A Novel Static Model in d-q Coordinates for a Common-Emitter

Amplifier
This paper was downloaded from TechRxiv (https://www.techrxiv.org).

LICENSE

CC BY 4.0

SUBMISSION DATE / POSTED DATE

28-07-2023 / 30-07-2023

CITATION

Campos, Jose (2023). A Novel Static Model in d-q Coordinates for a Common-Emitter Amplifier. TechRxiv.
Preprint. https://doi.org/10.36227/techrxiv.23798628.v1

DOI

10.36227/techrxiv.23798628.v1
 A Novel Static Model in d-q Coordinates for a
Common-Emitter Amplifier
José M. Campos-Salazar
Electronic Engineering Department
Universitat Politècnica de Catalunya
Barcelona, Spain
jose.manuel.campos@upc.edu

Abstract—This paper introduces a static model in dq This paper introduces a novel model of an amplifier in the
coordinates for a common-emitter amplifier configuration to common emitter configuration (ECA) utilizing variables with
address low-power signal amplification. The model is derived static dynamics, represented in dq coordinates. The
through transformation from sinusoidal to rotational transformation process involves converting the original
coordinates. Linear compensators face challenges with system, initially described in sinusoidal coordinates, into an
sinusoidal waveforms, hence the need for static modeling. - rotational coordinate system using the Clarke transform.
Simulations validate the model's reliability. Overall, this
Subsequently, the Park transform is applied to this rotating
contribution improves the common-emitter amplifier
system, resulting in a static system represented in dq
configuration modeling and analysis.
coordinates, which are situated within a rotating reference
Keywords—BJT, common-emitter amplifier configuration, frame [22]. This approach allows for a comprehensive
dq-model, modeling, small-signal model, state-space model understanding and analysis of the ECA's behavior in steady-
state conditions.
I. INTRODUCTION
To validate the accuracy and proper functioning of the dq
The theory of low-power signal amplification is a model, an inverse transformation is applied to convert the dq
prominent topic in the field of microelectronics, highlighting model back to the standard model. This process allows for a
significant research interest [1]–[15]. Among the commonly comparison of variables between the original standard model
used devices for amplification in this domain are bipolar and the reconverted model. By evaluating the correlation
junction transistors (BJTs) and operational amplifiers (OP- between these variables, the suitability of the derived model
Amps) [1]–[15]. On the other hand, field-effect transistors can be determined.
(JFETs and MOSFETs) find primary utility in switching Furthermore, both the standard model and the model in dq
applications [16] and are less prevalent in amplifier setups. coordinates are subjected to simulations to evaluate and
OP-Amps have gained considerable prominence for confirm the reliability of the proposed model. Through these
amplifier applications across a broad frequency spectrum, simulations, the performance and behavior of the ECA can be
whereas the use of BJTs for amplification purposes has thoroughly analysed and compared in different scenarios and
declined, except in power amplifier applications [17], [18]. operating conditions.
For medium-frequency voltage amplification, OP-Amps Taking into account these verification steps, the paper
have become the preferred choice [17], [18]. However, in ensures the properly of the dq model and establishes its
such applications, the amplification mechanisms are not practicality and effectiveness, making it a simple contribution
regulated and rely on resistors in the bias circuit [17], [18]. to the field of ECA modeling and analysis.
This differs from amplifiers based on BJTs, where the gain
expression is directly influenced by the load resistor RL, The paper is organized as follows: section I introduces the
leading to complications in regulating the output voltage due problem and the motivation of the proposal. Sections II and III
to potential disturbances in RL. introduce the ECA topology and present the modeling of the
ECA in both dc and ac regimes, respectively. Section IV is the
To address the drawback of gain dependence on RL, a core of the paper and introduces the proposed model in dq
fundamental concept from control theory is introduced here. coordinates. Finally, Sections V and VI present the simulation
By incorporating a feedback loop that compares the output results and conclusions, respectively.
voltage with a reference value, an error signal is generated.
This error signal is then fed to a controller or compensator, II. ECA TOPOLOGY
which produces a signal to drive the actuator of the amplifier The topology of the ECA is depicted in Fig. 1. This
[19], [20]. The challenge lies in determining the appropriate configuration is the most common within the transistor
actuator and compensator to achieve the desired operation of amplifier family [23]. In this circuit the BJT is represented by
the amplifier. Q1. The topology also shows the bias resistors R1 and R2, the
However, due to the sinusoidal nature of the waveforms collector resistor RC and the emitter resistor RE. The bias
with which the amplifiers operate, their output dynamics - in voltage is defined as VCC. In the ac operating regime, the CC
particular the output voltages - remain sinusoidal even in coupling capacitor connects to the circuit, the input stage,
steady-state conditions. This sinusoidal behavior provides a configured by a sinusoidal voltage source vi and an inductor
challenge for the use of linear compensators, even when the L operating as an input filter. The coupling capacitor CB
amplifier models themselves are linear in nature [19]–[21]. allows to connect the ECA with the output stage comprised
by a capacitor C operating as an output filter and the load
To overcome this limitation, a viable solution is to derive resistor RL. The currents ii, iC, iB, and iE are the input,
an amplifier model with static dynamics. This provides a collector, base, and emitter currents respectively. The
model from which a linear compensator can be designed.
 Fig. 1. ECA topology.
(a)

