (Hz) 5 6 20 5 0.25 10e3
fp2 (Hz) 3e6 17e6 20e6 10e6 35e6 100e6
CMRR(0) dB 90 125 96 100 85 100
fcm (Hz) 200 2e3 100e3 500 3e3 2e3
GBP (Hz) 1e6 8e6 10e6 8e6 13e6 35e6
Rout (ohm) 75 70 50 10 10 8
Slew rate (V per micro sec.) 0.5 2.8 50 20 75 300
Overdrive recovery time (S) 5e-6 700e-9 0.5e-6
DC supply current (A) 1.4e-3 2.5e-3 5.1e-3 4e-3 7e-3 6.6e-3
Short circuit output current(A) 34e-3 32e-3 30e-3 40e-3 25e-3 90e-3
Common-mode input res. (ohm) 1.3e8 2e9 1e13 2.5e11
Common-mode input cap. (F) 5e-12 5.5e-12

Table 7.1: Typical OP AMP parameters taken from device data sheets.

222
7.9 Obtaining OP AMP macromodel parameters
from published device data
The OP AMP modular macromodel has one very distinct advantage when com-
pared to other amplifier models namely that it is possible to derive the macromodel
parameters directly from a common set characteristics found on the majority of
manufacturer’s data sheets. The data given in Table. 8.1 shows a typical range
of values found on OP AMP data sheets. In cases where a particular parameter
is not given then a starting point is to use a value obtained from a data sheet
of an equivalent device. The macromodel element values are then calculated us-
ing the equations presented in the previous sections of this tutorial. As a rule of
thumb it is good practice to test each block in the modular macromodel prior to
constructing a complete OP AMP macromodel.

7.10 More complete design examples.

In this section two larger design examples are presented. These demonstrate the
characteristics of the various OP AMP macromodels introduced in the previous
text and attempt to give readers guidance as to the correct model to choose for a
particular simulation.

7.10.1 Example 1: State variable filter design and simulation

The circuit given in Fig. 7.31 is a state variable filter which simultaneously gener-
ates band-pass, high-pass and low-pass responses. The circuit consists of an OP
AMP adder and two integrator circuits and requires three OP AMPS, two capaci-
tors and a number of resistors. The selection of the type of OP AMP for successful
operation of this filter is critical because devices with high offset voltage will cause
the integrators to saturate and the circuit will not function correctly. For opera-
tion below 20 kHz the OP27 is a good choice of OP AMP because of it’s low offset
voltage in the µV region. In this simulation both the DC characteristics and small
signal AC transfer characteristics are needed to check the filter design, hence the
AC macromodel with the DC parameters embedded in the input stage should allow
accurate modelling of the filter performance.9 The insert in Fig. 7.31 list the DC
output voltages for each of the OP AMP stages indicating that the integrators are
not saturated. The design of the state variable filter uses the following equations:
9
The magnitude of the output signals from the filter should also be checked to ensure that these
signals do not exceed the power supply voltages.

223
 1. The superposition principle yields
 
R1 R1 R1 R4
vhp = − vin − vlp + 1 + vbp (7.17)
R6 R7 R7 k R6 R4 + R5
When R1 = R6 = R7

3R4
vhp = −vin − vlp + vbp (7.18)
R4 + R5
2. Also
1
vbp = − vhp (7.19)
j ff0
where
1 1
f0 = = (7.20)
2πR2 C1 2πR3 C2
3. Similarly
1 1
vlp = − vbp = − vhp (7.21)
j ff0 ( ff0 )2

4. Hence
vhp ( ff0 )2
= (7.22)
vin 1 − ( ff0 )2 + ( Qj )( ff0 )
Where
1 R5
Q = (1 + ) (7.23)
3 R4
5. Also
f
vbp j f0
= (7.24)
vin 1 − ( ff0 )2 + ( Qj )( ff0 )

6. Also
vlp −1
= (7.25)
vin 1 − ( f0 ) + ( Qj )( ff0 )
f 2

Assuming f0 = 1 kHz and the required bandwidth of the band pass filter is 10
Hz, on setting R1 = R6 = R7 = 47kΩ and C1 = C2 = 2.2nF , calculation yields
R2 = R3 = 72.33kΩ10 In this design Q = 1k/10 = 100. Hence setting R4 = 1kΩ
yields R5 = 294kΩ (1 % tolerance). The simulation waveforms for the band pass
output are given in Fig. 7.32 11 . When the circuit Q factor is reduced to lower
values the other filter outputs act as traditional high and low pass filters. The
simulation results for Q factor one are shown in Fig. 7.33.
10
The values of R2 and R3 need to be trimmed if the filter center frequency and bandwidth are

224
 R7 number V1.I vbp.V vhp.V vlp.V
R=47k
1 8.56 e-11 -0.00149 -0.00149 0.000514
vin

R6 R1 R2 C1 R3 C2
R=47k R=47k R=72.33k C=2.2n R=72.33k C=2.2n
V1
U=0.1
In-
In- In- OP27 vlp
OP27 vhp OP27 vbp ICMZ
ICMZ ICMZ P1P2O
P1P2O P1P2O
In+
In+ In+
SUB3
SUB1 SUB2

R4 R5
R=1k R=294 k

dc simulation ac simulation Equation

Eqn1
DC1 AC1 Av_BP=dB(vbp.v/vin.v)
Type=lin Av_phase=phase(vbp.v/vin.v)
Start=100Hz
Stop=1900 Hz
Points=500

Figure 7.31: Three OP AMP state variable filter.

7.10.2 Example 2: Sinusoidal signal generation with the Wien

bridge oscillator
The Wien bridge sinusoidal oscillator has become a classic due to it’s simplicity and
low distortion capabilities. It is an ideal vehicle for demonstrating the properties of
OP AMP macromodels and indeed the performance of circuit simulators. Shown
in Fig. 7.34 is the basic Wien bridge oscillator which consists of a single OP AMP
with negative and positive feedback circuits. The design equations for this circuit
are

1. Non-inverting amplifier.
vout R3
=1+ (7.26)
v+ R4
2. Feedback factor
vout 1
b= = (7.27)
v+ 3 + j( ff0 − f0
f
)
1 1
Where f0 = =
2πR1C1 2πR2C2
required to high accuracy.
11
Note that the input signal vin has been set at 0.1 V peak. The circuit has a Q factor of 100
which means that the band pass output voltage is 10 V peak. Input signals of amplitude
much greater than 0.1 V are likely to drive the output signal into saturation when the power
supply voltages are ±15V .

225
 40

20
Av_BP in dB

-20

200 400 600 800 1e3 1.2e3 1.4e3 1.6e3 1.8e3

Frequency Hz

100

50
AV_phase in degrees

-50

-100
200 400 600 800 1e3 1.2e3 1.4e3 1.6e3 1.8e3
Frequency Hz

Figure 7.32: Simulation waveforms for current state variable filter circuit shown in
Fig. 7.31.

226
 0.1
vhp.v

0.05

1 10 100 1e3 1e4 1e5 1e6

Frequency Hz

0.1
vlp.v

0.05

1 10 100 1e3 1e4 1e5 1e6

Frequency Hz

Figure 7.33: State variable low pass and high pass response for Q = 1, R5 = 2kΩ.

3. Loop gain
The oscillator loop gain bAv must equal one for stable oscillations. Hence,
R3
1+ R4
bAv = (7.28)
3 + j( ff0 − f0
f
)

Moreover, at f = f0 ,
1 + R3
R4
bAv = (7.29)
3
Setting R3/R4 slightly greater than two causes oscillations to start and in-
crease in amplitude during each oscillatory cycle. Furthermore, if R3/R4 is
less than two oscillations will never start or decrease to zero.

Figure 7.35 shows a set of Wien bridge oscillator waveforms. In this example
the OP AMP is modelled using the OP27 AC macromodel. This has been done
deliberately to demonstrate what happens with a poor choice of OP AMP model.
The oscillator frequency is 10 kHz with both feedback capacitors and resistors
having equal values. Notice that the oscillatory output voltage continues to grow
with increasing time until it’s value far exceeds the limit set by a practical OP
AMP power supply voltages. The lower of the two curves in Fig. 7.35 illustrates
the frequency spectrum of the oscillator output signal. The data for this curve

227
has been generated using the Time2Freq function. Adding slew rate and voltage
limiting to the OP27 macromodel will limit the oscillator output voltage excursions
to the OP AMP power supply values. The waveforms for this simulation are shown
in Fig. 7.36. When analysing transient response data using function Time2Freq
it is advisable to restrict the analysis to regions of the response where the output
waveform has reached a steady state otherwise the frequency spectrum will include
effects due to growing, or decreasing, transients. The voltage limiting network
clips the oscillator output voltage restricting its excursions to below the OP AMP
power supply voltages. The clipping is very visible in Fig. 7.36. Notice also that
the output waveform is distorted and is no longer a pure sinusoidal waveform of 10
kHz frequency. Odd harmonics are clearly visible and the fundamental frequency
has also decreased due to the signal saturation distortion. In a practical Wien
bridge oscillator the output waveform should be a pure sinusoid with zero or little
harmonic distortion. One way to achieve this is to change the amplitude of the
OP AMP gain with changing signal level: as the output signal increases so Av
is decreased or as the output signal level decreases Av is increased. At all times
the circuit parameters are changed to achieve the condition bAv = 1. The circuit
shown in Fig. 7.37 uses two diodes and a resistor to automatically change the OP
AMP closed loop gain with changing signal level. Fig. 7.38 shows the corresponding
waveforms for the Wien bridge circuit with automatic gain control. Changing the
value of resistor R5 causes the amplitude of the oscillator output voltage to stabilise
at a different value; decreasing R5 also decreases vout. The automatic gain control
version of the Wien bridge oscillator also reduces the amount of harmonic distortion
generated by the oscillator. This can be clearly observed in Fig. 7.38. Changing
the oscillator frequency can be accomplished by either changing the capacitor or
resistor values in the feedback network b. To demonstrate how this can be done
using Qucs, consider the circuit shown in Fig. 7.39. In this circuit time controlled
switches change the value of both capacitors as the simulation progresses. The
recorded output waveform for this circuit is shown in Fig. 7.40.

228
 C2 R2
C=1 nF R=15.8k C1 R1
C=1nF R=15.8k
SUB2
File=OP27_ICMZP1P2O.sch

In+
vout
OP27
dc simulation ICMZ
P1P2O
DC1 In-

transient
simulation
R4 R3
TR1 R=10k R=21k
Type=lin
Start=0
Stop=10 ms

Equation

Eqn1
y=1
fscan=Time2Freq(vout.Vt,time[700:1000])

Figure 7.34: Classic Wien bridge sinusoidal oscillator.

1e11
vout.Vt

-1e11
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time
Frequency Spectrum

1e9

0 5e3 1e4 1.5e4 2e4 2.5e4 3e4 3.5e4 4e4 4.5e4 5e4
Frequency Hz

Figure 7.35: Simulation waveforms for the circuit shown in Fig. 7.34: OP27 AC
macromodel.

229
 vout.Vt 10

-10

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

6
Freq-Specrum

0 3e3 6e3 9e3 1.2e4 1.5e4 1.8e4 2.1e4 2.4e4 2.7e4 3e4 3.3e4 3.6e4 3.9e4 4.2e4 4.5e4 4.8e4
Frequency

Figure 7.36: Simulation waveforms for the circuit shown in Fig. 7.34: OP27 AC +
slew rate + vlimit macromodel.

C2 R2
C=1 nF R=15.8k C1 R1
C=1nF R=15.8k

SUB1
V1
In+ VEE
U=15 V
OP27 vout
ICMZ SLWRT
dc simulation P1P2 VLIM V2
O U=15 V
In- VCC
DC1

transient
simulation
R4 R3
TR1 R=10k R=21k
Type=lin
Start=0
Stop=10 ms
IntegrationMethod=Trapezoidal
R5 D1
R=50k
Equation

Eqn1 D2
fscan=Time2Freq(vout.Vt,time[700:1000])

Figure 7.37: Wien bridge oscillator with automatic gain control.

230
 1
vout.Vt

-1

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

0.4
Freq-Spectrum

0.2

0 3e3 6e3 9e3 1.2e4 1.5e4 1.8e4 2.1e4 2.4e4 2.7e4 3e4 3.3e4 3.6e4 3.9e4 4.2e4 4.5e4 4.8e4
Frequency Hz

Figure 7.38: Simulation waveforms for the circuit shown in Fig. 7.37: OP27 AC +
slew rate + vlimit macromodel.

dc simulation S6
C8
C=0.0625nF
DC1

S5
C7
transient C=0.125nF
simulation
S4
TR1 C6
Type=lin C=0.25nF
Start=0 R1
Stop=10 ms R=15.8k
IntegrationMethod=Trapezoidal C1
C=0.5nF
SUB1
S3 S1 S2
time=8 ms time=7 ms time=6 ms V1
In+ VEE
U=15 V
OP27 vout
C5 C3 C4 C2 R2 ICMZ SLWRT
C=0.0625nF C=0.125nF C=0.25nF C=0.5 nF R=15.8k P1P2 VLIM V2
O U=15 V
In- VCC

R4 R3
R=10k R=21k

R5 D1
R=50k

D2

Figure 7.39: Wien bridge oscillator with switched capacitor frequency control.

231
 1
vout.Vt

-1

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

1
vout.Vt

-1

0.005 0.0051 0.0052 0.0053 0.0054 0.0055 0.0056 0.0057 0.0058 0.0059
time

1
vout.Vt

-1

0.006 0.0061 0.0062 0.0063 0.0064 0.0065 0.0066 0.0067 0.0068 0.0069
time

1
vout.Vt

-1

0.007 0.0071 0.0072 0.0073 0.0074 0.0075 0.0076 0.0077 0.0078 0.0079
time

1
vout.Vt

-1

0.008 0.0081 0.0082 0.0083 0.0084 0.0085 0.0086 0.0087 0.0088 0.0089
time

Figure 7.40: Simulation waveforms for the circuit shown in Fig. 7.39: OP27 AC +
slew rate + vlimit macromodel.

232
7.11 Update number one: March 2007
In this first update to the operational amplifier tutorial readers will be intro-
duced to Qucs macromodel model building using schematics and SPICE to Qucs
conversion techniques, secondly to procedures for constructing Qucs operational
amplifier libraries, and finally to two different approaches which allow existing OP
AMP models to be extended to include new amplifier performance parameters,
for example power supply rejection. This update is very much a report on the
OP AMP modelling work that has been done by the Qucs development team since
version 0.0.10 of the package was released in September 2006. Future Qucs releases
will offer many significant improvements in OP AMP modelling particularly via
SPICE to Qucs netlist conversion, subcircuit passing and equation embedding in
Qucs schematics and library development. Following the release of Qucs 0.0.11,
and a suitable period of time for new feature debugging, many of the ideas intro-
duced in this update will be developed to include OP AMP model building using
embedded equations in Qucs schematics.

7.11.1 Building a library component for the modular OP AMP

macromodel

One of the main strengths of the modular macromodel approach to device mod-
elling is the fact that the parameters implicit in each section of a macromodel are
essentially independent, allowing subcircuit blocks to be easily connected together
to form an overall device model. Taking this idea further one can construct a
complete schematic for an OP AMP model from the circuitry that represents indi-
vidual macromodel subcircuit blocks. The diagram shown in Fig. 7.41 illustrates
a typical circuit schematic for a modular OP AMP macromodel. In this schematic
the component values are for the UA741 OP AMP. By attaching a symbol to the
modular macromodel schematic the UA741 modular OP AMP model is ready for
general use and can be placed in an existing12 or a user defined library. Moreover,
by recalculating the component values further library elements can be constructed
and the development of a more extensive Qucs OP AMP library undertaken13 .

12
Qucs 0.0.10, and earlier releases, were distributed with an OP AMP library called OpAmps.
However, this only contained a component level model for the 741 OP AMP. Many of the
models discussed in this text have been added to the Qucs OpAmps library. These should
assist readers who wish to experiment with their own OP AMP circuits.
13
One of the important future tasks is the development of component libraries for use with Qucs
- this will take time but should be possible given enough effort by everyone interested in Qucs.

233
 Voff1
U=0.35m V

Ib1 R1 RSUM1
P_INN I=80nA R=1M R=1

Ioff1
I=10nA Cin1
C=1.4 pF SRC1
G=1 S
Ib2
I=80nA R2
Voff2 R=1M
U=0.35m V

P_INP
RCM1 SRC2
RDCMZ R=1M RCM2 G=1 S
R=650M R=1

CCM1
ECM1 C=795.8 pF
G=31.623

GMSRT1
G=0.01 S

RSCALE1 VSR1 RSRT1 RADO CP1

R=100 U=7.26 V R=1 R=200k C=159.15nF

SRC3 GMP1
G=1 S D1 G=1 S
Is=1e-12 A
Bv=14.5
Ibv=20 mA

RP2 CP2
R=1 C=53.05nF ROS1
R=75
P_VCC

GMP2 EOS1 VLIM1

G=1 S G=1 U=2 V

D2 D3 DVL1
Is=1e-15 A Is=1e-15 A Is=8e-16 A

HCL1 P_OUT
G=35 DVLM2
RDCCL1 Is=8e-16 A
R=100M

ECL
G=1 VLIM2
U=2 V

P_VEE

Figure 7.41: Modular OP AMP macromodel in schematic form - this model does
not include signal overloading.

234
7.11.2 Changing model parameters: use of the SPICEPP
preprocessor
Changing the component data in Fig. 7.41 allows users to generate modular macro-
models for different operational amplifiers. Although this is a perfectly viable
approach to model generation it is both tedious and error prone. A more straight-
forward way is to get the computer to do the tedious work involving component
value calculation from device data. With this approach users are only required
to enter the device data; as a simple list derived from manufacturers data sheets.
One way to do this is to write a SPICE preprocessor template14 and let a SPICE
preprocessor generate the model for a specific OP AMP. The PS2SP template file
for an OP27 OP AMP modular macromodel is given in Fig. 7.42. The resulting
SPICE file is shown in Fig. 7.43. After construction of the SPICE OP27 netlist the
Qucs OP27 model is generated via the schematic capture SPICE netlist facility.15

7.11.3 The Boyle operational amplifier SPICE model

The Boyle16 operational amplifier model was one of the earliest attempts at con-
structing an OP AMP macromodel that achieved significantly reduced simulation
times, when compared to those times obtained with discrete transistor level mod-
els17 , while maintaining acceptable functional properties and simulation accuracy.
The Boyle macromodel was designed to model differential gain versus frequency,
DC common-mode gain, device input and output characteristics, slew rate limit-
ing, output voltage swing and short-circuit limiting. The circuit schematic for the
Boyle macromodel of a bipolar OP AMP is illustrated in Fig. 7.44. This model
consists of three connected stages: the input stage, the intermediate voltage gain
stage and the output stage. Calculation of individual component values is complex,
relying on a set of equations derived from the physical properties of the semicon-
ductor devices and the structure of the electrical network. These equations are
derived in the Boyle paper and summerised in the following list. Starting with
IS1 =8.0e-16, the emitter base leakage current of transistor T1, and by assuming
14
The use of the SPICE preprocessors SPICEPP and SPICEPRM are described in Qucs tutorial
Qucs simulation of SPICE netlists. Since both SPICEPP and SPICEPRM were first written,
Friedrch Schmidt has developed a PSpice to SPICE3/XSPICE preprocessor which combines,
and extends, the features found in both SPICEPP and SPICEPRM. This preprocessor is
called PS2SP. The Perl script version of PS2SP is licensed under GPL and may be downloaded
from http://members.aon.at/fschmid7/.
15
See the tutorial Qucs simulation of SPICE netlist for instructions on how this can be done.
16
G.R. Boyle, B.M. Cohn, D. Pederson, and J.E. Solomon, Macromodelling of integrated circuit
operational amplifiers, IEEE Journal of Solid State Circuits, vol. SC-9, pp. 353-364, 1974.
17
See Fig. 7.3. Tests show that the Boyle macromodel reduces simulation times for common
amplifier, timer and filter circuits by a factor between six and ten.

235
R2 = 100k the model component values can be calculated using:
 
V OS ∼

V OS
1. IS2 = IS1 · exp V t = IS1 1 + ,where V t = 26e-3 V.
Vt

C2 SR+
2. IC1 = , where SR+ is the positive slew rate.
2
3. IC2 = IC1
IOS IOS
4. IB1 = IB − and IB2 = IB +
2 2
IC1 IC2
5. B1 = and B2 =
IB1 IB2
 
B1 + 1 B2 + 1
6. IEE = + IC1
B1 B2
1
7. RC1 =
2πGBP C2
8. RC2 = RC1
 
B1 + B2 1 IC1
9. RE1 = RC1 − , where gm1 = , and RE2 = RE1
2 + B1 + B2 gm1 Vt

C2  π 
10. CEE = · tan 4φ , where 4φ = 90o − Φm and Φm is the phase
2 180
margin.
1
11. GCM =
CM M RRC1
1
12. GA =
RC1
AvOLRC1
13. GB =
R2RO2
14. ISD1 = IX · exp (T M P 1)+1e-32, where IX = 2 · IC1 · R2 · GB − IS1 ,
−1
and T M P 1 =
IS1
RO1
Vt
Vt IX
15. RC = ln (T EM P 2), where T EM P 2 =
100 · IX ISD1

236
 
ISCP
16. V C = abs (V CC) − V OU TP + V t · ln
IS1
 
ISCN
17. V E = abs (V EE) + V OU TN + V T · ln
IS1
(V CC − V EE) (V CC − V EE)
18. RP =
PD
Rather than calculate the Boyle macromodel component values by hand using
a calculator it is better to use a PS2SP preprocessor template that does these
calculations and also generates the Boyle SPICE netlist. A template for this task
is given in Fig. 7.45. The parameters at the beginning of the listing are for the
UA741 OP AMP. In Fig. 7.45 the macromodel internal nodes are indicated by
numbers and external nodes by descriptive names. This makes it easier to attach
the macromodel interface nodes to a Qucs schematic symbol. The SPICE netlist
shown in Fig. 7.46 was generated by SP2SP.

7.11.4 Model accuracy

The modular and Boyle OP AMP macromodels are examples of typical device
models in common use with todays popular circuit simulators. A question which
often crops up is which model is best to use when simulating a particular circuit?
This is a complex question which requires careful consideration. One rule of thumb
worth following is always validate a SPICE/Qucs model before use. Users
can then check that a specific model does simulate the circuit parameters that con-
trol the function and accuracy of the circuit being designed18 . One way to check
the performance of a given model is to simulate a specific device parameter. The
simulation results can then be compared to manufacturers published figures and
the accuracy of a model easily determined. By way of an example consider the
simulation circuit shown in Fig. 7.47. In this circuit the capacitors and inductors
ensure that the devices under test are in ac open loop mode with stable dc con-
ditions. Figure 7.48 illustrates the observed simulation gain and phase results for
four different OP AMP models. Except at very high frequencies, which are outside
device normal operating range, good agreement is found between manufacturers
data and that recorded by the open loop voltage gain test for both the modular
and Boyle macromodels.
18
An interesting series of articles by Ron Mancini, on verification and use of SPICE models
in circuit design can be found in the following editions of EDN magazine:Validate SPICE
models before use, EDN March 31, 2005 p.22; Understanding SPICE models, EDN April 14,
p 32; Verify your ac SPICE model, EDN May 26, 2005; Beyond the SPICE model’s dc and
ac performance, EDN June 23 2005, and Compare SPICE-model performance, EDN August
18, 2005.

237
∗ s u b c i r c u i t p o r t s : in+ in− p out p v c c p v e e
. s u b c k t opamp ac in p in n p out p v c c p v e e
∗ OP27 OP AMP p a r a m e t e r s
. param v o f f = 30 . 0u i b = 15n i o f f = 12n
. param rd = 4meg cd = 1 . 4p cmrrdc = 1 . 778 e6
. param fcmz = 2000 . 0 a o l d c = 1 . 778 e6 gbp = 8meg
. param f p 2 = 17meg p s l e w r=2 . 8 e6 n s l e w r=2 . 8 e6
. param vccm=15 vpoutm=14 veem=−15
. param vnoutm=−14 idcoutm=32m r o=70 . 0
. param p1={ ( 1 0 0 ∗ p s l e w r ) / ( 2 ∗ 3 . 1412∗ gbp ) −0 . 7}
. param p2={ ( 1 0 0 ∗ n s l e w r ) / ( 2 ∗ 3 . 1412∗ gbp ) −0 . 7}
∗ input stage
v o f f 1 in n 6 { v o f f /2 }
v o f f 2 7 in p { v o f f /2 }
ib1 0 6 {ib}
ib2 7 0 {ib}
i o f f 1 7 6 { i o f f /2 }
r1 6 8 { rd /2 }
r2 7 8 { rd /2 }
c i n 1 6 7 { cd }
∗ common−mode z e r o s t a g e
ecm1 12 0 8 0 {1 e6 / cmrrdc }
rcm1 12 13 1meg
ccm1 12 13 { 1 / ( 2 ∗ 3 . 1412∗1 e6 ∗ fcmz ) }
rcm2 13 0 1
∗ d i f f e r e n t i a l and common−mode s i g n a l summing s t a g e
gmsum1 0 14 7 6 1
gmsum2 0 14 13 0 1
rsum1 14 0 1
∗ slew rate stage
g s r c 1 0 15 13 0 1
r s c a l e 1 15 0 100
d s l 15 16 { d s l e w r a t e }
. model d s l e w r a t e d ( i s=1 e −12 bv= { p1+p2 } )
v s r 1 16 0 {p1}
gmsrt1 0 17 15 0 0 . 01
r s r t 1 17 0 1
∗ voltage gain stage 1
gmp1 0 9 17 0 1
rado 9 0 { a o l d c }
cp1 9 0 { 1 / ( 2 ∗ 3 . 1412∗ gbp ) }
∗ voltage gain stage 2
gmp2 0 11 9 0 1
rp2 11 0 1
cp2 11 0 { 1 / ( 2 ∗ 3 . 1412∗ f p 2 ) }
∗ out put s t a g e
e o s 1 10 0 11 0 1
r o s 1 10 50 { r o }
∗ out put c u r r e n t l i m i t e r s t a g e
r d c l 1 50 0 100meg
d c l 1 21 50 d c l i m
d c l 2 50 21 d c l i m
. model d c l i m d ( i s=1 e −15 c j 0=0 . 0 )
v c l 1 50 p out 0v
h c l 1 0 22 v c l 1 {0 . 9/ idcoutm }
e c l 1 21 22 50 0 1
∗ voltage limiting stage
d v l 1 p out 30 d v l i m i t
. model d v l i m i t d ( i s=8 e −16)
d v l 2 40 p out d v l i m i t
v l i m 1 p v c c 30 { vcc−vccm+1}
v l i m 2 40 p v e e {−v e e +veem+1}
. ends
. end

238 Figure 7.42: PS2SP template for the OP27 modular macromodel.
∗ s u b c i r c u i t p o r t s : in+ in− p out p v c c p v e e
∗ i n f i l e=op27 . pp d a t e=Tue Feb 13 17 : 32 : 37 2007 Converted with p s 2 s p . p l V4 . 11
∗ o p t i o n s : −sp3=0 − l t s p i c e=0 −fromsub=0 −f r o m l i b=0 −c h e c k=0 ( t i n y l i n e s=1 )
∗ c o p y r i g h t 2007 by F r i e d r i c h Schmidt − terms of Gnu L i c e n c e
. s u b c k t opamp ac in p in n p out p v c c p v e e
v o f f 1 in n 6 1 . 5 e −05
v o f f 2 7 in p 1 . 5 e −05
i b 1 0 6 1 . 5 e −08
i b 2 7 0 1 . 5 e −08
i o f f 1 7 6 6 e −09
r 1 6 8 2000000
r 2 7 8 2000000
c i n 1 6 7 1 . 4 e −12
ecm1 12 0 8 0 0 . 56 2429 6962 8796 4
rcm1 12 13 1meg
ccm1 12 13 7 . 95874188208328 e −11
rcm2 13 0 1
gmsum1 0 14 7 6 1
gmsum2 0 14 13 0 1
rsum1 14 0 1
g s r c 1 0 15 13 0 1
r s c a l e 1 15 0 100
d s l 15 16 0
. model d s l e w r a t e d ( i s=1 e −12 bv= 9 . 7422386349166 )
v s r 1 16 0 4 . 8711193174583
gmsrt1 0 17 15 0 0 . 01
r s r t 1 17 0 1
gmp1 0 9 17 0 1
rado 9 0 1778000
cp1 9 0 1 . 98968547052082 e −08
gmp2 0 11 9 0 1
rp2 11 0 1
cp2 11 0 9 . 36322574362739 e −09
e o s 1 10 0 11 0 1
r o s 1 10 50 70
r d c l 1 50 0 100meg
d c l 1 21 50 d c l i m
d c l 2 50 21 d c l i m
. model d c l i m d ( i s=1 e −15 c j 0=0 . 0 )
v c l 1 50 p out 0v
h c l 1 0 22 v c l 1 28 . 125
e c l 1 21 22 50 0 1
d v l 1 p out 30 d v l i m i t
. model d v l i m i t d ( i s=8 e −16)
d v l 2 40 p out d v l i m i t
v l i m 1 p v c c 30 −14
v l i m 2 40 p v e e −14
. ends
. end

Figure 7.43: SPICE netlist for the OP27 modular macromodel.

239
 P_VCC

VC
RC1 RC2 RP

D3

C1

D1 D2
C2 RO1 P_OUT
GCM D4
T1 T2 RO2 GC

R2 RC
P_INN VE
RE1 RE2
GA GB

CEE IEE REE

P_VEE

P_INP

Figure 7.44: Boyle macromodel for a BJT OP AMP

7.11.5 The PSpice modified Boyle model

One of the most widely used OP AMP simulation models is a modified version
of the Boyle macromodel. This was originally developed for use with the PSpice
circuit simulator. Many semiconductor manufacturers provide models for their de-
vices based on the modified Boyle macromodel19 . A typical modified Boyle macro-
model SPICE netlist is shown in Fig. 7.49. The circuit structure and performance
are very similar, but significantly different to the original Boyle model. Some of
the common OP AMP parameters NOT modeled are (1) input offset voltage, (2)
temperature coefficient of input offset voltage, (3) input offset current, (4) equiva-
lent input voltage and noise currents, (5) common-mode input voltage range, and
(6) temperature effect on component stability. Items (2), (4), (5) and (6) are also
not modeled by the standard Boyle macromodel. Although the modified Boyle
macromodel is similar to the original Boyle model it is not possible to use this
model as it is defined with Qucs; due to the fact that SPICE 2G nonlinear con-
trolled sources, egnd and fb, are included in the SPICE netlist. Controlled source
egnd is employed to model the OP AMP reference voltage as the average of the
VCC and VEE power rail voltages rather than the ground voltage assumed in the
original Boyle macromodel20 . Current conrolled current source fb is used to model

19
See for example the OP AMP section of the Texas Instruments (TI) Web site and the TI
Operational Amplifier Circuits, Linear Circuits, Data Manual, 1990.
20
Taking the OP AMP reference voltage to be the average of VCC and VEE allows devices with
non-symmetrical power supply voltages to be simulated.

240
∗ Boyle macromodel t e m p l a t e f o r Qucs .
∗ D e s i g n p a r a m e t e r s ( For UA741 )
. param v t=26 e−3 $ Thermal v o l t a g e a t room temp .
. param c2=30 e −12 $ Compensation c a p a c i t a n c e
. param p o s i t i v e s l e w r a t e=0 . 625 e6 n e g a t i v e s l e w r a t e=0 . 50 e6 $ Slew r a t e s
. param i s 1=8 . 0 e −16 $ T1 l e a k a g e c u r r e n t
. param v o s=0 . 7 e−3 i b=80n i o s=20n $ I n p u t v o l t a g e and c u r r e n t p a r a m e t e r s
. param va=200 $ Nominal e a r l y v o l t a g e
. param gbp=1 . 0 e6 $ Gain bandwidth p r o d u c t
. param pm=70 $ E x c e s s phase a t u n i t y g a i n .
. param cmrr=31622 . 8 $ Common−mode r e j e c t i o n r a t i o ( 9 0 dB)
. param a v o l=200 k $ DC open l o o p d i f f e r e n t i a l g a i n
. param r o 2=489 . 2 $ DC outpu t r e s i s t a n c e
. param r o 1=76 . 8 $ High f r e q u e n c y AC out put r e s i s t a n c e
. param r 2=100 k
. param vout p=14 . 2 $ P o s i t i v e s a t u r a t i o n v o l t a g e − f o r VCC=15 v
. param vout n=−13 . 5 $ N e g a t i v e s a t u r a t i o n v o l t a g e − f o r VCC=−15v
. param v c c=15 $ P o s i t i v e power s u p p l y v o l t a g e
. param v e e=−15 $ N e g a t i v e power s u p p l y v o l t a g e
. param i s c p=25m $ S h o r t c i r c u i t out put c u r r e n t
. param i s c n=25m $ S h o r t c i r c u i t out put c u r r e n t
. param pd=59 . 4m $ T y p i c a l power d i s s i p a t i o n
∗ Design e q u a t i o n s
. param i s 2={ i s 1 ∗(1+ v o s / v t ) }
. param i c 1={0 . 5∗ c2 ∗ p o s i t i v e s l e w r a t e } i c 2={ i c 1 }
. param i b 1={ ib −0 . 5∗ i o s } i b 2={ i b +0 . 5∗ i o s }
. param b1={ i c 1 / i b 1 } b2={ i c 2 / i b 2 }
. param i e e={ ( ( b1 +1)/ b1+(b2 +1)/ b2 ) ∗ i c 1 }
. param gm1={ i c 1 / v t } r c 1={ 1 / ( 2 ∗ 3 . 1412∗ gbp ∗ c2 ) } r c 2=r c 1
. param r e 1={ ( ( b1+b2 )/(2+ b1+b2 ) ) ∗ ( rc1 −1/gm1 ) } r e 2=r e 1
. param r e e={va / i e e } c e e={ ( 2 ∗ i c 1 / n e g a t i v e s l e w r a t e )−c2 }
. param dphi={90−pm} c1={ ( c2 / 2 ) ∗ tan ( dphi ∗3 . 1 4 1 2 / 1 8 0 ) }
. param gcm={ 1 / ( cmrr ∗ r c 1 ) } ga={ 1/ r c 1 } gb={ ( a v o l ∗ r c 1 ) / ( r 2 ∗ r o 2 ) }
. param i x={ 2∗ i c 1 ∗ r 2 ∗gb−i s 1 } tmp1={−1 . 0 / ( r o 1 ∗ i s 1 / vt ) } i s d 1={ i x ∗ exp ( tmp1)+1e −32}
. param tmp2={ i x / i s d 1 } r c={ v t / ( 1 0 0 ∗ i x ) ∗ l n ( tmp2 ) }
. param gc={ 1/ r c }
. param vc={ abs ( v c c )−vout p+vt ∗ l n ( i s c p/ i s 1 ) } ve={ abs ( v e e )+ vout n+vt ∗ l n ( i s c n/ i s 1 ) }
. param rp={ ( vcc−v e e ) ∗ ( vcc−v e e ) / pd}
∗ Nodes : I n p u t n i n p n i n n n v c c n v e e Output n out
Q1 8 n i n n 10 qmod1
Q2 9 n i n p 11 qmod2
RC1 n v c c 8 { r c 1 }
RC2 n v c c 9 { r c 2 }
RE1 1 10 { r e 1 }
RE2 1 11 { r e 2 }
RE 1 0 { r e e }
CE 1 0 { c e e }
IEE 1 n v e e { i e e }
C1 8 9 { c1 }
RP n v c c n v e e { rp }
GCM 0 12 1 0 {gcm}
GA 12 0 8 9 { ga }
R2 12 0 { r 2 }
C2 12 13 30p
GB 13 0 12 0 {gb}
RO2 13 0 { r o 2 }
RO1 13 n out { r o 1 }
D1 13 14 dmod1
D2 14 13 dmod1
GC 0 14 n out 0 { gc }
RC 14 0 { r c }
D3 n out 15 DMOD3
D4 16 n out DMOD3
VC n v c c 15 { vc }
VE 16 n v e e { ve }
. model dmod1 d ( i s={ i s d 1 } r s=1 ) 241
. model dmod3 d ( i s=8 e −16 r s=1 )
. model qmod1 npn ( i s={ i s 1 } BF={b1} )
. model qmod2 npn ( i s={ i s 2 } BF={b2} )
. end

Figure 7.45: PS2SP template for the Boyle macromodel with UA741 parameters
listed.
∗ b o y l e macromodel t e m p l a t e f o r q u c s .
∗ i n f i l e=ua741 b o y l e . p s 2 s p d a t e=Tue Feb 6 20 : 58 : 12 2007 Converted with p s 2 s p . p l V4 . 11
∗ o p t i o n s : −sp3=0 − l t s p i c e=0 −fromsub=0 −f r o m l i b=0 −c h e c k=0 ( t i n y l i n e s=1 )
∗ c o p y r i g h t 2007 by F r i e d r i c h Schmidt − terms of Gnu L i c e n c e
q1 8 n i n n 10 qmod1
q2 9 n i n p 11 qmod2
r c 1 n v c c 8 5305 . 82792138885
r c 2 n v c c 9 5305 . 82792138885
r e 1 1 10 1820 . 05072213971
r e 2 1 11 1820 . 05072213971
r e 1 0 13192612 . 1372032
c e 1 0 7 . 5 e −12
i e e 1 n v e e 1 . 516 e −05
c1 8 9 5 . 4588124089082 e −12
rp n v c c n v e e 15151 . 5151515152
gcm 0 12 1 0 5 . 96000354174836 e −09
ga 12 0 8 9 0 . 000188472
r 2 12 0 100000
c2 12 13 30p
gb 13 0 12 0 21 . 6918557701915
r o 2 13 0 489 . 2
r o 1 13 n out 76 . 8
d1 13 14 dmod1
d2 14 13 dmod1
gc 0 14 n out 0 1621 . 78603105575
r c 14 0 0 . 0 0 0 6 1 6 6 0 4 1 5 1 7 5 0 5 3 9
d3 n out 15 dmod3
d4 16 n out dmod3
vc n v c c 15 1 . 60789905279489
ve 16 n v e e 2 . 30789905279488
. model dmod1 d ( i s=1 e −32 r s=1 )
. model dmod3 d ( i s=8 e −16 r s=1 )
. model qmod1 npn ( i s=8 e −16 b f=107 . 1 4 2 8 5 7 1 4 2 8 5 7 )
. model qmod2 npn ( i s=8 . 21538461538461 e −16 b f=83 . 3 3 3 3 3 3 3 3 3 3 3 3 3 )
. end

Figure 7.46: SPICE netlist for the Boyle UA741 macromodel.

242
 vout_mod

L1 RL
R1
L=1000H R=0.00001 R=2k

C1
- VCC
V2
U=15 V
C=1000F UA741
(MOD)

+ VEE
V3
V1 U=15 V
U=1 V
SUB1

vout_boyle Equation

Eqn1
gain_mod_27_dB=dB(vout_mod_27.v)
C2 L2 gain_mod_dB=dB(vout_mod.v)
R2
C=1000F L=1000H gain_boyle_dB=dB(vout_boyle.v)
R=0.00001
gain_boyle_27_dB=dB(vout_boyle_27.v)
phase_boyle_deg=rad2deg(unwrap(angle(vout_boyle.v)))
phase_boyle_27_deg=rad2deg(unwrap(angle(vout_boyle_27.v)))
- VCC
phase_mod_deg=rad2deg(unwrap(angle(vout_mod.v)))
phase_mod_27_deg=rad2deg(unwrap(angle(vout_mod_27.v)))

UA741(Boyle)

VEE
+
SUB2
vout_mod_27

RL1
L3 R3
L=1000H R=2k
R=0.00001

C3
- VCC
dc simulation
C=1000F OP27 DC1
vin (MOD)

VEE ac simulation
+
AC1
SUB5 Type=log
Start=1Hz
Stop=100MHz
Points=200

vout_boyle_27

C4 L4 RL2
R4
C=1000F L=1000H R=0.00001 R=2k

- VCC

OP27
(Boyle)
+ VEE

SUB6

Figure 7.47: Test circuit for simulating OP AMP model open loop voltage gain.

243
 100 0

phase_mod_deg
gain_mod_dB

-50

0
-100

-150
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency acfrequency

0
100

phase_boyle_deg
gain_boyle_dB

0 -100

-100 -200
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency acfrequency

0
100
phase_mod_27_deg
gain_mod_27_dB

-100

-200
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency acfrequency

0
100
phase_boyle_27_deg
gain_boyle_27_dB

-100

-200
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8 1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency acfrequency

Figure 7.48: Open loop voltage gain simulation waveforms for the modular and
Boyle UA741 and OP27 macromodels.

244
OP AMP output current limiting. The nonlinear polynomial21 form of controlled
sources were included in the 2G series of SPICE simulators to allow behavioural
models of summers, multipliers, buffers and other important functional compo-
nents to be easily constructed. Single and multidimensional polynomial forms of
controlled sources are defined by SPICE 2G. Taking (1) the voltage controlled volt-
age source and (2) the current controlled current sources as examples the syntax
is as follows:
Ename N(+) N(-) POLY(n) NC1(+) NC1(-) NC2(+) NC2(-)..... P0 P1 P2......,
where n indicates the order of the polynomial with coefficients P0 .....Pn, and
NCn(+), NCn(-) etc are the control node pairs.
This becomes:

• For POLY(1) 22 : Ename N(+) N(-) P0 P1 P2.........

• For POLY(2): Ename N(+) N(-) POLY(2) NC1(+) NC1(-) NC2(+) NC(-) P0 P1 P2......

• For POLY(3): Ename N(+) N(-) POLY(3) NC1(+) NC1(-) NC2(+) NC2(-) NC3(+) NC3(-
) P0 P1 P2....
and so on.

Similarly: Fname N(+) N(-) POLY(n) V1 V2 V3 ...... P0 P1 P2 ....., where V1, V2 .... are
independent voltage sources whose current controls the output. This becomes:

• For POLY(1): Fname N(+) N(-) V1 P0 P1 P2........

• For POLY(2): Fname N(+) N(-) POLY(2) V1 V2 P0 P1 P2.......

• For POLY(3): Fname N(+) N(-) POLY(3) V1 V2 V3 P0 P1 P2 P3......., and so on.

The meaning of the coefficients in the nonlinear controlled source definitions de-
pends on the dimension of the polynomial. The following examples indicate how
SPICE calculates current or voltage values.

• For POLY(1): The polynomial function f v is calculated using

f v = P 0 + (P 1 ∗ f a) + (P 2 ∗ f a2 ) + (P 3 ∗ f a3 ) + (P 4 ∗ f a4 ) + .........,
where f a is either a voltage or current independent variable.
21
The definition of these polynomial functions was changed in the SPICE 3 series simulators
to a more conventional algebraic form when specifying the B type source components. This
often gives compatibility problems when attempting to simulate SPICE 2 models with circuit
simulators developed from SPICE 3f4 or earlier simulators. Most popular SPICE based circuit
simulators now accept both types of nonlinear syntax.
22
If only one P coefficient is given in the single dimension polynomial case, then SPICE assumes
that this is P1 and that P0 equals zero. Similarly if the POLY keyword is not explicitly
stated in a controlled source definition then it is assumed by SPICE to be POLY(1).

245
 • For POLY(2): The polynomial function f v is calculated using
f v = P 0 + (P 1 ∗ f a) + (P 2 ∗ f b) + (P 3 ∗ f a2 ) + (P 4 ∗ f a ∗ f b) + (P 5 ∗
f b2 ) + (P 6 ∗ f a3 ) +(P 7 ∗ f a2 ∗ f b) + .........., where f a and f b are both either
voltage or current independent variables.

• For POLY(3): The polynomial function f v is calculated using

f v = P 0 + (P 1 ∗ f a) + (P 2 ∗ f b) + (P 3 ∗ f c) + (P 4 ∗ f a2 ) + (P 5 ∗ f a ∗ f b)
+(P 6 ∗ f a ∗ f c) + (P 7 ∗ f b2 ) + (P 8 ∗ f b ∗ f c) + (P 9 ∗ f c2 ) + (P 10 ∗ f a3 )
+(P 11 ∗ f a2 ∗ f b) + (P 12 ∗ f a2 ∗ f c) + (P 13 ∗ f a ∗ f b2 ) + (P 14 ∗ f a ∗ f b ∗ f c)
+(P 15 ∗ f a ∗ f c2 ) + (P 16 ∗ f b3 ) + (P 17 ∗ f b2 ∗ f c) + (P 18 ∗ f b ∗ f c2 )
+(P 19 ∗ f c3 )............., where f a, f b, and f c are all either voltage or cur-
rent independent variables.

From Fig. 7.49 the controlled generators egnd and fb are:

• egnd 99 0 poly(2) (3,0) (4,0) 0 .5 .5

Which is the same as egnd 99 0 poly(2) 3 0 4 0 0 0.5 0.5
By comparison with the SPICE polynomial equations for controlled sources,
V (3) V (4)
V (egnd) = + implying that the controlled voltage source V (egnd)
2 2
is the sum of two linear voltage sources.

• fb 7 99 poly(5) vb vc ve vlp vln 0 10.61E6 -10E6 10E6 10E6 -10E6

By comparison with the SPICE polynomial equations for controlled sources
I(fb) = 10.61e6*I(vb) - 10e6*I(vc) + 10e6*I(ve) + 10e6*I(vlp) -10e6*I(vlp)
implying that the controlled current I(f b) is the sum of five linear controlled
current sources.

SPICE sources engd and fb can therefore be replaced in the modified Boyle model
by the following SPICE code23 :

* egnd 99 0 poly(2) (3,0) (4,0) 0 .5 .5

* Forms voltage source with output
* V=0.5*V(4)+0.5*V(3)
egnd1 999 0 4 0 0.5
23
It is worth noting that the code for the polynomial form of controlled sources can only be
replaced by a series connection of linear controlled voltage sources or a parallel connection of
linear controlled current sources provided no higher order polynomial coefficients are present
in the original SPICE code. Some SPICE models use these higher order coefficients to generate
multiply functions. Such cases cannot be converted to code which will simulate using Qucs
0.0.10. Sometime in the future this restriction will be removed when nonlinear voltage and
current sources are added to Qucs.

246
egnd2 99 999 3 0 0.5
*
*fb 7 99 poly(5) vb vc ve vlp vln 0 10.61e6 -10e6 10e6 10e6 -10e6
*
* Forms current source with output
* I=10.61e6*i(vb)-10e6*i(vc)+10e6*i(ve)+10e6*i(vlp)-10e6*i(vln)
*
*Sum 5 current sources to give fb.
fb1 7 99 vb 10.61e6
fb2 7 99 vc -10e6
fb3 7 99 ve 10e6
fb4 7 99 vlp 10e6
fb5 7 99 vln -10e6

Modified Boyle macromodels are often generated using the PSpice Parts24 program.
Such models have similar structured SPICE netlists with different component val-
ues. However, changes in technology do result in changes in the input stage that
reflect the use of npn, pnp and JFET input transistors in real OP AMPs. Hence to
use manufacturers published modified Boyle models with Qucs all that is required
is the replacement of the SPICE polynomial controlled sources with linear sources
and the correct component values. Again this is best done using a SPICE prepro-
cessor template. The templates for OP AMPS with npn and PJF input transistors
are shown in Figures 7.50 and 7.51. The SPICE netlists shown in Figs. 7.52 and
7.53 were generated by the PS2SP preprocessor. For OP AMPS with pnp input
transistors simply change the BJT model reference from npn to pnp and use the
same template.

24
The Parts modelling program is an integral component in the PSpice circuit simulation software
originally developed by the MicroSim Corporation, 1993, The Design Centre:Parts (Irvine,
Calif.). It now forms part of Cadence Design Systems OrCad suite of CAD software.

247
∗ connections : non−i n v e r t i n g i n p u t
∗ | i n v e r t i n g input
∗ | | p o s i t i v e power s u p p l y
∗ | | | n e g a t i v e power s u p p l y
∗ | | | | out put
∗ | | | | |
. s u b c k t uA741 1 2 3 4 5
∗
c1 11 12 8 . 661E−12
c2 6 7 30 . 00E−12
dc 5 53 dx
de 54 5 dx
d l p 90 91 dx
d l n 92 90 dx
dp 4 3 dx
egnd 99 0 p o l y ( 2 ) ( 3 , 0 ) ( 4 , 0 ) 0 . 5 . 5
fb 7 99 p o l y ( 5 ) vb vc ve v l p v l n 0 10 . 61E6 −10E6 10E6 10E6 −10E6
ga 6 0 11 12 188 . 5E−6
gcm 0 6 10 99 5 . 961E−9
iee 10 4 dc 15 . 16E−6
hlim 90 0 v l i m 1K
q1 11 2 13 qx
q2 12 1 14 qx
r2 6 9 100 . 0E3
rc1 3 11 5 . 305E3
rc2 3 12 5 . 305E3
r e 1 13 10 1 . 836E3
r e 2 14 10 1 . 836E3
r e e 10 99 13 . 19E6
ro1 8 5 50
ro2 7 99 100
rp 3 4 18 . 16E3
vb 9 0 dc 0
vc 3 53 dc 1
ve 54 4 dc 1
v l i m 7 8 dc 0
v l p 91 0 dc 40
vln 0 92 dc 40
. model dx D( I s=800 . 0E−18 Rs=1 )
. model qx NPN( I s=800 . 0E−18 Bf=93 . 7 5 )
. ends

Figure 7.49: PSpice modified Boyle macromodel for the UA741 OP AMP.

248
∗ M o d i f i e d Boyle OP AMP model t e m p l a t e
∗ npn BJT i n p u t d e v i c e s .
∗
∗UA741C OP AMP p a r a m e t e r s , m a n u f a c t u r e r Texas I n s t r u m e n t s
. param c1=4 . 664 p c2=20 . 0p
. param ep1=0 . 5 ep2=0 . 5
. param f p 1=10 . 61 e6 f p 2=−10e6 f p 3=10 e6 f p 4=10 e6 f p 5=−10e6
. param vc=2 . 6 ve=2 . 6 v l p=25 v l n=25
. param ga=137 . 7 e−6 gcm=2 . 57 e−9
. param i e e=10 . 16 e−6 hlim=1k
. param r 2=100 k
. param r c 1=7 . 957 k r c 2=7 . 957 k
. param r e 1=2 . 74 k r e 2=2 . 74 k
. param r e e=19 . 69 e6 r o 1=150 r o 2=150
. param rp=18 . 11 k
∗
. s u b c k t ua741 TI P INP P INN P VCC P VEE P OUT
c1 11 12 { c1 }
c2 6 7 { c2 }
dc P OUT 53 dx
de 54 P OUT dx
d l p 90 91 dx
d l n 92 90 dx
∗ egnd 99 0 p o l y ( 2 ) ( 3 , 0 ) ( 4 , 0 ) 0 0 . 5 0 . 5
∗ Qucs m o d i f i c a t i o n , Mike Brinson , Feb 2007
egnd1 999 0 P VCC 0 { ep1 }
egnd2 99 999 P VEE 0 { ep2 }
∗ f b 7 99 p o l y ( 5 ) vb vc ve v l p v l n 0 10 . 61 e6 −10e6 10 e6 10 e6 −10e6
∗ Forms c u r r e n t s o u r c e with out put
∗ I=10 . 61 e6 ∗ i ( vb)−10 e6 ∗ i ( vc )+10 e6 ∗ i ( ve )+10 e6 ∗ i ( v l p )−10 e6 ∗ i ( v l n )
∗ Qucs m o d i f i c a t i o n , Mike Brinson , Feb 2007 .
∗Sum 5 c u r r e n t s o u r c e s t o g i v e f b .
f b 1 7 99 vb { f p 1 }
f b 2 7 99 vc { f p 2 }
f b 3 7 99 ve { f p 3 }
f b 4 7 99 v l p { f p 4 }
f b 5 7 99 v l n { f p 5 }
∗
ga 6 0 11 12 { ga }
gcm 0 6 10 99 {gcm}
i e e 10 P VEE { i e e }
hlim 90 0 v l i m { hlim }
q1 11 P INN 13 qx
q2 12 P INP 14 qx
r 2 6 9 100 k
r c 1 P VCC 11 { r c 1 }
r c 2 P VCC 12 { r c 2 }
r e 1 13 10 { r e 1 }
r e 2 14 10 { r e 2 }
r e e 10 99 { r e e }
r o 1 8 P OUT { r o 1 }
r o 2 7 99 { r o 2 }
rp P VCC P VEE { rp }
vb 9 0 dc 0
vc P VCC 53 dc { vc }
ve 54 P VEE dc { ve }
v l i m 7 8 dc 0
v l p 91 0 dc { v l p }
v l n 0 92 dc { v l p }
. model dx d ( i s=800 . 0 e −18)
. model qx npn ( i s=800 . 0 e −18 b f=62 . 5 )
. ends
. end

Figure 7.50: Modified Boyle PS2SP netlist for the UA741 OP AMP.
249
∗ M o d i f i e d Boyle OP AMP model t e m p l a t e
∗ JFET i n p u t d e v i c e s .
∗
∗TL081 OP AMP p a r a m e t e r s , m a n u f a c t u r e r Texas I n s t r u m e n t s
. param c1=3 . 498 p c2=15 . 0p
. param ep1=0 . 5 ep2=0 . 5
. param f p 1=4 . 715 e6 f p 2=−5e6 f p 3=5 e6 f p 4=5 e6 f p 5=−5e6
. param vc=2 . 2 ve=2 . 2 v l p=25 v l n=25
. param ga=282 . 8 e−6 gcm=8 . 942 e−9
. param i s s=195 . 0 e−6 hlim=1k
. param r 2=100 k
. param rd1=3 . 536 k rd2=3 . 536 k
. param r s s=1 . 026 e6 r o 1=150 r o 2=150
. param rp=2 . 14 k
∗
. s u b c k t ua741 TI P INP P INN P VCC P VEE P OUT
c1 11 12 { c1 }
c2 6 7 { c2 }
dc P OUT 53 dx
de 54 P OUT dx
d l p 90 91 dx
d l n 92 90 dx
∗ egnd 99 0 p o l y ( 2 ) ( 3 , 0 ) ( 4 , 0 ) 0 0 . 5 0 . 5
∗ Qucs m o d i f i c a t i o n , Mike Brinson , Feb 2007
egnd1 999 0 P VCC 0 { ep1 }
egnd2 99 999 P VEE 0 { ep2 }
∗ f b 7 99 p o l y ( 5 ) vb vc ve v l p v l n 0 10 . 61 e6 −10e6 10 e6 10 e6 −10e6
∗ Forms c u r r e n t s o u r c e with out put
∗ I=10 . 61 e6 ∗ i ( vb)−10 e6 ∗ i ( vc )+10 e6 ∗ i ( ve )+10 e6 ∗ i ( v l p )−10 e6 ∗ i ( v l n )
∗ Qucs m o d i f i c a t i o n , Mike Brinson , Feb 2007 .
∗Sum 5 c u r r e n t s o u r c e s t o g i v e f b .
f b 1 7 99 vb { f p 1 }
f b 2 7 99 vc { f p 2 }
f b 3 7 99 ve { f p 3 }
f b 4 7 99 v l p { f p 4 }
f b 5 7 99 v l n { f p 5 }
∗
ga 6 0 11 12 { ga }
gcm 0 6 10 99 {gcm}
i s s P VCC 10 { i s s }
hlim 90 0 v l i m { hlim }
j 1 11 P INN 10 j x
j 2 12 P INP 10 j x
r 2 6 9 100 k
rd1 P VEE 11 { rd1 }
rd2 P VEE 12 { rd2 }
r o 1 8 P OUT { r o 1 }
r o 2 7 99 { r o 2 }
rp P VCC P VEE { rp }
r s s 10 99 { r s s }
vb 9 0 dc 0
vc P VCC 53 dc { vc }
ve 54 P VEE dc { ve }
v l i m 7 8 dc 0
v l p 91 0 dc { v l p }
v l n 0 92 dc { v l p }
. model dx d ( i s=800 . 0 e −18)
. model j x p j f ( i s=15 . 0 e −12 b e t a=270 . 1 e−6 v t o=−1)
. ends
. end

Figure 7.51: Modified Boyle PS2SP netlist for the TL081 OP AMP.

250
∗ m o d i f i e d b o y l e op amp model t e m p l a t e
∗ i n f i l e=Mod b o y l e t e m p l a t e npn . pp d a t e=Thu Feb 8 23 : 54 : 59 2007 Converted with p s 2 s p . p l V4 . 11
∗ o p t i o n s : −sp3=1 − l t s p i c e=0 −fromsub=0 −f r o m l i b=0 −c h e c k=0 ( t i n y l i n e s=1 )
∗ c o p y r i g h t 2007 by F r i e d r i c h Schmidt − terms of Gnu L i c e n c e
. s u b c k t ua741 t i p i n p p i n n p v c c p v e e p out
c1 11 12 4 . 664 e −12
c2 6 7 2 e −11
dc p out 53 dx
de 54 p out dx
d l p 90 91 dx
d l n 92 90 dx
egnd1 999 0 p v c c 0 0 . 5
egnd2 99 999 p v e e 0 0 . 5
f b 1 7 99 vb 9 . 42507068803016 e −08
f b 2 7 99 vc −1e −07
f b 3 7 99 ve 1 e −07
f b 4 7 99 v l p 1 e −07
f b 5 7 99 v l n −1e −07
ga 6 0 11 12 0 . 0001377
gcm 0 6 10 99 2 . 57 e −09
i e e 10 p v e e 1 . 016 e −05
hlim 90 0 v l i m 0 . 001
q1 11 p i n n 13 qx
q2 12 p i n p 14 qx
r 2 6 9 100 k
r c 1 p v c c 11 7957
r c 2 p v c c 12 7957
r e 1 13 10 2740
r e 2 14 10 2740
r e e 10 99 19690000
r o 1 8 p out 150
r o 2 7 99 150
rp p v c c p v e e 18110
vb 9 0 0
vc p v c c 53 2 . 6
ve 54 p v e e 2 . 6
vlim 7 8 0
v l p 91 0 25
v l n 0 92 25
. model dx d ( i s=800 . 0 e −18)
. model qx npn ( i s=800 . 0 e −18 b f=62 . 5 )
. ends
. end

Figure 7.52: Modified Boyle SPICE netlist for the TI UA741 OP AMP.

251
∗ m o d i f i e d b o y l e op amp model t e m p l a t e
∗ i n f i l e=TL081 TI . pp d a t e=Sun Feb 11 16 : 04 : 22 2007 Converted with p s 2 s p . p l V4 . 11
∗ o p t i o n s : −sp3=0 − l t s p i c e=0 −fromsub=0 −f r o m l i b=0 −c h e c k=0 ( t i n y l i n e s=1 )
∗ c o p y r i g h t 2007 by F r i e d r i c h Schmidt − terms of Gnu L i c e n c e
. s u b c k t ua741 t i p i n p p i n n p v c c p v e e p out t i m e s
c1 11 12 3 . 498 e −12
c2 6 7 1 . 5 e −11
dc p out 53 dx
de 54 p out dx
d l p 90 91 dx
d l n 92 90 dx
egnd1 999 0 p v c c 0 0 . 5
egnd2 99 999 p v e e 0 0 . 5
f b 1 7 99 vb 4715000
f b 2 7 99 vc −5000000
f b 3 7 99 ve 5000000
f b 4 7 99 v l p 5000000
f b 5 7 99 v l n −5000000
ga 6 0 11 12 0 . 0002828
gcm 0 6 10 99 8 . 942 e −09
i s s p v c c 10 0 . 000195
hlim 90 0 v l i m 1000
j 1 11 p i n n 10 j x
j 2 12 p i n p 10 j x
r 2 6 9 100 k
rd1 p v e e 11 3536
rd2 p v e e 12 3536
r o 1 8 p out 150
r o 2 7 99 150
rp p v c c p v e e 2140
r s s 10 99 1026000
vb 9 0 dc 0
vc p v c c 53 dc 2 . 2
ve 54 p v e e dc 2 . 2
v l i m 7 8 dc 0
v l p 91 0 dc 25
v l n 0 92 dc 25
. model dx d ( i s=800 . 0 e −18)
. model j x p j f ( i s=15 . 0 e −12 b e t a=270 . 1 e−6 v t o=−1)
. ends
. end

Figure 7.53: Modified Boyle SPICE netlist for the TI TL081 OP AMP.

252
7.12 Constructing Qucs OPAMP libraries
Qucs release 0.0.10 includes a facility which allows users to build their own com-
ponent libraries. This facility can be used to construct any library which contains
device models formed using the standard schematic entry route provided the indi-
vidual components that make up a model do not contain components that require
file netlists. Qucs, for example, converts SPICE netlists to Qucs formated netlists
when a simulation is performed but does not retain the converted netlists. Hence,
to add OP AMP macromodels that are based on SPICE netlist to a Qucs library
a slightly modified procedure is required that involves users copying the converted
SPICE netlist into a Qucs library. One way for generating SPICE netlist based
OP AMP models is as follows25 :

1. Construct a Qucs OP AMP model using the procedure described on page 3

of the Qucs Simulation of SPICE Netlists tutorial.

2. Add this model to a user defined library using the Qucs Create Library
facility (short cut Ctrl+Shift+L).

3. Place a copy of the OP AMP model on a drawing sheet and undertake a

DC analysis. NOTE: drag and drop the model symbol of the device you are
simulating from your current work project and NOT from the newly created
Qucs library.

4. Copy the section of the Qucs netlist that has been converted from the model’s
SPICE netlist and paste this into the newly created library model. The
converted SPICE netlist can be displayed by pressing key F6. User generated
library files are held in directory user_lib.26

To demonstrate the procedure consider the following example based on the UA741
Boyle model:
Steps 1 and 2 result in the following entry in a user created library:

<Component ua741 ( b o y l e )>

<D e s c r i p t i o n >
UA741 Boyle macromodel
</ D e s c r i p t i o n >
<Model>
. Def : Lib OPAMP ua741 b o y l e net0 net1 net2 net3 net4
Sub : X1 n e t 0 n e t 1 n e t 2 n e t 3 n e t 4 gnd Type=”ua741 b o y l e c i r ”
. Def : End

25
The procedure presented here must be considered a work around and may change as Qucs
develops.
26
The location of the user created libraries will differ from system to system depending where
.qucs is installed.

253
 </Model>
<Symbol>
<. ID −20 74 SUB>
<L i n e −20 60 0 −125 #00007 f 2 1>
<L i n e −20 −65 100 65 #00007 f 2 1>
<L i n e −20 60 100 −60 #00007 f 2 1>
<L i n e −35 −35 15 0 #00007 f 2 1>
<L i n e −35 40 15 0 #00007 f 2 1>
<L i n e 80 0 15 0 #00007 f 2 1>
<. PortSym −35 −35 1 0>
<. PortSym −35 40 2 0>
<. PortSym 95 0 3 180>
<L i n e 60 50 0 −40 #00007 f 2 1>
<L i n e 60 −15 0 −40 #00007 f 2 1>
<Text −15 −55 30 #000000 0 ”−”>
<Text −15 30 20 #000000 0 ”+”>
<Text −15 −5 12 #000000 0 ”UA741 ( Boyle ) ”>
<. PortSym 60 −55 4 180>
<. PortSym 60 50 5 180>
<Text 65 −30 12 #000000 0 ”VCC”>
<Text 65 20 12 #000000 0 ”VEE”>
</Symbol>
</Component>

Note that the model requires a subcircuit of type ua741_boyle_cir which is not
included when the library is created by Qucs. After completing the cut and paste
operation described in steps 3 and 4 above the resulting library entry becomes the
Qucs netlist shown next.
<Component ua741 ( b o y l e )>
<D e s c r i p t i o n >
UA741 Boyle macromodel
</ D e s c r i p t i o n >
<Model>
. Def : Lib OPAMP ua741 b o y l e net0 net1 net2 net3 net4
Sub : X1 n e t 0 n e t 1 n e t 2 n e t 3 n e t 4 gnd Type=”ua741 b o y l e c i r ”
. Def : End
. Def : ua741 b o y l e c i r netN INN netN INP netN OUT netN VCC netN VEE r e f
Vdc :VE n e t 1 6 netN VEE U=”2 . 3079 ”
Vdc :VC netN VCC n e t 1 5 U=”1 . 6079 ”
Diode : D4 netN OUT n e t 1 6 I s=”8 e −16 ” Rs=”1 ” N=”1 ” M=”0 . 5 ” Cj0=”1 e −14 ” Vj=”0 . 7 ”
Diode : D3 n e t 1 5 netN OUT I s=”8 e −16 ” Rs=”1 ” N=”1 ” M=”0 . 5 ” Cj0=”1 e −14 ” Vj=”0 . 7 ”
R:RC n e t 1 4 r e f R=”0 . 000616604 ”
VCCS:GC netN OUT r e f net14 r e f G=”1621 . 79 ”
Diode : D2 n e t 1 3 n e t 1 4 I s=”1 e −32 ” Rs=”1 ” N=”1 ” M=”0 . 5 ” Cj0=”1 e −14 ” Vj=”0 . 7 ”
Diode : D1 n e t 1 4 n e t 1 3 I s=”1 e −32 ” Rs=”1 ” N=”1 ” M=”0 . 5 ” Cj0=”1 e −14 ” Vj=”0 . 7 ”
R:RO1 n e t 1 3 netN OUT R=”76 . 8 ”
R:RO2 n e t 1 3 r e f R=”489 . 2 ”
VCCS:GB n e t 1 2 n e t 1 3 ref r e f G=”21 . 6919 ”
C: C2 n e t 1 2 n e t 1 3 C=”30p ”
R: R2 n e t 1 2 r e f R=”100000 ”
VCCS:GA n e t 8 n e t 1 2 ref n e t 9 G=”0 . 000188472 ”
VCCS:GCM n e t 1 ref net12 r e f G=”5 . 96 e −09 ”
R:RP netN VCC netN VEE R=”15151 . 5 ”
C: C1 n e t 8 n e t 9 C=”5 . 45881 e −12 ”
I d c : IEE netN VEE n e t 1 I=”1 . 516 e −05 ”
C:CE n e t 1 r e f C=”0 ”
R:RE n e t 1 r e f R=”1 . 31926 e+07 ”
R: RE2 n e t 1 n e t 1 1 R=”1820 . 05 ”
R: RE1 n e t 1 n e t 1 0 R=”1820 . 05 ”
R: RC2 netN VCC n e t 9 R=”5305 . 83 ”

254
 R: RC1 netN VCC n e t 8 R=”5305 . 83 ”
BJT : Q2 netN INP n e t 9 n e t 1 1 r e f Type=”npn ” I s=”8 . 21538 e −16 ” Bf=”83 . 3333 ” Nf=”1 ” Nr=”1 ” I k f=”0 ”
I k r=”0 ” Vaf=”0 ” Var=”0 ” I s e=”0 ” Ne=”1 . 5 ” I s c=”0 ” Nc=”2 ” Br=”1 ” Rbm=”0 ” I r b=”0 ” Cje=”0 ” Vje=”0 . 75 ”
Mje=”0 . 33 ” Cjc=”0 ” Vjc=”0 . 75 ” Mjc=”0 . 33 ” Xcjc=”1 ” C js=”0 ” Vjs=”0 . 75 ” Mjs=”0 ” Fc=”0 . 5 ” Vtf=”0 ”
Tf=”0 ” Xtf=”0 ” I t f=”0 ” Tr=”0 ”
BJT : Q1 netN INN n e t 8 n e t 1 0 r e f Type=”npn ” I s=”8 e −16 ” Bf=”107 . 143 ” Nf=”1 ”
Nr=”1 ” I k f=”0 ” I k r=”0 ” Vaf=”0 ” Var=”0 ” I s e=”0 ” Ne=”1 . 5 ” I s c=”0 ” Nc=”2 ” Br=”1 ”
Rbm=”0 ” I r b=”0 ” Cje=”0 ” Vje=”0 . 75 ” Mje=”0 . 33 ” Cjc=”0 ” Vjc=”0 . 75 ” Mjc=”0 . 33 ”
Xcjc=”1 ” C js=”0 ” Vjs=”0 . 75 ” Mjs=”0 ” Fc=”0 . 5 ” Vtf=”0 ” Tf=”0 ” Xtf=”0 ” I t f=”0 ” Tr=”0 ”
. Def : End
</Model>
<Symbol>
<. ID −20 74 SUB>
<L i n e −20 60 0 −125 #00007 f 2 1>
<L i n e −20 −65 100 65 #00007 f 2 1>
<L i n e −20 60 100 −60 #00007 f 2 1>
<L i n e −35 −35 15 0 #00007 f 2 1>
<L i n e −35 40 15 0 #00007 f 2 1>
<L i n e 80 0 15 0 #00007 f 2 1>
<. PortSym −35 −35 1 0>
<. PortSym −35 40 2 0>
<. PortSym 95 0 3 180>
<L i n e 60 50 0 −40 #00007 f 2 1>
<L i n e 60 −15 0 −40 #00007 f 2 1>
<Text −15 −55 30 #000000 0 ”−”>
<Text −15 30 20 #000000 0 ”+”>
<Text −15 −5 12 #000000 0 ”UA741 ( Boyle ) ”>
<. PortSym 60 −55 4 180>
<. PortSym 60 50 5 180>
<Text 65 −30 12 #000000 0 ”VCC”>
<Text 65 20 12 #000000 0 ”VEE”>
</Symbol>
</Component>

7.13 Extending existing OP AMP models

The modular, Boyle and modified Boyle OP AMP models are three popular macro-
models selected from a large number of different models that are in common use
today. Most device manufacturers provide similar macromodels, or extended ver-
sions which more accurately model the performance of specific devices. Indeed, a
growing trend has developed which mixes Boyle type models with modular struc-
tures27 . Often, in practical design projects specific OP AMP properties must be
simulated which are not modelled with an available OP AMP model. Two ap-
proaches can be used to overcome such deficiencies; firstly, a macromodel itself
can be modified so that it models the required additional attributes, or secondly
external components can be added which again extend model performance.
One important OP AMP parameter that the standard and modified Boyle models
do not model is the frequency dependence of amplifier common-mode gain. Only
27
See for example: Ray Kendall, User-friendly model simplifies SPICE OP-AMP simulation,
EDN magazine, January 4, 2007, pp. 63-69.

255
the dc value of the CMRR is modelled. Such frequency dependency can be added
by a simple modification28 , requiring one extra node, that simulates ac CMRR
and gives close agreement between macromodel performance and data sheet spec-
ifications. Components CEE, REE and GCM, see Fig. 7.44, are replaced by the
network shown in Fig. 7.54. Data sheets for the UA741 show the CMRR falling
above a break frequency of about 200 Hz, due to the zero generated by CEE caus-
ing the common-mode gain to increase. This effect can be simulated in the Boyle
macromodel by the addition of one extra node and two extra resistors and changes
to REE and controlled source GCM as in Fig. 7.44. In this modified network, the
common-mode voltage is detected at the junction of RE4 and CEE, introducing a
zero into the response and attenuating the signal. The frequency of the zero is set
1
by 2∗π∗CEE∗RE3 . The new value of CEE must have the same value as the original
29
CEE value if the same slew-rate is to be maintained, so for a 200 Hz cut-off this
gives RE3=106.1M. RE4 is arbitrarily fixed at 10 Ω, which introduces another
pole at about 2 GHz, well outside the frequency of interest. The value of REE is
increased to RE5 (15.06meg), so that RE5 in parallel with RE3 equals the original
RE3
value of REE. GCM is also increased by the factor RE4 maintaining the correct
low frequency common-mode gain. Differential frequency response and slew rate
are unchanged by these modifications. The simulation results for the common-
mode test circuit shown in Fig. 7.20 are given in Fig. 7.55. These indicate close
agreement between the modular and ac Boyle macromodels.

Modifying the circuit of an existing OP AMP macromodel is at best a complex

process or at worst impossible because the model details are either not known
or well understood. One way to add features to an existing model is to add an
external circuit to a model’s terminals. This circuit acts as a signal processing
element adding additional capabilities to the original macromodel. One circuit
feature not modelled by any of the macromodels introduced in earlier sections
is power supply rejection. By adding a simple passive electrical network to the
terminals of a macromodel it is possible to model OP AMP power supply rejection.
Power supply rejection(PSRR) is a measure of the ability of an OP AMP to reject
unwanted signals that enter at the power terminals. It is defined as the ratio of
28
This section is based on unpublished work by David Faulkner and Mike Brinson., Department
of Computing, Communications Technology and Mathematical Sciences, London Metropoli-
tan University, UK.
29
In the Boyle macromodel the value of CEE is set by the OP AMP slew rate. Adjusting both
the positive and negative slew rates changes the value of CEE. For the UA741 these have been
set at 0.625e6 and 0.500e6 respectively. This gives CEE=7.5pF which is commonly quoted
for the UA741 value of CEE, see Andrei Vladimirescu, The SPICE Book,1994, John Wiley
and Sons, Inc., ISBN 0-471-60926-9, pp 228-239. Also note care must be taken when choosing
values for the two slew rates because negative values of CEE can occur which are physically
not realisable.

256
 CEE RE3 RE5
C=7.5p R=106.1M R=15.06M
GCM
G=6.32e-2

RE4
R=10

Figure 7.54: AC CMRR modification for Boyle macromodel

differential-mode gain to power supply injected signal gain. The simulation of OP

AMP power supply rejection30 is possible using the test circuit given in Fig. 7.56,
where
V +V−
 + 
± −
+
Vout (f ) = AD (f ) [V − V ] + ACM (f ) + AP S (f )± VS ,
2
where AD (f ) is the OP AMP differential-mode gain, ACM (f ) is the OP AMP
common-mode gain, and AP S (f )± is the OP AMP power supply injected gain.
The superscript ± indicates that the ac signal source is connected to either the OP
AMP positive or negative power supply terminals but not simultaneously to both.
Assuming that the OP AMP power supply injected gain has a single dominant
zero at fP±SZ1 , analysis yields
  
1 f
1+j
VOU T (f )± αPSRR (0)± fP±SZ1
=   
VS f
1+j
αGBP
  
f
1+j
± ±  f± R2
Where AP S(f ) = AP S(0) P SZ1 , α = = 1e − 4
f R1 + R2
1+j
fp1
30
M. E. Brinson and D. J. Faulkner, Measurement and modelling of operational amplifier power
supply rejection, Int. J. Electronics, 1995, vol. 78, NO. 4, 667-678.

257
 2e-4

1.5e-4
vout_boyle_orig.v

1e-4

5e-5

0
1 10 100 1e3
acfrequency

2e-4

1.5e-4
vout_boyle_ac.v

1e-4

5e-5

0
1 10 100 1e3
acfrequency

2e-4

1.5e-4
vout_mod.v

1e-4

5e-5

0
1 10 100 1e3
acfrequency

Figure 7.55: AC common-mode simulation results for (1) Boyle macromodel, (2)
ac Boyle macromodel and (3) the modular macromodel

258
 V3
U=1 V
R1
R=100k

- VCC vout
V1
U=15 V
R2
R=10 UA741(Boyle)
V2
VEE U=15 V
+
R3
R=10
SUB1
R4
R=100k
dc simulation ac simulation
DC1 AC1
Type=log
Start=1 Hz
Stop=10 MHz
Points=500

Figure 7.56: Test circuit for the simulation of PSRR(f) voltage transfer function
characteristic

P SRR(0)± AD (0)
P SRR(f )± =    , P SRR(0)+ = and P SRR(0)− =
f AP S (0)+
1+j
fP±SZ1
AD (0)
.
AP S (0)−
Typical values for the UA741 are P SRR(0)+ = 110000, P SRR(0)− = 170000, fP+SZ1 =
685Hz, fP−SZ1 = 6.2Hz. The considerable difference in the dominant zero frequen-
cies of the injected power supply gains is normally due to the fact that the OP
AMP circuits are not symmetric when viewed from the power supply signal in-
jection terminals. By adding external components to an OP AMP macromodel
power supply rejection effects can be easily simulated. The schematic shown in
Fig. 7.57 shows the TI UA741 model with RC networks connected between the
power supply terminals and earth. The voltage controlled voltage sources probe
the voltages at the center nodes of the additional RC networks. These networks
generate the power supply injected signals at dc. They also generate the dominant
zero in the power supply rejection characteristic.

The values for the passive components can be calculated using:

106 1 106 1
RA = , CA = , RB = , CA =
P SRR(0)+ 6 +
2 · 10 · π · fP SZ1 P SRR(0)− 2 · 10 · π · fP−SZ1
6

Which gives, for the example UA741 device data, RA = 9Ω, CA = 232pF , RB =
5.9Ω and CB = 25.7pF . Simulation waveforms for the small signal frequency
response of the test circuit are shown in Fig. 7.58. In the case of the modular

259
 Vout_mod

dc simulation R1
VS
R=100k
U=1 V
DC1

V1

R2 - VCC U=15 V

R=10 UA741
(MOD)
V2
U=15 V
+ VEE
R3
R=10 R4
R=100k SUB5
ac simulation
AC1 R11
Type=log R=100k
Start=1 Hz
Stop=10 kHz
Points=400
CA
C=232p
RA
R=9
R5
EP2 EN2 R=1M
R9 G=0.5 G=0.5
R=10
- vout_TI
VCC
UA741(TI)

VEE
+
SUB4

EP1 EN1
G=0.5 G=0.5
R10 R6
R=10 R=1M
RB
R=5.9
R12
R=100k CB
Equation C=25.7nF

Eqn1
PSRR_P=dB(p1/(vout_TI.v*p2*alpha))
fpz1=685
alpha=1e-4
gbp=1e-6
p1=mag(1+j*acfrequency/fpz1)
p2=mag(1+j*acfrequency/alpha*gbp)

Figure 7.57: Test circuit showing OP AMP with external power supply rejection
modelling network

UA741 model the simulation signal plot clearly demonstrates the fact that the
model does not correctly represent the effects due to power supply injected signals.

260
 0.03

0.025
vout_TI.v

0.02

0.015

0.01
1 10 100 1e3 1e4
acfrequency

110
PSRR_P

105

100
1 10 100 1e3 1e4
acfrequency

3e-21

2e-21
Vout_mod.v

1e-21

0
1 10 100 1e3 1e4
acfrequency

Figure 7.58: Simulation waveforms for the circuit illustrated in Fig. 7.57

261
7.14 End note
While writing this tutorial I have tried to demonstrate how practical models of
operational amplifiers can be constructed using basic electronic concepts and the
range of Qucs built-in components. The modular OP AMP macromodel was de-
liberately chosen as the foundation for the tutorial for two reasons; firstly Qucs
is mature enough to easily simulate such models, and secondly the parameters
which determine the operation of the macromodel can be be calculated directly
from information provided on device data sheets. Recent modelling development
by the Qucs team has concentrated on improving the SPICE to Qucs conversion
facilities. This work has had a direct impact on Qucs ability to import and simu-
late manufacturers OP AMP models. The tutorial upgrade explains how SPICE
Boyle type OP AMP macromodels can be converted to work with Qucs. The Qucs
OP AMP library (OpAmps) has been extended to include models for a range of
popular 8 pin DIL devices. If you require a model with a specific specification that
is not modelled by an available macromodel then adding extra functionality may
be the only way forward. Two procedures for extending models are outlined in the
tutorial upgrade. Much work still remains to be done before Qucs can simulate
a wide range of the macromodels published by device manufacturers. With the
recent addition of subcircuit/component equations to Qucs it is now possible to
write generalised macromodel macros for OP AMPs. However, before this can
be done time is required to fully test the features that Stefan and Michael have
recently added to Qucs release 0.0.11. This topic and the modelling of other OP
AMP properties such as noise will be the subject of a further OP AMP tutorial
update sometime in the future. My thanks to David Faulkner for all his help and
support during the period we were working on a number of the concepts that form
part of the basis of this tutorial. Once again a special thanks to Michael Margraf
and Stefan Jahn for all their help and encouragement over the period that I have
been writing this tutorial and testing the many examples it includes.

262
8 Modelling the 555 Timer
8.1 Introduction
The 555 timer was designed by Hans R. Camenzind in 19701 and first produced
by Signetics during the period 1971-19722 . The device was originally called ”The
IC time machine” and given the part number SE555/NE555. Over the last 30 plus
years more than ten different semiconductor chip production companies have made
555 parts, making it one of the most popular ICs of all time3 . Today it is still used
in a wide range of circuit applications.

The 555 timer is one of the first examples of a mixed mode IC circuit that includes
both analogue and digital components. The primary purpose of the 555 timer is
the generation of accurately timed single pulse or oscillatory pulse waveforms. By
adding one or two external resistors and one capacitor the device can function as
a monostable or astable pulse oscillator.

The 555 timer is a difficult device to simulate. During circuit operation it switches
rapidly between two very different DC states4 . Such rapid changes can be the
cause of simulator DC convergence and transient analysis errors. Most of the
popular simulators include some form of 555 timer model, either built-in or as a
subcircuit, which functions to some degree. These models usually include a number
of p-n junctions and non-linear controlled sources, making simulation times longer
than those obtained with simpler models. At the heart of the 555 timer are two
comparators and a set-reset flip flop. A block diagram of the main functional
elements that comprise the 555 timer is illustrated in Fig. 8.1.

The current Qucs release does not include a model for the 555 timer. The pur-
pose of the work reported in this tutorial note has been to develop a 555 timer
1
See ”The 555 Timer IC. An interview with Hans Camenzind - The designer of the most success-
ful integrated circuit ever developed”, http://semiconductormuseum.com/Transistors/
LectureHall/Camenzind/
2
Now part of the Philips organisation.
3
Recent manufacturing volumes indicate that the 555 timer is as popular as ever, with for
example, Samsung (Korea) producing over one billion devices in 2003; see Wikipedia entry
at http://en.wikipedia.org/
4
Typically between ground and a voltage close to power rail VCC.

263
model from scratch which simulates efficiently, and is based only on the circuit
components implemented in Qucs 0.0.10. Moreover, while developing the Qucs
555 model every attempt has been made to reduce the number of p-n junctions to
a minimum, yielding both model simplicity and reduced circuit simulation times.
The approach adopted is centred on established macromodelling techniques where
signals at the timer device pins accurately model real device signals but internal
macromodel signals often bare no relation to those found in an actual device. In-
ternally, the macromodel simply processes input signal information and outputs
signals, in the correct format, to the device output pins. In no way is an attempt
made to simulate the actual 555 timer circuitry.

555

GND VCC

P_VCC1
TRIG DIS

P_RESET1 OUT TRESH

R1
R=5k
RES CON

+ DIGITAL
THRESH SUB6
P_THRESH1 LOGIC File=timer_555.sch

Reset Q +
- P_OUTPUT1
P_CONTROL1 Thresh
R2 SUB3
AMP
R=5k Trig QB
_
+ SUB5
TRIG
P_GND1
SUB2
Discharge
- Switch
P_TRIGGER1 SUB4
R3 P_DISCHARGE1
R=5k

SUB1

Figure 8.1: 555 Timer functional block diagram.

8.2 The Qucs 555 timer model

Fig. 8.1 illustrates the new Qucs 555 timer model. In this model each of the major
functional blocks have been separated into macromodel subcircuits, grouping sim-
ilar types of component together. Essentially, the model only includes standard
Qucs components which all work together to produce the correct output signals

264
through careful selection of threshold parameters, voltage limits, logic levels and
rise and fall times. These notes concentrate on explaining the structure and pa-
rameters of the macromodel subcircuits that form the 555 timer model, rather
than describing the function of the device5 . The 555 timer is an 8 pin device with:

• Pin 1 Ground [GND] - Most negative supply connected to the device, nor-
mally this is common ground (0V).

• Pin 2 Trigger [TRIG] - Input pin to the lower comparator. Used to set the
RS latch.

• Pin 3 Output [OUT] - The 555 timer output signal pin.

• Pin 4 Reset [RES] - Used to reset the RS latch.

• Pin 5 Control [CON] - Direct access point to the (2/3)VCC divider node.
Used to set the reference voltage for the upper comparator.

• Pin 6 Threshold [THRESH] - Input pin to upper comparator. Used to reset

the RS latch.

• Pin 7 Discharge [DIS] - Collector output of an npn BJT switch. Used to

discharge the external timing capacitor.

• Pin 8 VCC [VCC] - Most positive supply connected to device, normally this
is 5V, 10V or 15V.

8.2.1 The trigger comparator macromodel

The trigger comparator input pins are connected between the (1/3)VCC divider
node and device package pin 2 (TRIG). Trigger input signals dropping below the
(1/3)VCC divider node voltage cause the trigger output voltage to switch, setting
the RS latch in the digital logic subcircuit. This action also causes the 555 timer
output signal to go high. The trigger input is level sensitive. Retriggering will
occur if the trigger pulse is held low longer than the 555 timer output pulse width.
The trigger comparator circuitry also has a storage time of several microseconds,
limiting the minimum monostable output pulse to around 10µS. A DC current,
popularly referred to as the trigger current, flows from device pin 2 (TRIG) into
the external circuit. This has a typical value of 500 nA, setting the upper limit
of resistance that can be connected from pin 2 to ground6 . The circuit diagram
5
A good tutorial guide to the operation of the 555 timer can be found at http://www.uoguelph.
ca/~antoon/gadgets/555/555.html
6
At VCC = 5V this resistance is roughly 3.3MΩ.

265
of the trigger comparator macromodel is shown in Fig. 8.2. The differential input
signal is sensed by operational amplifier OP1. This has it’s gain set to 1e6, giving
a differential input signal resolution of 1µV. OP1 output voltages are limited to
±1V. Note the upper +1V signal level corresponds to a logic ’1’ signal. Finally,
the trigger comparator output voltage rise and fall times are set by time constant
R1 ∗ C1. This network also adds a time delay to the comparator macromodel.

comp_vout1
Pcomp_vp1
+
R1 C1 TRIG
Pcomp_vn1 R=1k C=1 nF
OP1
G=1e6
I1
Umax=1 V -
I=500 nA SUB1
File=timer_trig.sch

Figure 8.2: Trigger comparator macromodel.

8.2.2 The threshold comparator macromodel

The threshold comparator macromodel is shown in Fig. 8.3. It is very similar to the
trigger comparator macromodel; one noteable difference is the size and direction of
pin 6 (THRES) threshold DC current which is typically 100nA and flows into pin
6 from the external circuitry7 . The threshold comparator is used to reset the RS
latch in the 555 timer digital logic block, causing the 555 timer output to go low.
Resetting occurs when the signal applied to external pin 6 (THRES) is driven from
below to above the (2/3)VCC divider node voltage. Again the threshold input is
level sensitive.

7
The threshold DC current sets the upper limit to the value of the external resistor that can be
connected between pin 6 and the VCC supply - for VCC = 5V this is approximately 16MΩ,
with VCC = 15 V this rises to roughly 20MΩ.

266
 I1
I=0.1 uA

POUT1
Num=3
PinP1 +
Num=1 THRESH
R1
R=1k C1
PinN1 OP1 C=1 nF
Num=2 G=1e6 -
Umax=1 V SUB1
File=timer_thresh.sch

Figure 8.3: Threshold comparator macromodel.

8.2.3 The digital logic macromodel

The digital logic macromodel consists of an SR latch with additional combinational
gates at the input of the model, see Fig. 8.4. The truth table for the SR latch is
listed in Table 8.1. All gates in the macromodel have logic ’1’ set at 1V and logic
’0’ set at 0V. RC timing networks have been added to the output of each gate,
ensuring that the gates have a finite rise and fall times rather than the Qucs default
value of zero seconds8 . Gate input signals with values less than the gate threshold
voltage (0.5V) are considered to be a logic ’0’ signal. A logic ’0’ signal on 555
timer pin 4 (RES) also resets the SR latch causing the output signal, pin 3 (OUT),
to move to a low state. The reset signal is an override signal in that it forces the
timer output to a low state regardless of the signals on other timer input pins.
Reset has a delay time of roughly 0.5µS, making the minimum reset pulse width
of approximately 0.5µS. The reset signal is inverted then ORed with the threshold
comparator output signal.
8
In mixed mode circuit simulation transient analysis problems can occur when devices change
state in zero seconds, see later notes for comments on this topic.

Set (S) Reset (R) Q (P-Q1) QB (P-QB1) Notes

1 0 1 0 Set state
0 0 1 0
0 1 0 1 Reset state
0 0 0 1
1 1 0 0 Undefined

Table 8.1: Truth table for an SR latch constructed using NOR gates.

267
 P_Q1

1 P_QB1
1
P_reset1
1 1
R1 C1
Y2 R=1k C=0.5nF R2 C2 C5
Y3 R=1k R4 C=0.05nF R5 C3
C=0.5nF Y4 Y1 C=1nF
R=1k R=1k

P_tresh1

1
C4
R3 C=0.09nF
P_trig1 Y5 R=1k

DIGITAL
LOGIC

Reset Q
Thresh
Trig QB

SUB1
File=timer_digital_comb.sch

Figure 8.4: Digital logic macromodel.

8.2.4 The 555 timer output amplifier macromodel

Illustrated in Fig. 8.5 is the macromodel for the timer output amplifier. This is
a simple model constructed from a voltage gain block plus a resistor to represent
the 555 timer output resistance. The voltage gain block has it’s value set to 3.5
in Fig. 8.5. This is the value needed to scale the logic ’1’ signal voltage to the
required external voltage at timer output pin 3 (OUT). This value is only correct
for power supply voltage VCC set to 5V, and must be changed for other voltages9 .

9
At this time Qucs does not allow parameters to be passed to subcircuits, making it difficult to
write generalised macromodels. Adding parameter passing to subcircuits and the calculation
of component values using equations is on the to-do list. Suggested values for the amplifier
gain are: (1) VCC = 5V, G = 3.5, (2) VCC = 10V, G = 8.5V and (3) VCC = 15V, G = 13.5.
These gain values correct for the voltage drop in the 555 timer totem-pole output stage.

268
 +
Pamp_P1 R1 P_vout1
R=7 AMP
_
SRC1
Pamp_N1 G=3.5
T=0

SUB1
File=timer_amp.sch

Figure 8.5: Output amplifier macromodel.

8.2.5 The discharge switch macromodel

The discharge switch macromodel is shown in Fig. 8.6. Like the actual 555 timer
the macromodel discharge switch is based on an npn transistor. A logic ’1’ signal
applied to terminal pin_control_in1 turns the npn transistor on causing the path
from the collector (555 timer pin DIS) to ground to become low resistance. It is
through this branch that the timer external capacitor is discharged. The reverse
characteristic is observed when the input control voltage is logic ’0’. In this case
the collector to ground branch has a very high resistance. Resistor R1 is included
in the macromodel to limit the npn base current when the BJT is turned on.
Similarly, resistor R2 has been added to the model to limit the external capacitor
discharge current10 .

10
Normally the external timing capacitor is discharged through a resistor in series with the
collector to ground path. However, if this series resistor is very small, or indeed does not
exist, it is theoretically possible for the discharge current to become very large, which in turn
leads to DC convergence errors or very long transient simulation times.

269
 T1
Type=npn
Is=1e-16
P_control_in1 R1 Nf=1
Vaf=0 Discharge
R=10K Switch
Bf=100

P_GND1

SUB1
P_Discharge1 R2 File=timer_Discharge.sch
R=200

Figure 8.6: The discharge switch macromodel.

8.3 Published 555 timer test circuits

The majority of manufacturers outline in their 555 timer specification sheets a
range of fundamental circuit applications11 . A number of these circuits are intro-
duced as a series of simulation test cases. The conditions chosen for the simulation
tests are as follows:

• Integration method Gear, order 6 (this method works well with circuits that
contain time constants that have widely different values)12 .

• Input driver signals have a finite rise and fall time, usually in nano seconds
(problems can occur when driver signals have either zero or very small rise
and fall times - often a simulator will reduce the transient analysis step
size in an attempt to reduce errors which in turn can significantly increase
simulation run times).

• Transient simulation parameter MinStep is set to one hundredth, or less, of

the smallest rise or fall time in the circuit (this is a good rule of thumb, giving
reasonable simulation times and accuracy, normally without DC convergence
or transient analysis time step problems).
11
See for example the ”Applications Information” section of the National Semiconductor LM555
Timer data sheet, July 2006, www.national.com.
12
One of the simulation tests also presents results using the standard trapezoidal second order
integration method.

270
8.3.1 The 555 timer monostable pulse generator
Figure 8.7 shows the basic 555 timer monostable pulse generator circuit. The
output pulse width is given by the equation T = 1.1 ∗ R5 ∗ C1; when R5 = 9.1k
and C1 = 0.01µF, T = 1ms. Figure 8.8 illustrates the simulation waveforms for
the monostable oscillator.

V1 555
U=5 V
R5
R=9.1k GND VCC
vtrig

TRIG DIS
V5
transient U1=5 V
vout
simulation U2=0 V OUT TRESH
T1=0.3ms reset
T2=0.35 ms
TR1 Tr=5 ns
Type=lin RES CON
C1 Tf=5 ns V4 C2
C=0.01 uF Start=0 U1=5 V C=0.01uF
Stop=0.6ms vdis
U2=0V
IntegrationMethod=Gear T1=0.1ms
Order=6 SUB1
T2=0.15ms

Figure 8.7: The basic 555 timer monostable pulse generator.

271
reset.Vt 5

0
0 5e-5 1e-4 1.5e-4 2e-4 2.5e-4 3e-4 3.5e-4 4e-4 4.5e-4 5e-4 5.5e-4 6e-4
time

5
vtrig.Vt

0
0 5e-5 1e-4 1.5e-4 2e-4 2.5e-4 3e-4 3.5e-4 4e-4 4.5e-4 5e-4 5.5e-4 6e-4
time

5
vdis.Vt

0 5e-5 1e-4 1.5e-4 2e-4 2.5e-4 3e-4 3.5e-4 4e-4 4.5e-4 5e-4 5.5e-4 6e-4
time
4
vout.Vt

0
0 5e-5 1e-4 1.5e-4 2e-4 2.5e-4 3e-4 3.5e-4 4e-4 4.5e-4 5e-4 5.5e-4 6e-4
time

Figure 8.8: Simulation waveforms for the basic monostable pulse generator.

8.3.2 The 555 timer astable pulse oscillator

Figure 8.9 shows the basic 555 timer astable pulse generator circuit. The charging
time for capacitor C1 is given by tc = 0.693(R5+R6)C1 seconds, and the discharge
time by td = 0.693(R6)C1 seconds. Hence, the period and frequency of oscillation
are:
1.44
T = tc + td = 0.693(R5 + 2R6)C1 seconds, and f = Hz.
(R5 + 2R6)C1
R6
The duty cycle for the timer output waveform is also given byD = .
R5 + 2R6
Figure 8.10 illustrates the simulation waveforms for the astable oscillator. When
resistor R6 is shunted by a diode, capacitor C1 charges via resistor R5 and dis-
charges via resistor R6. On setting R5 = R6 a 50 percent duty cycle results13 , see
13
The value of R6 needs to be trimmed to set the duty cycle to exactly 50 percent.

272
Figures 8.11 and 8.12.

R5
V1
R=3.9k
U=5 V 555

GND VCC
vtrig vdis
R6 TRIG DIS
R=3k
transient vout
simulation OUT TRESH
reset
TR1
Type=lin RES CON
C1 Start=0 V4 C2
C=0.01 uF Stop=0.3ms U1=5 V C=0.01uF
Points=1000 U2=0V
IntegrationMethod=Gear T1=0 SUB1
Order=6 T2=0.02ms File=timer_555.sch

Figure 8.9: The basic 555 timer astable pulse generator.

273
 5
reset.Vt

0
0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time
4
vtrig.Vt

0
0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time
5
vout.Vt

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time
5
vdis.Vt

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time

Figure 8.10: Simulation waveforms for the basic astable pulse generator.

R5
V1
R=3k
U=5 V 555

GND VCC
vtrig vdis
D1
R6 TRIG DIS
R=3.6k
transient vout
simulation OUT TRESH
reset
TR1
Type=lin RES CON
C1 Start=0 V4 C2
C=0.01 uF Stop=0.3ms U1=5 V C=0.01uF
Points=4000 U2=0V
IntegrationMethod=Gear T1=0 SUB1
Order=6 T2=0.02ms File=timer_555.sch

Figure 8.11: 555 timer astable pulse generator with 50 percent duty cycle.

274
 5
reset.Vt

0
0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time
5
vtrig.Vt

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time
5
vout.Vt

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time

5
vdis.Vt

0 2e-5 4e-5 6e-5 8e-5 1e-4 1.2e-4 1.4e-4 1.6e-4 1.8e-4 2e-4 2.2e-4 2.4e-4 2.6e-4 2.8e-4 3e-4
time

Figure 8.12: Simulation waveforms for 50 percent duty cycle astable pulse
generator.

8.3.3 Pulse width modulation

Triggering the 555 timer in monostable mode with a continuous sequence of pulses
allows the output pulse width to be modulated by changing the amplitude of a
signal applied to the control input pin 5 (CON). An example pulse width modulator
circuit is given in Fig. 8.13. In this circuit components C2, R6 and D1 convert
the 555 trigger signal into a falling edge triggering signal. This can be seen in
Fig. 8.14 which illustrates the trigger, discharge and resulting output waveform.
The 555 timer control pin is driven from a voltage pulse source. The specification
of the control waveform has been chosen to generate a triangular shaped signal so
that the modulation of the pulse width can be clearly seen as the control signal
amplitude changes.

275
 R5 R6 D1 transient
R=20k V1 R=4.7k simulation
U=5 V 555
TR1
Type=lin
GND VCC Start=0
vsig Stop=20ms
IntegrationMethod=Gear
TRIG DIS Order=6
V7 vout
C2 vtrig
U=5 V C=0.01uF
TH=0.75 ms OUT TRESH
TL=0.5 ms reset
Tr=20 ns vcon vdis
Tf=20 ns RES CON
C1 V4
C=0.01 uF U1=5 V
U2=0V V8
T1=0.2ms SUB1 U1=1 V
T2=0.5ms File=timer_555.sch U2=5 V
Tr=10 ns T1=0
Tf=10 ns T2=20ms
Tr=10 ms
Tf=10 ms

Figure 8.13: Pulse width modulator 555 timer circuit.

276
 5
reset.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time

5
vcon.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time

5
vsig.Vt

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time

10
vtrig.Vt

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time

5
vdis.Vt

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time
5
vout.Vt

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02
time

Figure 8.14: Simulation waveforms for pulse width modulator.

277
8.3.4 Pulse position modulation
A pulse position modulator can be constructed from the astable waveform gen-
erator given in Fig. 8.9. A modulating signal is applied to the control input pin
5 (CON); see Fig. 8.15. This signal causes the pulse position to vary with the
amplitude of the applied modulating signal. A typical set of simulation waveforms
for this circuit are shown in Fig. 8.16. This is a very difficult circuit to simulate.
It is one case where the trapezoidal integration method works successfully whereas
the 6th order Gear integration method appears to fail14 . Note that the trapezoidal
results were obtained using 30000 points, Initial step = 0.001 nS, MinStep = 1e-16,
MaxIter = 5000, abstol = 10uA and vntol = 10uV.

R5
V1
R=3.9k
U=5 V 555

GND VCC
vtrig vdis
R6 TRIG DIS
R=3k
transient vout
simulation OUT TRESH
reset vcon
TR1
Type=lin RES CON
C1 Start=0 V4 V5
C=0.01 uF Stop=10ms U1=5 V U1=5V
Points=30000 U2=0V U2=4 V
IntegrationMethod=Trapezoidal T1=0 SUB1 T1=0
Order=2 T2=0.02ms File=timer_555.sch T2=10 ms
Tr=5 ms
Tf=5 ms

Figure 8.15: Pulse position modulator 555 timer circuit.

14
The transient simulation never finishes and can only be terminated by clicking the simulation
abort button.

278
 5
reset.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

5
vtrig.Vt

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

5
vdis.Vt

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time
5
vout.Vt

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

5
vcon.Vt

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

Figure 8.16: Simulation waveforms for pulse position modulator obtained using
trapezoidal integration.

8.4 Multiple 555 timer simulation examples

Having established in the last section that the new Qucs 555 timer model can
simulate the standard application circuits listed in a typical device data sheet,
this part of the tutorial introduces two further, more complex, examples that
demonstrate how the 555 timer is used in practice.

8.4.1 Sequential pulse train generation

A very practical application of the 555 timer is the generation of timing pulses for
control purposes. The circuit illustrated in Fig. 8.17 shows a set of monostable
pulse generators connected in series and parallel. After circuit reset the falling
edge of input pulse vin triggers the start of pulse sequence generation. The time

279
duration of each monostable pulse is set by external capacitors C1 to C415 . The
specification of the monostable pulse generator subcircuit is given in Fig. 8.18.
The sequential pulse generator is a complex circuit with:

60 R instances, 40 C instances, 4 VCVS instances, 1 Vdc instances,

8 Idc instances, 2 Vpulse instances, 8 OpAmp instances, 4 Diode instances,
4 BJT instances, 8 Inv instances, 8 NOR instances and 4 OR instances.

VCC VCC VCC

V1
U=5 V
vin vout1 vout2 vout3
IN OUT IN OUT IN OUT

V3
U1=0 V vres RES CAP RES CAP RES CAP
U2=5V C1 C2 C3
T1=1ms GND C=0.01uF GND C=0.02uF GND C=0.05uF
V2 SUB1 SUB2 SUB3
T2=1.3 ms U1=5 V
Tr=10 ns U2=0 V
Tf=10 ns T1=0.2ms
T2=0.5ms ms
Tr=10 ns
Tf=10 ns

transient
simulation VCC

TR1
Type=lin vout4
Start=0 IN OUT
Stop=5 ms
IntegrationMethod=Gear
Order=6
MinStep=1e-15 RES CAP
C4
GND C=0.1uF

SUB4

Figure 8.17: Sequential pulse generator circuit.

15
The pulse duration times set by C1 to C4, in Fig. 8.17, have simply been chosen for demon-
stration purposes and do not represent any particular control timing sequence.

280
 P_VCC

R5 R6 D1
R=20k R=4.7k 555

GND VCC

P_CAP TRIG DIS

P_IN C2
C=0.01uF OUT TRESH

P_GND
RES CON
C3
P_OUT C=0.01uF
SUB1

P_RES

VCC

IN OUT

RES CAP

GND

SUB2
File=555_timer_mono.sch

Figure 8.18: Monostable pulse generator subcircuit.

281
The large number of components, and indeed the complexity of the circuit, tend
to make the simulation time of the pulse train generator circuit much greater than
typical times recorded when simulating single 555 timer circuits. Also, circuit
DC convergence and transient analysis time step errors can be a problem, due to
switching discontinuities, making careful selection of the non-linear diode param-
eters and the transient analysis conditions essential. In Fig. 8.18 a diode is used
to clamp the 555 timer trigger input at five volts when the signal attempts to rise
above 5 volts. The default Qucs diode parameters are similar to those specified by
SPICE16 . By default the diode emission constant is set to 1 and the diode series
resistance to zero ohms. Neither of these values are particularly representative for
silicon diodes. For silicon devices, rather than germanium diodes, n needs to be
between roughly 1.5 and 2. Similarly, all diodes have some series resistance, often
in the range 0.1 to 10 ohms depending on the power rating of the diode. To aid
simulation these parameters have been set to n = 2 and Rs = 10Ω. Figure. 8.19
illustrates a typical set of signal waveforms obtained from the simulation of the
sequential pulse generator: the simulation conditions employed to generate these
results are; Integration method = Gear, Order = 6, initialStep = 1 ns, MinStep
= 1e-15, reltol = 0.001, abstol = 10µA, vntol = 10µV, Solver = CroutLU and
initialDC = yes.

16
The default values were set in an early version of SPICE, probably version 1, and appear to
have not been changed as the simulator was developed.

282
 5
vres.Vt

0
0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time

5
vin.Vt

0
0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time
5
vout1.Vt

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time
5
vout2.Vt

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time

5
vout3.Vt

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time

5
vout4.Vt

0 5e-4 1e-3 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
time

Figure 8.19: Simulation waveforms for the monostable pulse generator circuit.

283
8.4.2 Frequency divider circuit
A common requirement in both digital and mixed mode circuit design is frequency
division, where a high frequency pulse train, often derived from a crystal controlled
clock, is divided down to a much lower frequency17 . The classical way of dividing
such signals is to use a chain of flip-flops each connected as a divide by two ele-
ment. The 555 timer can also be used for pulse train frequency division18 . The
schematic shown in Fig. 8.20 shows a basic monostable mode 555 circuit with a
train of pulses applied to the 555 trigger input pin 2 (TRIG). In an earlier section
of these notes it was explained that the 555 trigger comparator input was signal
level sensitive and retriggering takes place if the duration of the low signal section
of the trigger waveform is greater than the monostable pulse duration. In Fig. 8.20
the monostable pulse length is 0.22ms and rectangular voltage generator param-
eter TL is 0.5ms which causes retriggering to occur. The effects of retriggering
can be seen in Fig. 8.21. Frequency division employing 555 timers is based on
the monostable circuit shown in Fig. 8.20 and hence circuit designers must make
sure that retriggering does not take place. Illustrated in Fig. 8.22 is a two stage
frequency division circuit where each stage divides the input pulse train by five
giving an overall division ratio of twenty five. The output waveforms for this cir-
cuit are shown in Fig. 8.23. When designing 555 timer frequency divider circuits
good performance can be achieved if the period of the 555 timer is set at (N-0.5)
times the period of the input pulse train19 , where N is the division ratio and is in
the range 2 ≤ N ≤ 10.