voltages vBE and vCE are the voltages measured between the
base and emitter and between collector and emitter,
respectively. Finally, vo and vC are the output and capacitor
voltages respectively. Also, vo = vC is verified.
III. MODELING OF THE ECA
In this section, the fundamental equations governing the
ECA are derived, taking into account both the dc and ac
operating conditions. It is assumed that transistor Q1 operates
in the linear region, ensuring accurate signal amplification
[17], [18]. (b)
Fig. 2. ECA operating in dc regime. (a) Topology in dc bias. The
A. dc operation capacitors CB and CC act as open circuits. (b) Topology in dc bias with
During dc operation, the coupling capacitors (CB and CC) equivalent Thévenin input network.
behave as open circuits, leading to a modified circuit The small-signal model shown in Fig. 3 is derived and
configuration depicted in Fig 2a, commonly known as a dc defined in (5) in a state-space format.
biased ECA [17], [18]. By combining resistors R1, R2, and the
𝐱 𝐀∙𝐱 𝐁∙𝑢
voltage source VCC, a Thévenin equivalent circuit can be . (5)
derived, resulting in a new ECA circuit specifically designed 𝐲 𝐂∙𝐱 𝐃∙𝑢
for the dc regime, as illustrated in Fig. 2b. It should be noted
that in Fig. 2, lowercase letters are replaced with uppercase In this model, the input is scalar and defined as u = vi. On
letters, as it is assumed that the variables represent solely dc the other hand, the state vector is defined as x = [ii,vo]T.
components. The definitions of VBB and RB, are defined in (1) Moreover, in this case the output vector is defined as y = x.
and (2) respectively. Symbolically, {x, y}  {ℝ2}. The model matrices are defined
in (6) and are the state matrix (A), the input matrix (B), the
𝑉 ∙𝑉 (1) output matrix (C) and the direct-transmission matrix (D).
Symbolically, {A,C}  ℳ2x2 and {B,D}  ℳ2x1.
𝑅 𝑅 ||𝑅 . (2)
0
Considering (1) and (2), the information provided in Fig. 𝐀 ,𝐁
2b and references [17], [18], the operating point of the ECA ∙
0
can be determined to achieve a maximum voltage output (6)
excursion, vo. This operating point is defined in (3) and (4). 1 0 0 0
𝐂 ,𝐃 .
𝑉 ∙𝑉 (3) 0 1 0 0
𝐼 . (4)
Also, from (5) hie and Ro are defined in (7) and (8)
respectively. From (7), VT is the thermal voltage, which is 26
Equation (4) indicates that β represents the amplification
mV [17], [18].
factor in dc [17]. As the assumption is made that Q1 operates
within the linear region, the expressions ICQ = βIB, IEQ  ICQ, ℎ𝑖𝑏 , ℎ𝑖𝑒 𝛽 ∙ ℎ𝑖𝑏 (7)
and VBE = 0.7 V can verify.
𝑅 𝑅 ||𝑅 . (8)
B. ac operation
During ac operation of the ECA, the capacitors CC and CB From equation (6) the expression of k and Ri are defined
are considered short circuits and the source VCC is short in (9) and (10).
circuited. In addition, Q1 is replaced with its small-signal ∙
model [17], [18], assuming it operates in the linear region. 𝑘 , (9)
∙
The small-signal model of the ECA is illustrated in Fig. 3. 𝑅 𝑅 || ℎ𝑖𝑒 𝛽∙𝑅 . (10)