17
Often the resulting frequency is in the region 1 to 5 Hz and is used to flash an LED, or some
other optical actuator, on/off.
18
555 timers are normally more efficient than flip-flops in this application because single devices
can have divisors greater than two.
19
E. A Parr, IC 555 Projects, Bernard Babani (publishing) Ltd, 1981, p. 109.

284
 R5 transient
R=20k V1 simulation
U=5 V 555
TR1
Type=lin
GND VCC Start=0
vtrig1 Stop=10ms
IntegrationMethod=Gear
TRIG DIS Order=6
V7 vout1
U=5 V
TH=0.75 ms OUT TRESH
TL=0.5 ms reset
Tr=20 ns
Tf=20 ns RES CON
C1 V4
C=0.01 uF U1=5 V
U2=0V
T1=0 SUB1
T2=0.2ms File=timer_555.sch
Tr=10 ns
Tf=10 ns

Figure 8.20: A monostable mode 555 timer circuit with a pulse train applied to
the trigger input.

5
reset.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

5
vtrig1.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

4
vout1.Vt

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

Figure 8.21: Simulation waveforms for the circuit given in Fig. 8.20: these show
555 retriggering.

285
 R5 transient
R=20k V1 simulation
U=5 V 555
TR1
Type=lin
GND VCC Start=0
vtrig1 Stop=10ms
IntegrationMethod=Gear
TRIG DIS Order=6
V7 vout1
U=5 V
TH=0.2 ms OUT TRESH
TL=0.1 ms reset
Tr=20 ns
Tf=20 ns RES CON
C1 V4
C=0.0525 uF U1=5 V
U2=0V
T1=0 SUB1
T2=0.2ms
Tr=10 ns
Tf=10 ns

R6
R=20k

555

GND VCC

TRIG DIS
vout2

OUT TRESH

C2 RES CON
C=0.26 uF

SUB2

Figure 8.22: A two stage 555 timer frequency division circuit.

286
 5
reset.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

5
vtrig1.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time
4
vout1.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time
4
vout2.Vt

0
0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

Figure 8.23: Simulation waveforms for the circuit given in Fig. 8.22.

8.5 End note

Developing a simulation model for the 555 timer is an interesting challenge. This
tutorial note attempts to describe the principles and macromodelling technology
needed for such a task. It also demonstrates how much Qucs has matured as a
universal simulator. The new Qucs 555 timer model is very much a first attempt
on my part at building a functional model of this complex device. Much more
work needs to be done in the future to improve the 555 timer model. Low power
555 timer models are also needed for these popular variants. Longer term a uni-
versal parameterised subcircuit model for the 555 timer should become possible
once passing parameters to Qucs subcircuits and calculation of component values
using equations are implemented. A special thanks to Stefan Jahn for all his en-
couragement and the many modifications he made to Qucs, which either corrected
bugs or added functionality, during the period I have been working on this topic.

287
9 Qucs Simulation of SPICE
Netlists

9.1 Introduction
During the 1960’s and 70’s, the academic community worked tirelessly to develop
computer simulation programs that could act as aids in the process of circuit de-
sign. One of the best known of these programs is SPICE1 . First released in 1972 by
the University of California at Berkeley, SPICE has become an industrial standard
circuit simulator. Qucs is a modern circuit simulation program which attempts to
bring together a range of established and emerging circuit simulation technologies
to form a ”Quite Universal Circuit Simulator”. Although not yet finished, a sub-
stantial part of the central core of the package is functioning, allowing it to be used
as a simulation engine for the analysis and design of real circuits. Many of the ba-
sic circuit components and simulation domains found in SPICE are also available
in Qucs. Over the last three decades the SPICE simulation circuit netlist language
has become a standard for describing, interchanging and publishing semiconductor
device models and circuit data. Today, most semiconductor device manufacturers
provide SPICE models or subcircuit netlists for their discreet components and in-
tegrated circuits. One area where Qucs and SPICE differ significantly is in their
circuit file netlist formats which are very different2 . Qucs cannot directly simulate
standard SPICE circuit netlists but requires them to be converted to their Qucs
equivalent prior to simulation. The purpose of this tutorial note is to introduce
readers to a number of techniques that allow SPICE netlists to be simulated by
Qucs, secondly to indicate the limitations of the current SPICE to Qucs netlist
conversion process, and finally to present a preview of how Qucs is likely develop
in the future in the area of SPICE netlist compatibility.

1
The origins and background to the development of the SPICE simulator are described by
Ronald A. Rohrer in Circuit Simulation - the early years, illuminating SPICE’s strengths,
uncovering weaknesses, and projecting its future, IEEE Circuits and Devices, 1992, pp 32-37.
2
The Qucs netlist grammar is defined in appendix A1, of the Qucs Technical Papers.

288
9.2 The basic SPICE netlist format
SPICE simulation input data are text files which describe circuit structure, com-
ponent data and requested simulation tasks for the circuit who’s performance is
being simulated. Such text files form the fundamental input data to the SPICE
simulation engine, and normally include:

• A title statement

• Circuit node names

• Circuit element values

• Voltage and current source descriptions

• Analysis command statements

• Output data statements

• Other command statements

In SPICE 23 circuit node names (nets) are identified by integers numbered from
0 to 9999. SPICE 34 allows a mixture of letters and numbers for node names.
All circuit nodes must have a DC path to ground. Ground node is always node 0
and is considered global. Circuit element values are expressed as integers or real
numbers in scientific notation, for example 5, 0.5e1 5.0, or in engineering notation
using suffixes. The available SPICE suffixes are f = 1e-15 (femto), p = 1e-12
(pico), n = 1e-9 (nano), u = 1e-6 (micro), mil = 25e-6, m = 1e-3 (milli), k =
1e3 (kilo), meg = 1e6 (mega), g = 1e9 (giga) and t = 1e12 (tera). Component
unit abbreviations are allowed in circuit value descriptions. However, these must
not be separated from their associated values by spaces. Commonly used unit
abbreviations are V = Volt, A = Amps. Hz = Hertz, ohm = Ohm(Ω), H = Henry,
F = Farad and deg = Degree. SPICE input data files have the following format:

1. Title

2. * starts a comment line

3
A guide to SPICE 2 features and simulation data format is given in SPICE Version 2G User’s
Guide, A Vladimirescu, Kaihe Zhang, A.R. Newton, D. O. Pederson and A. Sangiovanni-
Vincentelli, August 1981, Department of Electrical Engineering and Computer Sciences, Uni-
versity of California, Berkeley, Ca., 94720, US.
4
See SPICE 3 Version 3F User’s Manual, B. Johnson, T. Quarles,A.R. Newton, D. O. Pederson
and A. Sangiovanni-Vincentelli, October 1992, Department of Electrical Engineering and
Computer Sciences, University of California, Berkeley, Ca., 94720, US.

289
 3. Circuit description

4. Simulation directives

5. Data output directives

6. .end

A typical SPICE input data file for a discreet component circuit is shown in
Fig. 9.1. In this netlist all nodes are shown numbered, following the SPICE 2
node naming convention. Also the power supply, AC input signal generator and
output load are not included. Essentially, the netlist shown in Fig. 9.1 represents
the amplifier without any external components connected to it. Although Qucs
cannot directly simulate SPICE netlists the software does contain a SPICE to
Qucs netlist conversion program called QUCSCONV. This routine takes as input
a SPICE netlist file and outputs an equivalent Qucs formatted netlist file. The
Qucs netlist file can be read and simulated by the Qucs simulation engine. To make
the process transparent, and indeed straightforward for users, the conversion stage
in simulating SPICE netlist files5 has been automated via the Qucs GUI simulate
command (F2 key). SPICE netlist files can be linked to a Qucs SPICE netlist
schematic symbol.6 These in turn can be connected, on a schematic, to any other
appropriate Qucs component symbol or user defined symbol. Figure 9.2 shows the
resulting schematic for the two stage BJT circuit. In this diagram the external
voltage sources and amplifier load have been added together with the usual Qucs
icons for DC and AC simulation of the circuit. During simulation Qucs treats
the SPICE netlist component as a subcircuit7 and generates the appropriate Qucs
netlist code. For example, the netlist shown in Fig. 9.3 illustrates the Qucs style
netlist code for the two stage BJT amplifier. Simulation of the two stage BJT
amplifier gives the output waveforms displayed in Fig. 9.4.

5
For convenience SPICE netlist files are often denoted with the extention cir and stored in a
Qucs project under the other category.
6
The schematic symbol SPICE netlist can be found in the file components section of the com-
ponents icon lists on the left hand side of the GUI. Its connection pin list may be setup and
edited via the Edit SPICE component properties dialogue.
7
Hence the need to separate the external voltage sources and amplifier load from the main
amplifier circuit.

290
∗ A two−s t a g e BJT a m p l i f i e r .
∗
∗ I n p u t node 2 , output node 9
∗ Power s u p p l y Vcc c o n n e c t e d t o node 10
∗
c1 2 3 10 u f
r 1 3 10 200 k
r 2 3 0 50 k
r 5 10 4 12 k
q1 4 3 5 qmod
r6 5 0 3.6 k
c2 4 6 10 u f
c4 5 0 15 u f
r 3 10 6 120 k
r 4 6 0 30 k
r 7 10 7 6 . 8 k
q2 7 6 8 qmod
r8 8 0 3.6 k
c5 8 0 25 u f
c3 7 9 10 u f
∗
. model qmod npn ( i s =2e −16 b f =50 br=1 rb=5 r c =1 r e =0
+ c j e =0.4 p f v j e =0.8 me=0.4 c j c =0.5 p f v j c =0.8 c c s =1p f va =100)
∗
. end

Figure 9.1: SPICE netlist for a simple two stage BJT amplifier.

dc simulation
X1
File=stoq_nl1.cir DC1
vin
2 9

V1
ac simulation
U=1m V vout
spice RL AC1
10
R=10k Ohm Type=log
Start=10 Hz
V2 Stop=100 MHz
U=15 V Ref Points=200

Equation

Eqn1
Phase=phase(vout.v)
gain=dB(vout.v/vin.v)

Figure 9.2: Qucs schematic for the two stage amplifier represented by the SPICE
netlist shown in Fig. 9.1.

291
. Def : s t o q n l 1 c i r n e t 2 n e t 9 n e t 1 0 ref
C : C3 n e t 7 n e t 9 C=”10uF ”
C : C5 n e t 8 r e f C=”25uF ”
R: R8 n e t 8 r e f R=”3. 6 k ”
BJT : Q2 n e t 6 n e t 7 n e t 8 r e f Type=”npn ” I s =”2e −16” Bf =”50” Br =”1”
Rb=”5” Rc=”1” Re=”0” Cje =”0. 4pF ”Vje = ”0 . 8 ” Mje = ”0 . 4 ” Cjc =”0. 5pF ”
Vjc = ”0 . 8 ” C j s =”1pF ” Vaf =”100” Nf =”1” Nr=”1” I k f =”0” I k r =”0” Var =”0”
I s e =”0” Ne = ”1 . 5 ” I s c =”0” Nc=”2” Rbm=”0” I r b =”0” Mjc = ”0 . 3 3 ” Xcjc =”1”
Vjs = ”0 . 7 5 ” Mjs =”0” Fc = ”0 . 5 ” Vtf =”0” Tf =”0” Xtf =”0” I t f =”0” Tr =”0”
R: R7 n e t 1 0 n e t 7 R=”6.8 k ”
R: R4 n e t 6 r e f R=”30k ”
R: R3 n e t 1 0 n e t 6 R=”120k ”
C : C4 n e t 5 r e f C=”15uF ”
C : C2 n e t 4 n e t 6 C=”10uF ”
R: R6 n e t 5 r e f R=”3. 6 k ”
BJT : Q1 n e t 3 n e t 4 n e t 5 r e f Type=”npn ” I s =”2e −16” Bf =”50” Br =”1”
Rb=”5” Rc=”1” Re=”0” Cje =”0. 4pF ”Vje = ”0 . 8 ” Mje = ”0 . 4 ” Cjc =”0. 5pF ”
Vjc = ”0 . 8 ” C j s =”1pF ” Vaf =”100” Nf =”1” Nr=”1” I k f =”0” I k r =”0” Var =”0”
I s e =”0” Ne = ”1 . 5 ” I s c =”0” Nc=”2” Rbm=”0” I r b =”0” Mjc = ”0 . 3 3 ” Xcjc =”1”
Vjs = ”0 . 7 5 ” Mjs =”0” Fc = ”0 . 5 ” Vtf =”0” Tf =”0” Xtf =”0” I t f =”0” Tr =”0”
R: R5 n e t 1 0 n e t 4 R=”12k ”
R: R2 n e t 3 r e f R=”50k ”
R: R1 n e t 3 n e t 1 0 R=”200k ”
C : C1 n e t 2 n e t 3 C=”10uF ”
. Def : End

Figure 9.3: Qucs format netlist for the two stage BJT amplifier: NOTE -In this
listing the entries for Q1 and Q2 have been edited so that they fit on
the text page.

292
 4
vout.v

0
10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency
200

60
Phase
gain

0
40

10 100 1e3 1e4 1e5 1e6 1e7 1e8 10 100 1e3 1e4 1e5 1e6 1e7 1e8
acfrequency acfrequency

Figure 9.4: Simulation waveforms for the two stage amplifier.

293
9.3 Defining symbols for Qucs SPICE netlist
components
Qucs automatically generates the symbol for a SPICE netlist component and does
not allow users to edit the resulting symbol. One of the disadvantage of this
feature is that the placement of the symbol input and output pins may be in a
position which is contrary to accepted use or signal flow direction. To overcome
this limitation a user defined symbol may be constructed where the SPICE netlist
component is embedded within the new symbol. Figure 9.5 illustrates such a
symbol for the two stage BJT amplifier and the resulting Qucs netlist for the new
symbol is shown in Fig. 9.6. From Fig. 9.6 we observe that embedding a SPICE
netlist symbol, within a user defined symbol, introduces an additional subcircuit
call in the resulting Qucs netlist; this is probably a small price to pay for the
convenience that a user defined symbol brings to the overall simulation process.

X1
File=stoq_nl1.cir
P_OUT1
2 9

P_IN1 VCC
spice

10

P_VCC1 Ref
SUB1

Figure 9.5: User defined symbol for the two stage BJT amplifier.

294
. Def : s t o q f i g 5 a m p net0 net1 net2
Sub : X1 n e t 0 n e t 1 n e t 2 gnd Type=”s t o q n l 1 c i r ”
. Def : End

. Def : s t o q n l 1 c i r n e t 2 n e t 9 n e t 1 0 ref
C : C3 n e t 7 n e t 9 C=”10uF ”
C : C5 n e t 8 r e f C=”25uF ”
R: R8 n e t 8 r e f R=”3. 6 k ”
BJT : Q2 n e t 6 n e t 7 n e t 8 r e f Type=”npn ” I s =”2e −16” Bf =”50” Br =”1”
Rb=”5” Rc=”1” Re=”0” Cje =”0. 4pF ”Vje = ”0 . 8 ” Mje = ”0 . 4 ” Cjc =”0. 5pF ”
Vjc = ”0 . 8 ” C j s =”1pF ” Vaf =”100” Nf =”1” Nr=”1” I k f =”0” I k r =”0” Var =”0”
I s e =”0” Ne = ”1 . 5 ” I s c =”0” Nc=”2” Rbm=”0” I r b =”0” Mjc = ”0 . 3 3 ” Xcjc =”1”
Vjs = ”0 . 7 5 ” Mjs =”0” Fc = ”0 . 5 ” Vtf =”0” Tf =”0” Xtf =”0” I t f =”0” Tr =”0”
R: R7 n e t 1 0 n e t 7 R=”6.8 k ”
R: R4 n e t 6 r e f R=”30k ”
R: R3 n e t 1 0 n e t 6 R=”120k ”
C : C4 n e t 5 r e f C=”15uF ”
C : C2 n e t 4 n e t 6 C=”10uF ”
R: R6 n e t 5 r e f R=”3. 6 k ”
BJT : Q1 n e t 3 n e t 4 n e t 5 r e f Type=”npn ” I s =”2e −16” Bf =”50” Br =”1”
Rb=”5” Rc=”1” Re=”0” Cje =”0. 4pF ”Vje = ”0 . 8 ” Mje = ”0 . 4 ” Cjc =”0. 5pF ”
Vjc = ”0 . 8 ” C j s =”1pF ” Vaf =”100” Nf =”1” Nr=”1” I k f =”0” I k r =”0” Var =”0”
I s e =”0” Ne = ”1 . 5 ” I s c =”0” Nc=”2” Rbm=”0” I r b =”0” Mjc = ”0 . 3 3 ” Xcjc =”1”
Vjs = ”0 . 7 5 ” Mjs =”0” Fc = ”0 . 5 ” Vtf =”0” Tf =”0” Xtf =”0” I t f =”0” Tr =”0”
R: R5 n e t 1 0 n e t 4 R=”12k ”
R: R2 n e t 3 r e f R=”50k ”
R: R1 n e t 3 n e t 1 0 R=”200k ”
C : C1 n e t 2 n e t 3 C=”10uF ”
. Def : End

Figure 9.6: Qucs format netlist for the two stage BJT amplifier represented by a
user defined symbol: NOTE -In this listing the entries for Q1 and Q2
have been edited so that they fit on the text page.

295
9.4 Handling SPICE subcircuits
Although Qucs treats SPICE netlist components as subcircuits the SPICE to Qucs
netlist conversion process still allows SPICE subcircuits to be defined within the
SPICE file being converted. Such subcircuits then become local subcircuits to the
SPICE netlist component to which they are attached. This allows complex circuits
consisting of many related, but often different, circuit blocks to be represented by
a single symbol in a Qucs schematic. In such cases the resulting symbol represents
a true subsection of an entire circuit rather than a simple single circuit function
subcircuit. To demonstrate this feature consider the following examples; (1) a
multisection LC delay line and (2) a CMOS ring counter.

9.4.1 Subcircuit example 1: a multisection LC delay line

The SPICE netlist for a ten section LC passive delay line is shown in Fig. 9.7. In
this listing each LC delay section is represented by a SPICE subcircuit and these
sections are connected in series to form the overall delay line. Figures 9.8 and 9.9
present the resulting Qucs netlist and generated waveforms obtained with the test
circuit shown in Fig. 9.10.

9.4.2 Subcircuit example 2: a two section CMOS ring counter

Subcircuit example one only contains a single local subcircuit. The next example
demonstrates how SPICE listings with more than one subcircuit are handled by
Qucs. Such circuits are representative of more complex electronic systems which
form easily identifiable subsystem blocks.8 Fig. 9.11 shows the SPICE netlist for
a simple two section CMOS ring counter. This circuit is modelled at discreet
component level and uses basic level one MOS parameters to define the MOS
transistors. These are then combined to form NAND and NOR subcircuits. Again
for completeness the resulting Qucs netlist is shown in Fig. 9.12 together with a
typical set of counter input and output signal waveforms, Fig. 9.13.

8
One significant advantage that Qucs has when compared to netlist entry only circuit simulators
is that it is possible the define schematic symbols for subsystem blocks that comprise discreet
components and one or more local subcircuits. These may then be employed like any other
Qucs symbols when constructing circuit schematics.

296
∗ Z0 = 320 Ohm.
∗
. s u b c k t l c n1 n2
l 1 n1 n2 10 uh
c1 n2 0 10 p f
. ends
∗
r s n9 n10 320ohm
x1 n10 n11 l c
x2 n11 n12 l c
x3 n12 n13 l c
x4 n13 n14 l c
x5 n14 n15 l c
x6 n15 n16 l c
x7 n16 n17 l c
x8 n17 n18 l c
x9 n18 n19 l c
x10 n19 n20 l c
r l n20 0 320ohm
. end

Figure 9.7: SPICE netlist for a ten section LC delay line..

. Def : s t o q f i g 1 0 a net0 net10 net1 net2 net3 net4

net5 net6 net7 net8 net9
Sub : X1 net0 net10 net1 net2 net3 net4
n e t 5 n e t 6 n e t 7 n e t 8 n e t 9 gnd Type=”t e s t 3 p p c i r ”
. Def : End

. Def : t e s t 3 p p c i rnetN9 netN11 netN12 netN13 netN14

netN15 netN16 netN17 netN18 netN19 netN20 ref
R: RL netN20 r e f R=”320Ohm”
Sub : X10 r e f netN19 netN20 Type=”LC”
Sub : X9 r e f netN18 netN19 Type=”LC”
Sub : X8 r e f netN17 netN18 Type=”LC”
Sub : X7 r e f netN16 netN17 Type=”LC”
Sub : X6 r e f netN15 netN16 Type=”LC”
Sub : X5 r e f netN14 netN15 Type=”LC”
Sub : X4 r e f netN13 netN14 Type=”LC”
Sub : X3 r e f netN12 netN13 Type=”LC”
Sub : X2 r e f netN11 netN12 Type=”LC”
Sub : X1 r e f netN10 netN11 Type=”LC”
R: RS netN9 netN10 R=”320Ohm”
. Def : LC r e f netN1 netN2
L : L1 netN1 netN2 L=”10uH”
C : C1 netN2 r e f C=”10pF ”
. Def : End
. Def : End

Figure 9.8: Qucs netlist for a 10 section LC delay line: NOTE -In this listing the
entries for the .Def statements have been edited so that they fit on the
text page.

297
 1

vin.Vt
0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.2

v10.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.2
v20.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.2
v30.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
0.2
v40.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
0.2
v50.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
0.2

0.1
v60.Vt

-0.1
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time
0.2
v70.Vt

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.2
v80.Vt

-0.2
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.1
v90.Vt

-0.1

0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

0.05
v100.Vt

0
0 1e-8 2e-8 3e-8 4e-8 5e-8 6e-8 7e-8 8e-8 9e-8 1e-7 1.1e-7 1.2e-7
time

Figure 9.9: Simulation waveforms for a 10 section LC delay line.

298
 v10
vin 10nS v20

20nS v30
V1
30nS v40
U1=0 V
U2=1 V 40nS v50
T1=0
T2=5 n 50nS v60
60nS v70
70nS v80
transient 80nS v90
simulation 90nS v100
100nS
TR1
Type=lin
Start=0
Stop=120 ns SUB1
IntegrationMethod=Gear
Order=6

Figure 9.10: LC delay line test circuit.

299
∗ Two s t a g e CMOS r i n g c o u n t e r c i r c u i t .
∗
x1 1 5 6 nand2
x2 1 6 7 nand2
x3 3 6 2 nand2
x4 2 7 3 nand2
x5 1 2 8 nor2
x6 1 8 9 nor2
x7 5 8 4 nor2
x8 4 9 5 nor2
∗
. model modp pmos ( v t o=−1 kp=10u
+ cgdo =0.2n c g s o =0.2n cgbo=2n )
. model modn nmos ( v t o=1 kp=10u
+ cgdo =0.2n c g s o =0.2n cgbo=2n )
∗
. s u b c k t nand2 1 2 3
m1 3 1 4 4 modp w=40u l =5u
m2 3 2 4 4 modp w=40u l =5u
m3 5 1 0 0 modn w=20u l =5u
m4 3 2 5 5 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
v c c 4 0 p u l s e ( 0 5 0 1 ns 1 ns 1 2 )
. ends
∗
. s u b c k t nor2 1 2 3
m1 4 1 7 7 modp w=40u l =5u
m2 3 2 4 4 modp w=40u l =5u
m3 3 2 0 0 modn w=20u l =5u
m4 3 1 0 0 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
v c c 7 0 p u l s e ( 0 5 0 1 ns 1 ns 1 2 )
. ends
. end

Figure 9.11: SPICE netlist for a two section CMOS ring counter.

300
# Qucs 0 . 0 . 1 1 / media / hda2 /OPAMP templates/ t e s t s t o q f i g 1 1 a . s c h
. Def : s t o q f i g 1 1 a c i r n e t 1 n e t 4 ref
. Def :NOR2 r e f n e t 1 n e t 2 n e t 3
Vpulse :VCC n e t 7 c n e t 0 U1=”0” U2=”5” T1=”0” Tr=”1 ns ” Tf=”1 ns ” T2=”1”
MOSFET:M1 n e t 1 n e t 4 n e t 7 n e t 7 Type=”p f e t ” W=”40u ” L=”5u ” Vt0=”−1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M2 n e t 2 n e t 3 n e t 4 n e t 4 Type=”p f e t ” W=”40u ” L=”5u ” Vt0=”−1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M3 n e t 2 n e t 3 ref r e f Type=”n f e t ” W=”20u ” L=”5u ” Vt0 =”1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M4 n e t 1 n e t 3 ref r e f Type=”n f e t ” W=”20u ” L=”5u ” Vt0 =”1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
C : C1 n e t 1 r e f C=”10p ”
C : C2 n e t 2 r e f C=”10p ”
Vdc :VCC c n e t 0 r e f U=”0”
. Def : End
. Def :NAND2 r e f n e t 1 n e t 2 n e t 3
Vpulse :VCC n e t 4 c n e t 1 U1=”0” U2=”5” T1=”0” Tr=”1 ns ” Tf=”1 ns ” T2=”1”
MOSFET:M1 n e t 1 n e t 3 n e t 4 n e t 4 Type=”p f e t ” W=”40u ” L=”5u ” Vt0=”−1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M2 n e t 2 n e t 3 n e t 4 n e t 4 Type=”p f e t ” W=”40u ” L=”5u ” Vt0=”−1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M3 n e t 1 n e t 5 ref r e f Type=”n f e t ” W=”20u ” L=”5u ” Vt0 =”1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
MOSFET:M4 n e t 2 n e t 3 n e t 5 n e t 5 Type=”n f e t ” W=”20u ” L=”5u ” Vt0 =”1”
Kp=”10u ” Cgdo =”0 .2 n ” Cgso =”0.2 n ” Cgbo=”2n ” I s =”1e −14” N=”1”
Lambda=”0” Gamma=”0” Phi = ”0 . 6 ”
C : C1 n e t 1 r e f C=”10p ”
C : C2 n e t 2 r e f C=”10p ”
Vdc :VCC c n e t 1 r e f U=”0”
. Def : End
Sub : X8 r e f n e t 4 n e t 9 n e t 5 Type=”NOR2”
Sub : X7 r e f n e t 5 n e t 8 n e t 4 Type=”NOR2”
Sub : X6 r e f n e t 1 n e t 8 n e t 9 Type=”NOR2”
Sub : X5 r e f n e t 1 n e t 2 n e t 8 Type=”NOR2”
Sub : X4 r e f n e t 2 n e t 7 n e t 3 Type=”NAND2”
Sub : X3 r e f n e t 3 n e t 6 n e t 2 Type=”NAND2”
Sub : X2 r e f n e t 1 n e t 6 n e t 7 Type=”NAND2”
Sub : X1 r e f n e t 1 n e t 5 n e t 6 Type=”NAND2”
. Def : End
Sub : X1 v i n vout gnd Type=”s t o q f i g 1 1 a c i r ”
V r e c t : V1 v i n gnd U=”5 V” TH=”1 us ” TL=”1 us ” Tr=”1 ns ” Tf=”1 ns ” Td=”0 ns ”
.TR: TR1 Type=”l i n ” S t a r t =”0” Stop =”30u ” P o i n t s =”1000” I n t e g r a t i o n M e t h o d =”T r a p e z o i d a l ”
Order =”2” I n i t i a l S t e p = ”0 . 0 1 ns ” MinStep =”1e −18” MaxIter =”150” r e l t o l = ”0 . 0 1 ”
a b s t o l =”1 uA” v n t o l =”100 uV” Temp= ”2 6 . 8 5 ” L T E r e l t o l =”1e −3” LTEabstol =”1e −4”
LTEfactor =”1” S o l v e r =”CroutLU ” relaxTSR=”no ” i n i t i a l D C =”y e s ” MaxStep =”0”

Figure 9.12: Qucs netlist for a two section CMOS ring counter: NOTE -In this
listing the entries for MOSFETs and transient analysis have been
edited so that they fit on the text page.

301
 6

4
vin.Vt

0 2e-6 4e-6 6e-6 8e-6 1e-5 1.2e-5 1.4e-5 1.6e-5 1.8e-5 2e-5 2.2e-5 2.4e-5 2.6e-5 2.8e-5 3e-5
time

4
vout.Vt

0 2e-6 4e-6 6e-6 8e-6 1e-5 1.2e-5 1.4e-5 1.6e-5 1.8e-5 2e-5 2.2e-5 2.4e-5 2.6e-5 2.8e-5 3e-5
time

Figure 9.13: Two stage CMOS ring counter signal waveforms.

302
9.5 Limitations when converting SPICE netlists
Not all SPICE netlists can be converted to Qucs netlist format and simulated by
Qucs9 . There are a number of reasons for this. The first and most obvious is due
to the fact that some SPICE components have not been implemented in Qucs yet.
Nonlinear controlled voltage and current sources are an example.10 There are also
a number of detailed differences between the SPICE and Qucs implementation of
components common to both simulators, one being the lack of PWL features in
the Qucs independent voltage and current sources. A second area that represents
a significant limitation, for those readers who regularly write SPICE netlists as
part of their simulation work, is the fact that Qucs contains a much greater range
of predefined primitive components that are not available in either the SPICE 2 or
SPICE 3 simulators. Perhaps this is not so much a limitation but an indication of
the current development effort being put into Qucs by the development team. As
the development of Qucs progresses it is expected that all the component features
found in SPICE will have a corresponding entry in Qucs11 .

9.6 Extending the SPICE netlist language

The standard SPICE 2 and SPICE 3 hardware description languages do not allow
(1) component values to be defined by algebraic equations12 or (2) parameters
to be passed to subcircuits. This makes writing universal subcircuit models very
difficult, forcing semiconductor device manufacturers to issue individual SPICE
models for each device they manufacture rather than a single generalised model13
for a given type of integrated circuit. A well known example being the SPICE
Boyle14 operational amplifier models. A number of current commercial circuit
9
A number of Qucs users have reported problems in the past when trying to simulate SPICE
netlists for components that have been published by device manufactures, see for example,
”Qucs SPICE error - please...”, William Flyn <WF215@ca...>, 29.8.2006, Qucs help forum.
10
SPICE 2 polynomal controlled voltage and current sources and SPICE 3 type B sources are
not implemented in any of the Qucs versions so far released. Their implementation is on the
to-do list but no date for their implementation has been fixed yet.
11
Future plans in this area are discussed in a later section of these notes.
12
Please note this is not strictly true as SPICE 3 B sources can be defined by equations involving
simulation variables and other data.
13
In a generalised model only one model description is provided for each generic component/-
circuit. Different component models are formed by passing parameters to the generalised
model. SPICE employs this approach to represent semiconductor devices through the use of
the .model statement. However, in the .model case the code for each type of semiconductor
device is hardwired into the simulator code rather than being defined by a subcircuit.
14
Boyle,G.R., B.M. Cohn, D.O. Pederson, and J.E. Solomon, 1974, Macromodeling of integrated
circuit amplifiers, IEEE Journal of Solid-State Circuits (December).

303
simulators15 have been extended to include the parameter based features outlined
above. In the case of those simulators based on the unextended Berkely SPICE
2G6 or SPICE 3F516 code a different approach is often adopted. This is based on
the use of a preprocessor, similar to that found in the C language, which takes
as input a parameter and equation style netlist and outputs a standard SPICE
netlist with the parameters and equations evaluated to give a numerical result.
The advantage of this approach is that the preprocessor can be used with any
SPICE simulator or indeed with Qucs. Two such preprocessors are SPICEPRM
and SPICEPP.17 The flow diagram for the Qucs simulation sequence including
a SPICE preprocessing stage is shown in Fig. 9.14. This diagram clearly shows
how both standard SPICE and parameterised netlists can be linked into the Qucs
simulation cycle. Of the two SPICE preprocessors introduced above SPICEPP is
probably the most useful from a Qucs users point of view18 as it adds more features
to the overall simulation process. Hence the notes that follow will concentrate on
describing how SPICEPP can be used with Qucs.

9.6.1 The SPICEPP preprocessor

SPICEPP19 is a preprocessor for Berkeley SPICE 3F5, adding support for a number
of structures found in commercial SPICE simulators, specifically SPICE commands
.param, .global, .lib, .temp, .meas and inline comments ($). The remainder of
these notes explain the use of commands .param, .global and the inline comment
as these add specific functionality to Qucs that is not provided by other sections
of the Qucs simulation software. The definition of these commands are:

• .param data=dataval <data2=dataval2> ............ The .param statement

adds the ability to parameterise SPICE data, including component values,
voltages, currents and equations.

• .globel node1 <node2> ............... The .global statement causes the named
nodes to override local subcircuit nodes of the same name.
15
For example PSPICE, HSPICE and IS-SPICE.
16
For example NGSPICE, TCLSPICE and WINSPICE.
17
(1) Andrew J. Borsa, SPICEPRM, A SPICE preprocessor for parameterised subcircuits,
V 0.11, 1996, <andy@moose.mv.com> (SPICEPRM can be downloaded from the Source-
forge.net ngspice project.) and (2) John Shaehen, SPICEPP, A SPICE proprocessor for
SPICE 3F5, V 1.5, 2000, <john@reptechnic.com.au>. (SPICEPP can be downloaded from
the Sourceforge.net tclspice project.)
18
SPICEPP was written after SPICEPRM and extends the facilities offered by SPICEPRM.
19
SPICEPP is written in PERL. The SPICEPP.pl script should be copied to a directory on your
search path. On my system I keep it in the Qucs bin directory. PERL must also be installed
on your system.

304
 SPICE SPICE
Generate SPICE
Parameterised Preprocessor netlist symbol
netlist
File XXXX File XXXX.cir

Predefined Qucs
component symbols

Qucs Qucs GUI

library
components
CIRCUIT
entered using Qucs
schematic capture
User defined subcircuit symbols

Generated using Qucs schematic

capture

Generate Qucs netlist code

from GUI schematic, including
conversion of SPICE code to SIMULATE
Qucs netlist format

Qucs netlist code

Run QUCSATOR

Simulation Qucs plots

output and tables
data

View

Figure 9.14: Flow diagram of Qucs simulator stages including SPICE

preprocessing.

• Algebraic statements are enclosed in quotes ‘ ‘20 .

• Inline comments start with the $ symbol and continue to the end of a line.

9.7 Circuit template models

When modelling devices or circuits for simulation a particularly productive ap-
proach is the use of a universal template that can be employed to generate models
for devices of the same type but with different characteristics. By simply changing
the parameters embedded in a universal template a new device model is generated
when the netlist code is passed through the SPICEPP preprocessor. Consider the
20
The ‘ character can be found on the most left key on the row of numerical keys (‘ 1 2 3 4 5 6
7 8 9 0 - .......) - this is the case on my keyboard.

305
SPICE template model shown in Fig. 9.15. This represents a simple modular AC
macromodel21 for an OP AMP. OP AMP internal pins are given by integers and
external pins by names in SPICE 3 format. The parameters for a UA741 OP AMP
are shown listed at the start of the SPICE preprocessor netlist. These are used
in the calculation of the component values in later sections of the netlist. In all
cases parameters must be defined before they are used in component calculations.
Passing this listing through the SPICEPP preprocessor22 and generating a Qucs
user defined symbol for the UA741 OP AMP results in the Qucs netlist and sym-
bol shown in Figures 9.16 and 9.17. An application of the generated UA741 OP
AMP model is shown in Fig. 9.18. This circuit is a notch filter. In Fig. 9.18 the
band rejection characteristic of the filter are realised by a twin-T RC network.
Figure 9.19 shows the simulated small signal transfer characteristics of this filter.

21
Details of the model derivation can be found in the Qucs Modelling Operational Amplifiers
tutorial, Qucs Web site.
22
The SPICEPP PERL script can be run from a shell using the command spicepp.pl name.pp
> name.cir , where name is the name of the file to be processed.

306
∗
∗ Device pins 1 . input in n , in p
∗ 2 . ou tput out
∗
∗ ua741 OP AMP p a r a m e t e r s
∗
. param v o f f = 0 . 7m
. param i b = 80n
. param i o f f = 20n
. param rd = 2meg
. param cd = 1 . 4 p
. param cmrrdc = 3 1 6 2 2 . 8
. param fcmz = 2 0 0 . 0
. param a o l d c = 199526
. param gbp = 1meg
. param f p 2 = 3meg
. param r o = 7 5 . 0
∗
∗ input stage
∗
v o f f 1 in n 6 ’ v o f f /2 ’
v o f f 2 7 in p ’ v o f f /2 ’
ib1 0 6 ib
ib2 7 0 ib
i o f f 1 7 6 ’ i o f f /2 ’
r1 6 8 ’ rd / 2 ’
r2 7 8 ’ rd / 2 ’
c i n 1 6 7 cd
∗
∗ common−mode z e r o s t a g e
∗
ecm1 12 0 8 0 ’ 1 e6 / cmrrdc ’
rcm1 12 13 1meg
ccm1 12 13 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ 1 e6 ∗ fcmz ) ’
rcm2 13 0 1
∗
∗ d i f f e r e n t i a l and common−mode
∗ s i g n a l summing s t a g e
∗
gmsum1 0 14 7 6 1
gmsum2 0 14 13 0 1
rsum1 14 0 1
∗
∗ voltage gain stage 1
∗
gmp1 0 9 14 0 1
rado 9 0 a o l d c
cp1 9 0 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ gbp ) ’
∗
∗ voltage gain stage 2
∗
gmp2 0 11 9 0 1
rp2 11 0 1
cp2 11 0 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ f p 2 ) ’
∗
∗ out put s t a g e
∗
e o s 1 10 0 11 0 1
r o s 1 10 out r o
∗

Figure 9.15: SPICE template preprocessor netlist for a UA741 AC modular OP

AMP model.

307
. Def : s t o q f i g 1 7 n e t 0 n e t 1 n e t 2
Sub : X1 n e t 0 n e t 1 n e t 2 gnd Type=”s t o q f i g 1 5 c i r ”
. Def : End

. Def : s t o q f i g 1 5 c i r netIN N netOUT n e t I N P ref

R: ROS1 n e t 1 0 netOUT R=”75”
VCVS: EOS1 n e t 1 1 n e t 1 0 ref r e f G=”1”
C : CP2 n e t 1 1 r e f C= ”5 . 3 0 5 8 3 e −08”
R: RP2 n e t 1 1 r e f R=”1”
VCCS:GMP2 n e t 9 r e f net11 r e f G=”1”
C : CP1 n e t 9 r e f C= ”1 . 5 9 1 7 5 e −07”
R:RADO n e t 9 r e f R=”199526”
VCCS:GMP1 n e t 1 4 r e f net9 r e f G=”1”
R:RSUM1 n e t 1 4 r e f R=”1”
VCCS:GMSUM2 n e t 1 3 r e f net14 r e f G=”1”
VCCS:GMSUM1 n e t 7 r e f n e t 1 4 n e t 6 G=”1”
R:RCM2 n e t 1 3 r e f R=”1”
C :CCM1 n e t 1 2 n e t 1 3 C= ”7 . 9 5 8 7 4 e −10”
R:RCM1 n e t 1 2 n e t 1 3 R=”1M”
VCVS:ECM1 n e t 8 n e t 1 2 ref r e f G= ”3 1 . 6 2 2 8 ”
C : CIN1 n e t 6 n e t 7 C= ”1.4 e −12”
R: R2 n e t 7 n e t 8 R=”1 e +06”
R: R1 n e t 6 n e t 8 R=”1 e +06”
I d c : IOFF1 n e t 7 n e t 6 I =”1e −08”
I d c : IB2 n e t 7 r e f I =”8e −08”
I d c : IB1 r e f n e t 6 I =”8e −08”
Vdc : VOFF2 n e t 7 n e t I N P U= ”0 . 0 0 0 3 5 ”
Vdc : VOFF1 netIN N n e t 6 U= ”0 . 0 0 0 3 5 ”
. Def : End

Figure 9.16: Qucs netlist for a UA741 AC modular OP AMP model.

X1
File=stoq_fig15.cir
IN_N OUT

P_IN_N P_OUT
-
spice
IN_P +
P_IN_P Ref SUB1

Figure 9.17: Qucs symbol for a UA741 AC modular OP AMP model.

308
 dc simulation ac simulation
DC1 AC1
Type=lin
Start=10 Hz
Stop=101 Hz
Equation Points=200
Eqn1
gain_dB=dB(vout.v)
phase_deg=phase(vout.v)
R1
R=100k
vin

V1
U=1 V
C4
C=0.175u
C3
C=0.175u - vout

+
R4 R3
R=20k R=22k SUB1
R6
R=15k
C2 R2
C=0.45u R=100

R5 C1
R=6.8k C=2.2u

Figure 9.18: A twin-T notch filter circuit.

15
vout.v

10

5
10 100
acfrequency

24

22
gain_dB

20

18

16
10 100
acfrequency

50
phase_deg

10 100
acfrequency

Figure 9.19: Small signal transfer characteristics for a twin-T notch filter circuit.

309
9.8 Building circuit design equations into netlists
Figure 9.20 illustrates a bandpass filter that has a bandwidth which is small com-
pared to it’s center frequency. The circuit is often referred to as the Dalyiannis-
Friend filter after its developers. The filter center frequency f0 , voltage gain mag-
nitude H0 , bandwidth B and Q factor are given by the following equations:

1
• f0 = p , where C = C1 = C2
2πC (R1 kR2 )R3

R3
• H0 =
2R1
1
• B=
πR3 C
r
f0 1 R3
• Q= =
B 2 R1 kR2

When designing a filter for a specific specification, for example say f0 = 1kHz,
B = 200Hz and H0 = 10, values for the filter resistor and capacitor values need
to be calculated. This can, of course, be done manually. However, this process is
often tedious, especially if a number of filters need to be designed each with differ-
ent specifications. Circuit simulators are by their very nature primarily designed
to analyse and simulate the performance of circuits who’s component values are
known. As such they are tools for analysis rather than design. In practice, of
course, engineers employ circuit simulators to check their circuit designs. Qucs is
attempting to bridge the gap between design and analysis by using add-on soft-
ware components for designing circuits with well understood structures and design
procedures23 .
In the previous section it was shown that the SPICEPP preprocessor could be
used to calculate model component values. By a simple extension of this concept
it is also possible to embed design equations into a netlist. Shown in Fig. 9.21 is a
SPICEPP netlist for the Dalyiannis-Friend filter. The UA741 OP AMP is modelled
with a SPICE subcircuit called opamp_ac and has its own set of parameters24 .
The first set of design parameters represent the filter specification and are used
in the SPICEPP conversion process to calculate the filter resistor and capacitor
component values. Note also the use of inline comments for documenting the
23
The Qucs Tools drop-down menu lists the currently available design functions that have been
implemented with release of Qucs you are using.
24
These are defined within a subcircuit and should have names unique to the subcircuit model
being defined.

310
 C2
R3

OP1
Vin
Vout

R1 C1

R2

Figure 9.20: The Dalyiannis-Friend bandpass filter circuit.

netlist code. Figures. 9.22 and 9.23 show a basic filter test circuit and the resulting
simulation transfer functions. Hence, not only can the SPICEPP preprocessor be
used for setting up device models but it can also aid the design of entire circuit
blocks provided design equations are available for a given circuit configuration. By
combining SPICEPP with Qucs a very significant design/analysis tool becomes
available opening up new possibilities for Qucs users.

311
∗ D e l y i a n n i s F r i e n d Bandpass f i l t e r d e s i g n
∗ Design parameters
. param f c = 2000.0 $ F i l t e r c e n t e r f r e q u e n c y ( Hz )
. param bw = 2 0 0 . 0 $ F i l t e r bandwidth ( Hz )
. param q = 10.0 $ F i l t e r q f a c t o r = f 0 /bw
. param r 3 i v = 200 k $ Assumed v a l u e f o r r f 3
. param h0 = 10.0 $ F i l t e r f 0 g a i n magnitude
∗
∗ F i l t e r c i r c u i t p i n s : i n p u t n1 , outpu t n3
∗
r 3 n3 n4 r 3 i v
c1 n2 n3 ’ q / ( 3 . 1 4 1 2 ∗ f c ∗ r 3 i v ) ’
c2 n2 n4 ’ q / ( 3 . 1 4 1 2 ∗ f c ∗ r 3 i v ) ’
r 1 n1 n2 ’ r 3 i v / ( 2 ∗ h0 ) ’
r 2 n2 0 ’ r 3 i v / ( ( 4 ∗ q∗q ) −(2∗ h0 ) ) ’
x1 0 n4 n3 opamp ac

∗ s u b c i r c u i t p o r t s : i n+ in− out
. s u b c k t opamp ac i n p i n n out
∗
∗ ua741 OP AMP p a r a m e t e r s
. param v o f f = 0 . 7m
. param i b = 80n
. param i o f f = 20n
. param rd = 2meg
. param cd = 1 . 4 p
. param cmrrdc = 3 1 6 2 2 . 8
. param fcmz = 2 0 0 . 0
. param a o l d c = 199526
. param gbp = 1meg
. param f p 2 = 3meg
. param r o = 7 5 . 0
∗ input stage
v o f f 1 in n 6 ’ v o f f /2 ’
v o f f 2 7 in p ’ v o f f /2 ’
ib1 0 6 ib
ib2 7 0 ib
i o f f 1 7 6 ’ i o f f /2 ’
r1 6 8 ’ rd / 2 ’
r2 7 8 ’ rd / 2 ’
c i n 1 6 7 cd
∗ common−mode z e r o s t a g e
ecm1 12 0 8 0 ’ 1 e6 / cmrrdc ’
rcm1 12 13 1meg
ccm1 12 13 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ 1 e6 ∗ fcmz ) ’
rcm2 13 0 1
∗ d i f f e r e n t i a l and common−mode s i g n a l summing s t a g e
gmsum1 0 14 7 6 1
gmsum2 0 14 13 0 1
rsum1 14 0 1
∗ voltage gain stage 1
gmp1 0 9 14 0 1
rado 9 0 a o l d c
cp1 9 0 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ gbp ) ’
∗ voltage gain stage 2
gmp2 0 11 9 0 1
rp2 11 0 1
cp2 11 0 ’ 1 / ( 2 ∗ 3 . 1 4 1 2 ∗ f p 2 ) ’
∗
∗ out put s t a g e
e o s 1 10 0 11 0 1
r o s 1 10 out r o
. ends

Figure 9.21: SPICEPP netlist for the Dalyiannis-Friend filter.

312
 X1
File=df_filter.cir
vin vout
N1 N3
ac simulation spice
V1 Ref
AC1 U=1 V
Type=lin
Start=1000Hz
Stop=3000Hz
Points=200

Equation
Eqn1
phase_deg=phase(vout.v)
gain_dB=dB(vout.v)

Figure 9.22: The Dalyiannis-Friend bandpass filter test circuit.

10 10
vout.v

vout.v

5 5

0 0

1e3 1.5e3 2e3 2.5e3 3e3 1.7e3 1.8e3 1.9e3 2e3 2.1e3 2.2e3
acfrequency acfrequency

200 200
phase_deg

phase_deg

0 0

-200 -200

1e3 1.5e3 2e3 2.5e3 3e3 1.7e3 1.8e3 1.9e3 2e3 2.1e3 2.2e3
acfrequency acfrequency

20
gain_dB

10

1e3 1.5e3 2e3 2.5e3 3e3

acfrequency

Figure 9.23: Simulated small signal AC transfer functions for the Dalyiannis-Friend
bandpass filter.

313
9.9 Global nodes
In the SPICE 2 and SPICE 3 hardware description languages only the earth node
is global. By convention this is given node name 0 and is assumed by the SPICE
language passer to be earth whenever it occurs in a circuit netlist. When connecting
discreet components with other subcircuit blocks there is often a need for other
nodes to be designated global; the classic example being power supply nodes.
SPICEPP allows nodes to designated as global. These are effectively connected
together to form one net covering both outside and inside subcircuits. The best way
to understand the use of global nodes is to consider an example. Figure 9.11 gives
the SPICE netlist for the two section CMOS ring counter. Many readers would
possibly have noticed that in this netlist both the NAND2 and NOR2 subcircuits
include internal voltage sources25 . This is, of course, not necessary and indeed
inefficient from a simulation point of view. A better approach would be to link
individual gates with a power supply net. The SPICEPP netlist given in Fig. 9.24
illustrates how the .global command can be used to define a global power supply
node. After passing this code through SPICEPP the SPICE netlist printed in
Fig. 9.25 results. Simulation with Qucs gives the same waveforms displayed in
Fig. 9.13.

25
The DC voltage supply for each logic block is generated by a pulse source. This has the effect of
simulating the rising edge of the power supply switch on transient and aids DC convergence.

314
∗ Two s t a g e CMOS r i n g c o u n t e r c i r c u i t .
∗
∗ E x t e r n a l nodes : i n p u t 1 , outpu t 4 , +ve s u p p l y nvcc
∗
∗ g l o b a l node
∗
. g l o b a l nvcc
∗
x1 1 5 6 nand2
x2 1 6 7 nand2
x3 3 6 2 nand2
x4 2 7 3 nand2
x5 1 2 8 nor2
x6 1 8 9 nor2
x7 5 8 4 nor2
x8 4 9 5 nor2
∗
. model modp pmos ( v t o=−1 kp=10u
+ cgdo =0.2n c g s o =0.2n cgbo=2n )
. model modn nmos ( v t o=1 kp=10u
+ cgdo =0.2n c g s o =0.2n cgbo=2n )
∗
. s u b c k t nand2 1 2 3
m1 3 1 nvcc nvcc modp w=40u l =5u
m2 3 2 nvcc nvcc modp w=40u l =5u
m3 5 1 0 0 modn w=20u l =5u
m4 3 2 5 5 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
∗ v c c 4 0 p u l s e ( 0 5 0 1 ns 1 ns 1 2 )
. ends
∗
. s u b c k t nor2 1 2 3
m1 4 1 nvcc nvcc modp w=40u l =5u
m2 3 2 4 4 modp w=40u l =5u
m3 3 2 0 0 modn w=20u l =5u
m4 3 1 0 0 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
∗ v c c 7 0 p u l s e ( 0 5 0 1 ns 1 ns 1 2 )
. ends

Figure 9.24: SPICEPP netlist for a two section CMOS ring counter with global
power supply net node nvcc.

315
∗ Two s t a g e CMOS r i n g c o u n t e r c i r c u i t .
x1 1 5 6 nvcc nand2
x2 1 6 7 nvcc nand2
x3 3 6 2 nvcc nand2
x4 2 7 3 nvcc nand2
x5 1 2 8 nvcc nor2
x6 1 8 9 nvcc nor2
x7 5 8 4 nvcc nor2
x8 4 9 5 nvcc nor2
. model modp pmos v t o=−1 kp=10u cgdo =0.2n c g s o =0.2n cgbo=2n
. model modn nmos v t o=1 kp=10u cgdo =0.2n c g s o =0.2n cgbo=2n
. s u b c k t nand2 1 2 3 nvcc
m1 3 1 nvcc nvcc modp w=40u l =5u
m2 3 2 nvcc nvcc modp w=40u l =5u
m3 5 1 0 0 modn w=20u l =5u
m4 3 2 5 5 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
. ends
. s u b c k t nor2 1 2 3 nvcc
m1 4 1 nvcc nvcc modp w=40u l =5u
m2 3 2 4 4 modp w=40u l =5u
m3 3 2 0 0 modn w=20u l =5u
m4 3 1 0 0 modn w=20u l =5u
c1 1 0 10p
c2 2 0 10p
. ends

Figure 9.25: SPICE netlist for a two section CMOS ring counter with global power
supply net node nvcc.

316
9.10 End Note
This tutorial note describes how SPICE netlists can be simulated using Qucs. The
text is much more than a basic outline of the processes needed to link SPICE cir-
cuit files to Qucs. While writing this note an attempt has been made to stress the
fact that topics like SPICE/Qucs netlist compatibility and conversion are impor-
tant to the future development of Qucs. So an interesting, and thought provoking
question, is how does Qucs develop next in relation to SPICE and indeed how
best is it to make sure that Qucs users can get the most from all the published
SPICE information and device models? After all there is no point in reinventing
the wheel! Complete compatibility with SPICE will not be possible until all the
basic SPICE 2 and SPICE 3 primitive components are added to Qucs. This will
take time but is happening as the Qucs team develops the package26 . Adding
equations to component calculations is a very much a current active topic in Qucs
development. Recently, Michael Magraf has added parameter passing to the Qucs
GUI. Stefan Jahn will add the necessary simulator routines for handling equa-
tions and parameter passing when time allows. In the long term not only will it
be possible to determine component values using calculations at the simulation
initialisation phase but it will also be possible to allow such components to be
dependent on simulation voltage and current variables. Qucs will then be able to
simulate circuits containing nonlinear voltage and current sources like the SPICE
3 B component. These notes are very much a report on some of the work on Qucs
device modelling I have been doing in recent months. Again if there is enough
interest in this area of Qucs development I will upgrade them in the future. My
thanks to Stefan Jahn for all his encouragement while I have been developing the
material reported in this tutorial note.

26
Michael Magraf has recently added a four terminal transmission line to Qucs. Future testing
will confirm if this is similar to the SPICE T component.

317
10 Biasing a BJT Transistor
10.1 Graphical methods
You can bias a bipolar junction transistor in several ways. Determining the best
method for your application is easy with a graphical technique.
Biasing an active device, such as a bipolar junction transistor (BJT), requires
that you set the dc voltages and currents of the device. To optimize the desired
result, you need various bias values. For instance, the input de-vice for a low-noise
amplifier may have its best noise performance at 50 µA of collector current and a
maximum of 5V of collector-to-emitter voltage, whereas later amplifier stages may
require 20-mA collector current and 18V collector-to-emitter voltage to generate
the necessary ac voltage at the output. When you determine the desired bias
conditions, you also need to make sure they are repeatable–within certain limits–
to ensure consistent performance.

Figure 10.1: Different feed back technics

318
Biasing-technique analysis for BJTs generally progresses in complexity from the
fixed-bias method (see fig 10.1, to the shunt circuit, to the stabilized circuit, .
Studies do not usually cover the shunt-divider and universal circuit. However,
questions still arise about the bias stability of the shunt bias circuit. It is usable
in some noncritical applications, but how inferior is it to the stabilized circuit?
Designers are generally taught that the stabilized circuit is the one to use for
repeatable biasing.

One way to analyze the stability of the various biasing methods is to use sta-
bility factors, which characterize the change in collector current due to changes
in the transistor’s HFE (current gain), ICBO (collector-to-base leakage current),
and VBE . Although these factors are useful, comparing bias circuits and bias-
resistor values requires tedious calculations. A visual presentation that compares
the stability of the various circuits is more useful.
Looking at the equation for IC in Figure 1b, note that much of the change in IC
is due to the differing voltages developed across R1 because of the range of HFE.
This difference leads to a question: If some of the current through R1 is fixed,
would the result be less voltage change across R1 and hence, less change in IC?
This thinking leads to the shunt-divider circuit (Figure 1c). Because VBE changes
little, R2 supplies a relatively fixed component of the current through R1, making
R1 a smaller value than it would be without R2. The equation for the shunt
divider shows that a smaller value of R1 in the denominator causes less change in
IC due to changes in HFE. However, along with RC and R2, R1 shows up in the
numerator as a multiplying factor for VBE.

You can next look at how strongly each of these factors influences IC. Because
you can derive all the circuits in Figure 1 from the universal circuit (Figure 1e)
by making the appropriate resistors either infinite (open circuits) or zero (short
circuits), the same universality is possible for the equations. Considering the circuit
equations and a range of parameters and bias-resistor values, you can produce
graphs in which the Y axis represents the change in IC.
To make valid comparisons of the circuits, you need a common parameter related
to the biasing for the X axis. The ratio of the collector current to the bias current
in R1 works. This ratio is common to the circuits and reflects how stiff the biasing
is. To show realistic conditions, the data also includes temperature effects on VBE
and HFE for a temperature range of 25 to 75◦ C and a 3-to-1 spread in HFE.
For comparison purposes, all the circuits use a 10V supply for VCC at a nominal
collector current of 1 mA, with HFE of 100 and VBE of 0.60V at 25◦ C. Calculating
resistors for 5V VCE and selecting RE to develop 1V at the emitter produces
the results for the graphical technique. The model for temperature effects of the
device is VBE=0.60?0.002?(T(actual) 25◦ C), representing the standard 2-mV/◦ C

319
coefficient for diodes. Calculations from the data sheet of the 2N2222A transistor
produce an average temperature coefficient for HFE of about 0.58% /◦ C, which
you can represent by

HF ET emp = HF EM ax × [1 + (T (actual)?25◦ C)0.0058] (10.1)

Calculating IC for a minimum HF E = 50 at 25◦ C and for the maximum HF E =
150 at 75◦ C yields an HF ET emp of 194 and VBE of 0.50V.
This analysis ignores the effects of ICBO. For the nominal collector current of 1
mA and a maximum temperature of 75◦ C, the contribution of ICBO to IC is a few
percent, at most, for the fixed-bias and shunt-bias circuits in Figures 1a and 1b
and less for the bias circuits of Figures 1c, 1d, and 1e.

10.1.1 Graphical approach shows trade-offs

The results of this analysis appear as a simple visual comparison of the current
stability of the various types of bias circuits (Figure 10.2). Using this figure, you
can select the type of bias circuit and the bias ratios for the necessary stability.

Figure 10.2: You can compare the performance of the BJT bias circuit by graphing
the change in collector current vs the ratio of the collector current to
the current in R1.

The horizontal axis is the ratio of the collector current, IC, to the current in resistor
R1. This bias ratio applies to all the circuits and indicates how much current is in

320
the base-biasing network compared with the collector current. Thus, a ratio of 1
indicates a stiff bias circuit, with as much current in R1 of the bias network as in
the collector, whereas a ratio of 50 indicates that the collector current is 50 times
the current in R1 of the bias network. Because some of the results are unexpected,
they give renewed consideration to some of the bias circuits previously ignored.

Figure 10.3: To eliminate the ac effects of feedback, split R1, and bypass the center
to ground.

The universal-bias method is obviously the best of the group. The price you pay for
its dc stability is the reduction in ac input resistance due to the negative feedback
on R1, a sort of Miller effect on resistors. R1 reduces by a factor of the voltage gain
plus 1. This feedback may improve distortion and bandwidth as well as reduce the
output impedance at the collector. If you don’t want these ac effects of feedback,
you can eliminate them by splitting R1 into two parts and bypassing the center
to ground (Figure 10.3). You can improve performance of this circuit at any bias
ratio by increasing the voltage drop across RE, increasing the voltage drop across
the collector resistor, or both.

The stabilized circuit has good stability to bias ratios as high as about 12. Above
this ratio, its stability rapidly decreases. The stabilized circuit relies on the voltage
changes fed back by the emitter current through RE, compared with the voltage,
VB, at the base. When the bias ratio becomes less stiff, changes in base current
flowing through R1 due to changes in HFE cause significant variations in VB.
These variations result in changes in IE and IC. As with the universal circuit, you
can improve performance of the stabilized circuit at any bias ratio by increasing
the voltage drop across RE. Keep in mind that these results are for a nominal
HFE range of 50 to 150 plus temperature effects. Lower minimum values of HFE

321
require stiffer bias ratios for the same performance.

The superior performance of the shunt-divider circuit at bias ratios greater than
12, compared with that of the stabilized circuit, is a surprise. When the shunt-
divider circuit’s bias is stiff, VC is strongly influenced by the ratio of R1 to R2
times VBE. As VBE changes because of temperature, VC and, thus, IC, change
approximately as the ratio of R1 to R2 times VBE changes.

Because IC plays the major role in determining VC, IC experiences wide varia-
tions for these stiff biasing ratios. As the ratio becomes less stiff, the changes
in VBE with temperature, multiplied by the voltage-divider action, become less
dominant, and performance improves until, at the ratio of about 12, the shunt
divider’s stability starts to surpass that of the stabilized circuit. You can account
for this performance by the negative feedback from the collector resistor through
R1. Because the collector resistor is usually much larger than the emitter resistor
of the stabilized circuit, the stability of the universal circuit holds up better for
less stiff bias ratios.

Because the shunt-divider circuit is more stable than the shunt circuit, consider
the divider circuit for applications that need less stability than the stabilized or
universal circuits offer. Because it saves the cost of the emitter-bypass capacitor
necessary in the universal and stabilized circuits, the shunt divider can be more
cost-effective. Negative feedback through R1 in the shunt-divider circuit reduces
the input resistance and may improve distortion and bandwidth, as well as reduce
the output impedance in the same manner as in the universal circuit. Again,
you can negate these effects with a bypass capacitor in the center of R1. This
bypass capacitor is typically much smaller than the emitter-bypass capacitor for
the stabilized circuit.

Because the bias current for the shunt-bias circuit consists of only the base current,
it has only one ratio of IC to IR1, namely HFE, and is plotted as a single point.
As the bias ratio for the universal and shunt-divider circuits increases, the value of
R2 increases until it becomes infinite at an HFE of 100. Under these conditions,
the circuits’ bias ratios converge with the shunt-circuit ratio.

Figure 10.2 leads you to several general conclusions. The universal circuit has the
best stability over the widest range of bias ratios. The stabilized circuit has good
stability for stiff bias ratios, but you should take care if biasing ratios exceed 12.
And, finally, the shunt-divider circuit is a significant improvement over the shunt
circuit and is better than the stabilized circuit for large bias ratios.

322
10.2 Simulation technics
The previous section deals with a graphical method, but a more common method
can be to use the simulators to determine all the possible variation for a given
schematic ( include hF E , Temperature, Voltage regulation, and so on ... ) ; so
the problem is more waht kind of feedback I can use or not. Sorry but there is no
striaght ansyert since this could a cost issu e for example, or a performance issue1 .

Anyway we need to evaluate the different biasing technics using the simulation
tool. One analysis will be done in the PA design chapter.

1
This point is obviously not understood in the same way when discussing with marketing or
development or research teams, who knows why ?

323
11 BJT Modeling and Verification
warning
This chapter will describe an RF design issue using QUCS. The author assume
that the basic manipulation of qucs is known. You will find herein mainly a Ma-
cOsX description that is close to a linux or unices architecture.

11.1 choice of transistor

The choice has been made to choose among the Philips RF wideband transistor
library. These components are easy to find, with resonnable prices.
This list could be found at http://www.semiconductors.philips.com/.
A resume of these transistors can be found in the figure 11.1
I will not discuss herein, the reason 1 why of the final choice, but the BF G425w
is the candidate. It offers high gain, with low figure noise ( if LNA consideration
) high transistion frequency ( 25 GHz ), its emitter is thermal lead, low feedback
capacitance. This device could be used in RF front end, analog or digital cellular,
radar detectors, pagers, SATV, oscillators. It is in a SOT343R package suitable
for small integration.
The maximum acheivable gain is 20 dB with 25 mA, Vce = 2 V at 2 GHz and
25◦ C. The third order intercept point in these conditions is typically 22dBm.
These parameter should be compatible with our need. Here are the spice parameter
of the device.

.SUBCKT BFG425W 1 2 3
L1 2 5 1.1E-09
L2 1 4 1.1E-09
L3 3 6 0.25E-09
Ccb 4 5 2.0E-15
Cbe 5 6 80.0E-15
Cce 4 6 80.0E-15
Cbpb 5 7 1.45E-13
1
regarding current, Ft , Vce , power dissipation, etc . . .

324
Figure 11.1: transistor table from philips semiconductor

325
Cbpc 4 8 1.45E-13
Rsb1 6 7 25
Rsb2 6 8 19
Q1 4 5 6 6 NPN

.MODEL NPN NPN

+ IS = 4.717E-17 + BF = 145 + NF = 0.9934
+ VAF = 31.12 + IKF = 0.304 + ISE = 3.002E-13
+ NE = 3 + BR = 11.37 + NR = 0.985
+ VAR = 1.874 + IKR = 0.121 + ISC = 4.848E-16
+ NC = 1.546 + RB = 14.41 + IRB = 0
+ RBM = 6.175 + RE = 0.1779 + RC = 1.780
+ CJE = 3.109E-13 + VJE = 0.9 + MJE = 0.3456
+ CJC = 1.377E-13 + VJC = 0.5569 + MJC = 0.2079
+ CJS = 6.675E-13 + VJS = 0.4183 + MJS = 0.2391
+ XCJC = 0.5 + TR = 0.0 + TF = 4.122E-12
+ XTF = 68.2 + VTF = 2.004 + ITF = 1.525
+ PTF = 0 + FC = 0.5501 + EG = 1.11
+ XTI = 3 + XTB = 1.5
.ENDS

Since the model used in SPICE and in QUCS rely on a gummel-poon modelisation,
and since the level of modelisation is the same, some quite direct conversion could
be used to create the library for QUCS.
To use directly this file, you will need to store the file in an other directory from
the project one ( a small bug taken into account ). Then it should work but some
there are still some issues on the parameters itselves, This is the reason why we
will proceed in an other way.
The data sheet could be found on the philips web site.

326
Figure 11.2: spice parameter extract from philips data sheet

327
11.2 library creation
Remember that when creating a device, it is almost always mandatory to read of
have a look at on how the model is done is the technical documentation. It is very
to understand the limitation, and how we can correct some data if needed. The
mian pity is that a lot of commercial software are quite obscure on the real model
they use and their limitation ; QUCS is quite exceptionnal on this point this the
complete modeling is explain theoretically in a special technical paper.

In order to conduct these test, we need to create a model of our component. To

perform this you should create the file that contain all the libraries, this file is
stored under

/usr/local/share/qucs/library/philips_RF_widebande_npn.lib

You can edit this file with vi. You need to add the following line :

<Qucs Library 0.0.7 "philips RF wideBand">

<Component BFG425W>
<Description>
RF wideband NPN 25GHz
2V, 25mA, 20dB , 2000MHz
Manufacturer: Philips Inc.
NPN complement: BFG425W
--------------------------
based on spice parameter from philips
--------------------------
sept 2005 thierry
</Description>
<Model>
<_BJT T_BFG425W_ 1 480 280 8 -26 0 0 "npn" 1 "47.17e-10"
1 "1" 1 "1" 1 "0.304" 1 "0.121" 1 "31.12" 1 "1.874" 0
"300.2e-15" 1 "3" 1 "484.8e-10" 1 "1.546" 1 "145" 1 "11.37"
1 "6.175" 1 "0" 1 "1.78" 1 "0177.9e-3" 1 "014.41" 1 "310.9e-15"
1 "0.900" 1 "0.346" 1 "137.7e-15" 1 "0.5569" 1 "0.207" 1 "0.500"
1 "667.5e-15" 1 "0.4183" 1 "0.239" 1 "0.550" 1 "4.122e-12" 1
"68.2" 1 "2.004" 1 "1.525" 1 "0.0" 1 "26.85" 1 "0.0" 0 "1.0" 0
"1.0" 0 "0.0" 0 "1.0" 0 "1.0" 0 "0.0" 0>
</Model>
</Component>

328
You can replace the 1 by 0, this will remove the visible checkbox, the fact to place
a 1 first enable the user to change and or view the parameters that are being used.
A trick to provide all the required syntax is to fill a NPN into the schematics,
perform a copy on the device, you should then have the model in the clipboard,
just paste into to file and add the description and the markup language boundaries.
The syntaxe is explained in the help at the topic description of the qucs file formats.
Then the device is visible in the Component Library Tool as mentionned in figure
11.3.

Figure 11.3: QUCS Component Library showing the new component

By doing this you haved the possibility to reuse the device as much as you want,
and you can debug devices in a more easy way.
Warning : in this section we have only describe the die of the device, for the
parasitic from the package, we will be obliged to describe this circuit, but later on.

11.3 device library verification

The first step, before using the device in a application, is to verify the model you
use. Especially since this model has been created by the user. In order to proceed,
you need to rely on exact data : that is to say the official datasheet.
it this step, you will need to create a project especially for the device verification.
It is good to keep a trace of the device verification, since you could have different
use of this device, so it is good to be able to redo some simulation around the
model itself.
The created project should look that the figure 11.4.

329
project name : model_verif_bfg425w
project location : $HOME/.qucs/

For the validation we will need to use a specific bias of the device : Ic should be
25mA, therefore Ib should be 300µA

330
Figure 11.4: QUCS project for model verification

331
 Figure 11.5: DC validation and temperature

332
11.4 parasitic description of the package
In order to simulate properly the device, you need to used the correct package,
that is to say the SOT 343R in our case, as mentionned on the philips web site (
see fig. 11.6).
Eventhough the device has two emitter, the model used has only one emitter. The
parasitic of this model are shoyn in the spice netlist described in the choice of
the transistor and reproduced in a schematic (see fig. 11.8). These parameter
are always critical to extract, either you have the knowledge to do it or then you
should rely on the piece of information given by the device manucfacturer. It is
also very difficult to figure out what have to be changed in such description of the
device. Some fitting have been performed using 3D electromagnetic software in
the time domain based on MOM methods to verify these parameters.
PhilipsÕ fifth generation double poly silicon wideband technology uses a steep
emitter doped profile resulting in transition frequencies over 20 GHz, and with
poly base contacts a low base resistance is obtained. Via the buried layer, the
collector contact is brought out at the top of the die. The substrate is connected
directly to the emitter package lead, resulting in improved thermal performance (
see fig 11.7).
From this schematics you can edit the symbol that could be used in the next
simulation file. To proceed type F 3 or edit circuit symbol from the file menu.
Simply drw a npn transistor and come back to the schematic by re-pressing F 3.

333
 Figure 11.6: SOT 343R package description

334
Figure 11.7: die connection if the fifth generation transistor from philips

335
 Figure 11.8: bf g425W in sot343R package description

336
11.5 small signal S parameter verification
In this section we will need to redraw a new schematics using the model we have
created, plus some extra components to place the measurements ports 2 .
You should have a schematics like the one mentionned in fig11.9.

Figure 11.9: schematics used for S parameters simulation

The components used to verify the model could be strange ( inductor of 1H and
capacitor of 1F ) It is normal since we need to have a very wide band response on
the circuit, and since we want to caracterize only the active device, and compare
with the datasheet. An other way is to use DC bloc or DC feed or bias Tee to
provide the power supply to the component. This is the right way to do it.
you should then create a display to visualize the S parameters : generally s11 and
s22 are in the smith and s12 and s21 are in polar
We have now to compare these results with the measured parameters from philips
:

! Filename: 225bfg425.001
! BFG425W Field C1
! V1=8.667E-001V,V2=2.000E+000V, I1=3.585E-004A, I2=2.496E-002A
! S11 S21 S12 S22
!Freq(GHz) Mag Ang Mag Ang Mag Ang Mag Ang
2
We will another method when we will use the device in a real project

337
 Figure 11.10: S parameters simulation for model verification

338
# GHz S MA R 50
0.040 0.325 -8.696 38.472 173.381 0.002 71.865 0.923 -3.072
0.100 0.331 -23.004 37.457 164.549 0.005 83.280 0.915 -9.551
0.200 0.315 -44.455 34.771 150.487 0.008 75.947 0.863 -18.965
0.300 0.296 -63.008 31.364 138.811 0.012 71.608 0.794 -26.449
0.400 0.278 -79.654 27.951 128.829 0.015 68.186 0.725 -32.076
0.500 0.265 -94.339 24.856 120.248 0.017 65.974 0.664 -36.332
0.600 0.254 -106.508 22.159 113.362 0.020 64.514 0.613 -39.533
0.700 0.246 -116.820 19.885 107.530 0.022 63.362 0.569 -42.071
0.800 0.240 -126.472 17.964 102.255 0.024 62.701 0.533 -44.121
0.900 0.235 -134.500 16.345 97.645 0.027 61.910 0.504 -45.968
1.000 0.232 -141.743 14.958 93.487 0.029 61.280 0.479 -47.614
1.100 0.230 -148.265 13.770 89.661 0.031 60.570 0.457 -49.172
1.200 0.230 -154.216 12.748 86.091 0.033 59.878 0.438 -50.696
1.300 0.230 -159.761 11.850 82.773 0.036 59.238 0.421 -52.103
1.400 0.231 -164.776 11.070 79.671 0.038 58.509 0.406 -53.483
1.500 0.233 -169.782 10.383 76.687 0.040 57.719 0.392 -54.842
1.600 0.234 -174.382 9.766 73.821 0.043 56.846 0.380 -56.285
1.700 0.236 -178.496 9.213 71.086 0.045 56.001 0.369 -57.740
1.800 0.238 177.334 8.725 68.404 0.047 54.999 0.358 -59.199
1.900 0.241 173.487 8.277 65.836 0.050 53.983 0.348 -60.790
2.000 0.244 169.856 7.874 63.295 0.052 52.923 0.338 -62.399
2.200 0.251 162.836 7.172 58.413 0.057 50.729 0.319 -65.657
2.400 0.259 156.208 6.578 53.682 0.062 48.414 0.301 -68.988
2.600 0.268 150.081 6.068 49.042 0.067 45.958 0.283 -72.558
2.800 0.277 144.221 5.628 44.575 0.072 43.380 0.266 -76.167
3.000 0.288 138.650 5.244 40.174 0.077 40.713 0.248 -80.054
3.500 0.319 125.843 4.470 29.452 0.090 33.634 0.204 -90.648
4.000 0.352 113.999 3.873 18.944 0.102 26.177 0.158 -103.541
4.500 0.389 103.406 3.406 8.713 0.113 18.415 0.113 -121.590
5.000 0.431 92.903 3.011 -1.792 0.123 9.782 0.071 -156.899
5.500 0.463 82.559 2.658 -11.364 0.131 2.534 0.054 148.652
6.000 0.506 73.164 2.374 -21.684 0.138 -6.413 0.095 100.575
6.500 0.516 66.705 2.179 -28.681 0.152 -10.089 0.112 92.309
7.000 0.551 59.664 2.011 -37.894 0.164 -17.920 0.164 82.321
7.500 0.610 50.773 1.808 -49.313 0.166 -29.630 0.246 65.957
8.000 0.644 43.502 1.653 -58.585 0.172 -37.580 0.300 56.971
8.500 0.683 35.816 1.496 -68.478 0.175 -46.984 0.361 47.167
9.000 0.709 27.972 1.338 -77.310 0.173 -55.176 0.412 37.289
9.500 0.736 20.858 1.212 -85.841 0.172 -63.448 0.449 29.117

339
 10.000 0.764 14.187 1.105 -95.600 0.173 -72.751 0.505 22.602
10.500 0.785 7.330 0.997 -104.961 0.171 -81.774 0.554 14.956
11.000 0.802 0.219 0.884 -113.744 0.164 -91.275 0.593 6.422
11.500 0.815 -6.751 0.791 -122.965 0.158 -100.952 0.631 -0.521
12.000 0.822 -13.843 0.690 -131.882 0.149 -111.108 0.667 -8.548
! DEEMBEDDED NOISE DATA
!FREQUENCY FMIN GAMMA OPT Rn
! (GHz) (dB) Mag Ang (NORMALIZED)

Using these parameter, we shoul compare on the sample display the modelised
results and the measurements results, or directly show the error using equations.
First we compare the results.

Figure 11.11: schematics used for S parameters from manufacturer

In the display that is used for the S parameters that we have simulated from
our modelisation, you can add the results from the meaurements files by adding
a measurement of Si,j using the right dataset with the combo box. You should
obtain the difference between the two.
By doing this, you should obtain the results presented in the figure 11.12.

IMPORTANT NOTE : The differences, you should obtain are still on investi-
gation for now.

340
Figure 11.12: Results from model and from meaures compared together

341
12 Power Amplifier Design
warning
This chapter will describe an RF design issue using QUCS. The author assume
that the basic manipulation of qucs is known. You will find herein mainly a Ma-
cOsX description that is close to a linux or unices architecture.

12.1 Field of interest

This power amplifier will be used in a more complex system taht I can not describe
herein, but the application is inside the 868M Hz ISM frequency band. This am-
plifier is considered as power amplifier since it is not a LNA, but its power is not
very high as well as you can see in the following system specification. It is more a
low input power amplifier driving relatively low current. An application note with
really high power level such several watts will be an other chapter.

12.2 System consideration

As a system point of view we need first to specify what kind of function we need.
this function will be defined as mentionned in table 12.1.

Table 12.1: System specification for the design of a PA

parameter description min typ max unit

Fo frequency of operation 863 868.6 870 MHz
Icc current consumption 20 25 mA
Zin input impedance 50 Ω
Zout output impedance 50 Ω
Pin input power −15 −10 −8 dBm
Pout input power 5 10 12 dBm
Vcc DC supply voltage 2.45 2.5 2.55 V

342
Cost issue is very important, therefore only one active component is allowed, and
the BOM 1 should be reduced as much as possible.

This design should work on a FR4 PCB used in a production line. The parameters
of such substrate is quite uncontrolled but can be caracterized, as long as you keep
the same supplier ( avoid strange suppliers who can change the FR4 composition
without notice ).
As mentionned previously you can describe a substrate inside the library with the
following lines :

<SUBST FR4_ 1 0 0 -30 24 0 0

"4.7" 1 "0.7 mm" 1 "35 um" 1 "2e-4" 1 "0.022e-6" 1 "0.15e-6" 1
>

The height of the substrate is 0.7mm but this describe only one RF layer of the
full implementation of the circuit which is a four layour board. The two inner layer
are power and ground, the top and bottom layer are RF layers.

12.3 Biasing consideration

In this section we will see how the biasing is made, especially using a emitter feed
back technic. If you remember well the data sheet of the transistor, there is a huge
dispersion on the hF E , and some other dispersion have to be taken into account :
resistance, supply voltage, . . . .
The used schematics is shown is fig 12.1. But we need to evaluate the component
first. Using small calculus it is easy to figure out the different resistance :
assuming that

Ic = βIb (12.1)

IbiasBridge  Ib (12.2)

Ic
IbiasBridge = (12.3)
10
Vcc − Vce
Re = (12.4)
Ic
10 × Vcc
R1 + R2 = (12.5)
Ic
1
Bill Of Material

343
 Figure 12.1: Schematics used for this study

344
 10
R2 = × (Vcc − Vce + Vbe ) (12.6)
Ic
The inputs are :
• Vcc = 2.5V
• Vbe = 0.412V
• Ic = 15mA
the results are :
• R1 = 1KΩ
• R2 = 600Ω
• Re = 33Ω
Using these values on the schematics, we can now see the stability of the design.
Adding the fact that the voltage regulator used in this case has an ondulation of 5
mV in the working domain. You need to simulate the DC schematics by modifying
the BF parameter of the transistor from 50 to 120 ( since this feature is not enabled
in the current version of Qucs 0.0.7 ).

Table 12.2: Variation of Ic in mA, due to the Vcc and β

Vcc vs β 50 80 120
2.45 12.21 13.34 14.07
2.50 12.62 13.78 14.54
2.55 13.03 14.23 15.01

From this table we can extract some stability factor :

∆Icc
|β=80 = 8.9µA/mV (12.7)
∆V
∆Icc
|V =2.5 = 30µA (12.8)
∆β cc
∆Icc
|β=seenote,Vcc =2.5 = . . . µA/C (12.9)
∆T
Note : For the temperature dependance, we need to take the minimum β for the
minimum temperature, and the maximum β for the maximum temperature.

345
12.4 Why thermal design ?
The objective of the thermal design in electronic equipment is to provide as low
a temperature rise, ∆T, above ambiant as is practical for a product’s electronic
components.
As a practical matter, a small 3C to 5C component temperature rise is almost
unavoidable, and actually has been found to be desirable. If the rise is less than
that, there can be more moistrure-related problems, particularly corrosion and
electrical leakage currents.

• Improves performance : avoids calibration drift, maintains phase lock loops,

stabilizes gain, ...

• Improves reliability : failure mechanisms accelerate rapidly at higher tem-

peratures through metal migration, increased ion mobility, ...
In most electronic components, the failure rate doubles for a 10C to 15C rise
in temperature and the slope is exponential ! temperature cycling is even
worse.
Temperature rise is particularly hard on components which depend on an
internal liquid, such as electrolytic capacitor, batteries, and lubricated bear-
ings.
Sophisticated thermal design is becoming a necessity as devices becomes
smaller and poxer density increase. Examples : VLSICs and surface mount
technology SMT.

• Improves life : higher ∆T increases mechanical stress, failures of connections,

metalisation contacts,...

12.4.1 Thermal management

The objective of thermal management is to design the internal thermal environ-
ment of the electronic equipment so the equipment performance will meet customer
expectations. Within the range of environmental conditions where the equipment
is expected to operate, the equipment should perform within specifications and
operate reliably. In general, the designer has little control over the external envi-
ronment, so he must design for an anticipated range. He does have more control
over the internal environment, but his attention should be directed toward the
ultimate goal ; maintaining a suitable environment for the critical components.
Analysis of the thermal environment can usually be divided into several distinct
parts because of almost–isothermal boundaries. Consider the typical enclosure
system, the isothermal boundaries are :

346
 • the enclosure at Te

• the interior at Tb

• the component at Tc

Because of these boundaries, ∆Tjc , ∆Tca and ∆Tja can be solved independently.
∆Tae and ∆Te∞ can also be solved independently for a sealed enclosure, but are
inter–dependent for a vented or forced air cooled enclosure.

approching the problem During the definition stage of a product, the choice of
enclosure is sometimes dictated by a competitor, the customer, or marketing. Fre-
quently the choice is ”as small as possible”, thus unwittingly passing judgment on
a particular choice, it is possible to make a thermal analysis of the proposed enclo-
sure. If the environment created for the component is unsuitable, then additional
cooling mechanisms must be developped.
One approch is to simplify the problem to one dimensionnal analysis. Heat energy
sources azre assumed to be evenly distributed throughout the volume. The enclo-
sure surface is assumed to be isothermal. The enclosure is assuemd to made of a
perfect thermal conductor. ( unfortunately, enclosures are more and more being
made of plastic, a thermal insulator, which complicates this sample approch).
The external environment is considered to be the walls of a large room of surface
emissivity ,  , of 1.0 at the same temperature, T∞ , as the surrounding air, and is
capable of absorbing an infinite amount of heat energy.
Heat transfert by conduction, radiation, free convection, venting, and forced con-
vection are basically representated by the equation :

Qt = Qk + Qr + Qc + Qv + Qf (12.10)
The most elusive component, thermal resistance Θx , can vary from simple to very
complex. Fortunately, most electronic enclosures do not have more than three
cooling paths and in many cases, the third path is minor one that can be neglected
for ease of calculation.
The following are some generally accepted guidelines that can be used to quickly
evaluate a design or configuration. These were obtained from notes provided by
[?].
Maximum power density :

• for small painted uniformly heated sealed enclosure

– naturally cooled < 4mW/cm3
– taller than 60cm < 2mW/cm3

347
 • for naturally cooled printed circuit boards < 16mW/cm2

• for forced air cooled printed circuit board < 110mW/cm2

• for small ( 60cm or less ) induced draft cooled enclosure < 20mW/cm3

forced air velocities :

• for PCB cages > 4m/sec

• for enclosures < 7.6m/sec

12.5 DC Power dissipation

An important issue in power amplifier design is the power dissipation. Even if in
this particular case the power dissipation is not that obvious, it is nice to see how
we can handle this anyway.

As a student you always learn that you can apply kirchoff law on temperature.
This only thing you have to know is the correspondance :

The temperature : is equivalent to the voltage

The power : is equivalent to the current

The thermal resistance : is equivalent to the resistance

You can also take into account some calorific capacity, and perturbation from near
effect due to the presence of other source of heating, in a dynamic design, but we
will only see the DC power dissipation here . . . from this start point you can then
imagine whatever you want.

In order to proceed, we need to create a model for this power dissipation. This
model can be very simple on its comprehension but very complex since all the
parameters are not well known. Therefore we will need to reduce the level of
modelisation that is used.
Here are the input parameters :

• The DC power dissipation is 15mA × 2.5V olts = 37.5mW

• the thermal resistance of the device is θjunctions older = 350degC/W

• the thermal resistance of the ambiante is θthpcba ir = 22degC/W

348
 • the ambiante temperature varies from −25degC to 75degC and 25degC typ-
ical

The schematics used for this simulation is shown is figure 12.22 .

Figure 12.2: Schematics used to simulate the DC power dissipation

12.6 Small signal analysis

The current version of QUCS do not include an Harmonic Balance solver, so we
need to do some other simualtions in order to have some ideas on the performances
of our design.

2
Note the possiblity to place the results of the simulation directly on the schematics, and some
comments on the schematics such as document name, revision, and so on.

349
13 Low Noise Amplifier Design
This section will describe a two stage LNA. The main goal is to see how we can
design this LNA using the QUCS software, but also to find innovative designs for
low power 1 solutions.

The main difference between as you should know, between PA and LNA, is that
in the design of a LNA the noise factor is crucial, and therefore a trade off has to
be made with the gain design. This design rule is well explained in all RF courses,
so I will go straightforward to the solution by explaining the ”pie” but not the
”recipie” !

As mentionned earlier, a particular attention will be placed on the DC study, since

the overall current consumtion is a crucial point, and the noise factor that we could
have.

13.0.1 System consideration

As a system point of view we need first to specify what kind of function we need.
this function will be defined as mentionned in table 13.1.

Table 13.1: System specification for the design of a LNA

parameter description min typ max unit

Fo frequency of operation 863 868.6 870 MHz
Icc current consumption 0.5 1 mA
Zin input impedance 50 Ω
Zout output impedance 50 Ω
Pin input power −120 −110 −90 dBm
Vcc DC supply voltage 1.4 1.5 1.6 V

note : for the DC supply voltage, we will have to find the correct ripple that is
acceptable on this design in order to be able to specify the voltage regulator and its
1
be careful when I usually use the term low power , I mean extremlly low power , below the
mA generally

350
PSRR regarding the other voltage in the design. To proceed, due to the fact that
some functionnalities are still missing on QUCS2 we will use some workaround for
the DC study.

13.0.2 Choice of transistor

In order to design a LNA, a particular attention has to be put in the this choice.
Therefore you will need to have a transistor that is well designed for very small
current and for LNA application.
I will use the BF G403AW from philips 3 . This transistor belongs to the 5th
generation.
To classify directly the different transistors that could be used, the different version
The parameter are the following :

TO BE UPDATED WITH THE CORRECT ONE

.SUBCKT BFG403W 1 2 3
L1 2 5 1.1E-09
L2 1 4 1.1E-09
L3 3 6 0.25E-09
Ccb 4 5 2.0E-15
Cbe 5 6 80.0E-15
Cce 4 6 80.0E-15
Cbpb 5 7 1.45E-13
Cbpc 4 8 1.45E-13
Rsb1 6 7 25
Rsb2 6 8 19
Q1 4 5 6 6 NPN

.MODEL NPN NPN

bla bla bla
bla bla bla
bla bla bla
bla bla bla
bla bla bla
.ENDS

2
normal it is still in development . . .
3
I do not have any stock option with philips, but they provide quite easily some prototypes and
the models of their transistors, further more their strategy is to continue to provide small
wideband RF transistor, so why not ?

351
In order to perform some simulation we should input this component in the de-
vice library as mentionned in the chapter on the BJT modeling, and create the
schematics thst uses this device. The parasitic element are the same since the
package used is the same as the BF G425W .

13.0.3 library creation

The major problem in this design is the fact that the needed current on the LNA
is not mentioned in the already measured S parameters from the manufacturer.
This is one of the reasons why, we need specicaly a non linear model to describe
the transistor. Of course a preliminary calculus could be done using these regular
parameters, but since we need also some other features such as distortion and so
on, a non linear model is mandatory.

In order to conduct these test, we need to create a model of our component. To

perform this you should create or edit the file that contain all the libraries, this
file is stored under

/usr/local/share/qucs/library/philips_RF_widebande_npn.lib

You can edit this file with vi. You need to add the following line :
<Qucs Library 0.0.7 "philips RF wideBand">
...
...
...
<Component BFG403W>
<Description>
RF wideband NPN 25GHz
2V, 3mA, 20dB , 2000MHz
Manufacturer: Philips Inc.
NPN complement: BFG403W
--------------------------
based on spice parameter from philips
--------------------------
sept 2005 thierry
</Description>
<Model>
<_BJT T_BFG403W_ 1 480 280 8 -26 0 0 "npn" bla bla bla bla>
</Model>
</Component>
...

352
...
...

13.0.4 DC study
13.0.5 SP study
13.0.6 Non linearities study
13.0.7 Possible improvement tips

353
14 Microstrip Design
14.1 10dB Directional Coupler Design
The below pictures shows two parallel conductor strips on a dielectric substrate
with a backplane metalization. Both the conductor strips have the width W , the
height t and the length l. There is a finite gap S between the conductors. The
substrates height is denoted by h. With the gap between the conductor strips
small enough a capacitive as well as inductive coupling occurs.

2 3

t
W

1 S 4

Figure 14.1: microstrip directional coupler

Such a microstrip structure is called “microstrip coupled lines”. Also defined in

figure 14.1 the port numbers 1. . . 4.

14.1.1 Some boring theory beforehand

There are two types of directional couplers: backward (coupling from port 1 to
port 4) and forward (coupling from port 1 to port 3) couplers.

The S-parameters of an ideal directional backward coupler are as follows – with C

354
denoting the coupling coefficient.
√
S21 = 1 − C2
S41 =C
S31 =0
S11 = S22 = S33 = S44 =0

In a three conductor system – as the microstrip coupled lines are – there are two
types of modes: even and odd. Thus such a system is described by odd and even
characteristic impedances (ZL,o and ZL,e ) and odd and even effective dielectric
constants (εr,ef f,o and εr,ef f,e ). The characteristic equations for an ideal backward
coupler are

εr,ef f,e = εr,ef f,o

ZL,e 6= ZL,o

and those for an ideal forward coupler are

6 εr,ef f,o
εr,ef f,e =
ZL,e = ZL,o

The S-parameters of the ideal directional forward coupler are as follows.

√
S21 = 1 − C2
S31 =C
S41 =0
S11 = S22 = S33 = S44 =0

For both ideal – forward and backward – couplers the reflection coefficients are
zero. Port 1 is called the injection port. Port 2 is the transmission port. In
a backward coupler port 4 is the coupled port and port 3 is called the isolated
port. In a forward coupler it’s the other way around.

Please note: The given S-parameters for forward and backward couplers are valid
for all side termination of each port with the reference impedance ZL – usually
50Ω.

355
14.1.2 Design equations
In microwave labs backward line couplers are most wide spread. The basic design
equations can be written as

ZL,e − ZL,o
C=
ZL,e + ZL,o
π
β ·l =
2
ZL2 = ZL,o · ZL,e
r
1+C
ZL,e = ZL ·
1−C
r
1−C
ZL,o = ZL ·
1+C

With
π
β ·l =
2
π π·c c λ
;l= = = =
2·β 2·ω 4·f 4

the length l of such a coupler is defined by a quarter wavelength. Both the char-
acteristic impedances can be computed by the reference impedance ZL , i.e. 50Ω,
and the coupling coefficient C.

14.1.3 Applying the design equations

With the previous definitions it’s easy to design the 10dB directional backward
coupler. We have the reference impedance ZL = 50Ω and the coupling coefficient
C in dB. First we linearize the coupling coefficient.

CdB = −10dB
; C = 10CdB /20 = 10−0.5 ≈ 0.316

Now we compute the even and odd impedances.

r
1+C
ZL,e = ZL · ≈ 69.4Ω
1−C
r
1−C
ZL,o = ZL · ≈ 36.0Ω
1+C

356
14.1.4 What next?
All grey theory you may think... With the impedances at hand the engineer had
to go into magic diagrams and find physical dimensions of his coupler. But now
there is Qucs. Things get easier.

Just select Tools → Line Calculation in the menubar or press Ctrl+3 to start
the transmission line calculator.

Then choose Coupled Microstrip in the Transmission Line Type selection

box. Something likely shown in figure 14.2 should appear.

Figure 14.2: Qucs Transcalc screenshot

Type in the calculated 69.4 in the Z0e field, 36.0 in the Z0o field and 90 in the
Ang l field of the Electrical Parameters panel. The Ang l field denotes the
desired electrical length of the line (remember: 90◦ ' π/2). Choose the Deg unit.

Our selected design frequency is 2GHz. Thus type in this value in the Freq field
of the Component Parameters panel.

Then press the Synthesize button or press F4. The program calculates the
physical parameters W, S and L in the Physical Parameters panel.

357
Please note: Depending on the substrate (shown in the Substrate Parameters
panel) the calculated values may vary.

Finally we got

W = 520µm
S = 199µm
L = 14.93mm

All done with designing... Feel any better?

14.1.5 Verification of the design

Ok. Let’s verify what we have designed so far. Choose Execute → Copy to Clip-
board from the menubar or press F2. This copies the currently shown microstrip
coupled line in Qucs Transcalc into the global clipboard.

Now switch to an empty Qucs schematic and press Ctrl+V. This inserts the
previously entered clipboard content – and click with the left mouse button in
order to place the selection into the schematic. This should give you something
likely shown in figure 14.3.

358
 Figure 14.3: coupled microstrip lines in a Qucs schematic

Now press the equation button (shown in figure 14.4) in Qucs’s toolbar.

359
 Figure 14.4: equation button

Place the equation into the schematic and enter the following equations. Press
Add in the equation dialog (see figure 14.5) to add new equations. Finally press
the OK button.

Figure 14.5: equation dialog

Also edit the properties of the MSTC1 component reducing the number of digits.
This will ensure that your technology is able to use these values when (if) they
decide to produce your design.

Now edit the S-parameter simulation properties. You can do that either by double
clicking the component and use the component dialog. Or you can directly click
on the values in the schematic and fill in 0.2 GHz for Start, 4.2 GHz for Stop
and 101 for Points.

Finally save your schematic by pressing Ctrl+S. Check whether all looks like as
shown in figure 14.6.

360
 P1 P2
Num=1 Num=2
Z=50 Ohm Z=50 Ohm
MSTC1
Subst=SubstTC1
P4 W=0.520 mm
Num=4 L=14.93 mm P3
Z=50 Ohm S=0.199 mm Num=3
Z=50 Ohm
SubstTC1
er=9.8
h=0.635 mm
S parameter t=17.5 um
simulation tand=0.0001
Equation rho=2.43902e-08
SPTC1 D=1.5e-07
Type=lin Eqn1
Start=0.2 GHz reflect=dB(S[1,1])
Stop=4.2 GHz isolated=dB(S[3,1])
Points=101 through=dB(S[2,1])
coupled=dB(S[4,1])

Figure 14.6: final microstrip coupler schematic

Now select Simulation → Simulate from the menubar or just press F2 to sim-
ulate the schematic.

When the simulation windows disappears then choose a Cartesian diagram from
the left hand selection view and place the diagram into the (yet empty) data
display area. Double click the through, reflect, isolated and coupled data
items in order to add it to the diagram within the diagram dialog as shown in
figure 14.7.

361
 Figure 14.7: diagram dialog

Press OK to finish the diagram dialog. Afterwards you will see the following
diagram.

362
 Figure 14.8: microstrip coupler simulation results

14.1.6 Suggested improvements

By use of the diagram dialog (double click the diagram) you may improve1 the
data visualization as you see it fit. I manually fixed the y-axis limits, set markers
and set curve thickness to 2 points. Also I entered a common x-axis label. See
figure 14.9 how it looks now.

1
... to feel even better.

363
 frequency: 2e+09
coupled: -10.3223
10

-10
coupled
isolated
through
reflect

-20

-30

-40

-50
0 1e9 2e9 3e9 4e9
frequency / Hz frequency: 2e+09
reflect: -32.0135

Figure 14.9: directional coupler simulation result diagram

The marker on the coupled curve shows a coupling factor of -10.32 at a frequency
of 2GHz (double click marker to change precision of the marker data). This is a
bit way off for which we tried to design it for.

Seems like coupling between the lines is a bit too weak. So we reduce the gap
between the strip conductors S by 16.5µm to be 0.1825 mm and simulate again.

364
 frequency: 2e+09
coupled: -10.0062
10

-10
coupled
isolated
through
reflect

-20

-30

-40

-50
0 1e9 2e9 3e9 4e9
frequency / Hz frequency: 2e+09
reflect: -31.6542

Figure 14.10: optimized directional coupler simulation result diagram

Finally a perfect2 10dB coupling as shown in figure 14.10.

14.1.7 Remaining thinkabouts

The diagram in figure 14.10 shows a reflection coefficient of about -31.7dB. The
isolation (about -22.2dB) is not as good as planned as well. So – what happened
with my design equations?

Have a look at figure 14.2. In the Calculated Results panel you see ErEff Even
and ErEff Odd differing significantly which is not what we expect from an ideal
backward coupler:
εr,ef f,e = εr,ef f,o
This “problem” arises from the fact that there are two dieletrica involved: air and
the substrate. Part of the electromagnetic fields cross air and part of them the
substrate. You can inhibit this by a dielectric overlay. It’s more expensive to
produce but improves your results.

2
... to feel great.

365
15 Measurement Expressions
Reference Manual

15.1 Introduction

This manual describes the measurement expressions available in ”Qucs”, the ”Quite
Universal Circuit Simulator”.

Measurement expressions come into play whenever the results of a ”Qucs” simula-
tion run need post processing. Examples would be the conversion of a simulated
voltage waveform from volts to dBV, the root mean square value of that waveform
or the determination of the peak voltage. The ”Qucs” measurement functions offer
a rich set of data manipulation tools.

If you are not familiar with the way how to enter those formulas, please refer
to chapter “Using Measurement Expressions”, which points out the possibilities
to create and change measurement expressions. Also the data types supported
are specified here. Chapter “Functions Syntax and Overview” introduces the basic
syntax of functions and a categorical list of all functions available. The core of
the document, a detailed compilation of all ”Qucs” functions divided into differ-
ent categories, is presented in chapter “Math Functions” and chapter “Electronics
Functions”. Finally, the Index contains an alphabetical list of all functions.

15.2 Using Measurement Expressions

The chapter describes the usage of mathematical expressions for post processing
simulation data in “Qucs”, how to enter formulas and modifying them. It gives a
brief description of the overall syntax of those expressions.

366
15.2.1 Entering Measurement Expressions

Measurement expressions generate new datasets by function or operator driven

evaluation of simulation results. Those new datasets are accessible in the data dis-
play tab after simulation. The related equations can be entered into the schematic
editor by the following means:

• Using the equation icon in the “Tools” bar (see fig. 15.1)

• Using menu item “Insert” → ”Insert equation”

Figure 15.1: Entering a new measurement expression via equation icon

You can now place the equation symbol by mouse click anywhere in the schematic.
Each mouse click creates a new equation instance each consisting of a variable
number of measurement expressions. Press the Esc key if you do not like further
equations.

Another option is to select an existing equation, copy it (either by menu item

“Edit” → ”Copy” or by Ctrl + C 1 ) and paste it (either by menu item “Edit” →
”Paste” or by Ctrl + V ).

1
Ctrl + C means that you have to press the Ctrl key and the C key simultaneously.

367
After having successfully created an equation instance, you are now able to modify
it.

15.2.2 Changing Measurement Expressions

For sake of simplicity we assume that you have just generated a new equation - if
you like to change an existing, more complicated equation the following steps are
the same.

Thus, the excerpt of your schematic surface looks like that in fig. 15.2.

Figure 15.2: Newly created equation

You can now manipulate the current name of the equation instance. Simply click
onto “Eqn1”, which becomes highlighted. Then type in a new name for it and
finalise your inputs with the Enter key.

After that, you can enter a new equation. Again, click onto “y=1”. Only the “1”
is marked, and you can enter a new expression there. Please use the variables, op-
erators and constants described in chapter “Syntax of Measurement Expressions”.
Note that you can also refer to results (dependents) of other equations. But how
to change the name of the current dependent “y”? Right click onto the equation,
and a context menu opens. Select the first item called “Edit properties”. A sub
window appears, which should look like the one in fig. 15.3. The alternative for
entering equations is to double click onto the equation.

You can now change the name of the dependent, the equation itself (which is “1”
in the example shown) and the name of the equation. If you do not want the
result to be exported into the data display tab, but temporarily need it for further
calculations, select “no” in the “Export value” cell.

368
 Figure 15.3: Editing equation properties

15.2.3 Syntax of Measurement Expressions

Function names, variable names, and constant names are all case sensitive in mea-
surement expressions - it is distinguished between lowercase and uppercase letters
such as ’a’ and ’A’.

In functions, commas are used to separate arguments.

Variable Names

User defined variable names consist of a letter, followed by any number of letters,
digits, or underscores.

The syntax of variable names created by the ”Qucs” simulator is as specified in

table 15.1. Please note that all voltages and currents in “Qucs” are peak values
except the noise voltages and currents which are RMS values at 1Hz bandwidth.

369
 Variable Name Description
nodename.V DC voltage at node nodename
name.I DC current through circuit component name
nodename.v AC voltage at node nodename
name.i AC current through circuit component name
nodename.vn AC noise voltage at node nodename
name.in AC noise current through circuit component name
nodename.Vt Transient voltage at node nodename
name.It Transient current through circuit component name
name.OP name = component name, OP = operating point (device dependent),
e.g. D1.Id
S[x,y] S-parameter, e.g. S[1,1]
Rn equivalent noise resistance
Sopt optimal reflection coefficient for minimum noise
Fmin minimum noise figure
F noise figure
nodename.Vb Harmonic balance voltage at node nodename

Table 15.1: Syntax of simulator generated variable names

Numbers

Numbers are written in conventional decimal way, with an optional decimal point
between the digits. For powers of ten, the familiar scientific notation with an ’e’
is used. In this way, ’1.234e6’ is an example for the real floating point number
1234000. Imaginary numbers can be entered by a multiplication factor ’i’ or ’j’
(see also table 15.3). An example would be ’1+2*i’ or - if you want to leave out
the multiplication sign - ’1+i2’.
Beside the scientific ’e’ notation the following number suffixes can be used (see
table 15.2):

Vectors and Matrices

You can enter vectors and matrices manually by enclosing columns and rows into
brackets. Columns are separated by commas, rows by semicolons. A valid matrix
entry in a measurement expression would be ’A=[1,2;3,4]’, defining the matrix

370
 Suffix Name Value Suffix Name Value
E exa 1E+18 m milli 1E-3
P peta 1E+15 u micro 1E-6
T tera 1E+12 n nano 1E-9
G giga 1E+9 p pico 1E-12
M mega 1E+6 f femto 1E-15
k kilo 1E+3 a atto 1E-18

Table 15.2: Number Suffixes

 
1 2

A= . The notation ’y=[1,2,3,4]’ configures the vector y = 1 2 3 4 .
3 4
You get access to components of matrices and vectors by writing its name followed
by brackets. Inside of the latter ranges (see table 15.6) or indices, separated by
commas, define the extract you desire. Examples are ’y=M’, accessing the whole
matrix M, ’y=M[2,3]’, extracting the value of the second row and third column of
M, or ’y=M[:,3]’, obtaining the complete third column.

Built-in Constants

The constants which can be used within measurement expressions are given in
table 15.3.

Constant Description Value

e Euler’s constant
√
2.718282
i,j Imaginary unit −1 i1
kB Boltzmann’s constant 1.380658e23 J/K
pi π 3.141593

Table 15.3: Built-in Constants

Operators

Operator Precedence Expressions are evaluated in the standard way, meaning

from left to right, unless there are parentheses. The priority of operators is also
handled familiarly, thus for example multiplication has precedence to addition.

371
Tables 15.4 and 15.5 specify sorted lists of all operators, the topmost having highest
priority. Operators on the same line have the same precedence.

Operator Description Example

() Parentheses, function call y=max(v)
ˆ Exponentiation y=3^4
* Multiplication y=3*4
/ Division y=3/4
% Modulo y=4%3
+ Addition y=3+4
- Subtraction y=3-4
: Range operator y=v[3:12]
+,- Unary plus, unary minus y=+x z=-y

Table 15.4: Arithmetic Operator Priorities

Operator Description Example

() Parentheses a=(x||y)&&z
! Negation z=!x
? : Abbreviation for conditional expression ”if x then y else z” a=x?y:z
&& And z=x&&y
|| Or z=x||y
ˆˆ Exclusive Or z=x^^y
== Equal z=x==y
!= Not equal z=x!=y
< Less than z=x<y
<= Less than or equal z=x<=y
> Larger than z=x>y
>= Larger than or equal z=x>=y

Table 15.5: Logical Operator Priorities

Ranges The general nomenclature of ranges is displayed in table 15.6. It shows

one-dimensional ranges, whereas also n-dimensional ranges are possible, if you
consider nested sweeps.

372
 Syntax Explanation
m:n Range from index m to index n
:n Range up to index n
m: Range starting from index m
: No range limitations

Table 15.6: Range definition

Post Processing of Simulation Data by Expressions

After a simulation has run the results are stored in datasets. Usually, such a
dataset is a vector or a matrix, but may also be a real or complex scalar. For
transient analysis, this dataset contains voltage or current information over time,
for Harmonic Balance it contains amplitudes at dedicated frequencies, while for S-
parameter analysis a vector of matrices (thus matrices in dependency of frequency)
is returned. In further generalisation the components of vectors and matrices
consist of complex numbers.

Additionally, datasets can be generated by using expressions. As an example

the linspace() function shall be named, which creates a vector of linearly spaced
elements.

15.3 Functions Syntax and Overview

This chapter introduces the basic syntax of the function descriptions and contains
a categorical list of all available functions.

15.3.1 Functions Reference Format

”Qucs” provides a rich set of functions, which can be used to generate and display
new datasets by function based evaluation of simulation results. Beside a large
number of mathematical standard functions such as square root (sqrt), exponential
function (exp), absolute value (abs), functions especially useful for calculation and
transformation of electronic values are implemented. Examples for the latter would
be the conversion from Watts to dBm, the generation of noise circles in an amplifier

373
design, or the conversion from S-parameters to Y-parameters.