Given the linear nature of the model in (5), further

 Fig. 3. Small-signal model of the ECA.

conversion of (5) to the Laplace domain (s) is performed. The

s-domain converted model is defined in (11).

𝑠 0 𝐼 𝑠
∙ . (11)
𝑠 𝑉 𝑠 0
∙

Taking (11) and assuming that s = j, the ECA gain

expression in the frequency domain is derived and defined in
(12).
𝑅i
𝑘 ∙
𝐀𝐯 j ∙ 𝜔 ∙ 𝐿
1 . (12)
∙
𝑅o ∙𝐶
Fig. 4. Bode diagram of Av(j). At low frequencies the Av(j)  18 dB
and phase reach 180°. At high frequencies, the Av(j)  54 dB.
In order to verify the amplifying effect of this device, the
Bode diagram of the gain Av(j) is plotted and shown in Fig. - coordinates, i.e. the Clarke transformation [24] has been
4. The parameters of this ECA given in Table 1 are used for
achieved. The definition of the - variables is described in
this plot.
(13).
From equation (12) and the information that can be taken
from Fig. 4, it can be seen that the ECA is indeed an inverter. 𝑖 𝑖 𝑡
⎧
On the other hand, at low frequencies the ECA has an 𝑖 𝑖 𝑡 90°
⎪
approximately constant gain of 18 dB and then at high ⎪ 𝑣
frequencies the gain is increased to 54 dB. It can also be seen 𝑣 𝑡
. (13)
that the phase of Av(j) remains constant at 180° (which ⎨𝑣 𝑣 𝑡 90°
justifies the inverter nature of the ECA) and after 10 kHz the ⎪ 𝑣 𝑣 𝑡
phase increases until it reaches a maximum of 255° at 300 ⎪
kHz. Then, as the frequency increases, the phase decreases ⎩ 𝑣 𝑣 𝑡 90°
until it returns to 180°.
Replacing the variables defined in (13) in (5), the state-
IV. MODELING OF THE ECA IN DQ COORDINATES
space model in - coordinates is derived and given in (14).
The conversion of the model in (5) to a model in dq
coordinates presents an interesting challenge, since by 𝐱 𝐀 ∙𝐱 𝐁 ∙𝐮
. (14)
definition this conversion applies to three-phase systems 𝐲 𝐂 ∙𝐱 𝐃 ∙𝐮
[22].
This conversion cannot be applied directly because the The matrices of the model in (14) are defined in (15). On
model in (5) can be considered as a single-phase system. the other hand, the vectors are defined as x =
However, this drawback can be overcome by the method [ii,ii,vo,vo]T, u = [vi,vi]T, and y = x. Symbolically,
proposed in [24]. This is done by shifting each of the state u  {ℝ2} and {x, y}  {ℝ4} and {A,C}  ℳ4x4 and
and input variables by 90°. The model with the phase-shifted {B,D}  ℳ4x2.
and non-phase variables then correspond to the system in Given the model in - coordinates already derived and
TABLE I. PARAMETERS OF THE ECA
defined in (14), Park's transformation is applied directly to it
[22]. The dq coordinate model of the system is defined in
Parameters Values (16). It should be known that, for the dq transformation to
R1 3 k operate properly, the unconverted variables must be pure
R2 15 k sinusoidal [22].
RC 2 k
RE 500  Finally, the matrices of the model in (16) are defined in
RL 1 k (17). The vectors are defined as xdq = [iid,iiq,vod,voq]T, udq =
L 10 mH [vid,viq]T, and ydq = xdq. Symbolically, udq  {ℝ2} and {xdq,ydq}
C 100 pF
Vcc 24 V  {ℝ4} and {Adq,Cdq}  ℳ4x4 and {Bdq,Ddq}  ℳ4x2. Also,
V 0.7 V from the matrices,  = 2f, where f is the input-voltage
 0 0 0 ⎤ simulated using MATLAB-Simulink.
⎡
⎢ 0 0 0 ⎥ The ECA is supplied with a sinusoidal voltage with an
⎢ ⎥ amplitude of 5 mv peak-to-peak and with a frequency of 5
𝐀 ⎢ ,
0 0 ⎥ kHz. The parameters of the ECA are outlined in Table 1.
⎢ ∙ ⎥
⎢ ⎥ The simulation results related to (5) and (16) are shown
⎣ 0 0
∙ ⎦ in Fig. 6 and Fig. 7 respectively. In addition, Fig. 8 shows the
simulation results after the transformation of the system in dq
0⎤ variables to the ECA variables defined in (5).
⎡
⎢0 ⎥ The dynamics of the current ii is shown in Fig. 5(a) and