Functions Reference Format

In the subsequent two chapters, each function is described using the following
structure:

<Function Name>

Outlines briefly the functionality of the function.

Syntax

Defines the general syntax of this function.

Arguments

Name, type, definition range and whether the argument is optional, are tabulated
here. In case of an optional parameter the default value is specified. “Type” is a
list defining the arguments allowed and may contain the following symbols:

Symbol Description
R Real number
C Complex number
Rn Vector consisting of n real elements
Cn Vector consisting of n complex elements
Rm×n Real matrix consisting of m rows and n columns
Cm×n Complex matrix consisting of m rows and n columns
Rm×n×p Vector of p real m × n matrices
Cm×n×p Vector of p complex m × n matrices

“Definition range” specifies the allowed range. Each range is introduced by a

bracket, either “[” or “]”, meaning that the following start value of the range is
either included or excluded. The start value is separated from the end value by a
comma. Then the end value follows, finished by a bracket again, either “[” or “]”.
The first bracket mentioned means “excluding the end value”, the second means
“including”.

374
If a range is given for a complex number, this specifies the real or imaginary value
of that number. If a range is given for a real or complex vector or matrix, this
specifies the real or imaginary value of each element of that vector or matrix. The
symbols mean “includes listed value” and “excludes listed value”.

Description

Gives a more detailed description on what the function does and what it returns.
In case some background knowledge is presented.

Examples

Shows an application of the function by one or several simple examples.

See also

Shows links to related functions. A mouse click onto the desired link leads to an
immediate jump to that function.

15.3.2 Functions Listed by Category

This compilation shows all “Qucs” functions sorted by category (an alphabetical
list is given in the appendix). Please click on the desired function to go to its
detailed description.

Math Functions

Vectors and Matrices: Creation

eye() ... Creates n x n identity matrix

linspace() ... Creates a real vector with linearly spaced components
logspace() ... Creates a real vector with logarithmically spaced components

375
Vectors and Matrices: Basic Matrix Functions

adjoint() ... Adjoint matrix

array() ... Read out single elements
det() ... Determinant of a matrix
inverse() ... Matrix inverse
transpose() ... Matrix transpose
length() ... Length of a vector

Elementary Mathematical Functions: Basic Real and Complex Functions

abs() ... Absolute value

angle() ... Phase angle in radians of a complex number. Synonym for “arg”
arg() ... Phase angle in radians of a complex number
conj() ... Conjugate of a complex number
deg2rad() ... Converts phase from degrees into radians
hypot() ... Euclidean distance function
imag() ... Imaginary value of a complex number
mag() ... Magnitude of a complex number
norm() ... Square of the absolute value of a vector
phase() ... Phase angle in degrees of a complex number
polar() ... Transform from polar coordinates into complex number
rad2deg() ... Converts phase from degrees into radians
real() ... Real value of a complex number
signum() ... Signum function
sign() ... Sign function
sqr() ... Square of a number
sqrt() ... Square root
unwrap() ... Unwraps a phase vector in radians

Elementary Mathematical Functions: Exponential and Logarithmic

Functions

exp() ... Exponential function

limexp() ... Limited exponential function
log10() ... Decimal logarithm
log2() ... Binary logarithm
ln() ... Natural logarithm (base e)

376
Elementary Mathematical Functions: Trigonometry

cos() ... Cosine function

cosec() ... Cosecant
cot() ... Cotangent function
sec() ... Secant
sin() ... Sine function
tan() ... Tangent function

Elementary Mathematical Functions: Inverse Trigonometric Functions

arccos() ... Arc cosine (also known as “inverse cosine”)

arccosec() ... Arc cosecant (also known as “inverse cosecant”)
arccot() ... Arc cotangent
arcsec() ... Arc secant (also known as “inverse secant”)
arcsin() ... Arc sine (also known as “inverse sine”)
arctan() ... Arc tangent (also known as “inverse tangent”)

Elementary Mathematical Functions: Hyperbolic Functions

cosh() ... Hyperbolic cosine

cosech() ... Hyperbolic cosecant
coth() ... Hyperbolic cotangent
sech() ... Hyperbolic secant
sinh() ... Hyperbolic sine
tanh() ... Hyperbolic tangent

Elementary Mathematical Functions: Inverse Hyperbolic Functions

arcosh() ... Hyperbolic area cosine

arcosech() ... Hyperbolic area cosecant
arcoth() ... Hyperbolic area cotangent
arsech() ... Hyperbolic area secant
arsinh() ... Hyperbolic area sine
artanh() ... Hyperbolic area tangent

377
Elementary Mathematical Functions: Rounding

ceil() ... Round to the next higher integer

fix() ... Truncate decimal places from real number
floor() ... Round to the next lower integer
round() ... Round to nearest integer

Elementary Mathematical Functions: Special Mathematical Functions

besseli0() ... Modified Bessel function of order zero

besselj() ... Bessel function of n-th order
bessely() ... Bessel function of second kind and n-th order
erf() ... Error function
erfc() ... Complementary error function
erfinv() ... Inverse error function
erfcinv() ... Inverse complementary error function
sinc() ... Sinc function
step() ... Step function

Data Analysis: Basic Statistics

avg() ... Average of vector elements

cumavg() ... Cumulative average of vector elements
max() ... Maximum value
min() ... Minimum value
rms() ... Root Mean Square of vector elements
runavg() ... Running average of vector elements
stddev() ... Standard deviation of vector elements
variance() ... Variance of vector elements
random() ... Random number between 0.0 and 1.0
srandom() ... Set seed for a new series of pseudo-random numbers

378
Data Analysis: Basic Operation

cumprod() ... Cumulative product of vector elements

cumsum() ... Cumulative sum of vector elements
interpolate() ... Equidistant spline interpolation of data vector
prod() ... Product of vector elements
sum() ... Sum of vector elements
xvalue() ... Returns x-value which is associated with the y-value nearest to a
specified y-value in a given vector
yvalue() ... Returns y-value of a given vector which is located nearest to the
specified x-value

Data Analysis: Differentiation and Integration

ddx() ... Differentiate mathematical expression with respect to a given variable

diff() ... Differentiate vector with respect to another vector
integrate() ... Integrate vector

Data Analysis: Signal Processing

dft() ... Discrete Fourier Transform

fft() ... Fast Fourier Transform
idft() ... Inverse Discrete Fourier Transform
ifft() ... Inverse Fast Fourier Transform
fftshift() ... Move the frequency 0 to the center of the FFT vector
Time2Freq() ... Interpreted Discrete Fourier Transform
Freq2Time() ... Interpreted Inverse Discrete Fourier Transform
kbd() ... Kaiser-Bessel derived window

Electronics Functions

Unit Conversion

dB() ... dB value

dbm() ... Convert voltage to power in dBm
dbm2w() ... Convert power in dBm to power in Watts
w2dbm() ... Convert power in Watts to power in dBm

379
Reflection Coefficients and VSWR

rtoswr() ... Converts reflection coefficient to voltage standing wave ratio (VSWR)
rtoy() ... Converts reflection coefficient to admittance
rtoz() ... Converts reflection coefficient to impedance
ytor() ... Converts admittance to reflection coefficient
ztor() ... Converts impedance to reflection coefficient

N-Port Matrix Conversions

stos() ... Converts S-parameter matrix to S-parameter matrix with different

reference impedance(s)
stoy() ... Converts S-parameter matrix to Y-parameter matrix
stoz() ... Converts S-parameter matrix to Z-parameter matrix
twoport() ... Converts a two-port matrix from one representation into another
ytos() ... Converts Y-parameter matrix to S-parameter matrix
ytoz() ... Converts Y-parameter matrix to Z-parameter matrix
ztos() ... Converts Z-parameter matrix to S-parameter matrix
ztoy() ... Converts Z-parameter matrix to Y-parameter matrix

Amplifiers

GaCircle() ... Circle(s) with constant available power gain Ga in the source plane
GpCircle() ... Circle(s) with constant operating power gain Gp in the load plane
Mu() ... Mu stability factor of a two-port S-parameter matrix
Mu2() ... Mu’ stability factor of a two-port S-parameter matrix
NoiseCircle() ... Generates circle(s) with constant Noise Figure(s)
PlotVs() ... Returns a data item based upon vector or matrix vector with
dependency on a given vector
Rollet() ... Rollet stability factor of a two-port S-parameter matrix
StabCircleL() ... Stability circle in the load plane
StabCircleS() ... Stability circle in the source plane
StabFactor() ... Stability factor of a two-port S-parameter matrix. Synonym for
Rollet()
StabMeasure() ... Stability measure B1 of a two-port S-parameter matrix
vt() ... Thermal voltage for a given temperature in Kelvin

380
15.4 Math Functions

15.4.1 Vectors and Matrices

Creation

eye()
Creates n x n identity matrix.

Syntax

y=eye(n)

Arguments

Name Type Def. Range Required

√
n N [1, +∞[

Description

This function creates the n x n identity matrix, that is

 
1 0··· 0 0

 0 1 0 ··· 0 

 .. .. ..
 . 0 . 0 .
 
 0 ··· 0 1 0 
0 0 ··· 0 1

Example

1 0
y=eye(2) returns .
0 1

See also

381
linspace()
Creates a real vector with linearly spaced components.

Syntax

y=linspace(xs,xe,n)

Arguments

Name Type Def. Range Required

√
xs R ]−∞, +∞[
√
xe R ]−∞, +∞[
√
n N [2, +∞[

Description

This function creates a real vector with n linearly spaced components. The first
component is xs, the last one is xe.

Example

y=linspace(1,2,3) returns 1, 1.5, 2.

See also

logspace()

382
logspace()
Creates a real vector with logarithmically spaced components.

Syntax

y=logspace(xs,xe,n)

Arguments

Name Type Def. Range Required

√
xs R ]−∞, +∞[
√
xe R ]−∞, +∞[
√
n N [2, +∞[

Description

This function creates a real vector with n logarithmically spaced components. The
first component is xs, the last one is xe.

Example

y=logspace(1,2,3) returns 1, 1.41, 2.

See also

linspace()

383
Basic Matrix Functions

adjoint()
Adjoint matrix.

Syntax

Y=adjoint(X)

Arguments

Name Type Def. Range Required

√
X Rm×n ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function calculates the adjoint matrix Y of a matrix X :

Y = X H = (X ∗ )T , where X ∗ is the complex conjugate matrix of X and X T is the

transposed of the matrix X.

Example

3+j1 0
X=eye(2)*(3+i) returns . Then,
0 3+j1

3-j1 0
Y=adjoint(X) returns .
0 3-j1

See also

transpose(), conj()

384
array()

Read out single elements.

Syntax

The “array()” function is an implicit command. Thus normally the respective first
expression (”preferred”) is used.

Syntax Preferred Alternative Preferred Alternative

1 y=VM[i,j] y=array(VM,i,j)
2 y=M[i,j] y=array(M,i,j)
3 y=VM[k] y=array(VM,k)
4 y=v[i] y=array(v,i) y=v[r] y=array(v,r)
5 y=v[i,r] y=array(v,i,r) y=v[r,j] y=array(v,r,j)
y=v[i,j] y=array(v,i,j) y=v[r1,r2] y=array(v,r1,r2)
6 y=s[i] y=array(s,i)

Arguments

Name Type Def. Range Required

m×n×p
√
VM R , Cm×n×p ]−∞, +∞[ (Syntax 1 and 3)
√
M Rm×n ,Cm×n ]−∞, +∞[ (Syntax 2)
√
v Rn ,Cn ]−∞, +∞[ (Syntax 4 and 5)
√
r, r1, r2 Rangexs : xe 0 ≤ xs ≤ n − 1, xs ≤ xe ≤ n − 1 (Syntax 4 and 5)
√
i N 0≤i≤m−1 (Syntax 1, 2, 4, 5, 6)
√
j N 0≤j ≤n−1 (Syntax 1, 2, 5)
√
k N 0≤k ≤p−1 (Syntax 3)
√
s String Arbitrary characters (Syntax 6)

Description

This function reads out real or complex vectors of matrices, matrices and vectors
or strings. Please refer to the following table for the return values:

385
 Syntax Argument 1 Argument 2 Argument 3 Result
y=VM[i,j] V M = (xijk ) i∈N j∈N Vector
(xij1 , · · · , xijK )
y=M[i,j] M = (xij ) i∈N j∈N Number xij
y=VM[k] V M = (xijk ) k∈N Matrix
 
x11k · · · x1nk
 .. ... .. 
 . . 
xm1k · · · xmnk
y=v[i] v = (vi ) i∈N Number vi
y=v[xs:xe] v = (vi ) xs, . . . , xe Vector
(vxs , · · · , vxe )
y=v[i,xs:xe] v = (vi ) i∈N xs, . . . , xe Vector
(vxs , · · · , vxe )
y=v[xs:xe,j] v = (vi ) xs, . . . , xe xs, . . . , xe Vector
(vxs , · · · , vxe )
y=v[i,j] v = (vi ) i∈N xs, . . . , xe Vector
(vxs , · · · , vxe )
y=v[xs1:xe1, v = (vi ) xs1, . . . , xe1 xs2, . . . , xe2 Vector
xs2:xe2] (vxs , · · · , vxe )
y=s[i] s = (si ) i∈N Character si

Again, v denotes a vector, M a matrix, VM a vector of matrices, s a vector of

characters and xs, xs1, xs2, xe, xe1, xe2 are range limiters.

Example

v=linspace(1,2,4) returns 1, 1.33, 1.67, 2. Then,

y=v[3] returns 2.

See also

386
det()
Determinant of a matrix.

Syntax

y=det(X)

Arguments

Name Type Def. Range Required

n×n n×n
√
X R ,C , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function calculates the determinant of a quadratical n x n matrix X. The

result is either a real or a complex number.

Example

3 0
X=eye(2)*3 returns . Then,
0 3

y=det(X) returns 9.

See also

eye()

387
inverse()
Matrix inverse.

Syntax

Y=inverse(X)

Arguments

Name Type Def. Range Required

n×n n×n
√
X R ,C , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function inverts a quadratical n x n matrix X. The generated inverted matrix

Y fulfills the equation

X · Y = X · X −1 = 1, where “ · ” denotes matrix multiplication and “1” the identity

matrix.

The matrix X must be regular, that means that its determinant ∆ 6= 0.

Example

3 0
X=eye(2)*3 returns . Then,
0 3

0.333 0
Y=inverse(X) returns .
0 0.333

See also

transpose(), eye(), det()

388
transpose()
Matrix transpose.

Syntax

Y=transpose(X)

Arguments

Name Type Def. Range Required

m×n m×n
√
X R ,C , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function transposes a m x n matrix X, which is equivalent to exchanging rows

and columns according to

Y = X T = (xij )T = (xji ) with 1 ≤ i ≤ m, 1 ≤ j ≤ n

The generated matrix Y is a n x m matrix.

Example

3 0
X=eye(2)*3 returns . Then,
0 3

3 0
Y=transpose(X) returns .
0 3

See also

eye(), inverse()

389
length()
Length of a vector.

Syntax

y=length(v)

Arguments

Name Type Def. Range Required

√
v R, C, R , Cn ]−∞, +∞[
n

Description

This function returns the length of vector v.

Example

length(linspace(1,2,3)) returns 3.

See also

390
15.4.2 Elementary Mathematical Functions

Basic Real and Complex Functions

abs()
Absolute value.

Syntax

y=abs(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function calculates the absolute value of a real or complex number, vector or
matrix.

x f or x ≥ 0
For x ∈ R: y =
−x f or x < 0

√
For C 3 x := a + i b ∧ a, b ∈ R: y = a2 + b 2

For x being a vector or a matrix the two equations above are applied to the
components of x.

Examples

y=abs(-3) returns 3,

y=abs(-3+4*i) returns 5.

See also

mag(), norm(), real(), imag(), conj(), phase(), arg(), hypot()

391
angle()
Phase angle in radians of a complex number. Synonym for “arg”.

Syntax

y=angle(x)

See also

abs(), mag(), norm(), real(), imag(), conj(), phase(), arg()

392
arg()

Phase angle in radians of a complex number.

Syntax

y=arg(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function returns the phase angle in degrees of a real or complex number,
vector or matrix.

0 f or x ≥ 0
For x ∈ R: y =
π f or x < 0

For C 3 x := a + i b ∧ a, b ∈ R:

Definition range Result

y = arctan
ab


a > 0, b > 0
a < 0, b > 0 y = arctan ab
+ π
a < 0, b < 0 y = arctan ab −
π
a > 0, b < 0 y = arctan ab
a = 0, b > 0 y = π2
a > 0, b > 0 y = − π2
a = 0, b = 0 y=0

In this case the arctan() function returns values in radians. The result y of the
phase function is in the range [−π, +π]. For x being a vector or a matrix the two
equations above are applied to the components of x.

Examples

y=arg(-3) returns 3.14,

393
y=arg(-3+4*i) returns 2.21.

See also

abs(), mag(), norm(), real(), imag(), conj(), phase()

394
conj()
Conjugate of a complex number.

Syntax

y=conj(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function returns the conjugate of a real or complex number, vector or matrix.

For x ∈ R: y = x

For C 3 x := a + i b ∧ a, b ∈ R: y = a − i b

For x being a vector or a matrix the two equations above are applied to the
components of x.

Example

y=conj(-3+4*i) returns -3-4*i.

See also

abs(), mag(), norm(), real(), imag(), phase(), arg()

395
deg2rad()
Converts phase from degrees into radians.

Syntax

y=deg2rad(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function converts a real phase, a complex phase or a phase vector given in
degrees into radians.

π
For x ∈ R: y = x
180

π
For x∈ C : y = Re {x}
180

For x being a vector the two equations above are applied to the components of x.

Example

y=deg2rad(45) returns 0.785.

See also

rad2deg(), phase(), arg()

396
hypot()
Euclidean distance function.

Syntax

z=hypot(x,y)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[
√
y R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the Euclidean distance z between two real or complex
numbers or vectors. For two numbers x, y ∈ C, this is
p
z= |x|2 + |y|2

For x, y being vectors (of same size) the equation above is applied componentwise.

Examples

z=hypot(3,4) returns 5,

z=hypot(1+2*i,1-2*i) returns 3.16.

See also

abs()

397
imag()
Imaginary value of a complex number.

Syntax

y=imag(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function returns the imaginary value of a real or complex number, vector or
matrix.

For x ∈ R: y = 0

For C 3 x := a + i b ∧ a, b ∈ R: y = b

For x being a vector or a matrix the two equations above are applied to the
components of x.

Example

y=imag(-3+4*i) returns 4.

See also

abs(), mag(), norm(), real(), conj(), phase(), arg()

398
mag()
Magnitude of a complex number.

Syntax

y=mag(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function calculates the magnitude (absolute value) of a real or complex num-
ber, vector or matrix.

x f or x ≥ 0
For x ∈ R: y =
−x f or x < 0

√
For C 3 x := a + i b ∧ a, b ∈ R: y = a2 + b 2

For x being a vector or a matrix the two equations above are applied to the
components of x.

Examples

y=mag(-3) returns 3,

y=mag(-3+4*i) returns 5.

See also

abs(), norm(), real(), imag(), conj(), phase(), arg()

399
norm()
Square of the absolute value of a vector.

Syntax

y=norm(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the square of the absolute value of a real or complex number,
vector or matrix.

For x ∈ R: y = x2

For C 3 x := a + i b ∧ a, b ∈ R: y = a2 + b2

For x being a vector or a matrix the two equations above are applied to the
components of x.

Example

y=norm(-3+4*i) returns 25.

See also

abs(), mag(), real(), imag(), conj(), phase(), arg()

400
phase()

Phase angle in degrees of a complex number.

Syntax

y=phase(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n , Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function returns the phase angle in degrees of a real or complex number,
vector or matrix.

0 f or x ≥ 0
For x ∈ R: y =
180 f or x < 0

For C 3 x := a + i b ∧ a, b ∈ R:

Definition range Result

y = arctan
ab


a > 0, b > 0
a < 0, b > 0 y = arctan ab
+ 180
a < 0, b < 0 y = arctan ab −
180
a > 0, b < 0 y = arctan ab
a = 0, b > 0 y = 90
a > 0, b > 0 y = −90
a = 0, b = 0 y=0

In this case the arctan() function returns values in degrees. The result y of the
phase function is in the range [−180, +180]. For x being a vector or a matrix the
two equations above are applied to the components of x.

Examples

y=phase(-3) returns 180,

401
y=phase(-3+4*i) returns 127.

See also

abs(), mag(), norm(), real(), imag(), conj(), arg()

402
polar()
Transform from polar coordinates into complex number.

Syntax

c=polar(a,p)

Arguments

Name Type Def. Range Required

√
a Rn , Cn ]−∞, +∞[
√
p Rn , Cn ]−∞, +∞[

Description

This function transforms a point given in polar coordinates (amplitude a and phase
p in degrees) in the complex plane into the corresponding complex number:

x + i y = a eip = a cos p + i a sin p

For a or p being vectors the equation above is applied to the components of a or

p.

Example

c=polar(3,45) returns 2.12+j2.12.

See also

abs(), mag(), norm(), real(), imag(), conj(), phase(), arg(), exp(), cos(), sin()

403
rad2deg()
Converts phase from degrees into radians.

Syntax

y=rad2deg(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function converts a real phase, a complex phase or a phase vector given in
radians into degrees.

180
For x ∈ R: y = x
π

180
For x∈ C : y = Re {x}
π

For x being a vector the two equations above are applied to the components of x.

Example

y=rad2deg(45) returns 0.785.

See also

deg2rad(), phase(), arg()

404
real()
Real value of a complex number.

Syntax

y=real(x)

Arguments

Name Type Def. Range Required

n n m×n
√
x R, C, R , C , R ,Cm×n ,Rm×n×p , Cm×n×p ]−∞, +∞[

Description

This function returns the real value of a real or complex number, vector or matrix.

For x ∈ R: y = x

For C 3 x := a + i b ∧ a, b ∈ R: y = a

For x being a vector or a matrix the two equations above are applied to the
components of x.

Example

y=real(-3+4*i) returns -3.

See also

abs(), mag(), norm(), imag(), conj(), phase(), arg()

405
signum()
Signum function.

Syntax

y=signum(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the sign of a real or complex number or vector.


 1 f or x > 0
For x ∈ R: y = 0 f or x = 0
−1 f or x < 0


( x
f or x 6= 0
For x ∈ C: y = |x|
0 f or x = 0

For x being a vector the two equations above are applied to the components of x.

Examples

y=signum(-4) returns -1,

y=signum(3+4*i) returns 0.6+j0.8.

See also

abs(), sign()

406
sign()
Sign function.

Syntax

y=sign(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the sign of a real or complex number or vector.


1 f or x >= 0
For x ∈ R: y =
−1 f or x < 0

( x
f or x 6= 0
For x ∈ C: y = |x|
1 f or x = 0

For x being a vector the two equations above are applied to the components of x.

Examples

y=sign(-4) returns -1,

y=sign(3+4*i) returns 0.6+j0.8.

See also

abs(), signum()

407
sqr()
Square of a number.

Syntax

y=sqr(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the square root of a real or complex number or vector.

y = x2

For x being a vector the two equations above are applied to the components of x.

Examples

y=sqr(-4) returns 16,

y=sqr(3+4*i) returns -7+j24.

See also

sqrt()

408
sqrt()
Square root.

Syntax

y=sqrt(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the square root of a real or complex number or vector.
 √
For x ∈ R: y = √x f or x ≥ 0
i −x f or x < 0

p ϕ
For x ∈ C: y = |x| ei 2 with ϕ = arg (x)

For x being a vector the two equations above are applied to the components of x.

Examples

y=sqrt(-4) returns 0+j2,

y=sqrt(3+4*i) returns 2+j1.

See also

sqr()

409
unwrap()
Unwraps a phase vector in radians.

Syntax

y=unwrap(x)

y=unwrap(x, t)

Arguments

Name Type Def. Range Required Default

√
x Rn , Cn ]−∞, +∞[
t R ]−∞, +∞[ π

Description

This function unwraps a phase vector x to avoid phase jumps. If two consecutive
values of x differ by more than tolerance t, ∓2π(depending on the sign of the
difference) is added to the current element of x. The predefined value of the
optional parameter t is π.

Examples

y=unwrap(3.15*linspace(-2,2,5)) returns -6.3, -9.43, -12.6, -15.7, -18.8,

y=unwrap(2*linspace(-2,2,5),1) returns -4, -8.28, -12.6, -16.8, -21.1,

y=unwrap(2*linspace(-2,2,5),3) returns -4, -2, 0, 2, 4.

See also

abs(), mag(), norm(), real(), imag(), conj(), phase(), arg()

410
Exponential and Logarithmic Functions

exp()

Exponential function.

Syntax

y=exp(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the exponential function of a real or complex number or

vector.

For x ∈ R: y = ex

For C 3 x := a + i b ∧ a, b ∈ R: y = ex = ea+i b = ea (cos b + i sin b)

For x being a vector the two equations above are applied to the components of x.

Examples

y=exp(-4) returns 0.0183,

y=exp(3+4*i) returns -13.1-j15.2.

See also

limexp(), ln(), log10(), log2(), cos(), sin()

411
limexp()
Limited exponential function.

Syntax

y=limexp(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function is equivalent to the exponential function exp(x), as long as x <= 80.
For larger arguments x, it limits the result to y = exp(80) · (1 + x − 80). The
argument can be a real or complex number or vector.

For x ∈ R: y = ex for x ≤ 80, y = e80 · (1 + x − 80) else.

For C 3 x := a+i b ∧ a, b ∈ R: y = limexp (x) = limexp (a + i b) = limexp (a) (cos b + i sin b)

For x being a vector the two equations above are applied to the components of x.

Examples

y=limexp(81) returns 1.11e+35, whereas y=exp(81) returns 1.51e+35, which

shows the limiting effect of the limexp() function.

y=limexp(3+4*i) returns -13.1-j15.2.

See also

exp(), ln(), log10(), log2(), cos(), sin()

412
log10()
Decimal logarithm.

Syntax

y=log10(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[ \ {0}

Description

This function calculates the principal value of the decimal logarithm (base 10) of
a real or complex number or vector.

ln (x)
f or x > 0


ln (10)

For x ∈ R: y =
ln (−x) π
+i f or x < 0


ln (10) ln (10)


ln (|x|) arg (x)

For x ∈ C: y = +i
ln (10) ln (10)

For x being a vector the two equations above are applied to the components of x.

Examples

y=log10(-4) returns 0.602+j1.36,

y=log10(3+4*i) returns 0.699+j0.403.

See also

ln(), log2(), exp(), arg()

413
log2()
Binary logarithm.

Syntax

y=log2(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[ \ {0}

Description

This function calculates the principal value of the binary logarithm (base 2) of a
real or complex number or vector.

ln (x)
f or x > 0


ln (2)

For x ∈ R: y =
ln (−x) π
+i f or x < 0


ln (2) ln (2)


ln (|x|) arg (x)

For x ∈ C: y = +i
ln (2) ln (2)

For x being a vector the two equations above are applied to the components of x.

Examples

y=log2(-4) returns 2+j4.53,

y=log2(3+4*i) returns 2.32+j1.34.

See also

ln(), log10(), exp(), arg()

414
ln()
Natural logarithm (base e).

Syntax

y=ln(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[ \ {0}

Description

This function calculates the principal value of the natural logarithm (base e) of a
real or complex number or vector.

ln (x) f or x > 0
For x ∈ R: y =
ln (−x) f or x < 0

For x ∈ C: y = ln (|x|) + i arg (x)

For x being a vector the two equations above are applied to the components of x.

Examples

y=ln(-4) returns 1.39+j3.14,

y=ln(3+4*i) returns 1.61+j0.927.

See also

log2(), log10(), exp(), arg()

415
Trigonometry

cos()
Cosine function.

Syntax

y=cos(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the cosine of a real or complex number or vector.

For x ∈ R: y = cos (x) with y ∈ [−1, 1]

1
For x ∈ C: y = 2
(exp (i x) + exp (−i x))

For x being a vector the two equations above are applied to the components of x.

Examples

y=cos(-0.5) returns 0.878,

y=cos(3+4*i) returns -27.0-j3.85.

See also

sin(), tan(), arccos()

416
cosec()
Cosecant.

Syntax

y=cosec(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[ \ {kπ} , k ∈ Z

Description

This function calculates the cosecant of a real or complex number or vector.

1
y = cosec x =
sin x

For x being a vector the equation above is applied to the components of x.

Example

y=cosec(1) returns 1.19.

See also

sin(), sec()

417
cot()
Cotangent function.

Syntax

y=cot(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[ \ {kπ} , k ∈ Z

Description

This function calculates the cotangent of a real or complex number or vector.

1
For x ∈ R: y = with y ∈ [−∞, +∞]
tan (x)
!
exp (i x)2 + 1
For x ∈ C: y = i
exp (i x)2 − 1

For x being a vector the two equations above are applied to the components of x.

Examples

y=cot(-0.5) returns -1.83,

y=cot(3+4*i) returns -0.000188-j1.

See also

tan(), sin(), cos(), arctan(), arccot()

418
sec()
Secant.

Syntax

y=sec(x)

Arguments

Name Type Def. Range Required

√
]−∞, +∞[ \ k + 21 π , k ∈ Z


x R, C, Rn , Cn

Description

This function calculates the secant of a real or complex number or vector.

1
y =sec x=
cos x

For x being a vector the equation above is applied to the components of x.

Example

y=sec(0) returns 1.

See also

cos(), cosec()

419
sin()
Sine function.

Syntax

y=sin(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the sine of a real or complex number or vector.

For x ∈ R: y = sin (x) with y ∈ [−1, 1]

For x ∈ C: y = 12 i (exp (−i x) − exp (i x))

For x being a vector the two equations above are applied to the components of x.

Examples

y=sin(-0.5) returns -0.479,

y=sin(3+4*i) returns 3.85-j27.

See also

cos(), tan(), arcsin()

420
tan()
Tangent function.

Syntax

y=tan(x)

Arguments

Name Type Def. Range Required

√
]−∞, +∞[ \ k + 21 π , k ∈ Z


x R, C, Rn , Cn

Description

This function calculates the tangent of a real or complex number or vector.

For x ∈ R: y = tan (x) with y ∈ [−∞, +∞]

!
exp (i x)2 − 1
For x ∈ C: y = −i
exp (i x)2 + 1

For x being a vector the two equations above are applied to the components of x.

Examples

y=tan(-0.5) returns -0.546,

y=tan(3+4*i) returns -0.000187+j0.999.

See also

cot(), sin(), cos(), arctan(), arccot()

421
Inverse Trigonometric Functions

arccos()
Arc cosine (also known as “inverse cosine”).

Syntax

y=arccos(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn [−1, +1]

Description

This function calculates principal value of the the arc cosine of a real or complex
number or vector.

For x ∈ R: y = arccos (x) with y ∈ [0, π]

√

For x ∈ C: y = −i ln x + x2 − 1

For x being a vector the two equations above are applied to the components of x.

Examples

y=arccos(-1) returns 3.14,

y=arccos(3+4*i) returns 0.937-j2.31.

See also

cos(), arcsin(), arctan(), arccot()

arccosec()
Arc cosecant (also known as “inverse cosecant”).

422
Syntax

y=arccosec(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn C\ {0}

Description

This function calculates the principal value of the the arc cosecant of a real or
complex number or vector.

For x ∈ R: y = arccosec (x) with y ∈ − π2 , π2

 

hq i
1 i
For x ∈ C: y = −i ln 1− x2
+ x

For x being a vector the two equations above are applied to the components of x.

Examples

y=arccosec(-1) returns -1.57,

y=arccosec(3+4*i) returns 0.119-j0.16.

See also

cosec(), arcsec()

423
arccot()
Arc cotangent.

Syntax

y=arccot(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the principal value of the arc cotangent of a real or complex
number or vector.

For x ∈ R: y =arccot(x) with y ∈ [0, π]

 
i x−i
For x ∈ C: y = ln
2 x+i

For x being a vector the two equations above are applied to the components of x.

Examples

y=arccot(-1) returns 2.36,

y=arccot(3+4*i) returns 0.122-j0.159.

See also

cot(), tan(), arccos(), arcsin(), arctan()

424
arcsec()
Arc secant (also known as “inverse secant”).

Syntax

y=arcsec(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn C\ {0}

Description

This function calculates the principal value of the arc secant of a real or complex
number or vector.

For x ∈ R: y = arcsec (x) with y ∈ [0, π]

hq i
π 1 i
For x ∈ C: y = 2
+ i ln 1− x2
+ x

For x being a vector the two equations above are applied to the components of x.

Examples

y=arcsec(-1) returns 3.14,

y=arcsec(3+4*i) returns 1.45+j0.16.

See also

sec(), arccosec()

425
arcsin()
Arc sine (also known as “inverse sine”).

Syntax

y=arcsin(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn [−1, +1]

Description

This function calculates the principal value of the arc sine of a real or complex
number or vector.

For x ∈ R: y = arcsin (x) with y ∈ − π2 , π2

 

 √ 
For x ∈ C: y = −i ln i x + 1 − x2

For x being a vector the two equations above are applied to the components of x.

Examples

y=arcsin(-1) returns -1.57,

y=arcsin(3+4*i) returns 0.634+j2.31.

See also

sin(), arccos(), arctan(), arccot()

426
arctan()
Arc tangent (also known as “inverse tangent”).

Syntax

z=arctan(x)

z=arctan(y,x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[
y R, C, Rn , Cn ]−∞, +∞[

Description

For the first syntax ( z =arctan(x ) ), this function calculates the principal value of
the arc tangent of a real or complex number or vector.

For x ∈ R: y = arctan (x) with y ∈ − π2 , π2

 

 
1 2i
For x ∈ C: y = − i ln −1
2 x+i

For x being a vector the two equations above are applied to the components of x.

If the second syntax ( z =arctan(y, x ) ) finds application, the expression

z = ± arctan (y/x)

(with the arctan() function defined above) is evaluated. The sign of z is determined
by

+ f or Re {x} > 0
sign(z)= .
− f or Re {x} > 0

Note that for the second syntax the case x = y = 0 is not defined.

427
Examples

z=arctan(-1) returns -0.785,

z=arctan(3+4*i) returns 1.45+j0.159,

z=arctan(1,1) returns 0.785.

See also

tan(), arccos(), arcsin(), arccot()

428
Hyperbolic Functions

cosh()

Hyperbolic cosine.

Syntax

y=cosh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the hyperbolic cosine of a real or complex number or

vector.

y = 21 (ex + e−x )

For x being a vector the equation above is applied to the components of x.

Examples

y=cosh(-1) returns 1.54,

y=cosh(3+4*i) returns -6.58-j7.58.

See also

exp(), sinh(), tanh(), cos()

429
cosech()
Hyperbolic cosecant.

Syntax

y=cosech(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[\ {0}

Description

This function calculates the hyperbolic cosecant of a real or complex number or

vector.

1
y=
sinh x

For x being a vector the equation above is applied to the components of x.

Examples

y=cosech(-1) returns -0.851,

y=cosech(3+4*i) returns -0.0649+j0.0755.

See also

exp(), sinh(), sech(), cosec()

430
coth()
Hyperbolic cotangent.

Syntax

y=coth(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[\ {0}

Description

This function calculates the hyperbolic cotangent of a real or complex number or

vector.

1 ex + e−x
y= = x
tanh x e − e−x

For x being a vector the equation above is applied to the components of x.

Examples

y=coth(-1) returns -1.31,

y=coth(3+4*i) returns 0.999-j0.0049.

See also

exp(), cosh(), sinh(), tanh(), tan()

431
sech()
Hyperbolic secant.

Syntax

y=sech(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the hyperbolic secant of a real or complex number or

vector.

1
y=
cosh x

For x being a vector the equation above is applied to the components of x.

Examples

y=sech(-1) returns 0.648,

y=sech(3+4*i) returns -0.0653+j0.0752.

See also

exp(), cosh(), cosech(), sec()

432
sinh()
Hyperbolic sine.

Syntax

y=sinh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the hyperbolic sine of a real or complex number or vector.

y = 21 (ex − e−x )

For x being a vector the equation above is applied to the components of x.

Examples

y=sinh(-1) returns -1.18,

y=sinh(3+4*i) returns -6.55-j7.62.

See also

exp(), cosh(), tanh(), sin()

433
tanh()
Hyperbolic tangent.

Syntax

y=tanh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the hyperbolic tangent of a real or complex number or

vector.

ex − e−x
y=
ex + e−x

For x being a vector the equation above is applied to the components of x.

Examples

y=tanh(-1) returns -0.762,

y=tanh(3+4*i) returns 1+j0.00491.

See also

exp(), cosh(), sinh(), coth(), tan()

434
Inverse Hyperbolic Functions

arcosh()
Hyperbolic area cosine.

Syntax

y=arcosh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn [1, +∞[

Description

This function calculates the hyperbolic area cosine of a real or complex number or
vector, which is the inverse function to the “cosh” function.
√

y = arcosh x = ln x + x2 − 1

For x being a vector the equation above is applied to the components of x.

Examples

y=arcosh(1) returns 0,

y=arcosh(3+4*i) returns 2.31+j0.937.

See also

arsinh(), artanh(), cosh(), arccos(), ln(), sqrt()

435
arcosech()
Hyperbolic area cosecant.

Syntax

y=arcosech(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn C\ {0}

Description

This function calculates the hyperbolic area cosecant of a real or complex number
or vector, which is the inverse function to the “cosech” function.
q 
1 1
For x ∈ C\ {0}: y = ln 1+ x2
+ x

For x being a vector the equation above is applied to the components of x.

Examples

y=arcosech(1) returns 0.881,

y=arcosech(i) returns -i1.57.

See also

cosech(), arsech(), ln(), sqrt()

436
arcoth()
Hyperbolic area cotangent.

Syntax

y=arcoth(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, −1[ ∪ ]+1, +∞[

Description

This function calculates the hyperbolic area cotangent of a real or complex number
or vector, which is the inverse function to the “cotanh” function.
 
1 x+1
y = arcoth x = ln
2 x−1

For x being a vector the equation above is applied to the components of x.

Examples

y=arcoth(2) returns 0.549,

y=arcoth(3+4*i) returns 0.118-j0.161.

See also

arsinh(), arcosh(), tanh(), arctan(), ln(), sqrt()

437
arsech()
Hyperbolic area secant.

Syntax

y=arsech(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn C\ {0}

Description

This function calculates the hyperbolic area secant of a real or complex number or
vector, which is the inverse function to the “sech” function.
q q 
1 1 1
For x ∈ C\ {0}: y = ln x
−1 x +1+ x

For x being a vector the equation above is applied to the components of x.

Examples

y=arsech(1) returns 0,

y=arsech(3+4*i) returns 0.16-j1.45.

See also

sech(), arcosech(), ln(), sqrt()

438
arsinh()
Hyperbolic area sine.

Syntax

y=arsinh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the hyperbolic area sine of a real or complex number or
vector, which is the inverse function to the “sinh” function.
√

y = arsinh x = ln x + x2 + 1

For x being a vector the equation above is applied to the components of x.

Examples

y=arsinh(1) returns 0.881,

y=arsinh(3+4*i) returns 2.3+j0.918.

See also

arcosh(), artanh(), sinh(), arcsin(), ln(), sqrt()

439
artanh()
Hyperbolic area tangent.

Syntax

y=artanh(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−1, +1[

Description

This function calculates the hyperbolic area tangent of a real or complex number
or vector, which is the inverse function to the “tanh” function.
 
1 1+x
y = artanh x = ln
2 1−x

For x being a vector the equation above is applied to the components of x.

Examples

y=artanh(0) returns 0,

y=artanh(3+4*i) returns 0.118+j1.41.

See also

arsinh(), arcosh(), tanh(), arctan(), ln(), sqrt()

440
Rounding

ceil()
Round to the next higher integer.

Syntax

y=ceil(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function rounds a real number x to the next higher integer value.

If x is a complex number both real part and imaginary part are rounded. For x
being a vector the operation above is applied to the components of x.

Examples

y=ceil(-3.5) returns -3,

y=ceil(3.2+4.7*i) returns 4+j5.

See also

floor(), fix(), round()

441
fix()
Truncate decimal places from real number.

Syntax

y=fix(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function truncates the decimal places from a real number x and returns an
integer.

If x is a complex number both real part and imaginary part are rounded. For x
being a vector the operation above is applied to the components of x.

Examples

y=fix(-3.5) returns -3,

y=fix(3.2+4.7*i) returns 3+j4.

See also

ceil(), floor(), round()

442
floor()
Round to the next lower integer.

Syntax

y=floor(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function rounds a real number x to the next lower integer value.

If x is a complex number both real part and imaginary part are rounded. For x
being a vector the operation above is applied to the components of x.

Examples

y=floor(-3.5) returns -4,

y=floor(3.2+4.7*i) returns 3+j4.

See also

ceil(), fix(), round()

443
round()
Round to nearest integer.

Syntax

y=round(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function rounds a real number x to its nearest integer value.

If x is a complex number both real part and imaginary part are rounded. For x
being a vector the operation above is applied to the components of x.

Examples

y=round(-3.5) returns -4,

y=round(3.2+4.7*i) returns 3+j5.

See also

ceil(), floor(), fix()

444
Special Mathematical Functions

besseli0()
Modified Bessel function of order zero.

Syntax

i0=besseli0(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the modified Bessel function of order zero of a real or
complex number or vector.

∞ x 2k


X
2
i0 (x) = J0 (i x) = ,
k=0
k! Γ (k + 1)

where J0 (x)is the Bessel function of order zero and Γ (x)denotes the gamma func-
tion.

For x being a vector the equation above is applied to the components of x.

Example

y=besseli0(1) returns 1.266.

See also

besselj(), bessely()

445
besselj()
Bessel function of n-th order.

Syntax

jn=besselj(n,x)

Arguments

Name Type Def. Range Required

√
n N [0, +∞[
√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the Bessel function of n-th order of a real or complex
number or vector.

∞
n+2k
X (−1)k x2
Jn (x) = ,
k=0
k! Γ (n + k + 1)

where Γ (x)denotes the gamma function.

For x being a vector the equation above is applied to the components of x.

Example

y=besselj(1,1) returns 0,44.

See also

besseli0(), bessely()

446
bessely()
Bessel function of second kind and n-th order.

Syntax

yn=bessely(n,x)

Arguments

Name Type Def. Range Required

√
n N [0, +∞[
√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the Bessel function of second kind and n-th order of a real
or complex number or vector.

Jm (x) cos mπ − J−m (x)

Yn (x) = lim ,
m→n sin mπ

where Jm (x)denotes the Bessel function of first kind and n-th order.

For x being a vector the equation above is applied to the components of x.

Example

y=bessely(1,1) returns -0.781.

See also

besseli0(), besselj()

447
erf()
Error function.

Syntax

y=erf(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the error function of a real or complex number or vector.
For x ∈ R,

Zx
2 2
y=√ e−t dt
π
0

If x is a complex number both real part and imaginary part are subjected to the
equation above. For x being a vector the equation is applied to the components
of x.

Example

y=erf(0.5) returns 0.520.

See also

erfc(), erfinv(), erfcinv(), exp()

448
erfc()
Complementary error function.

Syntax

y=erfc(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the complementary error function of a real or complex

number or vector. For x ∈ R,

Zx
2 2
y =1− √ e−t dt
π
0

If x is a complex number both real part and imaginary part are subjected to the
equation above. For x being a vector the equation is applied to the components
of x.

Example

y=erfc(0.5) returns 0.480.

See also

erf(), erfinv(), erfcinv(), exp()

449
erfinv()
Inverse error function.

Syntax

y=erfinv(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−1, +1[

Description

This function evaluates the inverse of the error function of a real or complex number
or vector. For −1 < x < 1,

y = erf −1 (x)

If x is a complex number both real part and imaginary part are subjected to the
equation above. For x being a vector the equation is applied to the components
of x.

Example

y=erfinv(0.8) returns 0.906.

See also

erf(), erfc(), erfcinv(), exp()

450
erfcinv()
Inverse complementary error function.

Syntax

y=erfcinv(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]0, +2[

Description

This function evaluates the inverse of the complementary error function of a real
or complex number or vector. For 0 < x < 2,

y = erfc−1 (x)

If x is a complex number both real part and imaginary part are subjected to the
equation above. For x being a vector the equation is applied to the components
of x.

Example

y=erfcinv(0.5) returns 0.477.

See also

erf(), erfc(), erfinv(), exp()

451
sinc()
Sinc function.

Syntax

y=sinc(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function evaluates the sinc function of a real or complex number or vector.

sin x
(
f or x 6= 0
y= x
1 f or x = 0

For x being a vector the equation above is applied to the components of x.

Examples

y=sinc(-3) returns 0.047,

y=sinc(3+4*i) returns -3.86-j3.86.

See also

sin()

452
step()
Step function.

Syntax

y=step(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function calculates the step function of a real or complex number or vector.
For x ∈ R,

 0 f or x < 0
y= 0.5 f or x = 0
1 f or x > 0


If x is a complex number both real part and imaginary part are subjected to the
equation above. For x being a vector the equation is applied to the components
of x.

Example

y=step(0.5) returns 1.

See also

453
15.4.3 Data Analysis

Basic Statistics

avg()
Average of vector elements.

Syntax

y=avg(x)

Arguments

Name Type Def. Range Required

√
x R, C, R , Cn , Range xs : xe
n
]−∞, +∞[

Description

This function returns the sum of the elements of a real or complex vector or range.

1P n
For x ∈Cn : y = xi , 1 ≤ i ≤ n (for vectors) or xs ≤ i ≤ xe (for ranges)
n i=1

For x being a real or complex number, x itself is returned.

Example

y=avg(linspace(1,3,10)) returns 2.

See also

sum(), max(), min()

454
cumavg()
Cumulative average of vector elements.

Syntax

y=cumavg(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the cumulative average of the elements of a real or complex
vector.

1P k
For x ∈Cn : yk = xi , 1 ≤ k ≤ n
k i=1

For x being a real or complex number, x itself is returned.

Example

y=cumavg(linspace(1,3,3)) returns 1, 1.5, 2.

See also

cumsum(), cumprod(), avg(), sum(), prod(), max(), min()

455
max()
Maximum value.

Syntax

y=max(x)

y=max(a,b)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn , Range xs : xe ]−∞, +∞[
√
a R, C ]−∞, +∞[
√
b R, C ]−∞, +∞[

Description

For the first syntax ( y=max(x) ), this function returns the maximum value of a
real or complex vector or range.

For x ∈Rn : y =max (xi ) , 1 ≤ i ≤ n (for vectors) or xs ≤ i ≤ xe (for ranges)

For x ∈ Cn : y = max (± |xi |) , 1 ≤ i ≤ n (for vectors) or xs ≤ i ≤ xe (for ranges),

+ f or |arg (xi )| ≤ π2

with sign
− else

For x being a real or complex number: that is the case n = 1.

The second syntax ( y=max(a,b) ) finds application, if two (generally complex)

numbers a and b need to be compared. In principle, the maximum of the absolute
values is selected, but it must be considered whether a and b are located in the
right or left complex half plane. If the latter is the case, the negative absolute
value of a and b needs to be regarded (for example, which is the case for negative
real numbers), otherwise the positive absolute value is taken:

y = max (± |a| , ± |b|),

456
 π π
 
+ f or |arg (a)| ≤ + f or |arg (b)| ≤
with |a| sign 2 and |b| sign 2
− else − else

Example

y=max(linspace(1,3,10)) returns 3.

y=max(1,3) returns 3.

y=max(1,1+i) returns 1+j1.

y=max(1,-1+i) returns 1.

See also

min(), abs()

457
min()

Minimum value.

Syntax

y=min(x)

y=min(a,b)

Arguments

Name Type Def. Range Required

√
x R, C, R , Cn , Range xs : xe
n
]−∞, +∞[
√
a R, C ]−∞, +∞[
√
b R, C ]−∞, +∞[

Description

For the first syntax ( y=min(x) ), this function returns the minimum value of a
real or complex vector or range.

For x ∈Rn : y =min (xi ) , 1 ≤ i ≤ n (for vectors) or xs ≤ i ≤ xe (for ranges)

For x ∈ Cn : y = min (± |xi |) , 1 ≤ i ≤ n (for vectors) or xs ≤ i ≤ xe (for ranges),

+ f or |arg (xi )| ≤ π2

with sign
− else

For x being a real or complex number: that is the case n = 1.

The second syntax ( y=min(a,b) ) finds application, if two (generally complex)

numbers a and b need to be compared. In principle, the maximum of the absolute
values is selected, but it must be considered whether a and b are located in the
right or left complex half plane. If the latter is the case, the negative absolute
value of a and b needs to be regarded (for example, which is the case for negative
real numbers), otherwise the positive absolute value is taken:

y = min (± |a| , ± |b|),

458
 π π
 
+ f or |arg (a)| ≤ + f or |arg (b)| ≤
with |a| sign 2 and |b| sign 2
− else − else

Example

y=min(linspace(1,3,10)) returns 1.

y=min(1,3) returns 1.

y=min(1,1+i) returns 1.

y=min(1,-1+i) returns -1+j1.

See also

max(), abs()

459
rms()
Root Mean Square of vector elements.

Syntax

y=rms(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the rms (root mean square) value of the elements of a real
or complex vector. By application of the trapezoidal integration rule,
r n

1 P 1 for 2 ≤ i ≤ n − 1
n
for x ∈C : y = ai xi x∗i , 1 ≤ i ≤ n, ai = 1
n i=1 2
for i = 1 or i = n

For x being a real or complex number, |x| itself is returned.

Example

y=rms(linspace(1,2,8)) returns 1.43.

See also

variance(), stddev(), avg()

460
runavg()
Running average of vector elements.

Syntax

y=runavg(x,m)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[
√
m N [1, +∞[

Description

This function returns the running average over m elements of a real or complex
vector.

1 k+m−1
For x ∈Cn : yk =
P
xi , 1 ≤ k ≤ n
m i=k

For x being a real or complex number, x itself is returned.

Example

y=runavg(linspace(1,3,6),2) returns 1.2, 1.6, 2, 2.4, 2.8.

See also

cumavg(), cumsum(), avg(), sum()

461
stddev()
Standard deviation of vector elements.

Syntax

y=stddev(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the stddev of the elements of a real or complex vector x.
p
For x ∈Cn : y = variance(x)

For x being a real or complex number, 0 is returned.

Example

y=stddev(linspace(1,3,10)) returns 0.673.

See also

stddev(), avg(), max(), min()

462
variance()
Variance of vector elements.

Syntax

y=variance(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the variance of the elements of a real or complex vector.

1 P n
For x ∈Cn : y = (xi − x)2 , where x denotes mean (average) value of x.
n − 1 i=1

For x being a real or complex number, 0 is returned.

Example

y=variance(linspace(1,3,10)) returns 0.453.

See also

stddev(), avg(), max(), min()

463
random()
Random number between 0.0 and 1.0.

Syntax

y=random()

Arguments

None.

Description

This function returns a pseudo-random real number between 0.0 (including) and
1.0 (excluding). The starting point of the random number generator can be set by
srandom().

Example

y=random()

See also

srandom()

464
srandom()
Set seed for a new series of pseudo-random numbers.

Syntax

y=srandom(x)

Arguments

Name Type Def. Range Required

√
x R ]−∞, +∞[

Description

This function establishes x as the seed for a new series of pseudo-random numbers.
Please note that only integer values for x are considered, so for example x = 1.1
will give the same seed as x = 1.

Example

y=srandom(100)

See also

random()

465
Basic Operation

cumprod()
Cumulative product of vector elements.

Syntax

y=cumprod(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the cumulative product of the elements of a real or complex
vector.

k
For x ∈Cn : yk =
Q
xi , 1 ≤ k ≤ n
i=1

For x being a real or complex number, x itself is returned.

Example

y=cumprod(linspace(1,3,3)) returns 1, 2, 6.

See also

cumsum(), cumavg(), prod(), sum(), avg(), max(), min()

466
cumsum()
Cumulative sum of vector elements.

Syntax

y=cumsum(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the cumulative sum of the elements of a real or complex
vector.

Xk
n
For x ∈C : yk = xi , 1 ≤ k ≤ n
i=1

For x being a real or complex number, x itself is returned.

Example

y=cumsum(linspace(1,3,3)) returns 1, 3, 6.

See also

cumprod(), cumavg(), sum(), prod(), avg(), max(), min()

467
interpolate()
Equidistant spline interpolation of data vector.

Syntax

z=interpolate(y,t,m)

z=interpolate(y,t)

Arguments

Name Type Def. Range Required Default

√
y Rn , Cn ]−∞, +∞[
√
t Rn , Cn ]−∞, +∞[
m N [3, +∞[ 64

Description

This function uses spline interpolation to interpolate between the points of a vector
y(t). If the number of samples n is not specified, a default value of n = 64 is
assumed.

Example

z=interpolate(linspace(0,2,3)*linspace(0,2,3),linspace(0,2,3))

returns a smooth parabolic curve:

Use the Cartesian diagram to display it.

See also

sum(), prod()

468
 4
2
000.5Interpolate11
00
.15
. 2
Figure 15.4: Interpolated curve

prod()
Product of vector elements.

Syntax

y=prod(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the product of the elements of a real or complex vector.

n
For x ∈Cn : y =
Q
xi
i=1

For x being a real or complex number, x itself is returned.

Example

y=prod(linspace(1,3,10)) returns 583.

469
See also

sum(), avg(), max(), min()

470
sum()
Sum of vector elements.

Syntax

y=sum(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the sum of the elements of a real or complex vector.

n
For x ∈Cn : y =
P
xi
i=1

For x being a real or complex number, x itself is returned.

Example

y=sum(linspace(1,3,10)) returns 20.

See also

prod(), avg(), max(), min()

471
xvalue()
Returns x-value which is associated with the y-value nearest to a specified
y-value in a given vector.

Syntax

x=xvalue(f,yval)

Arguments

Name Type Def. Range Required

√
f Rn , Cn ]−∞, +∞[
√
yval R, C ]−∞, +∞[

Description

This function returns the x -value which is associated with the y-value nearest
to yval in the given vector f ; therefore the vector f must have a single data
dependency.

Example

x=xvalue(f,1).

See also

yvalue(), interpolate()

472
yvalue()
Returns y-value of a given vector which is located nearest to the specified
x-value.

Syntax

y=yvalue(f,xval)

Arguments

Name Type Def. Range Required

√
f Rn , Cn ]−∞, +∞[
√
xval R, C ]−∞, +∞[

Description

This function returns the y-value of the given vector f which is located nearest to
the x-value xval ; therefore the vector f must have a single data dependency.

Example

y=yvalue(f,1).

See also

xvalue(), interpolate()

473
Differentiation and Integration

ddx()
Differentiate mathematical expression with respect to a given variable.

Syntax

y=ddx(f(x),x)

Arguments

Name Type Def. Range Required Default

√
f(x)
√
x R, C, Rm , Cm ]−∞, +∞[

Description

This function executes a symbolic differentiation on a function f(x) with respect

to a variable x. The result is evaluated at the contents x0 of x.

df 
y=
dx 

x0

If x is a vector, the differential quotient is evaluated for all components of x, giving

a result vector y.

Example

Create a vector x by setting x=linspace(0,2,3), thus x = [0, 1, 2]T . Entering

y=ddx(sin(x),x returns 1, 0.54, -0.416.

df d sin(x)
Why? = = cos(x), and cos(x) evaluated at x = [0, 1, 2]T gives the
dx dx
result above.

See also

diff()

474
diff()
Differentiate vector with respect to another vector.

Syntax

z=diff(y,x,n)

Arguments

Name Type Def. Range Required Default

√
y Rk , Ck ]−∞, +∞[
√
x Rm , Cm ]−∞, +∞[
n N 1

Description

This function numerically differentiates a vector y with respect to a vector x.

If the optional integer parameter n is given, the n-th derivative is calculated.
Differentiation is executed for N=min(k,m) elements. For n=1,
  
1 yi − yi−1 yi+1 − yi

 + for N − 1 > i > 0
2 xi − xi−1 xi+1 − xi



∆yi yi+1 − yi

= for i = 0
∆xi  xi+1 − xi
yi − yi−1


for i = N − 1



xi − xi−1

If n>1, the result of the differentiation above is assigned to y and the aforemen-
tioned differentiation step is repeated until the number of those steps is equal to
n.

Example

z=diff(linspace(1,3,3),linspace(2,3,3)) returns 2, 2, 2.

See also

integrate(), sum(), max(), min()

475
integrate()
Integrate vector.

Syntax

z=integrate(y,h)

Arguments

Name Type Def. Range Required

√
y R, C, Rn , Cn ]−∞, +∞[
√
h R, C ]−∞, +∞[

Description

This function numerically integrates a vector x with respect to a differential h.

The integration method is according to the trapezoidal rule:
y yn 
R 0
f (t) dt ≈ h + y1 + y2 + . . . + yn−1 +
2 2

Example

R3
Calculate an approximation of the integral t dt using 101 points:
1

z=integrate(linspace(1,3,101)) returns 4.

See also

diff(), sum(), max(), min()

476
Signal Processing

dft()
Discrete Fourier Transform.

Syntax

y=dft(v)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[

Description

This function computes the Discrete Fourier Transform (DFT) of a vector v. The
advantage of this function compared to fft() is that the number n of components
of v is arbitrary, while for the latter n must be a power of 2. The drawbacks are
that dft() is slower and less accurate than fft().

Example

This calculates the spectrum y of a DC signal:

y
1
y=dft(linspace(1,1,7)) returns -1.59e-17+j1.59e-17
..
.
2.22e-16-j1.11e-16

Please note that in this example 7 points are used for the time vector v. Since
7 is not a power of 2, the same expression used together with the fft() function
would lead to wrong results. Note also the rounding errors where “0” would be the
correct value.

See also

idft(), fft(), ifft(), Freq2Time(), Time2Freq()

477
fft()
Fast Fourier Transform.

Syntax

y=fft(v)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[

Description

This function computes the Fast Fourier Transform (FFT) of a vector v. The
number n of components of v must be a power of 2.

Example

This calculates the spectrum y of a DC signal:

y
1
y=fft(linspace(1,1,8)) returns 0
..
.
0

See also

ifft(), dft(), idft(), Freq2Time(), Time2Freq(), fftshift()

478
idft()
Inverse Discrete Fourier Transform.

Syntax

y=idft(v)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[

Description

This function computes the Inverse Discrete Fourier Transform (IDFT) of a vector
v. The advantage of this function compared to ifft() is that the number n of
components of v is arbitrary, while for the latter n must be a power of 2. The
drawbacks are that idft() is slower and less accurate than ifft().

Example

This calculates the time function y belonging to a white spectrum:

y
7
y=idft(linspace(1,1,7)) returns -1.11e-16-j1.11e-16
..
.
1.55e-15+j7.77e-16

Please note that in this example 7 points are used for the spectrum vector v. Since
7 is not a power of 2, the same expression used together with the ifft() function
would lead to wrong results. Note also the rounding errors where “0” would be the
correct value.

See also

dft(), ifft(), fft(), Freq2Time(), Time2Freq(), fftshift()

479
ifft()
Inverse Fast Fourier Transform.

Syntax

y=ifft(v)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[

Description

This function computes the Inverse Fast Fourier Transform (IFFT) of a vector v.
The number n of components of v must be a power of 2.

Example

This calculates the time function y belonging to a white spectrum:

y
8
y=ifft(linspace(1,1,8)) returns 0
..
.
0

See also

fft(), dft(), idft(), Freq2Time(), Time2Freq(), fftshift()

480
fftshift()

Move the frequency 0 to the center of the FFT vector.

Syntax

y=fftshift(v)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[

Description

This function shuffles the FFT values of vector v in order to move the frequency 0
to the center of the vector. Below of it the components with negative frequencies
are located, above those with positive frequencies. Herewith the ”classical” look of
a spectrum as gained by a spectrum analyzer is obtained.

Example

x
1
Suppose x to be the result of a FFT of 8 elements, e.g. 2
..
.
8

The result of the FFT is sorted in such a way that the component with frequency
zero is the first element (1) of the vector. The components with positive frequency
follow (2,3,4). After that, the components with negative frequency (5,6,7,8) are
arranged, starting from the most negative value. This pattern can be exemplarily
generated in Qucs by writing x=linspace(1,8,8). Then

481
 y
5
6
7
y=fftshift(x) returns 8
1
2
3
4

As you can see, the component with frequency 0 (element 1) is moved to the middle
of the spectrum vector. Beneath of it the components with negative frequencies
appear (5,6,7,8), above those with positive frequencies (2,3,4).

See also

fft(), ifft(), dft(), idft()

482
Time2Freq()
Interpreted Discrete Fourier Transform.

Syntax

y=Time2Freq(v,t)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[
√
t Rk , Ck ]−∞, +∞[

Description

This function computes the Discrete Fourier Transform (DFT) of a vector v with
respect to a time vector t.

Example

This calculates the spectrum y(f) of a DC signal:

y=Time2Freq(linspace(1,1,7),linspace(0,1,2)) returns

Frequency y
0 1
0.167 -1.59e-17+j1.59e-17
.. ..
. .
1 2.22e-16-j1.11e-16

Please note that in this example 7 points are used for the time vector v. Note also
the rounding errors at t>0, where “0” would be the correct value.

See also

idft(), fft(), ifft(), Freq2Time()

483
Freq2Time()
Interpreted Inverse Discrete Fourier Transform.

Syntax

y=Freq2Time(v,f)

Arguments

Name Type Def. Range Required

√
v Rn , Cn ]−∞, +∞[
√
f Rk , Ck ]−∞, +∞[

Description

This function computes the Inverse Discrete Fourier Transform (IDFT) of a vector
v with respect to a frequency vector f.

Example

This calculates the time function y(t) belonging to a white spectrum:

y=Freq2Time(linspace(1,1,7),linspace(0,1,2)) returns

Frequency y
0 7
0.167 -1.11e-16-j1.11e-16
.. ..
. .
1 1.55e-15+j7.77e-16

Please note that in this example 7 points are used for the spectrum vector v. Note
also the rounding errors at t>0, where “0” would be the correct value.

See also

dft(), ifft(), fft(), Time2Freq()

484
kbd()

Kaiser-Bessel derived window.

Syntax

y=kbd(a,n)

y=kbd(a)

Arguments

Name Type Def. Range Required Default

√
a R ]−∞, +∞[
n N [1, +∞[ 64

Description

This function generates a Kaiser-Bessel window according to

v
u k
uP  q 4i


u I0 π a 1 − n
−1
u i=0
yk = u
uP n ,
t 2  q
4i


I0 π a 1 − n
−1
i=0

yn−k−1 = yk

n
for 0 ≤ k < 2

If the parameter n is not specified, n=64 is assumed.

Example

y=kbd(0.1,4) returns .

See also

dft(), ifft(), fft()

485
15.5 Electronics Functions

15.5.1 Unit Conversion

dB()
dB value.

Syntax

y=dB(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function returns the dB value of a real or complex number or vector.

y = 20 log |x|

For x being a vector the equation above is applied to the components of x.

Example

y=db(10) returns 20.

See also

log10()

486
dbm()
Convert voltage to power in dBm.

Syntax

y=dBm(u,Z0)

y=dBm(u)

Arguments

Name Type Def. Range Required Default

√
u R, C, Rn , Cn ]−∞, +∞[
Z0 R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function returns the corresponding dBm power of a real or complex voltage
or vector u. The impedance Z0 referred to is either specified or 50Ω.

|u|2
y = 10 log
Z0 0.001W

For u being a vector the equation above is applied to the components of u.

Please note that u is considered as a rms value, not as an amplitude.

Example

y=dbm(1) returns 13.

See also

dbm2w(), w2dbm(), log10()

487
dbm2w()
Convert power in dBm to power in Watts.

Syntax

y=dBm2w(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function converts the real or complex power or power vector, given in dBm,
to the corresponding power in Watts.

x
y = 0.001 10 10

For x being a vector the equation above is applied to the components of x.

Example

y=dbm2w(10) returns 0.01.

See also

dbm(), w2dbm()

488
w2dbm()
Convert power in Watts to power in dBm.

Syntax

y=w2dBm(x)

Arguments

Name Type Def. Range Required

√
x R, C, Rn , Cn ]−∞, +∞[

Description

This function converts the real or complex power or power vector, given in Watts,
to the corresponding power in dBm.

x
y = 10 log
0.001W

For x being a vector the equation above is applied to the components of x.

Example

y=w2dbm(1) returns 30.

See also

dbm(), dbm2w(), log10()

489
15.5.2 Reflection Coefficients and VSWR

rtoswr()
Converts reflection coefficient to voltage standing wave ratio (VSWR).

Syntax

s=rtoswr(r)

Arguments

Name Type Def. Range Required

√
r R, C, Rn , Cn |r| ≤ 1

Description

For a real or complex reflection coefficient r, this function calculates the corre-
sponding voltage standing wave ratio (VSWR) s according to

1 + |r|
s=
1 − |r|

VSWR is a real number and if usually given in the notation “s : 1”.

For r being a vector the equation above is applied to the components of r.

Examples

s=rtoswr(0) returns 1.

s=rtoswr(0.1+0.2*i) returns 1.58.

See also

ytor(), ztor(), rtoy(), rtoz()

490
rtoy()
Converts reflection coefficient to admittance.

Syntax

y=rtoy(r)

y=rtoy(r, Z0)

Arguments

Name Type Def. Range Required Default

√
r R, C, Rn , Cn |r| ≤ 1
Z0 R, C ]−∞, +∞[ 50

Description

For a real or complex reflection coefficient r, this function calculates the corre-
sponding admittance y according to

1 1−r
y=
Z0 1 + r

If the reference impedance Z0 is not provided, the function assumes Z0 = 50Ω.

For r being a vector the equation above is applied to the components of r.

Example

y=rtoy(0.333) returns 0.01.

See also

ytor(), ztor(), rtoswr()

491
rtoz()
Converts reflection coefficient to impedance.

Syntax

z=rtoz(r)

z=rtoz(r, Z0)

Arguments

Name Type Def. Range Required Default

√
r R, C, Rn , Cn |r| ≤ 1
Z0 R, C ]−∞, +∞[ 50

Description

For a real or complex reflection coefficient r, this function calculates the corre-
sponding impedance Z according to

1−r
Z = Z0
1+r

If the reference impedance Z0 is not provided, the function assumes Z0 = 50Ω.

For r being a vector the equation above is applied to the components of r.

Example

z=rtoz(0.333) returns 99.9.

See also

ztor(), ytor(), rtoswr()

492
ytor()
Converts admittance to reflection coefficient.

Syntax

r=ytor(Y)

r=ytor(Y, Z0)

Arguments

Name Type Def. Range Required Default

√
Y R, C, Rn , Cn ]−∞, +∞[
Z0 R, C ]−∞, +∞[ 50

Description

For a real or complex admittance y, this function calculates the corresponding

reflection coefficient according to

1 − Y Z0
r=
1 + Y Z0

For Y being a vector the equation above is applied to the components of Y.

If the reference impedance Z0 is not provided, the function assumes Z0 = 50Ω.

Often a dB measure is given for the reflection coefficient, the so called “return
loss”:

RL = −20 log |r| [dB]

Example

r=ytor(0.01) returns 0.333.

See also

rtoy(), rtoz(), rtoswr(), log10(), dB()

493
ztor()
Converts impedance to reflection coefficient.

Syntax

r=ztor(Z)

r=ztor(Z, Z0)

Arguments

Name Type Def. Range Required Default

√
Z R, C, Rn , Cn ]−∞, +∞[
Z0 R, C ]−∞, +∞[ 50

Description

For a real or complex impedance Z, this function calculates the corresponding

reflection coefficient according to

Z − Z0
r=
Z + Z0

For Z being a vector the equation above is applied to the components of Z.

If the reference impedance Z0 is not provided, the function assumes Z0 = 50Ω.

Often a dB measure is given for the reflection coefficient, the so called “return
loss”:

RL = −20 log |r| [dB]

Example

r=ztor(100) returns 0.333.

See also

rtoz(), rtoy(), rtoswr(), log10(), dB()

494
15.5.3 N-Port Matrix Conversions

stos()

Converts S-parameter matrix to S-parameter matrix with different reference

impedance(s).

Syntax

y=stos(S, Zref)

y=stos(S, Zref, Z0)

Arguments

Name Type Def. Range Required Default

|Sij | ∈ ]−∞, +∞[ , 1 ≤ i, j ≤ n √
S Rn×n , Cn×n
|Sii | ≤ 1, 1 ≤ i ≤ n
√
Zref R, C, Rn , Cn ]−∞, +∞[
Z0 R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function converts a real or complex scattering parameter matrix S into a scat-
tering matrix Y. S has a reference impedance Zref, whereas the created scattering
matrix Y has a reference impedance Z0.

If the reference impedance Z0 is not provided, the function assumes Z0 = 50Ω.

Both Zref and Z0 can be real or complex numbers or vectors; in the latter case
the function operates on the elements of Zref and Z0.

Example

Conversion of 50Ω terminated S-parameters to 100Ω terminated S-parameters:

-0.241 0
S2=stos(eye(2)*0.1,50,100) returns .
0 -0.241

See also

495
twoport(), stoy(), stoz()

496
stoy()
Converts S-parameter matrix to Y-parameter matrix.

Syntax

Y=stoy(S)

Y=stoy(S, Zref)

Arguments

Name Type Def. Range Required Default

|Sij | ∈ ]−∞, +∞[ , 1 ≤ i, j ≤ n √
S Rn×n , Cn×n
|Sii | ≤ 1, 1 ≤ i ≤ n
Zref R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function converts a real or complex scattering parameter matrix S into an

admittance matrix Y. S has a reference impedance Zref, which is assumed to be
Zref = 50Ω if not provided by the user.

Zref can be real or complex number or vector; in the latter case the function
operates on the elements of Zref.

Example

0.00818 0
Y=stoy(eye(2)*0.1,100) returns .
0 0.00818

See also

twoport(), stos(), stoz(), ytos()

497
stoz()
Converts S-parameter matrix to Z-parameter matrix.

Syntax

Z=stoz(S)

Z=stoz(S, Zref)

Arguments

Name Type Def. Range Required Default

|Sij | ∈ ]−∞, +∞[ , 1 ≤ i, j ≤ n √
S Rn×n , Cn×n
|Sii | ≤ 1, 1 ≤ i ≤ n
Zref R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function converts a real or complex scattering parameter matrix S into an

impedance matrix Z. S has a reference impedance Zref, which is assumed to be
Zref = 50Ω if not provided by the user.

Zref can be real or complex number or vector; in the latter case the function
operates on the elements of Zref.

Example

122 0
Z=stoz(eye(2)*0.1,100) returns .
0 122

See also

twoport(), stos(), stoy(), ztos()

498
twoport()
Converts a two-port matrix from one representation into another.

Syntax

U=twoport(X, from, to)

Arguments

Name Type Def. Range Required

√
X R , C2×2
2×2
]−∞, +∞[
√
from Character {0 Y 0 , 0 Z 0 , 0 H 0 , 0 G0 , 0 A0 , 0 S 0 , 0 T 0 }
√
to Character {0 Y 0 , 0 Z 0 , 0 H 0 , 0 G0 , 0 A0 , 0 S 0 , 0 T 0 }

Description

This function converts a real or complex two-port matrix X from one representa-
tion into another.

Example

Transfer a two-port Y matrix Y1 into a Z matrix:

Y1=eye(2)*0.1

10 0
Z1=twoport(Y1,’Y’,’Z’) returns .
0 10

See also

stos(), ytos(), ztos(), stoz(), stoy(), ytoz(), ztoy()

499
ytos()
Converts Y-parameter matrix to S-parameter matrix.

Syntax

S=ytos(Y)

S=ytos(Y, Z0)

Arguments

Name Type Def. Range Required Default

√
Y R , Cn×n
n×n
]−∞, +∞[
Z0 R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function converts a real or complex admittance matrix Y into a scattering

parameter matrix S. Y has a reference impedance Z0, which is assumed to be Z0
= 50Ω if not provided by the user.

Z0 can be real or complex number or vector; in the latter case the function operates
on the elements of Z0.

Example

-0.818 0
S=ytos(eye(2)*0.1,100) returns .
0 -0.818

See also

twoport(), stos(), ztos(), stoy()

500
ytoz()
Converts Y-parameter matrix to Z-parameter matrix.

Syntax

Z=ytoz(Y)

Arguments

Name Type Def. Range Required

√
Y R , Cn×n
n×n
]−∞, +∞[

Description

This function converts a real or complex admittance matrix Y into an impedance

matrix Z.

Example

10 0
Z=ytoz(eye(2)*0.1) returns .
0 10

See also

twoport(), ztoy()

501
ztos()
Converts Z-parameter matrix to S-parameter matrix.

Syntax

S=ztos(Z)

S=ztos(Z, Z0)

Arguments

Name Type Def. Range Required Default

√
Z R , Cn×n
n×n
]−∞, +∞[
Z0 R, C, Rn , Cn ]−∞, +∞[ 50

Description

This function converts a real or complex impedance matrix Z into a scattering

parameter matrix S. Z has a reference impedance Z0, which is assumed to be Z0
= 50Ω if not provided by the user.

Z0 can be real or complex number or vector; in the latter case the function operates
on the elements of Z0.

Example

-0.998 0
S=ztos(eye(2)*0.1,100) returns .
0 -0.998

See also

twoport(), twoport(), stos(), ytos(), stoz()

502
ztoy()
Converts Z-parameter matrix to Y-parameter matrix.

Syntax

Y=ztoy(Z)

Arguments

Name Type Def. Range Required

√
Z R , Cn×n
n×n
]−∞, +∞[

Description

This function converts a real or complex impedance matrix Z into an admittance

matrix Y.

Example

10 0
Y=ztoy(eye(2)*0.1) returns .
0 10

See also

twoport(), ytoz()

503
15.5.4 Amplifiers

GaCircle()
Circle(s) with constant available power gain Ga in the source plane.

Syntax

y=GaCircle(X,Ga,v)

y=GaCircle(X,Ga,n)

y=GaCircle(X,Ga)

Arguments

Name Type Def. Range Required Default

√
X R2×2×p , C2×2×p ]−∞, +∞[
v Rn [0, 360]o
√
Ga R, Rm [0, +∞[
n N [2, +∞[ 64

Description

This function generates the points of the circle of constant available power gain GA
in the complex source plane (rS ) of an amplifier. The amplifier is described by a
two-port S-parameter matrix S. Radius r and center c of this circle are calculated
as follows:
q
1 − 2 · K · gA · |S12 S21 | + gA2 · |S12 S21 |2 ∗
gA (S11 − S22 ∆∗ )
r= 1 + gA · |S11 |2 − |∆|2 
 and c =
,
1 + gA |S11 |2 − |∆|2



GA
where gA = and K Rollet stability factor. ∆ denotes determinant of S.
|S21 |2

The points of the circle can be specified by the angle vector v, where the angle
must be given in degrees. Another possibility is to specify the number n of angular
equally distributed points around the circle. If no additional argument to X is
given, 64 points are taken. The available power gain can also be specified in a
vector Ga, leading to the generation of m circles, where m is the size of Ga.

504
Please also refer to “Qucs - Technical Papers”, chapter 1.5.

Example

v=GaCircle(S)

See also

GpCircle(), Rollet()

505
GpCircle()
Circle(s) with constant operating power gain Gp in the load plane.

Syntax

y=GpCircle(X,Gp,v)

y=GpCircle(X,Gp,n)

y=GpCircle(X,Gp)

Arguments

Name Type Def. Range Required Default

2×2×p
√
X R , C2×2×p ]−∞, +∞[
v Rn [0, 360]o
√
Gp R, Rm [0, +∞[
n N [2, +∞[ 64

Description

This function generates the points of the circle of constant operating power gain
GP in the complex load plane (rL ) of an amplifier. The amplifier is described by a
two-port S-parameter matrix S. Radius r and center c of this circle are calculated
as follows:
q
1 − 2 · K · gP · |S12 S21 | + gP2 · |S12 S21 |2 ∗
gA (S22 − S11 ∆∗ )
r= and c =
,
1 + gP · |S22 |2 − |∆|2  1 + gP |S22 |2 − |∆|2



GP
where gA = and K Rollet stability factor. ∆ denotes determinant of S.
|S21 |2

The points of the circle can be specified by the angle vector v, where the angle
must be given in degrees. Another possibility is to specify the number n of angular
equally distributed points around the circle. If no additional argument to X is
given, 64 points are taken. The available power gain can also be specified in a
vector Gp, leading to the generation of m circles, where m is the size of Gp.

Please also refer to “Qucs - Technical Papers”, chapter 1.5.

506
Example

v=GpCircle(S)

See also

GaCircle(), Rollet()

507
Mu()
Mu stability factor of a two-port S-parameter matrix.

Syntax

y=Mu(S)

Arguments

Name Type Def. Range Required

2×2×p 2×2×p
√
S R ,C ,R2×2 , C2×2 ]−∞, +∞[

Description

This function returns the Mu stability factor µ of an amplifier being described by

a two-port S-parameter matrix S :

1 − |S11 |2
µ= ∗
|S22 − S11 ∆| + |S21 S12 |

∆ denotes determinant of S.

The amplifier is unconditionally stable if µ > 1.

For S being a vector of matrices the equation above is applied to the sub-matrices
of S.

Example

m=Mu(S)

See also

Mu2(), Rollet(), StabCircleS(), StabCircleL()

508
Mu2()
Mu’ stability factor of a two-port S-parameter matrix.

Syntax

y=Mu2(S)

Arguments

Name Type Def. Range Required

2×2×p 2×2×p
√
S R ,C ,R2×2 , C2×2 ]−∞, +∞[

Description

This function returns the Mu’ stability factor µ0 of an amplifier being described
by a two-port S-parameter matrix S :

0 1 − |S22 |2
µ = ∗
|S11 − S22 ∆| + |S21 S12 |

∆ denotes determinant of S.

The amplifier is unconditionally stable if µ0 > 1.

For S being a vector of matrices the equation above is applied to the sub-matrices
of S.

Example

m=Mu2(S)

See also

Mu2(), Rollet(), StabCircleS(), StabCircleL()

509
NoiseCircle()

Generates circle(s) with constant Noise Figure(s).

Syntax

y=NoiseCircle(Sopt,Fmin,Rn,F,v)

y=NoiseCircle(Sopt,Fmin,Rn,F,n)

y=NoiseCircle(Sopt,Fmin,Rn,F)

Arguments

Name Type Def. Range Required Default

√
Sopt Rn , Cn ]−∞, +∞[
√
Fmin Rn [1, +∞[
√
Rn Rn , Cn [0, +∞[
√
F R, Rn [1, +∞[
v Rn [0, 360]o
n N [2, +∞[ 64

Description

This function generates the points of the circle of constant Noise Figure (NF) F
in the complex source plane (rS ) of an amplifier. Generally, the amplifier has its
minimum NF Fmin , if the source reflection coefficient rS = Sopt (noise matching).
Note that this state with optimum source reflection coefficient Sopt is different
from power matching ! Thus power gain under noise matching is lower than the
maximum obtainable gain. The values of Sopt , Fmin and the normalised equivalent
noise resistance Rn /Z0 can be usually taken from the data sheet of the amplifier.

Radius r and center c of the circle of constant NF are calculated as follows:

q
N 2 + N · 1 − |Sopt |2


Sopt F − Fmin
r= 1+N and c = , with N = · Z0 · |1 + Sopt |2
1+N 4 Rn
.

The points of the circle can be specified by the angle vector v, where the angle
must be given in degrees. Another possibility is to specify the number n of angular

510
equally distributed points around the circle. If no additional argument to X is
given, 64 points are taken.

Please also refer to “Qucs - Technical Papers”, chapter 2.2.

Example

v=NoiseCircle(Sopt,Fmin,Rn,F)

See also

GaCircle(), GpCircle()

511
PlotVs()
Returns a data item based upon vector or matrix vector with dependency on
a given vector.

Syntax

y=PlotVs(X, v)

Arguments

Name Type Def. Range Required

√
X R , C , Rm×n×p , Cm×n×p
n n
]−∞, +∞[
√
v Rn , Cn ]−∞, +∞[

Description

This function returns a data item based upon a vector or matrix vector X with
dependency on a given vector v.

Example

PlotVs(Gain,frequency/1E9).

See also

512
Rollet()
Rollet stability factor of a two-port S-parameter matrix.

Syntax

y=Rollet(S)

Arguments

Name Type Def. Range Required

2×2×p 2×2×p
√
S R ,C ,R2×2 , C2×2 ]−∞, +∞[

Description

This function returns the Rollet stability factor K of an amplifier being described
by a two-port S-parameter matrix S :

1 − |S11 |2 − |S22 |2 + |∆|2

K=
2 |S21 | |S12 |

∆ denotes determinant of S.

The amplifier is unconditionally stable if K > 1 and |∆| < 1.

Note that a large K may be misleading in case of a multi-stage amplifier, pretending

extraordinary stability. This is in conflict with reality where a large gain amplifier
usually suffers from instability due to parasitics.

For S being a vector of matrices the equation above is applied to the sub-matrices
of S.

Example

K=Rollet(S)

See also

Mu(), Mu2(), StabCircleS(), StabCircleL()

513
StabCircleL()
Stability circle in the load plane.

Syntax

y=StabCircleL(X)

y=StabCircleL(X,v)

y=StabCircleL(X,n)

Arguments

Name Type Def. Range Required Default

2×2×p
√
X R , C2×2×p ]−∞, +∞[
v Rn [0, 360]o
n N [2, +∞[ 64

Description

This function generates the stability circle points in the complex load reflection
coefficient (rL ) plane of an amplifier. The amplifier is described by a two-port S-
parameter matrix S. Radius r and center c of this circle are calculated as follows:

∗ ∗
 
 and c = S22 − S11 · ∆
 S21 S12 
r=

|S22 |2 − |∆|2  |S22 |2 − |∆|2

∆ denotes determinant of S.

The points of the circle can be specified by the angle vector v, where the angle
must be given in degrees. Another possibility is to specify the number n of angular
equally distributed points around the circle. If no additional argument to X is
given, 64 points are taken.

If the center of the rL plane lies within this circle and |S11 | ≤ 1 then the circuit is
stable for all reflection coefficients inside the circle. If the center of the rL plane lies
outside the circle and |S11 | ≤ 1 then the circuit is stable for all reflection coefficients
outside the circle (please also refer to “Qucs - Technical Papers”, chapter 1.5).

Example

514
v=StabCircleL(S)

See also

StabCircleS(), Rollet(), Mu(), Mu2()

515
StabCircleS()
Stability circle in the source plane.

Syntax

y=StabCircleS(X)

y=StabCircleS(X,v)

y=StabCircleS(X,n)

Arguments

Name Type Def. Range Required Default

2×2×p
√
X R , C2×2×p ]−∞, +∞[
v Rn [0, 360]o
n N [2, +∞[ 64

Description

This function generates the stability circle points in the complex source reflection
coefficient (rS ) plane of an amplifier. The amplifier is described by a two-port S-
parameter matrix S. Radius r and center c of this circle are calculated as follows:

∗ ∗
 
 and c = S11 − S22 · ∆
 S21 S12 
r=

|S11 |2 − |∆|2  |S11 |2 − |∆|2

∆ denotes determinant of S.

The points of the circle can be specified by the angle vector v, where the angle
must be given in degrees. Another possibility is to specify the number n of angular
equally distributed points around the circle. If no additional argument to X is
given, 64 points are taken.

If the center of the rS plane lies within this circle and |S22 | ≤ 1 then the circuit is
stable for all reflection coefficients inside the circle. If the center of the rS plane lies
outside the circle and |S22 | ≤ 1 then the circuit is stable for all reflection coefficients
outside the circle (please also refer to “Qucs - Technical Papers”, chapter 1.5).

Example

516
v=StabCircleS(S)

See also

StabCircleL(), Rollet(), Mu(), Mu2()

517
StabFactor()
Stability factor of a two-port S-parameter matrix. Synonym for Rollet()

Syntax

y=StabFactor(S)

See also

Rollet()

518
StabMeasure()
Stability measure B1 of a two-port S-parameter matrix.

Syntax

y=StabMeasure(S)

Arguments

Name Type Def. Range Required

2×2×p 2×2×p
√
S R ,C ,R2×2 , C2×2 ]−∞, +∞[

Description

This function returns the stability measure B1 of a two-port S-parameter matrix

S:

B1 = 1 + |S11 |2 − |S22 |2 − |∆|2

∆ denotes determinant of S.

The amplifier is unconditionally stable if B1 > 0 and the Rollet factor K > 1.

For S being a vector of matrices the equation above is applied to the sub-matrices
of S.

Example

B1=StabMeasure(S)

See also

Rollet(), Mu(), Mu2(), StabCircleS(), StabCircleL()

519
vt()
Thermal voltage for a given temperature in Kelvin.

Syntax

y=vt(t)

Arguments

Name Type Def. Range Required Default

√
t R [0, +∞[

Description

This function returns the corresponding thermal voltage Vt in Volt of a real absolute
temperature (vector) T in Kelvin according to

kT
Vt =
e

where k is the Boltzmann constant and e denotes the electrical charge on the
electron. For t being a vector the equation above is applied to the components of
k.

Please note that t is always larger than or equal to zero.

Example

y=vt(300) returns 0.0259.

520
16 Component, compact device and
circuit modelling using symbolic
equations

16.1 Introduction

Qucs releases 0.0.11 and 0.0.12 mark a turning point in the development of the
Qucs component and circuit modelling facilities. Release 0.0.11 introduced com-
ponent values defined by equations and for the first time allowed subcircuits with
parameters. Release 0.0.12 extends these features to add model development using
symbolic equations that are similar to compact device code written in the Verilog-
A modelling language. In designing the latest Qucs modelling features the Qucs
team has made a central focus of their work the need to provide the package with
an interactive and easy to use modelling system which allows fast model prototype
construction. Much of these new aspects have up to now been undocumented and
are likely to be very new to most Qucs users. The aim of this tutorial note is to
outline the background to these important package extensions and to provide real
help to Qucs users who are interested in writing and experimenting with their own
models. The text includes a number of illustrative examples for readers to try and
experiment with.

16.2 Qucs electronic device and circuit modelling

Circuit simulation packages are complex software systems which often take years
to mature to a stage where they are capable of analysing the current generation of
integrated and discrete electronic circuits. Most circuit simulators have a number
of common basic attributes; firstly circuits are represented by a textual netlist or
a schematic diagram which contains all the information required by a simulator

521
to analyse the performance of a circuit, and secondly a simulation engine which
undertakes the calculation of circuit performance in one or more different circuit
domains such as DC, AC or transient, and thirdly a post simulation processing
system which structures and displays the simulation data in both tabular and
graphical forms. All circuit simulators have one other important attribute, namely
that they represent individual electronic components by a model, or abstraction,
in a way that can be understood and analysed by the simulation engine when
undertaking a simulation task. Without component models the science of circuit
simulation would not have developed to the stage it has today. From a users point
of view component models are the key to simulator productivity; the greater the
number of different models the easier it becomes to analyse mixed analogue and
digital electronic systems.

Shown in Fig. 16.1 is a block diagram of the analogue component modelling and
simulation facilities currently provided by the Qucs package. The diagram is struc-
tured as a flow chart which emphasises the different device modelling routes. When
Qucs was first released only two of these were available for users to develop new de-
vice models. The first of these has been used extensively by the package developers
to construct the built-in models that are distributed with each Qucs release. This
fundamental route involves hand coding the C++ code for a new model1 , its com-
pilation and linking with the core Qucs C++ code. Obviously, this does require
a specialised knowledge of the Qucs model programming interface2 , the necessary
C++ skills, including a good working knowledge of the Trolltech Qt toolkit3 . At
the time of writing these notes the latest device to be added to Qucs using this
approach is the exponential pulse source4 . Models based on hand written C++
code are normally restricted to basic devices that form the fundamental compo-
nent core of a simulator - particularly where simulation computational efficiency
is important. One disadvantage of this approach, is the obvious one, in that the
time to implement a new model increases disproportionately with increasing model
complexity. For most Qucs users this route would not be the most natural to use
when developing new models. However, for the specialist who spends a significant
amount of time researching new device models this has always in the past, been the

1
The technical details of the built-in models are described in: Qucs Technical Papers, Stefan
Jahn, Michael Margraf, Vincent Habchi and Raimund Jacob, http://qucs.sourceforge.
net/technical.html.
2
Writing the documentation for the Qucs model programming interface is on the to do list and
will be completed, when time allows, sometime in the future.
3
Qt is a registered trademark of Trolltech, Norway; http://www.trolltech.com/copyright.
4
Added by Gunther Kraut on 15 April 2007. This device has been added for compatibility with
SPICE.

522
 Verilog-A Hand coded Nonlinear equation Component SPICE SPICE SPICE
Compact device model defined devices data processing netlist preprocessor parameterised
device C++ code using Qucs netlist
code equations

User
defined
ADMS subcircuit
compiler schematic
QUCS GUI capture
Schematic symbol
capture Circuit
symbols entered using Qucs
schematic capture Library
components
C++ code

User defined Qucs

subcircuits components
Symbolic
equations
Qucs Tools:
Line Calculator
Attenuator Design
Matching Circuits
Generate Qucs netlist code Filter Design
from GUI schematic, including
Simulate conversion of SPICE code to
Qucs format

Post simulation
data processing
QUCSATOR using Qucs
equations
C++ component code
compiled and linked
to Qucsator core C++
code via API

Qucs plots
Simulation output data and tables

Figure 16.1: Qucs analogue component modelling and simulation block diagram
(not including optimisation)

route of choice. Unfortunately, modern semiconductor device models are becom-

ing so complex that the model development time can stretch into months or even
years and requires typically thousands of lines of C or C++ code to characterise a
model5 . With the more complex models the problem of finding bugs in the model
code also acts as a limit to fast model development.

For the average Qucs user their first introduction to the software is probably
through constructing circuit schematics made entirely from the standard com-
5
A good introduction to writing compact device models is given in “How to (and how not to)
write a compact model in Verilog-A”, Geoffrey J. Coram, 2004, Proc. 2004 IEEE International
Behavioral Modeling and Simulation Conference (BMAS 2004), pp 97- 106.

523
ponent models built into the package and the testing of their performance by
launching the simulator from one of the Qucs simulation icons.6 The next natural
stage in the Qucs modelling and simulation learning curve is the use of subcircuits
where groups of built-in components are collected together to form a higher level
circuit block. These blocks are often arranged with a common theme, forming
a Qucs library. The process of modelling new devices/circuits is normally done
by connecting existing component models and user defined subcircuits. With this
type of modelling higher level functional models can only be constructed from ex-
isting fundamental components or previously constructed subcircuits. Engineers
often call this approach to modelling, macromodelling. Qucs releases up to 0.0.10
relied on macromodelling for functional model development via the Qucs schematic
interface. This route remains popular amongst most Qucs users because it is easy
to understand, is fully interactive and allows straight forward testing of new mod-
els. One feature that is common to all components included in Qucs releases up
to 0.0.10 may not be immediately obvious to readers, namely that, with the ex-
ception of sweep variables, component values could only be numbers, for example
R1 = 1k, and were not allowed to be represented by algebraic expressions like R1
= Value1, where Value1 = 100.0+50 · X. Its also worth pointing out at this point
that during simulation, again performed by Qucs releases up to 0.0.10, component
values were required to remain constant and could not be a function of the circuit
variables such as voltage, current or charge.

One way to remove the component value restrictions imposed by early Qucs releases
is to model devices and circuits using preprocessor extended forms of the SPICE
netlist language. Circuit design equations can then be embedded in SPICE netlists
and the calculation of component values completed by the SPICE preprocessor.
Both the SPICE to Qucs and OP AMP tutorials7 outline in detail the steps required
to merge circuit design and simulation in this way. This modelling route is a very
important and powerful model development tool. So much so that ongoing tests to
identify how compatible Qucs is with the industrial standard SPICE 2g6 and 3f5
syntax are currently being undertaken as part of the Qucs development schedule8 .
Although perfectly viable as a model development tool the use of an extended
SPICE netlist language has a number of serious disadvantages, namely that not all
the Qucs built-in component models have equivalent SPICE models and secondly

6
The “Getting Started with Qucs” tutorial by Stefan Jahn outlines a number of basic simulation
techniques; http://qucs.sourceforge.net/docs.html.
7
Qucs simulation of SPICE netlists and Modelling Operational Amplifiers, Mike Brinson, http:
//qucs.sourceforge.net/docs.html.
8
Qucs: Report Book; SPICE to Qucs test reports, Mike Brinson, http://qucs.sourceforge.
net/docs.html.

524
text netlists are the only entry medium for describing models.

The previous paragraphs give a brief statement of the different component mod-
elling routes that were available up to release 0.0.11. Qucs 0.0.11 is very much a
modelling water shed in that symbolic equations were introduced for the calcula-
tion of component values, previously equations were only allowed when structuring
simulation output data for post simulation listing or plotting. Release 0.0.11 allows
the following types of variable;

1. sweep variables,

2. equations left hand side,

3. component parameter’s left hand side (e.g. R1.R),

4. subcircuit parameters and

5. simulation output data.

With each Qucs release the number of analysis functions, and other data process-
ing features, included in the Qucs equation set continues to expand9 . From release
0.0.11 parameters are also allowed with subcircuits so that data can be passed
to a model. This allows generalised subcircuit/macromodels to be developed for
popular devices such as operational amplifiers. Through the use of embedded
design equations within subcircuits and parameter passing it became possible to
construct powerful models that mix both circuit design procedures and the cal-
culation of individual component values. Qucs 0.0.11 still imposed the restriction
that equations could not be functions of voltage, current or charge.

With the release of Qucs 0.0.12 the voltage, current and charge restrictions imposed
on equations will finally be relaxed. The introduction of a new device modelling
component called the equation defined device (EDD) allows firstly device current to
be formulated as a function of voltage, and secondly device charge to be calculated
as a function of voltage and current. The syntax adopted for the new model
borrows heavily on the compact device modelling approach taken by the Verilog-A
modelling language.

9
See Measurement Expressions Reference Manual, Gunther Kraut and Stefan Jahn, http:
//qucs.sourceforge.net/docs.html.

525
Some readers will probably have noted that so far these notes make no reference
to the ADMS model development route illustrated in Fig. 16.1. ADMS stands for
Automated device model synthesizer10 and includes a Verilog-A to C/C++ com-
piler. It allows compact device models to be described in the Verilog-A language
then compiled to C/C++ and the resulting code linked with the Qucs core simula-
tion code11 . Model development using ADMS is similar to the fundamental hand
coded C++ model development route except that model development is greatly
simplified by the power of the high level Verilog-A language. A strong relation-
ship exists between the ADMS and EDD modelling procedures in that EDD can
be considered a fast interactive model prototyping method whose equations can
easily be expressed in Verilog-A and compiled into C/C++ code for permanent
inclusion in the Qucs simulator12 .

The opening paragraphs attempt to outline the available device modelling tech-
niques that are central to the functioning of the Qucs package. The remaining
sections of this tutorial note are devoted to illustrating the power of Qucs mod-
elling through the introduction of a number of illustrative examples. Initially these
start from a simple, and hopefully familiar, point and then proceed to more com-
plex examples which present many of the concepts lightly touched upon in the
opening text.

16.3 Extending circuit simulation capabilities with

equations

Just adding component value calculations, via equations, to a circuit simulator

immediately increases the underlying design and simulation capabilities way be-
yond that found in earlier generation simulators. Consider the simple RC circuit
shown in Fig. 16.2. Capacitor Cap is stepped from 0.1µF to 1.1µF and the small
signal AC response of the network calculated. In this example the values for both
R1 and Cap are given as numeric values. The simulation test shows the effect of
stepping the value of one component through a series of values and recording the
effect of component changes on circuit performance. In other words this is a clas-
10
L.Lemaitre, C.C. McAndrew, and S. Hamm, ADMS - Automated Device Model Synthesizer,
Proc. IEEE CICC, 2002.
11
For more details see, Qucs Description: Verilog-AMS interface, Stefan Jahn and Hélène Par-
ruitte, http://qucs.sourceforge.net/docs.html.
12
Appendix A gives an operator and function comparison table for Qucs and Verilog-A.

526
sical circuit analysis use of a circuit simulator. In a real design situation different
data is often required. Most designers would prefer to find the value of Cap that
gives a specific RC cut-off frequency (fc ) for a specified value of R1. This is the
type of investigative problem where adding equations into the simulation process
generates more informative results. Shown in Fig. 16.3 is a similar RC network to
that illustrated in Fig. 16.2.

Capacitor voltage V Cap is given by:

V1
VCap = p (16.1)
1 + ω 2 · C12 · R12
where the cut-off frequency in the voltage transfer function is
1
fc = (16.2)
2π · R1 · C1

Hence, by expressing Cap as a function of fc and stepping fc through a range of

frequencies, the effect of capacitance changes on the voltage transfer function can
be found. More importantly a nomogram of Cap values against fc can be plotted
giving the circuit designer a visual aid for determing the value of Cap required
for given values of R1 and fc . Although the circuits shown in Figs. 16.2 and 16.3
are very basic they do demonstrate how much more powerful a circuit simulator
becomes when component values are calculated using equations.

16.3.1 Low pass active filter design with embedded design

equations

In this section a more advanced circuit design example is introduced to illustrate

the power of embedded design equations in a Qucs simulation schematic. A second
order Sallen-Key low pass filter is employed for this task because it is so well known
and most readers are likely to have met it’s design in the past. A second order low
pass filter is represented by the voltage transfer function:

Vout A0
A(S) = = (16.3)
Vin (1 + a2 · S + b2 · S 2 )

where A0 is the passband DC gain and coefficients a2 , b2 are for Bessel, Butter-
worth, Tschebyscheff or similar polynomials.

527
 VCap

Parameter V1 C1
sweep U=1 V R1 C=Cap
R=1k

SW1
Sim=AC1
Type=lin
Param=Cap
Start=0.1u
Stop=1.1u
Points=11 1
VCap.v

ac simulation 0.5

AC1
Type=log
Start=1Hz
Stop=1 MHz
Points=61 0

1 10 100 1e3 1e4 1e5 1e6

acfrequency

1e-6
Cap

5e-7

0
2 4 6 8 10
number

Figure 16.2: A simple RC circuit simulation using numerical component values

528
 VCap

ac simulation Parameter V1 C1
sweep U=1 V R1 C=Cap
R=Rvalue
AC1
Type=log
Start=1Hz SW1
Stop=1 MHz Sim=AC1
Points=61 Type=log
Param=fc
Start=10
Stop=1000 1
Points=21
VCap.v

0.5
Equation

Eqn1
Rvalue=1000
0
Cap=1/(2*pi*Rvalue*fc)
1 10 100 1e3 1e4 1e5 1e6
acfrequency

1e-5
Cap

1e-6

1e-7
10 100 1e3
fc

Figure 16.3: A simple RC circuit simulation employing equation determined com-

ponent values

529
The following list13 gives the second order coefficients for the Bessel → 1.3617,
0.618; Butterworth → 1.4142, 1.000; and 3dB ripple Tschebyscheff → 1.065,
1.9305, polynomials. The second order Sallen-Key low pass filter circuit is shown
in Fig. 16.4. This circuit has a voltage gain transfer function given by:

A0
A(S) =
1 + ωc · [C1 · (R1 + R2 ) + (1 − A0) · R1 · C2 ] · S + ωc2 · R1 · R2 · C1 · C2 · S 2
(16.4)
where
R3
A0 = 1 + (16.5)
R4

This can be simplified by letting R1 = R2 = R and C1 = C2 = C; the transfer

function then becomes:

A0
A(S) = . (16.6)
1 + [ωc · R · C · (3 − A0)] · S + (ωc · R · C)2 · S 2
 

By comparison
a2 = ωc · R · C · (3 − A0) (16.7)
and
b2 = (ωc · R · C)2 (16.8)

Fixing C and solving for R and A0, yields

√
b2 a2
R= , and A0 = 3 − √ . (16.9)
ωc · C b2
Also once A0 is known the value for R4 can be calculated using equation
R3
A0 = 1 + . (16.10)
R4
Hence by providing values for C and R3 the values for R and A0, and of course
R4, can be determined for a specified cut off frequency f c. Figure 16.5 shows the
final design schematic and the simulation results for this example. A number of
important observations can be made from Fig. 16.5:
13
See OP Amps for everyone, Chapter 16: Active filter design technology, Texas Instruments,
August 2002, SL0D006B, PP 16.1,16.63.

530
 C2

SUB1

+ VEE V2
V1 C1 U=15 V
R1 R2 Vout
U=1 V
OPA27(TI)

VCC
- V3
U=15 V

R3 R4

Figure 16.4: The Sallen-Key lowpass active filter circuit

1. One or more equation blocks hold both design and post simulation data
processing equations plus assignments for named items: C, f c and R3 are
given numerical values, the a and b polynomial coefficients are set to the
values introduced in the text, and finally the design equations for R, A0 and
R4 calculations are listed.

2. The order of entries in equation blocks is not important because Qucs auto-
matically sorts out the data it requires when calculating equations.

3. The lefthand quantities in the assignment entries in the equation blocks are
linked to the component values in the schematic, see for example C and R.

4. The OP27 operational amplifier model is from the modified Qucs 0.0.11
OPAMP library. This model was generated using the SPICE to Qucs mod-
elling route.

5. To design and simulate a Sallen-Key low pass filter with a different cut off
frequency14 simply change the value of f c and rerun the Qucs simulator.

6. On completion of a simulation, pressing key F5 (Show last messages) causes

the simulation log to be displayed. This includes the calculated values of the
14
If the design calculations result in impractical values for the filter components then the value
of C should be changed and the simulation repeated.

531
 C2
C=C

SUB1

+ VEE V2
V1 C1 U=15 V
R2 R1 Vout
U=1 V C=C
R=R R=R OPA27(TI)

VCC
dc simulation - V3
U=15 V
DC1

ac simulation
Equation R3 R4
AC1 R=R3_calc R=R4_calc
Type=log
Eqn1 Start=1 Hz
C=22e-9 Stop=100 kHz 4
a2=1.065 Points=101
b2=1.9305
fc=3000
R=sqrt(b2)/(2*pi*fc*C)
A0=3-a2/(sqrt(b2))
Vout.v

R3_calc=4700 2
R4_calc=(A0-1)*R3_calc
gain_dB=dB(Vout.v)
gain_phase=rad2deg(unwrap(angle(Vout.v)))

1 10 100 1e3 1e4 1e5

acfrequency

0
0
gain_phase
gain_dB

-100

-50
-200
1 10 100 1e3 1e4 1e5 1 10 100 1e3 1e4 1e5
acfrequency acfrequency

Figure 16.5: The Sallen-Key lowpass active filter schematic with embedded design
equations

532
 components and the netlist for the circuit, see Fig. 16.6.

7. One final point of significance that some readers may have noticed - all
numerical values in equation blocks must be specified in scientific notation;
electronic notation like 1k or 3nF is not allowed15 .

15
In long term it is expected that electronic notation will be allowed. The changes for this are
on the to do list but at the moment the work has a low priority.

533
Output :
−−−−−−−
n e t l i s t content
13 R i n s t a n c e s
5 C instances
2 VCCS i n s t a n c e s
5 CCCS i n s t a n c e s
2 VCVS i n s t a n c e s
1 CCVS i n s t a n c e s
8 Vdc i n s t a n c e s
1 Idc i n s t a n c e s
1 Vac i n s t a n c e s
4 Diode i n s t a n c e s
2 BJT i n s t a n c e s
1 DC i n s t a n c e s
1 AC i n s t a n c e s
creating netlist . . .
c h e c k e r n o t i c e , v a r i a b l e ‘ Vout . v ’ in e q u a t i o n ‘ g a i n dB ’ not y e t d e f i n e d
c h e c k e r n o t i c e , v a r i a b l e ‘ Vout . v ’ in e q u a t i o n ‘ g a i n phase ’ not y e t d e f i n e d
kB = 1 . 38065 e −23
e = 2 . 71828
p i = 3 . 14159
C = 2 . 2 e −08
a2 = 1 . 065
b2 = 1 . 9305
f c = 3000
R = 3350 . 51
A0 = 2 . 2335
R3 c a l c = 4700
R4 c a l c = 5797 . 43
kB = 1 . 38065 e −23
e = 2 . 71828
p i = 3 . 14159
kB = 1 . 38065 e −23
e = 2 . 71828
p i = 3 . 14159
kB = 1 . 38065 e −23
e = 2 . 71828
p i = 3 . 14159

Figure 16.6: Message output log for the simulation of the Sallen-key low pass cir-
cuit: for brevity only the component value section is given

534
16.4 Introduction to Qucs subcircuit parameters

Subcircuits are a concept that has been part of the simulation scene for a long
time. All circuit simulators based on SPICE have subcircuits as part of their
basic device compliment. This is not surprising because they form a natural way
of breaking an electronic system down into a number of smaller self contained
functional blocks. What is surprising however, is the fact that a significant number
of simulators, including SPICE 2g6 and 3f516 , do not allow parameters to be passed
to a subcircuit. Parameter passing appears to have been first introduced when a
number of the popular commercial circuit simulators were being developed17 . Qucs
releases up to version 0.0.10 are similar to SPICE in that they also did not allow
parameters with subcircuits.