𝐁 ⎢ ⎥, the output voltage vC is shown in Fig. 5(b). It is clear from
⎢0 0⎥ Fig. 5(a) that the ECA dynamics are sinusoidal, since ii is
⎣0 0⎦ (15)
sinusoidal. On the other hand, Fig. 5(b) shows the gain and
phase shift that the ECA produces in the output voltage vC.
1 0 0 0
0 1 0 0 Regarding the conversion of the ECA model into dq
𝐂 , coordinates, the dynamics of the state variables in dq are
0 0 1 0
0 0 0 1 shown in Fig. 7. Figure 7(a) depicts the dynamics of iid and iiq
and Fig. 7(b) the dynamics of vCd and vCq. From Fig. 7 it can
0 0 0 0 be seen the static nature of the model in dq, where the
0 0 0 0 variables in dq coordinates reach their steady state values
𝐃 . very quickly (at about 2 s) after system start-up.
0 0 0 0
0 0 0 0 In Fig. 8, the proper operation of the model in dq
coordinates, namely the model presented in equation (16), is
𝐱 𝐀 ∙𝐱 𝐁 ∙𝐮 clearly illustrated. Evidently, for effective functioning
. (16)
𝐲 𝐂 ∙𝐱 𝐃 ∙𝐮 according to (16), the pairs of variables (ii, iiconv) and (vC,
vCconv) must exhibit identical dynamics. Notably, these
dynamics are entirely shared, thereby validating the correct
⎡ 𝜔 0 0 ⎤ derivation of the ECA's dq model.
⎢ 𝜔 0 𝜔 ⎥
⎢ ⎥ VI. CONCLUSION
𝐀 ⎢ ,
0 0 ⎥ The paper introduces and investigates a novel static model
⎢ ∙ ⎥
⎢ ⎥ in dq coordinates for a common-emitter amplifier configured
⎣ 0 𝜔 ∙ ⎦ as a common emitter. The study begins by examining the
standard model of the common-emitter amplifier in both the
⎡ 0⎤ dc and ac regimes. Essential equations are derived for both
⎢0 ⎥ regimes, showing the linear behavior of the common-emitter
𝐁 ⎢ ⎥, amplifier. In particular, the small-signal model is presented
(17)
⎢0 0⎥ specifically for the ac operation of the common-emitter
⎣0 0⎦
amplifier.
1 0 0 0 Using the small-signal model, this paper then proceeds to
0 1 0 0 obtain a static model in dq coordinates through the
𝐂 ,
0 0 1 0 application of Park's transformation. This transformation
0 0 0 1 allows a comprehensive representation of the behavior of the
0 0 0 0
common-emitter amplifier in steady state conditions.
0 0 0 0 Finally, the standard model and the dq-coordinate model
𝐃 .
0 0 0 0 are subjected to simulations for thorough evaluation and
0 0 0 0 comparison. The results of the simulations lead to the
conclusion that the dq-coordinate model of the common-
frequency. emitter amplifier is adequately derived, allowing static and
fast response operation after system start-up.
The linear model in dq coordinates of the ECA defined in
(16) is illustrated diagrammatically in Fig. 5. The ECA model Overall, the results of this article show the viability and
in dq coordinates is configured by two decoupled networks effectiveness of the proposed dq-coordinate model, providing
associated with the d and q channels. a valuable contribution to the field of common-emitter
amplifier modeling and analysis.
It can also be seen that each of the channels is configured
by a voltage circuit and a current circuit. The voltage circuits REFERENCES
contain current-controlled voltage sources and the current [1] B. Perez-Verdu, J. L. Huertas, and A. Rodriguez-Vazquez, “A new
circuits contain current- and voltage-controlled current nonlinear time-domain op-amp macromodel using threshold
sources. functions and digitally controlled network elements,” IEEE Journal
of Solid-State Circuits, vol. 23, no. 4, pp. 959–971, Aug. 1988.
V. SIMULATION RESULTS [2] D. F. Bowers, “A fast precision operational amplifier featuring two
separate control loops,” in 2014 IEEE Bipolar/BiCMOS Circuits and
The lossless models defined by equations (5) and (10) are
 (a)