This very important limitation has been removed with release 0.0.11, which allows
parameters to be attached to component symbols and used in subcircuit equa-
tion calculations. Shown in Fig. 16.7 are the circuit schematic and user generated
symbol for a simple harmonic generator with a fundamental and three harmonic
sinusoidal components. Parameters f1 to f4 determine these frequency compo-
nents. Notice that an equation block, at the circuit schematic level, is used to
calculate the harmonic frequencies. Parameters ph1 to ph4 set the phase of the
individual sinusoidal oscillators. The process of attaching parameters, and their
default values, to a subcircuit symbol is straightforward; simply right click on the
symbol subcircuit name, SUB1 in Fig. 16.7, and an Edit Subcircuit Properties
dialog box appears allowing parameter names and their default values to be en-
tered18 . Subcircuit parameters and their values are normally displayed as a list
underneath the subcircuit name. Changing parameter values is done in a similar
fashion to changing the values of the standard built-in components. The diagram
and simulation results illustrated in Fig. 16.8 show a waveform formed from a
fundamental and two harmonics.

An equation block is employed to calculate and plot the amplitude and power spec-
tral densities of the harmonic waveform. By changing the fundamental frequency,
signal amplitudes and phases different wave shapes can be generated by resimulat-

16
One of the reasons SPICE preprocessors were developed was to allow parameter passing to
subroutines, for more details see Qucs Tutorial: Qucs simulation of SPICE netlists, Mike
Brinson, http://qucs.sourceforge.net/.
17
See, for example, the extended netlist format originally designed by the MicroSim Corporation
for the PSpice circuit simulator.
18
See Appendix B for a more detailed description of the procedure.

535
 V1
U=f1_amp P_sig
f=f1 HG1
Phase=ph1

Equation
V2 SUB1
U=f2_amp Eqn1 f1=1000
f=f2 f2=2*f1 f1_amp=1.0
Phase=ph2 f3=3*f1 f2_amp=0.0
f4=4*f1 ph1=0.0
ph2=0.0
V3 f3_amp=0.0
U=f3_amp f4_amp=0.0
f=f3 ph3=0.0
Phase=ph3 ph4=0.0

V4
U=f4_amp
f=f4
Phase=ph4

Figure 16.7: Harmonic generator subcircuit schematic and symbol

ing the circuit. In this example transient analysis is used to generate the harmonic
waveform with the run time set to 10ms and the number of points equal to 50019 .
This gives a sampling time of 20µs and a sampling frequency of 50kHz. Equation
block Eqn1 demonstrates how the Qucs functions20 can be used to postprocess
simulation generated data - in this example they are used to compute the DFT of
the harmonic generator waveform, convert the resulting spectra from double sided
to single sided form, compute and plot the amplitude and power spectral densities.

19
Qucs function length() determines the correct data length in equation block Eqn1 calculations.
20
If you have used a program like Octave, or indeed Matlab, many of these functions should
be familiar to you. These functions provide Qucs with powerful numerical resource which
significantly extends the range of problems that Qucs can analyse.

536
 Equation

hg_sig Eqn1
ts=(max(time)-min(time))/length(time)
transient fs=1/ts
simulation HG1 Adft=dft(hg_sig.Vt)
LAdft=length(hg_sig.Vt)
R1 Amp2=2*Adft[1:(LAdfto2)-1]
TR1 R=50 Ohm LAdfto2=LAdft/2
Type=lin Amp_squared=Adft[:LAdfto2]*conj(Adft[:LAdfto2])
SUB1 Amp=sqrt(Amp_squared)
Start=0 f1=1000
Stop=10 ms f_bin=linspace(1, LAdfto2, LAdfto2)
f1_amp=5.0 f=(f_bin-1)*fs/LAdft
Points=500 f2_amp=2.0 PLAmp=PlotVs(2*Amp/LAdft,f)
ph1=0 PLPower=PlotVs(4*Amp*Amp/(LAdft*LAdft),f)
ph2=0
f3_amp=2
f4_amp=0
ph3=0
ph4=90

5
hg_sig.Vt

-5

0 1e-3 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
time

6 30
Power Spectral density (V^2) PLPower
Amplitude Spectral density (V) PLAmp

4 20

2 10

0 0

0 5e3 1e4 1.5e4 2e4 2.5e4 0 5e3 1e4 1.5e4 2e4 2.5e4
Frequency Hz Frequency Hz

Figure 16.8: Harmonic generator subcircuit test circuit and simulation waveforms

537
16.5 Building universal macromodels using
subcircuits and parameters

Passing parameters to subcircuits allows universal macromodels to be built. One

obvious application of this technique is the modelling of operational amplifiers
(OP AMP) and other integrated circuits. The approach adopted is similar to that
outlined in the last section. However, because of the complexity of the models
it is advisable to break a model into a series of smaller blocks. These are then
combined to form a complete subcircuit macromodel. Two techniques are possible
when partitioning models, these are demonstrated next. Shown in Fig. 16.9 is a
simple AC OP AMP model21 consisting of an input stage, an intermediate gain
stage and an output stage22 . An equation block, if needed, is associated with each
stage. These blocks contain the equations for calculating the component values
in a given stage. A single schematic symbol represents the model. This has a
list of parameters attached. The flow of information into a macromodel starts
with parameters passed into a subcircuit, via a schematic symbol, then onto the
equation blocks, where it is finally used to calculate the component values. Hence,
by simply changing the subcircuit parameters different OP AMPs can be simu-
lated using a single generalised macromodel. However, please note that different
OP AMP circuit structures, or indeed technologies, naturally result in a series of
generalised subcircuit macromodels to cover all possible types in a given device
family. The second technique involves breaking a model down into smaller blocks
and associating subcircuit symbols with each block. This approach is illustrated
in Fig. 16.10. Again parameters are passed from the top level symbol (called AC
in the schematic) to the inner subcircuits. These pass their own parameters down
a subcircuit level where the component calculations are completed. The second
technique results in two levels of subcircuit, accounting for the change in param-
eter name when passing a parameter from top to lower hierarchy. A second more
detailed example showing how to construct nested subcircuits is presented later in
these notes.

In reality the macromodel for a typical OP AMP that models DC, AC and transient

21
The term AC here refers to the fact that the OP AMP model chosen for demonstration purposes
is a simplified version of a multi-domain OP AMP model. It only models small signal AC
parameters and device input stage bias and offset properties.
22
The schematic shown in Fig. 16.9 forms part of a modular OP AMP macromodel. A de-
tailed description of the function of individual networks and the derivation of the compo-
nent equations is given in Qucs tutorial Modelling Operational Amplifiers, Mike Brinson,
http://qucs.sourceforge.net/docs.html.

538
 Voff1
U=voff1

Ib1 R1 GMSRT1
Intermediate gain stage
P_INN1 I=ib R=r1 G=0.01 S

Ioff1 RSRT1 RADO1 CP1

I=ioff1 Cin1 R=1 R=aoldc C=cp1
C=cd
Ib2
I=ib R2
Voff2 GMP1
R=r2 G=1 S Equation
U=voff2
Eqn2
cp1=1/(2*pi*gbp)
P_INP1
Equation

Eqn1
voff1=voff/2 -
voff2=voff/2 AC
ioff1=ioff/2 ROS1 P_OUT1
r1=rd/2 + R=ro
r2=rd/2
SUB1 Output stage
voff=0.7e-3 EOS1
ioff=80e-3 G=1
rd=2e6
Input Stage cd=1.4e-12
aoldc=200e3
gbp=1e6
ro=75

Figure 16.9: Expanded AC OP AMP model showing circuitry and equation blocks

domains is much more complex than the model given in Fig. 16.9. The schematic
for a typical multi-domain OP AMP modular macromodel is shown in Fig. 16.11,
where each section of the macromodel is represented, if needed, by it’s own equation
block.

The test schematics shown in Figures. 16.12 and 16.13 show two OP AMPs with
different subcircuit parameters. In Fig. 16.12 the small signal characteristics of
unity gain closed loop amplifiers clearly show the difference in performance of
the OP AMPs. Figure 16.13 is particularly interesting in that it illustrates how
Qucs can be used to determine the effect of amplifier offset voltage on integrator
DC saturation by stepping resister rp through a series of values. The low offset
voltage of the OP27 makes this device much more suitable for integrator circuits
when compared to the popular UA741. These results can be confirmed by a simple
calculation: the offset voltage for the UA741 is set at 0.7 mV and the amplifier
open loop DC gain at roughly 200, 000. The UA741 goes into saturation when
rp is approximately 20 MΩ. In saturation the OP AMP gain becomes open loop
giving a DC output voltage of roughly 0.7e-3 · 2e5 or 14 V, which agrees with the
Qucs simulation results.

539
 - AC P_INN1
IN ON IN output
Input Inter O IN O
+ Stage stage
Stage P_OUT1
IP OP IP
SUB1 SUB4
v_off=0.7e-3 P_INP1 SUB3
SUB2 ro=r_o
i_off=20e-9 voff=v_off gbp=g_bp
r_d=2e6 ioff=i_off aoldc=a_oldc
c_d=1.4e-12 rd=r_d
i_b=80e-9 ib=i_b
g_bp=1e6 cd=c_d
a_oldc=200e3
r_o=75

Voff1
U=voff1 P_INN3

Ib1 R1
P_INN2 I=ib R=r1
IN ON
Input Ioff1
Stage I=ioff1 Cin1
C=cd
IP OP
Ib2
I=ib R2
SUB7 Voff2
voff=v_off R=r2
U=voff2 P_INP3
ioff=i_off
rd=r_d
ib=i_b
cd=c_d P_INP2
Equation

Eqn1
voff1=voff/2
voff2=voff/2
ioff1=ioff/2
r1=rd/2
r2=rd/2

P_INP4 GMSRT1
G=0.01 S
IN RSRT1 RADO1 CP1
Inter R=1 R=aoldc C=cp1
O P_OUT2
stage
IP
GMP1
SUB8 G=1 S Equation
gbp=g_bp
aoldc=a_oldc Eqn2
cp1=1/(2*pi*gbp)
P_INN4

output
IN O PO1 ROS1 P_OUT3
Stage R=ro

SUB9 EOS1
ro=r_o G=1

Figure 16.10: Modular AC OP AMP model showing subcircuits

540
 Equation Equation Equation
Eqn1 Eqn2 Eqn3
voff1=voff/2 ecm1=1e6/cmrrdc p1=(100*pslewr)/(2*pi*gbp)-0.7
voff2=voff/2 ccm1=1/(2*pi*1e6*fcmz) p2=(100*nslewr)/(2*pi*gbp)-0.7
ioff1=ioff/2 psum=p1+p2
r1=rd/2
Voff1 r2=rd/2
U=voff1

Ib1 R1 RSUM1
P_INN I=ib R=r1 R=1

Ioff1
I=ioff1 Cin1
C=cd SRC1
G=1 S
Ib2
I=ib R2
Voff2 R=r2
U=voff2

P_INP
RCM1 SRC2
RDCMZ R=1M RCM2 G=1 S
R=650M R=1
Equation
CCM1 Eqn4
ECM1 C=ccm1
G=ecm1 cp1=1/(2*pi*gbp)

GMSRT1
G=0.01 S

RSCALE1 VSR1 RSRT1 RADO CP1

R=100 U=p1 R=1 R=aoldc C=cp1

SRC3 GMP1
G=1 S D1 G=1 S
Is=1e-12 A
Bv=psum Equation
Ibv=20 mA
Eqn5
cp2=1/(2*pi*fp2)

RP2 CP2
R=1 C=cp2 ROS1
R=ro

P_VCC
GMP2 EOS1
G=1 S G=1 VLIM1
U=vlim1

Equation
D2 D3
Is=1e-15 A Is=1e-15 A DVL1
Eqn7
Cj0=0.0 Cj0=0.0 Is=8e-16 A vlim1=vcc-vccm+1
vlim2=-vee+veem+1

HCL1
G=hcl1 P_OUT
RDCCL1 DVLM2
R=100M Is=8e-16 A

ECL Equation
G=1 VLIM2
Eqn6 U=vlim2
hcl1=0.9/idcoutm

P_VEE

Figure 16.11: Modular OP AMP subcircuit schematic with embedded component

calculation equations

541
 R1 vout_op27
R=4.7k

V1
U=15 V

V3 - VCC

U=1 V R2 MOD
R=4.7k
V2
+ VEE
U=15 V

SUB1
voff=30e-6
ib=15e-9
ioff=12e-9
rd=4e6 vout_ua741
cd=1.4e-12
cmrrdc=1778279.4
fcmz=2009.0
aoldc=1778279.4
gbp=8e6
fp2=17e6
pslewr=2.8e6 R3
nslewr=2.8e6 R=4.7k
vcc=15
vee=-15
vccm=14
veem=-14
ro=75
idcoutm=32e-3
- VCC

R4 MOD
R=4.7k

+ VEE

SUB2
voff=0.7e-3
ib=80e-9
ioff=10e-9
rd=2e6
cd=1.4e-12
cmrrdc=31622.77
Equation
fcmz=200.0
aoldc=199526.3
Eqn1 gbp=1e6
dc simulation gain_ua741=dB(vout_ua741.v) fp2=3e6
phase_ua741=phase(vout_ua741.v) pslewr=0.5e6
phase_op27=phase(vout_op27.v) nslewr=0.5e6
DC1 gain_op27=dB(vout_op27.v) vcc=15
vee=-15
vccm=14
veem=-14
ro=75
idcoutm=34e-3
number vout_op27.V vout_ua741.V ac simulation
1 -3.87e-05 0.001
AC1
Type=log
Start=1 Hz
Stop=100MHz
Points=161

1
vout_ua741.v
vout_op27.v

0.5

1 10 100 1e3 1e4 1e5 1e6 1e7 1e8

Frequency Hz

-20
gain_ua741
gain_op27

-40

-60
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz

200
phase_ua741
phase_op27

100

-100
1 10 100 1e3 1e4 1e5 1e6 1e7 1e8
Frequency Hz

Figure 16.12: Unity gain OP AMP test circuit and waveforms

542
 R3 R4 vout_op27
R=rp R=rp
vout_ua741

C1 C2
C=1 uF C=1 uF

V1 V3
- VCC
U=15 V
- VCC U=15 V

R1 MOD R2
R=1k MOD
R=1k
V2 V4
+ VEE
U=15 V
+
U=15 V
VEE

SUB1 SUB2
voff=0.7e-3 voff=30e-6
ib=80e-9 ib=15e-9
ioff=10e-9 ioff=12e-9
rd=2e6 rd=4e6
cd=1.4e-12 cd=1.4e-12
cmrrdc=31622.77 Parameter dc simulation cmrrdc=1778279.4
fcmz=200.0 sweep fcmz=2009.0
aoldc=199526.3 aoldc=1778279.4
gbp=1e6 DC1 gbp=8e6
fp2=3e6 SW1 fp2=17e6
pslewr=0.5e6 Sim=DC1 pslewr=2.8e6
nslewr=0.5e6 Type=log nslewr=2.8e6
vcc=15 Param=rp vcc=15
vee=-15 Start=1e3 vee=-15
vccm=14 Stop=1e9 vccm=14
veem=-14 Points=31 veem=-14
ro=75 ro=75
idcoutm=34e-3 idcoutm=32e-3

15

10
vout_ua741.V
vout_op27.V

1e3 1e4 1e5 1e6 1e7 1e8 1e9

rp

Figure 16.13: Integrator test circuits for determining DC saturation

543
16.6 More complex nested subcircuit models

In the previous two sections the example circuits only included subcircuits nested to
one or two levels. Qucs does however, allow subcircuits to be nested to an arbitrary
level and parameters can be passed down the nested chain to any depth required.
Some care is needed when setting up the parameter passing sequence. Shown in
Fig. 16.14 is a top level subcircuit with temperature swept between 10 and 110
centigrade. A simple resistor voltage divider network is at the bottom of a series of
linked subcircuits, three levels down. R2 in the divider is a function of temperature.
A schematic representation of the coupled subcircuits parameter passing sequence
is also given in the right hand side of Fig. 16.14. Each level passes the value of
temperature to it’s next lower member in the hierarchy. The Qucs generated netlist
given in Fig. 16.15 clearly shows the parameter passing mechanism employed by
Qucs. The ability to nest subcircuits and pass parameters down a hierarchy is an
important feature in Qucs because it allows both circuit design and device data
to be passed to different sections of the circuit/system being simulated. These
parameters can, of course, be at different levels in a problem hierarchy providing
a very flexible and powerful design/analysis tool.

544
 vp01
SUB 3
dc simulation IN OUT
SUB3
V1 IN OUT
DC1 U=1 V
SUB1
sp1=tsweep
Sp1 = tsweep

Parameter
sweep 1

SW1 SUB2
Sim=DC1 0.8
vp01.V

IN OUT
Type=lin
Param=tsweep
Start=10
Stop=110 0.6 Sp2=Sp1
Points=100

0.4
20 40 60 80 100
tsweep SUB1
IN OUT

tscan=Sp2

R1
P1 R2 R=10k P2
R=10k Temp=tscan
Temp=26.85 Tc1=0.01
Tc2=0.015

Figure 16.14: A nested subcircuit showing parameter passing sequence

545
# Qucs 0 . 0 . 12 / media / hda2 /Qucs e q u a t i o n m o d e l l i n g p r j / r d i v t e s t tsweep 3 l . s c h
. Def : r d i v sub1 temp n e t 1 n e t 0 t s c a n=”27 ”
R: R2 gnd n e t 0 R=”10 k ” Temp=” t s c a n ” Tc1=”0 . 01 ” Tc2=”0 . 015 ” Tnom=”26 . 85 ”
R: R1 n e t 1 n e t 0 R=”10 k ” Temp=”26 . 85 ” Tc1=”0 . 0 ” Tc2=”0 . 0 ” Tnom=”26 . 85 ”
. Def : End

. Def : r d i v t e s t 6 temp n e t 1 n e t 0 sp2=”27 ”

Sub : SUB1 n e t 1 n e t 0 Type=” r d i v sub1 temp ” t s c a n=”sp2 ”
. Def : End

. Def : r d i v sub3 temp n e t 0 n e t 1 sp1=”27 ”

Sub : SUB1 n e t 0 n e t 1 Type=” r d i v t e s t 6 temp ” sp2=”sp1 ”
. Def : End

Vdc : V1 n e t 0 gnd U=”1 V”

.DC:DC1 Temp=”26 . 85 ” r e l t o l=”0 . 001 ” a b s t o l=”1 pA” v n t o l=”1 uV”
saveOPs=”no ” MaxIter=”150 ” s a v e A l l=”no ” c o n v H e l p e r=”none ” S o l v e r=”CroutLU ”
.SW:SW1 Sim=”DC1” Type=” l i n ” Param=”tsweep ” Start=”10 ” Stop=”110 ” P o i n t s=”100 ”
Sub : SUB1 n e t 0 vp01 Type=” r d i v sub3 temp ” sp1=”tsweep ”

Figure 16.15: Qucs netlist for nested subcircuit showing parameter passing
sequence

546
16.7 Introduction to equation defined devices
(EDD)

Although adding symbolic equations to a simulator merges circuit design and anal-
ysis, it is by making these equations functions of circuit variables that the real
power of modern circuit simulator is fully exploited. Equations that are functions
of voltage, current and charge have to be continuously evaluated as a simulation
progresses. This is in contrast to the type of equations previously introduced,
which are only evaluated at the start of a simulation sequence. When component
properties are functions of circuit variables considerable complexity is added to
a simulation engine and as a result most simulators restrict such properties to
a small number of component types, the most common being controlled current
and voltage generators23 . Qucs version 0.0.12 introduces an equation defined de-
vice (EDD) which allows it’s terminal currents to be functions of voltage, and it’s
stored charge to be functions of voltage and current. The EDD is similar, but more
advanced, to the B type controlled source implemented in SPICE 3f5. It is capable
of realising the same models as the SPICE B type device plus an extensive range
of more complex compact device models. At this stage in Qucs development only
the explicit form of EDD is implemented24 . EDD is an advanced component that
allows Qucs users to construct their own device models from a set of equations de-
rived from the physical properties that characterise a device. The explicit form of
EDD can only be used to develop models for devices where their defining equations
can be transformed into the explicit analysis form required by Qucs25 . A range
of functions similar to those defined in the Verilog-A compact device modelling
language are provided by Qucs, making the equation modelling language easy to
use and powerful. The ternary ? : form of the C language if statement has also
been implemented to allow selection of model equations that change with differing
device voltage, current and charge conditions. Before introducing the EDD symbol
23
Probably the most well known non-linear controlled generators are the SPICE 2g6 and 3f5
forms, see A. Vladimirescu, Kaihe Zhang, A.R. Newton, D.O. Pederson and A. Sangiovanni-
Vincentelli, SPICE Version 2G User’s Guide, 1981, Department of Electrical Engineering and
Computer Sciences, University of California, Berkeley, Ca. 94720, section 11, Appendix B:
Nonlinear dependent sources., and B. Johnson, T. Quarles, A.R. Newton, D.O. Pederson and
A. Sangiovanni-Vincentelli, SPICE3 Version f User’s Manual, 1992, Department of Electrical
Engineering and Computer Sciences, University of California, Berkeley, Ca. 94720, section
3.2.2.4, Non-linear dependent sources.
24
See Qucs Technical Papers, Section 10.7: Equation defined models, Stefan Jahn, Michael
Margraf, Vincent Habchi and Raimund Jacob, http://qucs.sourceforge.net/technical.
html.
25
The Y parameters of the device being modelled must also exist for the explicit form of the
EDD to be valid.

547
and it’s properties consider the following circuit simulation modelling problem: a
model for a device is required where the output voltage is a function of two input
voltages V IN1 and V IN2 , such that

Vout (V IN1 , V IN2 ) = V IN1 · V IN2 , (16.11)

where V IN1 and V IN2 can be arbitrary varying voltages.

This type of model is difficult to simulate at functional level26 using the pre-version
0.0.12 built-in devices. A linear voltage controlled voltage source can be used to
multiply a voltage by a constant. Multiplying by a second voltage is not possible
with the linear controlled sources. Qucs AM modulated and PM modulated sources
are the nearest that Qucs has to the source defined above. These sources however,
only allow sinusoidal carrier signals. Illustrated in Fig. 16.16 is a four quadrant
multiplier EDD which allows multiplication of two varying signals27 . The EDD
device generates current I1 = V 2 · V 3. This in turn is transformed to the output
voltage by a unity gain current controlled voltage source SRC1. An EDD device
can consist of up to 8 branches. The branches have currents, I1 to I8, voltages V1
to V8 and internal charges Q1 to Q8 respectively. Overall the total device current
depends how these branches are connected. A similar comment applies to the total
device charge. In Fig. 16.16 currents I2 and I3 are set to zero, charges Q2 and Q3
are also zero, and voltages V 2 = V IN1 and V 3 = V IN2 . Hence current I1 becomes
the multiplication of V IN1 and V IN2 . The fact that currents I2 and I3 are set to
zero implies that the terminals connected to the external input voltages have high
impedance and act as voltage probes. The test circuit in Fig. 16.16 is shown with
signal inputs generated by sinusoidal oscillators; V1 acts as a modulating signal
and V2 as a carrier signal. The bottom right hand corner of Fig. 16.16 includes a
second graph which illustrates the effect of changing signal V2 to a square wave
source with 0.05ms period.

26
It is, of course, possible to model the multiplier operation at discrete component level e.g.
using a Gilbert cell mixer circuit.
27
This model is based on an idea suggested by Stefan Jahn, during the EDD development phase.

548
 transient 5
simulation V1
U=1 V
f=1 kHz
TR1

Out.Vt
Type=lin
Start=0 0
Stop=1 ms
Out
Points=401

-5
R1
V2 R=50 Ohm 0 2e-4 4e-4 6e-4 8e-4 1e-3
U=5 V
f=10 kHz time
VMULT1

D1
I1=V2*V3
Q1=0
I2=0
Q2=0
I3=0
Q3=0

Out1
Num=1
1
SRC1 1
G=1
vmul_2_tb:Out.Vt

In1 0
Num=2
3

In2 -1
Num=3
0 2e-4 4e-4 6e-4 8e-4 1e-3
time

Figure 16.16: Qucs EDD four quadrent multiplier model and test circuit

16.8 The Qucs EDD component

A two terminal model for a universal non-linear component with resistive, capaci-
tive and inductive parallel branches is shown in Fig. 16.17. All three branches have
elements that can be functions of either voltage or current or charge28 . The Qucs
EDD component can be used to model this nonlinear device. One EDD element is
needed to model the resistive and capacitive branches. A second EDD device, plus
a gyrator, models the inductive branch. The total terminal current is the sum of
the individual branch currents. Equations for the three branch currents are given

28
Each branch can be a function of one or more of these circuit variables but not necessarily all
three at the same time.

549
by the following equations:

I = I1 + IC + IL, (16.12)

where
dV 1 dQ1
I1 = f (V ), IC = C(V, I) · = (16.13)
dt dt
Also
dV 2 dIL
V 1 = i2, V 2 = −IL, i2 = −L(I) · , V 1 = L(I) · (16.14)
dt dt
Giving Z
1
IL = · V 2 · dt (16.15)
L(I)
and
dΦ
VL=V2=V1= (16.16)
dt
Hence Z
dV 1 1
I = f (V ) + C(V, I) · + · V 1 · dt (16.17)
dt L(I)

The EDD is characterised by eight parallel branches each comprising a current

component In and a charge component Qn, where n ranges from 1 to 8. The
currents may be constants or defined by equations that are functions of the EDD
branch voltages (these are designated V 1 to V 8). This form of the EDD compo-
nent is known as the explicit EDD model. Please note, EDD currents cannot be
functions of current. However, with release 0.0.12 implementation of the explicit
EDD the device charge can be a function of either voltage or current29 . The cur-
rent in the resistive branch being a function of EDD voltage allows a range of two
terminal30 devices to be modelled, allowing, for example, nonlinear resistors and
diode models to be easily developed. Similarly, the fact that the EDD charge can
be a function of voltage or current extends the range of allowed Qucs capacitor
types opening new areas of application. The same comments apply to the nonlin-
ear inductors where components that have inductance values which are functions
of current allow modelling of nonlinear transformer and coupled inductor effects.
This was not possible with earlier Qucs releases. The EDD current and charge
values may be defined by symbolic equations that include the operators and func-
tions listed in the “Short description of mathematical functions“ entry in the Qucs
help index31 .
29
This allows modelling of semiconductor capacitive effects where the amount of stored charge is
either a function of voltage (depletion layer capacitance), or a function of current (diffusion

550
 capacitance).
30
The number of device terminals can be increased to model transistors and other devices.
31
The Qucs operators and functions are a superset of those defined in the Verilog-A language
manual. However, in some cases the name of the operator or function differs slightly. For
example Verilog-A uses pow(x, y) for the power function whilst Qucs uses ∧ to denote xy .
An example of differing function names are the inverse trigonometric functions. A list of the
available functions is given in Appendix A.

551
 I

I1 IC IL

R1 Q C1 L1
V1 R=f(V) C=f(V,I) L=f(I)

I
I1+IC IL

i2

D1 D2
V1 I1=I1 I1=0
Q1=L(I)*V2 V2
Q1=C(V,I)*V1
1
1

X1
R=1

Equation Equation
defined defined
device Gyrator device
(EDD) (EDD)

Figure 16.17: A non-linear two terminal branch with parallel resistive, capacitive
and inductive components

552
16.9 Modelling nonlinear resistors

In many measurement applications a transducer is employed to transform changing

values of a physical quantity to, say, changes in resistance. Often the resistive
characterstics of these devices are nonlinear. To demonstrate how the EDD can be
used to model a nonlinear resistance the example shown in Fig. 16.18 is introduced.
In this schematic an EDD represents a resistance that is a function of the applied
voltage across it’s terminals. This example deliberately shows an extreme case
where the resistance changes in a resistive pulse like fashion as the terminal voltage
increases. The example also introduces for the first time the ternary ? : operator
and illustrates how it can be nested to give an ”if then else“ structure to define the
component properties. A point of note with these very nonlinear devices centres
around the fact that it is possible to define components that have discontinuities in
their I-V characteristics32 . The EDD current equation defines how the resistance
of this device changes with changing terminal voltage. This equation is given by

I1=V1/((V1<1.0) ? 1000 : (V1<2.0)

? 1000+4000*(V1-1) : (V1<5.0)
? 5000 : ((V1 >=5.0) && (V1<6.0))
? 5000-4500*(V1-5.0) : 500)

Which in terms of an ”if then else“ type statement is equivalent to:

I1 = V1/( if (V1 < 1.0) then 1000

else if (V1 < 2.0) then 1000 + 4000*(V1-1)
else if (V1 < 5.0) then 5000
else if ((V1 >= 5.0) && (V1 < 6.0)) then 5000 - 4500*(V1-5.0)
else 500 )

32
One effect of such a discontinuity is the introduction of rapidly changing circuit conditions
which can cause the simulator difficulties in converging to a correct solution. Sometimes, if this
happens, simulation run times may be dramatically increased or simulation fails altogether.

553
554
Parameter
sweep

dc simulation Equation
SW1
Eqn1 Sim=DC1
DC1 R=Vs/Pr1.I Type=lin
Param=Vs
Start=0
Stop=7
Vs Points=100

V1
U=Vs Pr1 D1
I1=V1/((V1<1.0) ? 1000 : (V1<2.0) ? 1000+4000*(V1-1) : (V1<5.0) ? 5000 : ((V1 >=5.0) && (V1<6.0)) ? 5000-4500*(V1-5.0) : 500)

1
0.1
5e3

0.01
4e3

1e-3
3e3
R

Pr1.I
1e-4
2e3

Figure 16.18: Qucs nonlinear resistor model

1e-5
1e3

1e-6 0
0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7
Vs Vs
16.10 Modelling nonlinear capacitors and inductors

Nonlinear capacitors, who’s C value is a function of terminal voltage, and nonlin-

ear inductors, who’s L value is a function of terminal current, commonly act as
control elements in electronic systems. SPICE 2g6 includes a nonlinear symbolic
polynomial form of C and L33 . The schematic shown in Fig. 16.19 illustrates how a
nonlinear capacitor can be modelled by an EDD. This model is based on a SPICE
like polynomial function with four coefficients; C0, C1, C2 and C334 . The test cir-
cuit is a simple RC network with nominally identical R and C component values
to those shown in Fig. 16.2. Increasing the value of DC source V1 also increases
C which in turn decreases the RC low pass filter -3dB frequency. This effect is
very visible in Fig. 16.19. The nonlinear changes in C are also clearly illustrated
in the output voltage and phase curves. The schematic symbol for the nonlinear
capacitor is shown in Fig. 16.19 with a red ring drawn around the normal capacitor
symbol. This denotes an EDD based component. An alternative convention is to
use red lettering within a symbol. The test circuit and simulation results for a
nonlinear inductance are shown in Fig. 16.20. The EDD model is similar to the
SPICE 2g6 nonlinear inductance model with four coefficients. This number can
be increased, if required, by extending the EDD polynomial expression. A gyra-
tor is employed with the EDD to model the nonlinear inductance. The effect of
nonlinear inductance on the inductance current is shown by the difference between
probe currents Pr1 and Pr2.

33
The details of these polynomial functions are presented in Test Reports 4 and 5 of the SPICE
to Qucs testing Series, Mike Brinson, http://qucs.sourceforge.net/docs.html.
34
SPICE 2g6 allows up to twenty coefficients. Simply add more higher order terms to the Qucs
polynomial if required.

555
 V2
U=1 V Vout
Vb Vout.V
1 1
V1 V 2 2
U=Vb R1
R=1k 3 3
4 4
5 5
SUB1 6 6
C0=1u 7 7
C1=0.5u 8 8
C2=0.2u 9 9
dc simulation C3=0.1u 10 10

DC1

Equation
Parameter
sweep Eqn1
ac simulation Ph_Vout=phase(Vout.v)
Vout_dB=dB(Vout.v)
AC1 SW1
Type=log Sim=AC1
Start=1 Hz Type=lin
Stop=10kHz D1
Param=Vb I1=0
Points=201 Start=1 Q1=C0*V1+(C1/2)*V1^2+(C2/3)*V1^3+(C3/4)*V1^4
Stop=10
Points=10 POUT1
1

PIN1

1
Vout.v

0.5

1 10 100 1e3 1e4

acfrequency

-20
Vout_dB

-40

-60

-80
1 10 100 1e3 1e4
acfrequency

200
Ph_Vout

150

100

1 10 100 1e3 1e4

acfrequency

Figure 16.19: Qucs nonlinear capacitor model

556
 D1
P_inp1 I1=0
Q1=L*V1+(L2/2)*V1^2+(L3/3)*V1^3+(L4/4)*V1^4

1
X1
P_inn1 R=1

in
IND=L+L2*I(L)+L3*I(L)^2+L4*I(L)^3
V1
U=vin Pr1 SUB1
f=1 MHz L=1e-6
L2=5e-7
transient L3=1e-7 Parameter
simulation L4=5e-8 sweep

TR1 SW1
Type=lin dc simulation Sim=TR1
Start=0 Type=lin
Stop=4 us Param=vin
DC1 Start=0
Stop=100
Points=3

Pr2 L1
L=1e-6

100
in.Vt

-100

0 5e-7 1e-6 1.5e-6 2e-6 2.5e-6 3e-6 3.5e-6 4e-6

time

4
Pr1.It

0
0 5e-7 1e-6 1.5e-6 2e-6 2.5e-6 3e-6 3.5e-6 4e-6
time
40
Pr2.It

20

0
0 5e-7 1e-6 1.5e-6 2e-6 2.5e-6 3e-6 3.5e-6 4e-6
time

Figure 16.20: Qucs nonlinear inductor model

557
16.11 Compact device modelling using EDD

Semiconductor device models are a corner stone of all circuit simulators. Often
they are characterised by the same parameters as those found in the SPICE 2g6 and
3f5 diode, BJT, FET and MOS models.35 . Since the original SPICE semiconductor
device models where first developed many new extensions to these models have
been proposed. Unfortunately, adding such models to a circuit simulator is a
complex process, being both time consuming and requiring specialised knowledge.
For the average Qucs user the hand coded C++ model generation route is one
that they would not contemplate attempting because of the depth of knowledge
and specialised skills required. The Qucs EDD was devised to promote fast, and
straight forward, prototyping of semiconductor compact models, allowing a wider
Qucs population the opportunity to try their hand at device model construction.
To demonstrate the stages needed to generate an EDD model of a semiconductor
device a compact model of a diode is introduced in this section36 .

The DC diode current Id is given by the following functions of diode voltage Vd 37 .

Id = Is · (exp (Vd /(n · V t) − 1) + Vd · GM IN, ∀ (−5 · n · V t ≤ Vd ) (16.18)

Id = −Is + Vd · GM IN, ∀ (−BV < Vd ) and (Vd < −5 · n · V t ≤ Vd ) (16.19)

Id = −IBV, ∀ (Vd = −BV ) (16.20)

Id = −Is · (exp (−(BV + Vd )/V t) − 1 + BV /V t) , ∀ (Vd < −BV ). (16.21)

35
The SPICE 2g6 and 3f5 device parameters are a subset of those commonly provided with
current generation of circuit simulators, including Qucs.
36
A second three terminal MESFET transistor example is available for downloading from the
Qucs Web site.
37
These equations are for the SPICE 2g6 diode model, see Giuseppe Massobrio, Chapter 1, Pn-
junction diode and Schottky diode, Semiconductor device modeling with SPICE, Edited by
Paolo Antognetti, Giuseppe Massobrio, 1988, McGraw-Hill,Inc, ISBN 0-07-002107-4.

558
In these equations:

• Is = the saturation current.

• n = the emission coefficient.

• GM IN = a small conductance in parallel with the diode38

• V t = kB · T /q, where T is the diode temperature in Kelvin, kB is Boltz-

mann’s constant and q the charge on the electron.

• BV = reverse breakdown voltage (positive number)

• IBV = reverse breakdown current (positive number).

Figure 16.21 gives the EDD model for the experimental semiconductor diode. The
ternary operator ?: is used to select the correct equation for each diode operating
region. The diode current Id is the sum of EDD branch currents I1 to I4, where
I1 represents the diode forward bias region, I2 the reverse bias region and I3
plus I4 the diode reverse bias breakdown region. When calculating diode current
a special form of the exponential function exp(), called limexp(), is employed to
assist Qucs to converge to a solution during DC and transient large signal analysis.
The function limexp() linearises the exponential function at large argument values
minimising the possibility of floating point overflow and generation of software
exceptions. The Id − Vd characteristic curves shown in Fig. 16.21 are for the
forward bias region with series resistance rs set to 0.01Ω. For completeness the
simulation data for the Qucs built-in diode are also given. Clearly the two sets of
results are very similar. The DC simulation results for the diode reverse breakdown
region of operation are shown in Fig. 16.22. Again for comparison an Id − Vd plot
for the Qucs built-in diode is also provided. In this region of operation some slight
differences are apparent: although for both devices the reverse breakdown is very
close to 100V the slope of the Id − V d curve at negative voltages beyond -BV is
different, emphasising that the SPICE diode model does not model breakdown or
zener effects well39 .

The next stage in the development of the diode model is to add capacitance effects:
depletion layer capacitance for the reverse bias region and diffusion capacitance
for the forward bias region. Diode capacitance is given by:

38
GMIN is added to help Qucs DC convergence. The SPICE default value is 1e-12S.
39
See Steven M. Sandler, SPICE subcircuit accurately models zener characteristics, Personal
Engineering, November 1998, pp 45-48 for more information on this subject.

559
 PANODE1

RS1
R=rs

D2
I1=(V1>-5.0*n*Vt) ? Is*(limexp(V1/(n*Vt))-1.0)+V1*GMIN : 0
Q1=0
I2=(-BV<V1) ? (V1<-5.0*n*Vt) ? -Is+V1*GMIN : 0 : 0

1
Q2=0
I3=(V1==-BV) ? -IBV : 0
Q3=0
Equation
I4=(V1<-BV) ? -Is*(limexp(-(BV+V1)/Vt)-1.0+BV/Vt) : 0
Q4=0
Eqn2
GMIN=1e-12
Vt=vt(300) PCATHODE1

Pr2

Vs D1
U=Vd Pr1 Is=1e-14 A
SUB1 N=1
n=1.0 Vj=1.0
rs=0.01 Rs=0.01
Is=1e-14 Bv=100.00
BV=100.0 Ibv=1e-3
IBV=1e-3
Vj=1.0

dc simulation Parameter Equation

sweep Eqn1
DC1
Id=Pr1.I
Id_Q=Pr2.I
SW1 lnId=ln(Pr1.I)
Sim=DC1 lnId_Q=ln(Pr2.I)
Type=lin
Param=Vd
Start=0
Stop=1
Points=190

10 0
Id (A)

ln(Id)

5
-20

0
-40
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Vd (V) Vd (V)

10 0
ID_Q (A)

ln(ID_Q)

5
-20

0
-40
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Vd (V) Vd (V)

Figure 16.21: Compact diode model DC test circuit and simulation results: SUB1
is the EDD diode model and D1 the Qucs diode model with the same
parameters as SUB1.

560
 0 0

ID_Q (A)
Id (A)
-5

-50

-10

-101 -100.5 -100 -99.5 -101 -100.5 -100 -99.5

Vd (V) Vd (V)

Figure 16.22: Compact diode model DC simulation results for the reverse break-
down region of operation

• Depletion layer capacitance

 −m
dQdep Vd
Cdep = = Area · Cj0 1 − (16.22)
dVd Vj

• Diffusion capacitance
dQdif f dId
Cdif f = = tt · (16.23)
dVd dVd

Where the total stored charge Qd = Qdep + Qdif f . Using the same notation as the
SPICE diode model:

Qdif f = tt · Id (16.24)
ZVd  −m
Vd
Qdep = Area · Cj0 1− dV, ∀ (Vd <= F C · Vj ) (16.25)
Vj
0

R 1 (ax + b)1+n
Using integration formula (ax + b)n dx = and simplifying yields:
a 1+n
"  1−m #
Area · Cj0 · Vj Vd
Qdep = 1− 1− (16.26)
1−m Vj

Also, in the forward bias region

ZVd  
Area · Cj0 m · Vd
Qdep = Area · Cj0 · F 1+ F3 + dV, ∀ (Vd >= F C · Vj )
F2 Vj
F C · Vj
(16.27)

561
On integrating

      
1 m 2 2


Qdep = Area · Cj0 F 1 + · F 3 · (Vd − F C · Vj ) + · Vd + (F C · Vj )
F2 2 · Vj
(16.28)

Where

Vj 
1 − (1 − F C)1−m , F 2 = (1 − F C)1+m , F 3 = 1 − F C · (1 + m)

F1 =
1−m
(16.29)

In these equations:

• F C = Coefficient for forward-bias depletion capacitance.

• m = Grading coefficient.

• tt = Transit time.

• Area = Device area.

• Cj0 = Zero-bias junction capacitance.

Figure 16.23 shows the extended diode model. The Cdep and Cdif f components of
the device capacitance have been included in the EDD model as stored charge Q1
and Q2. Again the ternary operator ?: is employed to select the correct equation
for each section of the diode DC operating range. An equation block is used to
simplify the charge equations through the use of factors F1, F2 and F3.40 . An area
factor has also been added to the EDD model in Fig. 16.23. This is introduced to
allow simulation of two or more equivalent parallel devices. The diode variables
scaled by area are:

Is (A) = Is · Area, Cj0(A) = Cj0 · Area, and rs (A) = rs/Area. (16.30)

40
In complex current and charge expressions precalculating subexpressions in equation blocks
ensures that they are only calculated once at the beginning of a simulation, ensuring minimum
run times for an EDD model.

562
The test circuit shown in Fig. 16.23 illustrates how device capacitance and resis-
tance can be determined as a function of diode bias voltage. Firstly, the diode S
parameters are determined at a given bias voltage, secondly these are converted
to Y parameters and the diode capacitance (Cap) and resistance (RD) extracted
from Y [1, 1], and finally the variation of Cap and RD with diode voltage Vd plotted
using the Qucs plotting function PlotVs. Notice that the value of Cap at Vd = 0V
agrees with the value of Cj0.

To complete the demonstration EDD diode model all that remains to do is to add
temperature dependence to the current and capacitance equations. Circuit simula-
tors normally use two temperatures to determine device temperature dependence;
the first called Tnom represents the temperature that the device parameters were
measured, and the second called Temp represents the current device temperature.
A high percentage of the diode parameters are temperature dependent. However,
to simplify the demonstration diode model only the temperature dependence of
parameters Is , V j and Cj0 will be included in the model. Adding extra tempera-
ture dependence to the diode model is left to readers as an exercise41 . One of the
great advantages of the EDD style of modelling is that it is interactive allowing
easy experimentation with models to any given level. The following equations list
the temperature dependence of Is , V j and Cj0.

Let T1 = Tnom and T2 = Temp, then

  XTn I   
T2 −q · Eg(300) T2
Is (T 2) = Is (T 1) exp 1− (16.31)
T1 kB · T 2 T1

 1.5  
T2 2 · kB · T 2 T2 T2
V j(T 2) = · V j(T 1) − ln − · Eg(T 1) − Eg(T 2)
T1 q T1 T1
(16.32)

  
−6 V j(T 2) − V j(T 1)
Cj0(T 2) = Cj0(T 1) 1 + m 400 · 10 (T 2 − T 1) −
V j(T 1)
(16.33)

In these equations:
41
For example, parameters m and BV are both temperature dependent.

563
 PANODE1

RS1
R=rs

D1
I1=(V1>-5.0*n*Vt) ? Is*(limexp(V1/(n*Vt))-1.0)+V1*GMIN : 0
Q1=(V1 < FC*Vj) ? tt*I1+Area*(Cj0*Vj/(1-m))*(1-(1-V1/Vj)^(1-m)) : 0
I2=(-BV<V1) ? (V1<-5.0*n*Vt) ? -Is+V1*GMIN : 0 : 0
4

1
Q2=(V1 >= FC*Vj) ? tt*I1+Area*Cj0*(F1+(1/F2)*(F3*(V1-FC*Vj)+(m/(2*Vj))*(V1*V1-FC*FC*Vj*Vj))) : 0
I3=(V1==-BV) ? -IBV : 0
Equation Q3=0
PCATHODE1 I4=(V1<-BV) ? -Is*(limexp(-(BV+V1)/Vt)-1.0+BV/Vt) : 0
Eqn2 Q4=0
GMIN=1e-12
F1=(Vj/(1-m))*(1-(1-FC)^(1-m))
F2=(1-FC)^(1+m)
F3=1-FC*(1+m)
Vt=vt(300)
X1 Vd

Vs1
U=Vs Pr1

P1 SUB1
Num=1 n=1.0
Z=50 Ohm rs=0.01
Is=1e-14
BV=100.0
IBV=1e-3
Vj=1.0
Parameter Cj0=1e-12
dc simulation FC=0.5
sweep tt=1e-12
DC1 Area=1
SW1 m=0.5
Sim=SP1
S parameter Type=lin
simulation Param=Vs
Start=-4
Stop=0.8 1e-11
SP1 Points=200
Type=const
Cap (F)

Values=[100 kHz]
5e-12
Equation

Eqn1
Y=stoy(S) 0
LN_RD=ln(RD)
RD=PlotVs(1/(real(Y[1,1])),Vs)
Cap=PlotVs(imag(Y[1,1])/Omega,Vs) -4 -3 -2 -1 0 1
Omega=2*pi*frequency Vd (V)

1e12

20
LN_RD
Rd ( )

5e11

0
0
-4 -3 -2 -1 0 1 -4 -3 -2 -1 0 1
Vd (V) VD (V)

Figure 16.23: Compact diode model capacitance and resistance simulation

564
 • XT I = Saturation current temperature exponent.

7.02e − 4 · T 2
• Eg(T ) = EG(0) − , the energy gap.
1108 + T
Figure 16.24 shows the extended EDD for the experimental diode model. Again the
limexp() function is used in preference to the standard exp() function in the tem-
perature calculations listed in equations block Eqn2. The test circuit in Fig. 16.24
sweeps the device temperature from 20 to 80 degrees Centigrade. The graph inlay
illustrates the experimental diode current Id plotted as a function of temperature.
The temperature of the built-in Qucs diode is held constant, at room tempera-
ture, and it’s current Id Q plotted as an overlay. The two curves cross at room
temperature, indicating identical currents at this temperature.

565
 Pr2
dc simulation
DC1

Vs D1
U=0.6 Pr1 Is=1e-14 A
N=1
SUB1 Cj0=1e-12 Parameter
n=1.0 Vj=1.0 sweep
rs=0.01 Rs=0.01
Is=1e-14 Bv=100.00
BV=100.0 Ibv=1e-3 SW2
IBV=1e-3 Temp=26.85 Sim=DC1
Vj=1.0 Xti=3.0 Type=lin
Equation Cj0=1e-12 Eg=1.11 Param=Temp_sw
m=0.5 Tnom=26.85 Start=-20
Eqn1 Area=1 Area=1 Stop=80
Id=Pr1.I FC=0.5 Points=100
Id_Q=Pr2.I tt=1e-12
lnId=ln(Pr1.I) XTI=3.0
lnId_Q=ln(Pr2.I) Tnom=26.85
Temp=Temp_sw
Eg=1.16 0.1
0.01
1e-3
Id_Q

1e-4
Id

1e-5
1e-6
1e-7
PANODE1 1e-8
-20 0 20 40 60 80
RS1 Temp (Centigrade)
R=rs_AREA

D2
I1=(V1>-5.0*n*Vt) ? Area*Is_T2*(limexp(V1/(n*Vt))-1.0)+V1*GMIN : 0
Q1=(V1 < FC*Vj) ? tt*I1+Area*(Cj0_T2*Vj_T2/(1-m))*(1-(1-V1/Vj_T2)^(1-m)) : 0
I2=(-BV<V1) ? (V1<-5.0*n*Vt) ? -Area*Is_T2+V1*GMIN : 0 : 0
4

Q2=(V1 >= FC*Vj) ? tt*I1+Area*Cj0_T2*(F1+(1/F2)*(F3*(V1-FC*Vj_T2)+(m/(2*Vj_T2))*(V1*V1-FC*FC*Vj_T2*Vj_T2))) : 0

I3=(V1==-BV) ? -IBV : 0
Q3=0
I4=(V1<-BV) ? -Area*Is_T2*(limexp(-(BV+V1)/Vt)-1.0+BV/Vt) : 0
Q4=0

PCATHODE1

Equation Equation
Eqn3 Eqn2
F1=(Vj/(1-m))*(1-(1-FC)^(1-m)) Cj0_T2=Cj0*(1+m*(400e-6*(T2-T1)-(Vj_T2-Vj)/Vj))
F2=(1-FC)^(1+m) rs_AREA=rs/AREA
F3=1-FC*(1+m) GMIN=1e-12
Vt=vt(300) A=7.02e-4
B=1108
T1=Tnom+273.15
Vj_T2=(T2/T1)*Vj-(2*kB*T2/q)*ln((T2/T1)^1.5)-((T2/T1)*Eg_T1-Eg_T2)
Is_T2=Is*(T2/T1)^(XTI/n)*limexp((-(q*Eg)/(kB*T2))*(1-T2/T1))
Eg_T1=Eg-A*T1*T1/(B+T1)
Eg_T2=Eg-A*T2*T2/(B+T2)
T2=Temp+273.15

Figure 16.24: Compact diode model with temperature dependence

566
16.12 Constructing EDD compact device models
and circuit macromodels

Component equations, subcircuits with parameters and EDD models are major
developments for the Qucs circuit simulator. They provide advanced modelling
capabilities with enough power and flexibility to allow a much greater range of
models to be developed than the ones currently provided with each Qucs release.
In the future it is proposed to add new models to the Qucs Web site. The Qucs
team is very keen to encourage all Qucs users to support the modelling effort. If
you have constructed a new model and would like to share it with other Qucs
users please post your model on the qucs-devel or qucs-help mailing lists. Both
the model schematic file and a brief outline of its operation and specification are
requested. An example model specification for the Curtice MESFET device can
be found on the Qucs Web site. Please use the same format when writing model
descriptions.

16.13 End Note

This tutorial note introduces a large number of new modelling concepts and shows
how equations, subcircuits with parameters and the new equation defined device
perform a central role in constructing Qucs models. The EDD approach to mod-
elling makes possible, for the first time, the construction of equation defined com-
pact device models and circuit macromodels using the Qucs schematic capture
facilities as an interactive modelling medium. This is a major step forward for
Qucs. Once again these notes are very much a record of work in progress: much
still remains to be done in the future to improve the modelling capabilities pro-
vided by Qucs. A major short term task will be the development of additional
models covering as wide a range of applications as possible. If Qucs is to fulfill it’s
mission to become a truly universal circuit simulator then it must be supported
by models. Some readers will have noticed that these notes include very little
information about the ADMS-Verlog-A and hand coded C++ model development
routes. This was a deliberate decision on my part. Sometime in the future I intend
to return to these subjects and update the tutorial. A very special thank you must
go to Stefan Jahn for all his hard work, skill, and dedication during the period he
has worked on programming the amazing modelling capabilities now embedded in
Qucs.

567
16.14 Appendix A: Qucs constants, operators and
functions

This appendix lists the constants, operators and a number of functions that are
available for constructing Qucs equations. Items in [...] indicate the equivalent
object in the Verilog-A language. The functions listed are common to Qucs and
Verilog-A. A number of other functions have been implemented in Qucs. The
full list can be found in the Qucs help system; ”Short Description of mathematical
Functions”or in the Qucs ”Measurement Expression Reference Manual“ by Gunther
Kraut and Stefan Jahn, http://qucs.sourceforge.net/docs.html.

• Constants

1. pi = 3.141593...

2. e = 2.718282...

3. kB = 1.380651e-23 J/K

4. -q = -1.602177e-19 C

• Operators

1. +x unary plus

2. -x unary minus

3. x+y addition

4. x-y subtraction

5. x*y multiplication

6. x/y division

7. x%y modulo (remainder)

8. x^y power [pow(x,y)]

9. ?: ternary (condition) ? (expression if true) : (expression if false)

10. || logical or

568
11. && logical and

12. == equal

13. < less than

14. <= less than or equal to

15. > greater than

16. >= greater than or equal to

17. != not equal to

18. ( ) brackets

569
 • Functions

1. ln(x) natural logarithm

2. log10(x) decimal logarithm [log(x)]

3. exp(x) exponential function base e

4. sqrt(x) square root

5. min(x,y) minimum

6. max(x,y) maximum

7. abs(x) absolute value

8. sin(x) sine

9. cos(x) cosine

10. tan(x) tangent

11. arcsin(x) inverse sine [asin(x)]

12. arccos(x) inverse cosine [acos(x)]

13. arctan(x[,y]) inverse tangent [atan2(x,y)]

14. sinh(x) hyperbolic sine

15. cosh(x) hyperbolic cosine

16. tanh(x) hyperbolic tangent

17. arsinh(x) inverse hyperbolic sine [asinh(x)]

18. arcosh(x) inverse hyperbolic cosine [acosh(x)]

19. artanh(x0 inverse hyperbolic tangent [atanh(x)]

20. limexp(x) argument limited exponential function

21. hypot(x,y) Euclidean distance function

570
16.15 Appendix B: Constructing subcircuits with
parameters

In this appendix a series of screen dumps illustrate the sequence needed to con-
struct a subcircuit with parameters. A simple series resonance circuit has been
chosen for the demonstration.

16.15.1 Enter the series resonance circuit and add input and
output pins

Figure 16.25: Stage 1: screen dump showing LCR circuit

571
16.15.2 Change the component names to Ls, Cs and Rs

Figure 16.26: Stage 2: screen dump showing LRC circuit

Figure 16.27: Stage 2: screen dump after name changes

572
16.15.3 Construct symbol for new subcircuit

Right click on the Qucs drawing area and select Edit Circuit symbol or press key
F9. Edit the drawing symbol to give the design shown in Fig. 16.28.

Figure 16.28: Stage 3: the subcircuit symbol

573
16.15.4 Add the names of the subcircuit parameters to the
LCR symbol

Right click on the SUB / File=name caption and enter names of subcircuit pa-
rameters with their default values.

Figure 16.29: Stage 4: entering subcircuit parameter names and default values

Figure 16.30: Stage 4: resulting subcircuit and parameter list with default values

574
16.15.5 Test the LCR subcircuit

Figure 16.31 gives a simple AC transfer function test circuit and resulting wave-
forms. Parameter R SW is swept over the range 1Ω to 10Ω and the AC transfer
function recorded and plotted.
Vin Vout
LCR
I O
V1
U=1 V
SUB1
Rs=R_SW
Cs=1u
Ls=1m

dc simulation ac simulation Parameter

sweep
DC1 AC1
Type=log
Start=100 Hz SW1
Stop=100kHz Sim=AC1
Points=100 Type=lin
Equation Param=R_SW
Start=1
Eqn1 Stop=10
gain=dB(Vout.v) Points=10
phase=phase(Vout.v)

1
Vout.v

0.5

100 1e3 1e4 1e5

acfrequency

0
100
phase
gain

-50
-100

100 1e3 1e4 1e5 100 1e3 1e4 1e5

acfrequency acfrequency

Figure 16.31: Stage 5: Subcircuit test circuit and output waveforms

575
17 Subcircuit and Verilog-A RF
Circuit Simulation Models for
Axial and Surface Mounted
Resistors

17.1 Introduction

Resistors are one of the fundamental building blocks in electronic circuit design.
In most instances conventional resistor circuit simulation models are characterized
by I/V characteristics specified by Ohm’s law. In reality the impedance of RF
resistors is frequency dependent, being determined by component physical prop-
erties, component manufacturing technology and how components are connected
in a circuit. At low frequencies fixed resistors have a nominal value at room tem-
perature and can be modelled accurately by Ohm’s law. At RF frequencies the
fact that a resistor acts more like an inductance or a capacitance can play a cru-
cial role in determining whether or not a circuit operates as designed. Similarly,
if a resistor is modelled as an ideal component at a frequency where it exhibits
significant reactive properties then the resulting simulation data are likely to be
incorrect. The subcircuit and Verilog-A compact resistor models introduced in
this Qucs note are designed to give good performance from low frequencies to RF
frequencies not greater than a few GHz.

17.2 RF Resistor Models

The schematic symbol, I/V equation and parameters of the Qucs linear resistor
model are shown in Figure 17.1. In contrast to this model Figure 17.2 illustrates
the structure of a printed circuit board (PCB) mounted metal film (MF) axial RF

576
resistor (a), its Qucs schematic symbol (b) and its equivalent circuit model (c). A
thin film surface mounted (SMD) resistor can also be represented by the model
shown in Figure 17.2 (c). At signal frequencies where the largest dimension of
an axial or SMD resistor is less than approximately 20 times the smallest signal
wavelength a resistor can be modelled by a lumped passive circuit consisting of
a resistor Rs in series with a small inductance Ls with the combination shunted
by parasitic capacitor Cp. In Figure 17.2 Rs is the nominal value of a resistor at
its parameter extraction temperature Tnom, Ls represents the inductance associ-
ated with Rs where the value of Ls is largely determined by the trimming method
employed during component manufacture to set the value of Rs to a specified tol-
erance. Similarly, capacitor Cp models a parasitic capacitance associated with Rs
where the value of Cp is a function of the physical size of Rs. At RF frequencies
it is important, for accurate operation, to add lead parasitic elements to the in-
trinsic equivalent circuit model shown within the red box draw in Figure 17.2. In
Figure 17.2 Llead and Cshunt represent resistor series lead inductance and shunt
capacitance to ground respectively. A typical set of model parameters for a 51 Ω

Figure 17.1: Qucs built-in resistor model.

Figure 17.2: PCB mounted resistor: (a) axial component mounting, (b) Qucs sym-
bol and (c) equivalent circuit model.

5 % MF axial resistor are (1) Ls=8nH, Cp=1pF, Llead =1nH and Cshunt=0.1pF.

577
Illustrated in Figure 17.3 is a basic S parameter test bench circuit for measuring
the S parameters of an RF resistor over a frequency range 1 MHz to 1.3 GHz. This
example also demonstrates how the real and imaginary parts of a resistor model
impedance can be extracted from S parameter simulation data. The graphs in
Figure 17.3 clearly demonstrate that the impedance of the typical MF RF resistor
described in previous text and modelled by the equivalent circuit shown in Figure
17.2 is a strong function of frequency at higher frequencies in the band 1 MHz to
1.3 GHZ.

Figure 17.3: Qucs S parameter simulation test circuit and plotted output data for
a MF axial resistor: Rs=51Ω, Ls=8nH, Cp=1pF, Llead =1nH and
Cshunt=0.1pF.

578
17.3 Analysis of the RF resistor model

A component level version of the proposed RF resistor model is shown in Figure

17.4, where
Z1 = j · ω · Llead (17.1)
Rs + j · ω · Ls · (1 − ω 2 · Cp · Ls) − j · ω · Cp · Rs2
Z2 = (17.2)
(1 − ω 2 · Cp · Ls)2 + (ω · Cp · Rs)2
j · ω · Llead
Z3 = (17.3)
ω2
(1 − · Llead · Cshunt)
Zseries = Z1 + Z2 = Rseries + j · Xseries (17.4)
Zseries
Zb = Zseries||XCshunt = = ZBR + j · ω · ZBI,
(1 + j · ω · Cshunt · Zseries)
(17.5)
Z = j · ω · Llead + Zb = ZR + j · ω · ZI. (17.6)
Figure 17.5 illustrates how a set of theoretical equations can be converted into

Figure 17.4: RF resistor model rotated through 90 degrees and connected with
one terminal grounded, similar to the test circuit in Figure 17.3. Sec-
tions of the model are shown grouped for calculation of the model
impedance Z.

Qucs equations for model simulation and post simulation data processing. In this
example Qucs equation Eqn1 holds values for RF resistor model parameters and
Qucs equation Eqn2 lists the model equations introduced at the start of this sec-
tion. Figure 17.5 also gives a set of cartesian graphs of post simulation output data

579
Figure 17.5: Theoretical analysis of RF resistance impedance Z using Qucs post
processing facilities: note a dummy simulation icon, in this example
DC simulation, is required to force Qucs to complete the analysis
calculations.

which illustrate how ZR and ZI, and other calculated items, vary with frequency
over the range 1 MHz to 1.3 GHz.

580
17.4 Direct measurement of RF resistor impedance
using a simulated impedance meter

A simple impedance meter for measuring the real and imaginary components of
component and circuit impedance, using small signal AC simulation, is shown
in Figure 17.6. The impedance measuring technique uses a 1 Amp AC constant
current source applied to one terminal of a two port electrical network. The second
terminal is grounded. A parallel high resistance resistor (1E9 Ω in Figure 17.6)
shunts the network under measurement to ensure that there is always a direct
current path to ground as required by the Qucs simulator during the calculation of
simulation results. If required the 1 Amp AC source can be set at a lower value. In
such cases the value of VRes must also be scaled to give the network impedance.

Figure 17.6: A simple Qucs test circuit for demonstrating the use of an AC constant
current source to measure electrical network impedance.

17.5 Extraction of RF resistor data from measured S

parameters

In the past the cost of Vector Network Analyser systems for measuring S pa-
rameters has been prohibitively expensive for individual engineers to purchase.
However, this scene is changing with the introduction of low cost systems like the

581
DGSAQ Vector Network Analyser (VNWA) 1 . This instrument operates over a
frequency band width of 1.3 GHz, providing a range of useful functions with high-
est accuracy at frequencies up to 500 MHz. This form of VNWA is particularly
suited to Radio Amateur requirements and Qucs users interested in RF circuit
analysis and design. Such equipment is ideal for measuring RF circuit S param-
eters and providing measured data for subcircuit and Verilog-A compact device
model parameter extraction. Shown in Figure 17.7 is a graph of measured S pa-
rameter data for a nominal 47 Ω resistor 2 . As well as displaying, and printing,
measured data the DGSAQ Vector Network Analyser software can output data
3
tabulated in Touchstone “SnP“
c file format. These files can be read by Qucs and
their contents attached to an S parameter file icon for inclusion in circuit schematic
diagrams. Figure 17.8 shows this process as part of an RF resistor model param-
eter extraction technique involving DGSAQ VNWA measured S parameter data
and Qucs simulated S parameter data. The brown “Test circuits” box shows test
circuits for firstly reading and processing the DGSAQ VNWA measured data listed
in file mike3.s1p, and for secondly generating simulated S parameter data for an
RF resistor specified by parameters Ls=L, Cp=C, Llead =LL, Cshunt=0.08 pF,
and Rs=47.3 Ω. Presented in Figure 17.9 are the Qucs Optimization controls”
which are used to set the range of L, C and LL values that optimizer ASCO will
select from to obtain the best fit between the measured and simulated S parameter
data. Note in this parameter extraction system that S[1,1] refers to measured S
parameter data and S[2,2] to simulated S parameter data. Two least squares cost
functions called CF1 and CF2 are used as targets in the minimisation process.
Values for CF1 and CF2 can be found in the red box called “Simulation Con-
trols“. In this parameter extraction example the least squares cost function CF1
is employed to minimize the square of the difference between the real values of the
S parameters and least squares cost function CF2 is employed to minimize the
square of the difference between the imaginary values of the S parameters. Qucs
post-simulation processing is also used to extract values for the real and imagi-
nary components of the RF resistor impedance. Both the S parameter data and
the impedance data are displayed as graphs in Figure 17.8. Notice in this exam-
ple the SPICE optimizer ASCO is used to find the values of L, C and LL which
minimize CF1 and CF2. Also note that Rs and Cshunt are held at fixed values
during optimization. In the case of Rs its nominal value can be found from DC
or low frequency AC measurements. Similarly the value selected for Cshunt has
1
DG8SAQ VNWA 3 & 3E- Vector Network Analysers, SDR Kits Limited, Grangeside Busi-
ness Centre, 129 Devizes Road, Trowbridge, Wilts BA14-7sZ, United Kingdom, 2014.
www.SDR-Kits.net.
2
See DG8SAQ VNWA 3 & 3E- Vector Network Analysers- Getting Started Manual for Windows
7, Vista and Windows XP.
3
http://en.wikipedia.org/wiki/Touchstone_filedata

582
been chosen to give a very small but representative value of the parasitic shunt
capacitance.. After optimization finishes the minimized values of L, C and LL are
given in the initial value column of the Qucs optimization Variables list, see Figure
17.9. For the 47 Ω resistor the post-minimization RF resistor model parameters
are Rs=47.3 Ω, Ls=10.43 nH, Cp=0.69 pF, Llead =1.46 nH and Cshunt=0.08 pF.
The theoretical simulation data illustrated in Figure 17.10 shows good agreement
with the measured and the optimized simulation data.

Figure 17.7: DGSAQ Vector Network Analyser S parameter measurements for a

47 Ω axial RF resistor.

583
Figure 17.8: Qucs device model parameter extraction system applied to a
nominal 47 Ω resistor represented by the subcircuit model
illustrated in Figure 17.2 (c). Fixed model parameter val-
ues: Rs=Rm=47.3Ω, CShunt=0.08pF; Optimised values:
Ls=L=10.43nH, Llead =LL=1.47nH, Cp=C =0.69pF. To reduce
simulation time the ASCO cost variance was set to 1e-3. The ASCO
method was set to DE/best/1/exp.

Figure 17.9: Qucs Minimization Icon drop down menus: left ”Variables“ and right
”Goals“.

584
Figure 17.10: Qucs simulation of nominal 47 Ω resistor based on theoretical
analysis.

Figure 17.11: Qucs device model parameter extraction system applied to a nominal
1000 Ω resistor represented by the subcircuit model illustrated in
Figure 17.2 (c).

585
Figure 17.12: Qucs simulation of nominal 1000 Ω resistor based on theoretical
analysis.

586
17.6 Extraction of RF resistor parameters from
measured S data for a nominal 1000 Ω axial
resistor

At low resistance values the impedance of an RF resistor becomes inductive as

the signal frequency is increased. This is due to the fact that the inductance Ls
contribution dominates any reactance effects by Cp, Llead and Cshunt. However,
as Rs is increased above a few hundred Ohm’s the reverse becomes true with
reactive effects dominated by contributions from Cp. Figures 17.11 and 17.12
demonstrate the dominance of Cp reactive effects at low to mid-range frequencies.

17.7 One more example: extraction of RF resistor

parameters from measured S data for a
nominal 100 Ω SMD resistor

Figure 17.13 is included in this Qucs note purely for comparison purposes. SMD
resistors are in general physically very small when compared to axial resistors. This
results in lower values for the inductive and capacative parasitics which in turn
ensures that the high frequency performance of SMD resistors is much improved.

587
Figure 17.13: Qucs device model parameter extraction system applied to a nominal
100 Ω SMD resistor represented by the subcircuit model illustrated
in Figure 17.2 (c).

588
17.8 A Verilog-A RF resistor model

Listed below is an example Verilog-A code model for the RF resistor model intro-
duced in Figure 17.2 (c). Due to the limitations of the Verilog-A language subset
provided by version 2.3.0 of the ”Analogue Device Model Synthesizer“ (ADMS)
4
inductors Ls and Llead are modelled by gyrators and capacitors with values
identical to Ls or Llead.

// Verilog-A module statement.

//
// RFresPCB.va RF resistor (Thin film resistor, axial type, PCB mounting)
//
// This is free software; you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation; either version 2, or (at your option)
// any later version.
//
// Copyright (C), Mike Brinson, mbrin72043@yahoo.co.uk, April 2014.
//
‘include "disciplines.vams"
‘include "constants.vams"
// Verilog-A module statement.
module RFresPCB(RT1, RT2);
inout RT1, RT2; // Module external interface nodes.
electrical RT1, RT2;
electrical n1, n2, n3, nx, ny, nz; // Internal nodes.
‘define attr(txt) (*txt*)
parameter real Rs = 50 from [1e-20 : inf)
‘attr(info="RF resistance" unit="Ohm’s");
parameter real Cp = 0.3e-12 from [0 : inf)
‘attr(info="Resistor shunt capacitance" unit="F");
parameter real Ls = 8.5e-9 from [1e-20 : inf)
‘attr(info="Series induuctance" unit="H");
parameter real Llead = 0.1e-9 from [1e-20 : inf)
‘attr(info="Parasitic lead induuctance" unit="H");
parameter real Cshunt = 1e-10 from [1e-20 : inf)
‘attr(info="Parasitic shunt capacitance" unit="F");
parameter real Tc1 = 0.0 from [-100 : 100]
4
http://sourceforge.net/projects/mot-adms/

589
 ‘attr(info="First order temperature coefficient" unit ="Ohm/Celsius");
parameter real Tc2 = 0.0 from [-100 : 100]
‘attr(info="Second order temperature coefficient" unit ="(Ohm/Celsius)^2");
parameter real Tnom = 26.85 from [-273.15 : 300]
‘attr(info="Parameter extraction temperature" unit="Celsius");
parameter real Temp = 26.85 from [-273.15 : 300]
‘attr(info="Simulation temperature" unit="Celsius");
branch (RT1, n1) bRT1n1; // Branch statements
branch (n1, n2) bn1n2;
branch (n1, n3) bn1n3;
branch (n2, n3) bn2n3;
branch (n3, RT2) bn3RT2;
real Rst, FourKT, n, Tdiff, Rn;
analog begin // Start of analog code
@(initial_model)
begin
Tdiff = Temp-Tnom; FourKT =4.0*‘P_K*Temp;
Rst = Rs*(1.0+Tc1*Tdiff+Tc2*Tdiff*Tdiff); Rn = FourKT/Rst;
end
I(n1) <+ ddt(Cshunt*V(n1)); I(bn1n2) <+ V(bn1n2)/Rst;
I(bn1n3) <+ ddt(Cp*V(bn1n3)); I(n3) <+ ddt(Cshunt*V(n3));
I(bRT1n1) <+ -V(nx); I(nx) <+ V(bRT1n1); // Llead
I(nx) <+ ddt(Llead*V(nx));
I(bn2n3) <+ -V(ny); I(ny) <+ V(bn2n3); // Ls
I(ny) <+ ddt(Ls*V(ny));
I(bn3RT2) <+ -V(nz); I(nz) <+ V(bn3RT2); // Llead
I(nz) <+ ddt(Llead*V(nz));
I(bn1n2) <+ white_noise(Rn, "thermal"); // Noise contribution
end // End of analog code
endmodule

590
Figure 17.14: Details of the proposed RF resistor model: equations, variables and
other data.

591
17.9 Extraction of Verilog-A RF resistor model
parameters from measured S data for a 100 Ω
axial resistor

This example demonstrates the use of ASCO for extracting Verilog-A model pa-
rameters from measured S parameter data. ASCO optimization yields a figure of
4nH forL in the model shown in Figure 17.2 (c). Other model parameter values
are given with the test circuit, see Figure 17.15.

Figure 17.15: Verilog-A models parameter data extraction for a 100 Ω axial
thin film resistor. Fixed model parameter values: Rs=Rm=101Ω,
CShunt=1e-15, Llead =LL=0.5nH, Cp=C =0.43pF; Optimised val-
ues: Ls=L=3.99nH. To reduce simulation time the ASCO cost vari-
ance was set to 1e-3. The ASCO method was set to DE/best/1/exp.

17.10 End Notes

This brief Qucs note outlines the fundamental properties of subicircuit and verilog-
A compact component models for RF resistors. The use of optimization for the
extraction of subcircuit and Verilog-A compact model parameters from measured
S parameters is also demonstrated. The presented techniques form part of the

592
simulation and device modelling capabilities available with the latest Qucs release
5
.

5
Qucs release 1.0.0 rc1, or greater.

593
18 Using Qucs in Textmode

18.1 Introduction

Qucs consists of two parts: The simulator backend and a frontend, provding a
GUI for drawing schematics, controlling simulation, and displaying the simula-
tion results. The operation of the simulation backend is controlled via a text file
(called netlist in the following) which describes the circuit to be simulated and the
simulation(s) to perform in a textual manner. The simulation backend outputs
simulation data. This document describes the syntax of netlist files, shows how
the netlists are actually used to control Qucs, and finally demonstrates how the
simulation data can be visualized via GNU Octave.

Controlling Qucs via netlists and using a separate program for visualizing simula-
tion data may seem complex at first glance; this approach, however, poses the ad-
vantage of allowing more flexible usage scenarios: The potentially cpu-consuming
circuit simulation could be executed on a powerful (even remote) server, while
the result visualization can be done locally on a different computer. By creating
netlists via other programs / scripts, it is easy to setup batch simulations.

18.2 Outline

After defining the prerequisites, Chaper 18.3 presents a basic example netlist and
shows how the simulation data can be visualized in Octave. Chapter 18.8 de-
scribes the various devices and schematic elements provided by Qucs; Chapter
18.12 describes the simulation commands.

594
18.3 Basics

18.4 Prerequisites

Qucs installed qucsator command available

The m-files located under qucs/qucs/octave.

Type:Name [node list] [parameters]

Every schematic element has a type and is instantiated with a specific name.

18.5 Example

In this chapter we will start with a simple example of performing an AC simulation

of the circuit shown in Fig. 18.1.

in R1
out

AC1
C1

gnd

Figure 18.1: Circuit

The texts in red denote the node names and can be chosen freely. The netlist
corresponding to the circuit is shown below.

Vac:V1 in gnd U="1 V" f="1 kHz"

R:R1 out in R="1 kOhm"
C:C1 out gnd C="10 nF"

595
It can be seen that the file is structured line-wise; every line instantiates one
schematic element. The netlist instantiates

• a resistor R1 with resistance R = 1kΩ,

• a capacitor C1 with capacityC = 10nF, and

• an AC source AC1 with ”1”V peak-peak voltage(??) and a frequency of 1kHz.

Storing this netlist in a file rc_ac.net and feeding it into the simulator

qucsator < rc_ac.net

yields the following result:

parsing netlist...
checking netlist...
checker error, no actions defined: nothing to do

Qucs does not know what to do as we did not define any simulations to perform!

18.6 AC Sweep

Deciding to perform an AC sweep, we add the another line to the netlist, yielding:

Vac:V1 in gnd U="1 V" f="1 kHz"

R:R1 out in R="1 kOhm"
C:C1 out gnd C="10 nF"

.AC:AC1 Type="log" Start="1 Hz" Stop="10 MHz" Points="300" Noise="no"

Using this modified netlist with qucs starts an AC analysis of the circuit. The
simulation data is written to stdout; for further processing we save the data in a
file called rc_ac.dat.

596
qucsator < rc_ac.net > rc_ac.dat

The saved data from the simulation can be processed via Octave or Python; the
next two subsections continue the example.

Output Voltage vs. Frequency

100
out.v

10-1

10-2
Voltage

10-3

10-4

10-5

10-6 0
10 101 102 103 104 105 106 107
Frequency

Figure 18.2: Simulation Result

18.6.1 Analysis with Octave

Starting GNU Octave and issuing the following commands

data=loadQucsDataSet(’temp.dat’)
loglog(data(1).data, data(3).data)

produces a log-log plot of the output voltage versus the frequency. A slightly more
polished plot is shown in Fig. 18.2.

597
18.6.2 Analysis with Python

Similar to the octave scripts, Qucs provides a Python script which allows parsing
of the generated simulation data file. The code example below shows how to load
and parse the data file with Python. The plot using Matplotlib is shown in Fig.
18.4.

import numpy as np
import matplotlib.pyplot as plt
import parse_result as pr

data = pr.parse_file(’rc_ac.dat’)

x = data[’acfrequency’]
y = np.abs(data[’out.v’])

plt.loglog(x, y, ’-r’)
plt.grid()
plt.xlabel(’Frequency’)
plt.xlabel(’Voltage’)
plt.legend([’out.v’])
plt.show()
plt.savefig(’rc_ac_python.eps’) # save plot as eps file

598
 Output Voltage vs. Frequency
100
out.v

10-1
Voltage

10-2

10-3 0
10 101 102 103 104 105 106 107
Frequency

Figure 18.3: Simulation Result

Nested Simulations. Qucs allows for nested simulations; as an example we con-

sider an AC analysis together with a parameter sweep. The AC analysis is set up
as before, but in addition the value of the capacitor C1 is increased in 5 steps from
10nF to 100nF. The netlist for this simulation is as follows.

Vac:V1 in gnd U="1 V" f="1 kHz"

R:R1 out in R="1 kOhm"
C:C1 out gnd C="Cx"

.SW:SW1 Sim="AC1" Type="lin" Param="Cx" Start="10 nF" Stop="100 nF" Points="5"

.AC:AC1 Start="1 Hz" Stop="10 MHz" Points="100" Type="log" Noise="no"

The provided Python script can parse data files produced by such a nested simula-
tion as well; in case of two nested simulations it returns not vectors, but matrices.
The Python script below parses the simulation data file and plots the output

599
voltage versus the frequency for two different capacitor values (10nF and 100nF,
respectively). The corresponding plot is shown in Fig. ??.

import numpy as np
import matplotlib.pyplot as plt
import parse_result as pr

data = pr.parse_file(’rc_ac_sweep.dat’)

x = data[’acfrequency’]
y = np.abs(data[’out.v’])
c = data[’Cx’]

plt.loglog(x,y[0,:],’-r’)
plt.loglog(x,y[4,:],’-g’)
plt.legend([’Cx=’ + str(c[0]), ’Cx=’ + str(c[4])])
plt.xlabel(’Frequency’)
plt.ylabel(’Voltage’)
plt.title(’Output Voltage vs. Frequency’)
plt.grid()
plt.savefig(’rc_ac_sweep_python.eps’)
plt.show()

600
 Output Voltage vs. Frequency
100
Cx=1e-08
Cx=1e-07

10-1
Voltage

10-2

10-3

10-4 0
10 101 102 103 104 105 106 107
Frequency

Figure 18.4: Simulation Result

18.7 Transient Simulation

We can use almost the same circuit to perform a transient analysis

Vpulse:V1 in gnd U1=0 U2=1 T1="1 ns" T2="1 ms"

R:R1 out in R="1 kOhm"
C:C1 out gnd C="100 nF"

.TR:TR Type=lin Start="0" Stop="1.5 ms" Points="51"

Here we have replaced the AC sourc with a voltage source generating pulses and
told qucs to perform a transient analysis from 0 to 1.5ms. Storing the simulation
results in a file and using octave to plot the results (analoguous to the previsou
Subsection) yields the plot shown in Fig. 18.5-

601
 Output Voltage vs. Time
1
in.v
out.v

0.8

0.6
Voltage

0.4

0.2

0
0 0.0005 0.001 0.0015 0.002
Time

Figure 18.5: Simulation Result

18.8 Qucs Devices

18.9 Passive Devices

18.9.1 Resistor

R:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

R ohmic resistance in Ohms n/a yes
Temp simulation temperature in degree Celsius 26.85 no
Tc1 first order temperature coefficient 0.0 no
Tc2 second order temperature coefficient 0.0 no
Tnom temperature at which parameters were 26.85 no
extracted

602
18.9.2 Capacitor

C:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

C capacitance in Farad n/a yes
V initial voltage for transient simulation n/a no

18.9.3 Inductor

L:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

L inductance in Henry n/a yes
I initial current for transient simulation n/a no

18.10 Nonlinear Components

18.10.1 Diode

Diode:Name Cathode-Node Anode-Node [Parameters]

603
 Parameter Name Default Value Mandatory
Is saturation current 1e-15 A yes
N emission coefficient 1 yes
Cj0 zero-bias junction capacitance 10 fF yes
M grading coefficient 0.5 yes
Vj junction potential 0.7 V yes
Fc forward-bias depletion capacitance coef- 0.5 no
ficient
Cp linear capacitance 0.0 fF no
Isr recombination current parameter 0.0 no
Nr emission coefficient for Isr 2.0 no
Rs ohmic series resistance 0.0 Ohm no
Tt transit time 0.0 ps no
Ikf high-injection knee current (0=infinity) 0 no
Kf flicker noise coefficient 0.0 no
Af flicker noise exponent 1.0 no
Ffe flicker noise frequency exponent 1.0 no
Bv reverse breakdown voltage 0 no
Ibv current at reverse breakdown voltage 1 mA no
Temp simulation temperature in degree Celsius 26.85 no
Xti saturation current temperature exponent 3.0 no
Eg energy bandgap in eV 1.11 no
Tbv Bv linear temperature coefficient 0.0 no
Trs Rs linear temperature coefficient 0.0 no
Ttt1 Tt linear temperature coefficient 0.0 no
Ttt2 Tt quadratic temperature coefficient 0.0 no
Tm1 M linear temperature coefficient 0.0 no
Tm2 M quadratic temperature coefficient 0.0 no
Tnom temperature at which parameters were 26.85 no
extracted
Area default area for diode 1.0 no

18.10.2 Bipolar Junction Transistor with Substrate

BJT:Name Base-Node Collector-Node Emitter-Node Substrate-Node [Parameters]

604
Parameter Name Default Value Mandatory
Type Polarity [npn, pnp] n/a yes
Is saturation current 1e-16 yes
Nf forward emission coefficient 1 yes
Nr reverse emission coefficient 1 yes
Ikf high current corner for forward beta 0 yes
Ikr high current corner for reverse beta 0 yes
Vaf forward early voltage 0 yes
Var reverse early voltage 0 yes
Ise base-emitter leakage saturation current 0 yes
Ne base-emitter leakage emission coefficient 1.5 yes
Isc base-collector leakage saturation current 0 yes
Nc base-collector leakage emission coefficient 2 yes
Bf forward beta 100 yes
Br reverse beta 1 yes
Rbm minimum base resistance for high cur- 0 yes
rents
Irb current for base resistance midpoint 0 yes
Rc collector ohmic resistance 0 yes
Re emitter ohmic resistance 0 yes
Rb zero-bias base resistance (may be high- 0 yes
current dependent)
Cje base-emitter zero-bias depletion capaci- 0 yes
tance
Vje base-emitter junction built-in potential 0.75 yes
Mje base-emitter junction exponential factor 0.33 yes
Cjc base-collector zero-bias depletion capaci- 0 yes
tance
Vjc base-collector junction built-in potential 0.75 yes
Mjc base-collector junction exponential factor 0.33 yes
Xcjc fraction of Cjc that goes to internal base 1.0 yes
pin
Cjs zero-bias collector-substrate capacitance 0 yes
Vjs substrate junction built-in potential 0.75 yes
Mjs substrate junction exponential factor 0 yes
Fc forward-bias depletion capacitance coef- 0.5 yes
ficient
Tf ideal forward transit time 0.0 yes
Xtf coefficient of bias-dependence for Tf 0.0 yes
Vtf voltage dependence of Tf on base- 0.0 yes
collector voltage
Itf high-current effect on Tf 0.0 yes
Tr ideal reverse transit time 0.0 yes
Temp simulation temperature in degree Celsius 26.85 no
605
Kf flicker noise coefficient 0.0 no
Af flicker noise exponent 1.0 no
Ffe flicker noise frequency exponent 1.0 no
Kb burst noise coefficient 0.0 no
Ab burst noise exponent 1.0 no
Fb burst noise corner frequency in Hertz 1.0 no
Ptf excess phase in degrees 0.0 no
18.10.3 Diac

Diac:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Vbo (bidirectional) breakover voltage 30 V yes
Ibo (bidirectional) breakover current 50 uA yes
Cj0 parasitic capacitance 10 pF no
Is saturation current 1e-10 A no
N emission coefficient 2 no
Ri intrinsic junction resistance 10 Ohm no
Temp simulation temperature 26.85 no

18.10.4 Silicon Controlled Rectifier (SCR)

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Vbo breakover voltage 400 V todo
Igt gate trigger current 50 uA todo
Cj0 parasitic capacitance 10 pF todo
Is saturation current 1e-10 A todo
N emission coefficient 2 todo
Ri intrinsic junction resistance 10 Ohm todo
Rg gate resistance 5 Ohm todo
Temp simulation temperature 26.85 todo

18.10.5 Triac (Bidirectional Thyristor)

xxx:Name Node1 Node2 [Parameters]

606
 Parameter Name Default Value Mandatory
Vbo (bidirectional) breakover voltage 400 V todo
Igt (bidirectional) gate trigger current 50 uA todo
Cj0 parasitic capacitance 10 pF todo
Is saturation current 1e-10 A todo
N emission coefficient 2 todo
Ri intrinsic junction resistance 10 Ohm todo
Rg gate resistance 5 Ohm todo
Temp simulation temperature 26.85 todo

18.10.6 Resonance Tunnel Diode

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Ip peak current 4 mA todo
Iv valley current 0.6 mA todo
Vv valley voltage 0.8 V todo
Wr resonance energy in Ws 2.7e-20 todo
eta Fermi energy in Ws 1e-20 todo
dW resonance width in Ws 4.5e-21 todo
Tmax maximum of transmission 0.95 todo
de fitting factor for electron density 0.9 todo
dv fitting factor for voltage drop 2.0 todo
nv fitting factor for diode current 16 todo
Cj0 zero-bias depletion capacitance 80 fF todo
M grading coefficient 0.5 todo
Vj junction potential 0.5 V todo
te life-time of electrons 0.6 ps todo
Temp simulation temperature in degree Celsius 26.85 todo
Area default area for diode 1.0 todo

18.10.7 Junction Field-effect Transistor

JFET:Name Gate-Node Drain-Node Source-Node [Parameters]

607
 Parameter Name Default Value Mandatory
Type polarity [nfet, pfet] n/a yes
Vt0 threshold voltage -2.0 V yes
Beta transconductance parameter 1e-4 yes
Lambda channel-length modulation parameter 0.0 yes
Rd parasitic drain resistance 0.0 yes
Rs parasitic source resistance 0.0 yes
Is gate-junction saturation current 1e-14 yes
N gate-junction emission coefficient 1.0 yes
Isr gate-junction recombination current pa- 1e-14 yes
rameter
Nr Isr emission coefficient 2.0 yes
Cgs zero-bias gate-source junction capaci- 0.0 yes
tance
Cgd zero-bias gate-drain junction capacitance 0.0 yes
Pb gate-junction potential 1.0 yes
Fc forward-bias junction capacitance coeffi- 0.5 yes
cient
M gate P-N grading coefficient 0.5 yes
Kf flicker noise coefficient 0.0 no
Af flicker noise exponent 1.0 no
Ffe flicker noise frequency exponent 1.0 no
Temp simulation temperature in degree Celsius 26.85 no
Xti saturation current temperature exponent 3.0 no
Vt0tc Vt0 temperature coefficient 0.0 no
Betatce Beta exponential temperature coefficient 0.0 no
Tnom temperature at which parameters were 26.85 no
extracted
Area default area for JFET 1.0 no

18.10.8 MOS field-effect transistor with substrate

MOSFET:Name Gate-Node Drain-Node Source-Node Bulk-Nod[Parameters]

608
Parameter Name Default Value Mandatory
Type polarity [nfet, pfet] n/a yes
Vt0 zero-bias threshold voltage 1.0 V yes
Kp transconductance coefficient in A/V2 2e-5 yes
Gamma bulk threshold in sqrt(V) yes
Phi surface potential 0.6 V yes
Lambda channel-length modulation parameter in 0.0 yes
1/V
Rd drain ohmic resistance 0.0 Ohm no
Rs source ohmic resistance 0.0 Ohm no
Rg gate ohmic resistance 0.0 Ohm no
Is bulk junction saturation current 1e-14 A yes
N bulk junction emission coefficient 1.0 yes
W channel width 1 um no
L channel length 1 um no
Ld lateral diffusion length 0.0 no
Tox oxide thickness 0.1 um no
Cgso gate-source overlap capacitance per me- 0.0 no
ter of channel width in F/m
Cgdo gate-drain overlap capacitance per meter 0.0 no
of channel width in F/m
Cgbo gate-bulk overlap capacitance per meter 0.0 no
of channel length in F/m
Cbd zero-bias bulk-drain junction capacitance 0.0 F no
Cbs zero-bias bulk-source junction capaci- 0.0 F no
tance
Pb bulk junction potential 0.8 V no
Mj bulk junction bottom grading coefficient 0.5 no
Fc bulk junction forward-bias depletion ca- 0.5 no
pacitance coefficient
Cjsw zero-bias bulk junction periphery capaci- 0.0 no
tance per meter of junction perimeter in
F/m
Mjsw bulk junction periphery grading coeffi- 0.33 no
cient
Tt bulk transit time 0.0 ps no
Nsub substrate bulk doping density in 1/cm3 0.0 no
Nss surface state density in 1/cm2 0.0 no
Tpg gate material type: 0 = alumina; -1 = 0 no
same as bulk; 1 = opposite to bulk
Uo surface mobility in cm2 /Vs 600.0 no
Rsh drain and source diffusion sheet resis- 0.0 no
tance in Ohms/square
Nrd number of equivalent drain squares 1 no
609
Nrs number of equivalent source squares 1 no
Cj zero-bias bulk junction bottom capaci- 0.0 no
tance per square meter of junction area
in F/m2
Js bulk junction saturation current per 0.0 no
square meter of junction area in A/m2
Ad drain diffusion area in m2 0.0 no
18.11 Sources

18.11.1 DC Voltage Source

V:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

U voltage in Volts yes

18.11.2 AC Voltage Source

Vac:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

U peak voltage in Volts n/a yes
f frequency in Hertz n/a no
Phase initial phase in degrees 0 no
Theta damping factor (transient simulation 0 no
only)

18.11.3 Voltage Controlled Voltage Source

VCVS:Name Node1 Node2 Node3 Node4 [Parameters]

Node1 is the input , Node2 is the output, Node3 is the ground for the input, and
Node4 is the ground for the output.

Parameter Name Default Value Mandatory

G forward transfer factor 1.0 todo
T delay time 0 todo

18.11.4 Voltage Controlled Current Source

VCVS:Name Node1 Node2 Node3 Node4 [Parameters]

610
Node1 is the input , Node2 is the output, Node3 is the ground for the input, and
Node4 is the ground for the output.

Parameter Name Default Value Mandatory

G forward transconductance 1.0 todo
T delay time 0 todo

18.11.5 Current Controlled Voltage Source

VCVS:Name Node1 Node2 Node3 Node4 [Parameters]

Node1 is the input , Node2 is the output, Node3 is the ground for the input, and
Node4 is the ground for the output.

Parameter Name Default Value Mandatory

G forward transfer factor 1.0 todo
T delay time 0 todo

18.11.6 Current Controlled Current Source

VCVS:Name Node1 Node2 Node3 Node4 [Parameters]

Node1 is the input , Node2 is the output, Node3 is the ground for the input, and
Node4 is the ground for the output.

Parameter Name Default Value Mandatory

G forward transfer factor 1.0 todo
T delay time 0 todo

18.11.7 Voltage Pulse

Vpulse:Name Node1 Node2 [Parameters]

611
 Parameter Name Default Value Mandatory
U1 voltage before and after the pulse 0V yes
U2 voltage of the pulse 1V yes
T1 start time of the pulse 0 yes
T2 ending time of the pulse 1 ms yes
Tr rise time of the leading edge 1 ns no
Tf fall time of the trailing edge 1 ns no

18.11.8 Current Pulse

Ipulse:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

I1 current before and after the pulse 0V yes
I2 current of the pulse 1V yes
T1 start time of the pulse 0 yes
T2 ending time of the pulse 1 ms yes
Tr rise time of the leading edge 1 ns no
Tf fall time of the trailing edge 1 ns no

18.11.9 Rectangle Voltage

Vrect:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

U voltage of high signal 1V yes
TH duration of high pulses 1 ms yes
TL duration of low pulses 1 ms yes
Tr rise time of the leading edge 1 ns todo
Tf fall time of the leading edge 1 ns todo
Td initial delay time 0 ns todo

18.11.10 Rectangle Current

Irect:Name Node1 Node2 [Parameters]

612
 Parameter Name Default Value Mandatory
I current at high pulse 1 mA yes
TH duration of high pulses 1 ms yes
TL duration of low pulses 1 ms yes
Tr rise time of the leading edge 1 ns todo
Tf fall time of the leading edge 1 ns todo
Td initial delay time 0 ns todo

18.11.11 Exponential Voltage Source

Vexp:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

U1 voltage before rising edge 0V todo
U2 maximum voltage of the pulse 1V todo
T1 start time of the exponentially rising edge 0 todo
T2 start of exponential decay 1 ms todo
Tr rise time of the rising edge 1 ns todo
Tf fall time of the falling edge 1 ns todo

18.11.12 Exponential Current Source

Vexp:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

I1 Current before rising edge 0V todo
I2 maximum current of the pulse 1A todo
T1 start time of the exponentially rising edge 0 todo
T2 start of exponential decay 1 ms todo
Tr rise time of the rising edge 1 ns todo
Tf fall time of the falling edge 1 ns todo

18.11.13 AC Voltage Source with Ammplitude Modulator

AM_Mod:Name Node1 Node2 Node3 [Parameters]

613
Node1 is the modulated output, Node2 is the common ground, and Node3 is the
modulation input.

Parameter Name Default Value Mandatory

U peak voltage in Volts 1V todo
f frequency in Hertz 1 GHz todo
Phase initial phase in degrees 0 todo
m modulation level 1.0 todo

18.11.14 AC Voltage Source with Ammplitude Modulator

AM_Mod:Name Node1 Node2 Node3 [Parameters]

Node1 is the modulated output, Node2 is the common ground, and Node3 is the
modulation input.

Parameter Name Default Value Mandatory

U peak voltage in Volts 1V todo
f frequency in Hertz 1 GHz todo
Phase initial phase in degrees 0 todo
m modulation level 1.0 todo

18.11.15 AC Voltage Source with Phase Modulator

PM_Mod:Name Node1 Node2 Node3 [Parameters]

Node1 is the modulated output, Node2 is the common ground, and Node3 is the
modulation input.

Parameter Name Default Value Mandatory

U peak voltage in Volts 1V todo
f frequency in Hertz 1 GHz todo
Phase initial phase in degrees 0 todo
M modulation index 1.0 todo

614
18.12 Simulation Commands

18.13 DC Simulation

.DC:Name [Parameters]

Parameter Name Default Value Mandatory

Temp simulation temperature in degree Celsius 26.85 no
reltol relative tolerance for convergence 0.001 no
absol absolute tolerance for currents 1 pA no
vntol absolute tolerance for voltages 1 uV no
saveOPs put operating points into dataset [yes,no] no
MaxIter maximum number of iterations until error 150 no
saveAll save subcircuit nodes into dataset [yes,no] no no
convHelper preferred convergence algorithm [none, none
gMinStepping, SteepestDescent, Line-
Search, Attenuation, SourceStepping]
Solver method for solving the circuit matrix CroutLU no
[CroutLU, DoolittleLU, HouseholderQR,
HouseholderLQ, GolubSVD]

18.14 AC Simulation

.AC:Name [Parameters]

Parameter Name Default Value Mandatory

Type sweep type [lin,log,list,const] n/a yes
Start start frequency in Hertz n/a yes
Stop stop frequency in Hertz n/a yes
Points number of simulation steps n/a yes
Noise calculate noise voltages no no

18.15 Parameter Sweep

.SW:Name [Parameters]

615
 Parameter Name Default Value Mandatory
Sim simulation to perform parameter sweep n/a yes
on
Type sweep type [lin,log,list,const] n/a yes
Param parameter to sweep n/a yes
Stop start value for sweep n/a yes
Start stop value for sweep n/a yes

18.16 Transient Simulation

.TR:Name [Parameters]

Parameter Name Default Value Mandatory

Type sweep type [lin,log,list,const] todo todo
Start start time in seconds todo todo
Stop stop time in seconds todo todo
Points number of simulation time steps 11(todo) todo
IntegrationMethod integration method [Euler, Trapezoidal, Trapezoidal todo
Gear, AdamsMoulton]
Order order of integration method 2 todo
InitialStep initial step size in seconds 1 ns (todo) todo
MinStep minimum step size in seconds 1e-16 todo
MaxIter maximum number of iterations until error 150 todo
reltol relative tolerance for convergence 0.001 todo
abstol absolute tolerance for currents 1 pA todo
vntol absolute tolerance for voltages 1 uV todo
Temp simulation temperature in degree Celsius 26.85 todo
LTEreltol relative tolerance of local truncation error 1e-3 todo
LTEabstol absolute tolerance of local truncation er- 1e-6 todo
ror
LTEfactor overestimation of local truncation error 1 todo
Solver method for solving the circuit matrix CroutLU todo
[CroutLU, DoolittleLU, HouseholderQR,
HouseholderLQ, GolubSVD]
relaxTSR relax time step raster [no, yes] yes todo
initialDC perform an initial DC analysis [yes, no] yes todo
MaxStep maximum step size in seconds 0 todo

616
18.17 S Parameter Simulation

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Type sweep type [lin,log,list,const] todo todo
Start start frequency in Hertz 1 GHz todo
Stop stop frequency in Hertz 10 GHz todo
Points number of simulation steps 19 todo
Noise calculate noise parameters no todo
NoiseIP input port for noise figure 1 todo
NoiseOP output port for noise figure 2 todo
saveCVs put characteristic values into dataset no todo
[yes,no]
saveAll save subcircuit characteristic values into no todo
dataset [yes,no]

18.17.1 xxx

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

L inductance in Henry n/a yes

18.17.2 Silicon Controlled Rectifier (SCR)

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Vbo breakover voltage 400 V todo
Igt gate trigger current 50 uA todo
Cj0 parasitic capacitance 10 pF todo
Is saturation current 1e-10 A todo
N emission coefficient 2 todo
Ri intrinsic junction resistance 10 Ohm todo
Rg gate resistance 5 Ohm todo
Temp simulation temperature 26.85 todo

617
18.17.3 Triac (Bidirectional Thyristor)

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Vbo (bidirectional) breakover voltage 400 V todo
Igt (bidirectional) gate trigger current 50 uA todo
Cj0 parasitic capacitance 10 pF todo
Is saturation current 1e-10 A todo
N emission coefficient 2 todo
Ri intrinsic junction resistance 10 Ohm todo
Rg gate resistance 5 Ohm todo
Temp simulation temperature 26.85 todo

18.17.4 Resonance Tunnel Diode

xxx:Name Node1 Node2 [Parameters]

Parameter Name Default Value Mandatory

Ip peak current 4 mA todo
Iv valley current 0.6 mA todo
Vv valley voltage 0.8 V todo
Wr resonance energy in Ws 2.7e-20 todo
eta Fermi energy in Ws 1e-20 todo
dW resonance width in Ws 4.5e-21 todo
Tmax maximum of transmission 0.95 todo
de fitting factor for electron density 0.9 todo
dv fitting factor for voltage drop 2.0 todo
nv fitting factor for diode current 16 todo
Cj0 zero-bias depletion capacitance 80 fF todo
M grading coefficient 0.5 todo
Vj junction potential 0.5 V todo
te life-time of electrons 0.6 ps todo
Temp simulation temperature in degree Celsius 26.85 todo
Area default area for diode 1.0 todo

618
19 Active Filters Design with Qucs
and Qucsactivefilter

19.1 Introduction

The purpose of this manual is to explain bases of active filter design methods using
qucs-activefilter tool. You can start qucs-activefilter by clicking in the
Qucs main menu Tools->Active Filters. Qucsactivefilter provides easy and power-
ful tool for manipulations with active filters. Qucsactivefilter can operate only ac-
tive filters. For passive filter use Qucsfilter instead. Basic explanations about active
filters could be found here: https://en.wikipedia.org/wiki/Active_filter

Qucsfilter builds active filters circuits based on RC-components and operational

amplifier (opamp). Qucsactivefilter uses ideal opamps. It uses 1-3 opamps per
filter section. The number of opamps depends on selected approximation type,
filter type and topology.

19.2 Interface description

The main window of this tool is shown in the Figure 19.1

619
 Figure 19.1: Qucs-Activefilter main window

Main windows contains five groups of controls:

1. Filter parameters input fields (left part of the window). You should filter
frequency response parameters here. These parameters are shown on the
frequency response preview. This part of the window also contains Calculate
and copy to clipboard push button.

2. Filter frequency response preview (middle part of the window).

3. Filter topology preview (right part of the window). You can see common
form of the filter section topology here.

4. Calculation console. Filter order, poles/zeros list and part list are printed
here if filter calculation is successful. Part list contains RC-elements values
for each section of the filter.

5. Menu bar. You can access File->Exit and get short Help here.

620
19.3 Filter transfer function

Active filters are characterized by transfer function in frequency domain. Common

form of the filter transfer function is given here:

bm sm + bm−1 sm−1 + . . . + b2 s2 + b1 s + b0
H(s) = (19.1)
an sn + an−1 sn−1 + . . . + a2 s2 + a1 s + a0

Filter order N is:

N = max(m, n) (19.2)

Filter order determines the number of filter sections and filter circuit complexity.
Active filter consists of k = N/2 2-nd order section and k = N %2 1-st order
sections.

Zeros of the transfer function are roots of numerator. Poles are the roots of de-
nominator. We need to know filter transfer function to determine components
(resistors and capacitors — RC) values of the active filter circuit.

Qucsactivefilter uses user-defined filter parameters in frequency domain to deter-

mine filter order, transfer function and RC-elements value for each section of the
filter.

Qucsactivefilter uses filter design algorithms provided by [D. Johnson, J. Johnson,

H. Moore A handbook of active filters — Prentice-Hall, Inc, Engewood Cliffs.,
N.J.07632, USA, — 1980].

19.4 Filters parameters explanation

We need to define following four groups of parameters to calculate active filter:

1. Frequency response approximation type. Butterworth, Chebyshev, Inverse

Chebyshev, Cauer (Elliptic) and Bessel filters are available.

2. Frequency response parameters: filter gain and bandwidth.

621
 3. Filter topology. Sallen-Key, Mutifeedback (MFB) and Cauer topologies are
available.

4. Filter type. Low-pass, high-pass, band-pass and band-stop filters are avail-
able.

All of these parameters are presented in the left side of Qucsactivefilter main
window.

Different filter topologies have different number of opamps, resistors and capacitors
per section. Sallen-Key and MFB topologies are the most suitable for Chebyshev
and Butterworth filters.

Frequency response parameters differ for various filter types (low-pass, high-pass,
band-pass, band-stop) and approximations.

Frequency response of low-pass and high-pass filters has following parameters:

1. Cutoff frequency Fc

2. Passband attenuation Ap

3. Passband ripple Rp (for Chebyshev and Cauer filters only)

4. Stopband attenuation As

5. Stopband frequency Fs

6. Passband gain K

Qucsactivefilter estimates filter order automatically for Chebyshev, Butterworth

and Cauer filter. Minimal order that provides required frequency domain parame-
ter is used. You don’t need define filter order manually for these approximations.
filtersOrder could not be determined automatically for Bessel filter. You should
define filer order for the Bessel filters manually.

Frequencies should be defined in Hertz (Hz). Attenuation and ripple should be

defined in decibels (dB).

Frequency response of band-pass and band-stop filters has following parameters:

1. Upper cutoff frequency Fu

622
 2. Lower cutoff frequency Fl

3. Transient bandwidth T W is gap between pass band and stop band.

4. Passband ripple Rp (for Chebyshev and Cauer filters only)

5. Stopband attenuation As

6. Passband gain K

Only Chebyshev, inverse Chebyshev and Cauer filters have ripple in pass band.
Butterworth and Bessel have no. Cauer filter has ripple in stop band too. Qucs-
activefilter suggests that stop band ripple less than stopband attenuation.

All of these parameters you can see in frequency response preview in the middle
part of the Qucsactivefilter main window.

19.5 Filter design example

For example, consider Chebyshev low-pass filter with following parameters:

1. Cutoff frequency: 2 kHz;

2. Passband ripple: 2 dB;

3. Stopband frequency: 2.2 kHz

4. Stopband attenuation: 20 dB

5. Passband gain: 0 dB

At first step we need to put these parameters into corresponding input fields. in the
left part of the main window. Then we need to select Chebyshev approximation in
the Approximation type combo box, select Sallen-Key topology in Filter topology
combo box and select low-pass filter in Filter type combo box. The next figure
presents main window with filled input fields for our example.

623
 Figure 19.2: Sallen-Key low-pass filter design example with Qucsactivefilter

As all input fields are filled, we can press Calculate and copy to clipboard button.
After this button is clicked, we can see calculation results (transfer function poles
and zeros list and part list) in the bottom part of the main window. Filter is
calculated successfully. If there was errors during filter calculation, calculation
process is aborted and warning message box appears. You should change frequency
response parameters and/or filter topology in such case.

You can use components values from the part list for active filter simulation with
external circuit simulation program.

Now filter schematic is in the system clipboard. We can switch back to the Qucs
schematic window and press Ctrl+V or Edit->Paste. Filter schematic appears
(Fig. 19.3)

624
 Figure 19.3: Sallen-Key filter schematic in Qucs

This schematic already contains AC and DC simulations and equation for the filter
gain calculation (K parameter). We can press F2 and simulate it. Simulation
completes and we can switch to the display page and place Cartesian plot on it. If
we place K graph on this plot, we can see frequency response of the filter (Fig.19.4).
This frequency response meets all required parameters.

0 acfrequency: 2.2e+03
K: -20.2

-20

-40
K

-60

-80

-100

-120
0 500 1e3 1.5e3 2e3 2.5e3 3e3 3.5e3 4e3 4.5e3 5e3 5.5e3 6e3
acfrequency

Figure 19.4: Filter simulation results (frequency response)

625
19.6 Manual transfer function definition

Using Qucsactivefilter you can define numerator and denominator coefficients man-
ually. It’s need to select User defined transfer function in Approximation type
combo box. Manually define transfer function button becomes available. Transfer
function setup dialog (Fig.19.5) appears after the click on this button.

Figure 19.5: Manual transfer function definition

You can fill two columns of the table and define numerator (ai ) and denominator
(bi ) transfer function coefficients. Then you can press Accept button and calculate
active filter with given topology.

Presented example (Fig.19.5) implements the following transfer function:

1
H(s) = (19.3)
s4 + 10s3 + 45s2 + 105s + 105

This transfer function implements 4th-order Bessel filter.

626
19.7 Conclusion

Qucsactivefilter tool was considered. You can easy design active filter with this
tool and Qucs. Report about any bugs for Qucsactivefilter to Vadim Kuznetsov
(E-mail: ra3xdh@gmail.com).

627