(b)
Fig. 5. dq-model of the ECA. (a) d-channel. (b) q-channel.

(a) (b)
Fig. 6. Simulation results regarding the dynamic variables from (5). (a) Input current ii. (b) Input and output voltages vi and vC.

(a) (b)
Fig. 7. Simulation results regarding the dynamic variables from (16). (a) d- and q-channel of ii. (b) d- and q-channel of vC.

(a) (b)
Fig. 8. Simulation results show the comparison between the input current and the output voltage derived from the model in (10), i.e. (ii, vC) and those
obtained from the inverse conversion of the model in (16) to the model in (10), i.e. (iiconv, vCconv). (a) Comparison between ii, and iiconv. (b) Comparison
between vC, and vCconv.

Technology Meeting (BCTM), pp. 72–75, Sep. 2014. [4] S. E. Hosseini and H. G. Dehrizi, “A new BJT-transistor with ability
[3] M. Gilasgar, A. Barlabé, and L. Pradell, “High-Efficiency of controlling current gain,” in International Multi-Conference on
Reconfigurable Dual-Band Class-F Power Amplifier with Harmonic Systems, Signals & Devices, pp. 1–4, Mar. 2012.
Control Network Using MEMS,” IEEE Microwave and Wireless [5] K.-J. De Langen and J. H. Huijsing, “Compact low-voltage power-
Components Letters, vol. 30, no. 7, pp. 677–680, Jul. 2020. efficient operational amplifier cells for VLSI,” IEEE Journal of
 Solid-State Circuits, vol. 33, no. 10, pp. 1482–1496, Oct. 1998. effects and optimization in inductively degenerated common-emitter
[6] L. Kong, H. Liu, X. Zhu, C. C. Boon, C. Li, Z. Liu, and K. S. low-noise amplifiers,” IEEE Trans. Microwave Theory Techn.,
Yeo,“Design of a Wideband Variable-Gain Amplifier with Self- vol.51, no. 4, pp. 1220–1226, Apr. 2003.
Compensated Transistor for Accurate dB-Linear Characteristic in 65 [14] E. Ozeren, C. Cahskan, M. Davulcu, H. Kayahan, and Y. Gurbuz,
nm CMOS Technology,” IEEE Transactions on Circuits and “4-Bit SiGe phase shifter using distributed active switches and
Systems I: Regular Papers, vol. 67, no. 12, pp. 4187–4198, Dec. variable gain amplifier for X-band phased array applications,” in
2020. 2014 9th European Microwave Integrated Circuit Conference, pp.
[7] I. M. Pandiev and E. Ch. Stoimenov, “Development of Single- 257–260, Oct. 2014.
Transistor Amplifier Modular Laboratory Kits for Electronic [15] J.-Y. Um, “A Compact Variable Gain Amplifier with Continuous
Engineering Education,” in 2020 XI National Conference with Time-Gain Compensation Using Systematic Predistorted Gain
International Participation (ELECTRONICA), pp. 1–4, Jul. 2020. Control,” IEEE Trans. Circuits Syst. II, vol. 69, no. 2, pp. 274–278,
[8] F. A. Farag and Y. A. Khalaf, “Digitally-Controlled Variable-Gain- Feb. 2022.
Amplifier based on current conveyor with opamp and inverters [16] R. W. Erickson and D. Maksimovic, Fundamentals of Power
only,” in 2010 International Conference on Microelectronics, pp. Electronics. Norwell, Mass, 2001.
224–227, Dec. 2010. [17] C. J. Savant, M. S. Roden, and G. L. Carpenter, Electronic Design:
[9] D. G. Makarov, V. V. Kryzhanovskyi, V. G. Krizhanovski, and A. Circuits and Systems, Subsequent edition. Redwood City, Calif:
V. Grebennikov, “Experimental investigation of class E power Benjamin-Cummings Pub Co, 1990.
amplifier with shunt filter for reduced duty ratio,” in 2018 [18] A. S. Sedra and K. C. Smith, Microelectronic Circuits. Oxford
International Conference on Information and Telecommunication University Press, 2004.
Technologies and Radio Electronics (UkrMiCo), pp. 1–4, Sep. 2018. [19] O. Katsuhiko, Modern Control Engineering. Boston, 2009.
[10] Y. Hou, W. Yu, Q. Yu, B. Wang, Y. Sun, W. Cheng, and M. Zhou, [20] The Control Handbook (three volume set). CRC Press, 2018.
“A 56–161 GHz Common-Emitter Amplifier with 16.5 dB Gain [21] B. C. Kuo, Automatic control systems, 6th edition. Englewood
Based on InP DHBT Process,” Electronics, vol. 10, no. 14, p. 1654, Cliffs, N.J: Prentice Hall, 1991.
Jul. 2021. [22] C. J. O’Rourke, M. M. Qasim, M. R. Overlin, and J. L. Kirtley Jr, “A
[11] H. Nam, D.-A. Nguyen, Y. Kim, and C. Seo, “Design of 6 GHz Geometric Interpretation of Reference Frames and Transformations:
Variable-Gain Low-Noise Amplifier Using Adaptive Bias Circuit for dq0, Clarke, and Park,” Colm O’Rourke, Dec. 2019.
Radar Receiver Front End,” Electronics, vol. 12, no. 9, p. 2036, Apr. [23] C. J. Savant, M. S. Roden, and G. L. Carpenter, Electronic Design:
2023. Circuits and Systems, Subsequent edition. Redwood City, Calif:
[12] D. L. Rode, “Output Resistance of the Common–Emitter Amplifier,” Benjamin-Cummings Pub Co, 1990.
IEEE Trans. Electron Devices, vol. 52, no. 9, pp. 2004–2008, Sep. [24] B. C. Trento, “Modeling and Control of Single Phase Grid-Tie
2005. Converters,” Masters Theses, Aug. 2012.
[13] P. Sivonen, S. Kangasmaa, and A. Parssinen, “Analysis of packaging