RD/RDC Series Converters Applications Manual MN-19220XX-001
RD/RDC Series Converters Applications Manual MN-19220XX-001
RD/RDC Series Converters Applications Manual MN-19220XX-001
Series Converters
Applications Manual
MN-19220XX-001
(RDC-19220/2, RDC-19220/2S, RDC-19224,
RD-19230, RD-19240 Series)
Please visit our Web site at www.ddc-web.com for the latest information
U
®
The Introduction gives a brief overview of different operating modes. Common applications are
shown and discussed.
The Basics and Series Overview sections detail the general operation of the converter including
required power supplies, input and reference signals, and timing.
A System Design section offers a design checklist and discusses some common parameters to
consider when designing with the RDC-19220/2/4, RDC-19220/2S, RD-19230, and RD-19240
series converters.
The Enhancements section contains discussions such as performance versus design decisions,
and ‘changing resolution on-the-fly.’ Sample calculations are provided to aid in modeling the
converter in specific applications.
Finally, the Mechanical Specifications section consists of outline drawings and pin-out tables for
the converters, as well as the relevant thin-film resistor networks and transformers
recommended.
Text Usage
BOLD–indicates important information and table, figure, and chapter references.
The Warning icon alerts the reader to hazards that will cause
damage to the product and possible injury to the user.
Trademarks
All trademarks are the property of their respective owners.
RESOLVER IN A SYSTEM
REF
REFERENCE
LEVEL
EQUIPMENT CONDITIONER
WITH RESOLVER 'N' BITS
ROTATING
SHAFT
θ
S1
INPUT
S2 R/D DIGITAL
LEVEL
CONVERTER PROCESSING
CONDITIONER S3
S4
The function of the RD/RDC series is to convert the amplitude signal from a
Synchro or Resolver (or LVDT/RVDT; See the Input Signal Configurations
discussion on page 9) to a digital angle and velocity. For static or slow
moving input changes the converter can be designed into the application
effortlessly. However, where rates of change at the input are high and the
device is part of a control loop, setting of dynamic parameters and bandwidth
should be assessed (see the Enhancements Section).
Please note:
external
components are
necessary for
the converter to
operate.
Cbw Rb
V V Cbw/10
E E Rb
L L
S S Cbw
RH RL BIT
J J
1 2 Cbw/10
VEL2 VEL1
SYNTHESIZED SHIFT
SIN REFERENCE
-S -
+S +
CONTROL
Please note: COS TRANSFORMER
GAIN DEMODULATOR - VEL
the external -C - +
+C + D1 D0
components are RV
D1
necessary for VDDP
UP/DOWN
HYSTERESIS
PCAP -5 V VCO
the converter to NCAP INVERTER
D0 COUNTER
&
-VCO
AGND
VDD INTERNAL
DATA R SET
GND LATCH ENCODER
EMULATION
VSS
The most critical portion of the circuit design is the external components. If
the external components are not connected the device will not function.
The RD/RDC component selection software is available to aid in selecting
these components. This is further discussed in the Mandatory External
Components section.
Power Supplies
The RDC-19220 and RDC-19220S converters require both +5 VDC and -5
VDC supplies. The RDC-19222/4, RDC-19222S, RD-19230, and the RD-
19240 converters can operate from a single +5 volt supply. This is
accomplished by using the four additional pins to connect two external
capacitors to develop the -5 volt needed for operation.
Configurations
Consult specific converter data sheet for the following configurations:
• direct/synchro/resolver
• DC
• LVDT position information bit weight
• encoder emulation
• velocity
Power Supplies
VDD
VDD
+ VDDP
47 μF .01μF
PCAP
+ RD-19230
10 μF/10V
NCAP or
VSSP RD-19240
VSS *
VSS
47 μF/10V
.01μF +
GND
AGND
Figure 5. Power Supply Set-Up for RD-19230 or RD-19240 (+5VDC P/S Operation)
To determine the charge pump rate, use an oscilloscope to measure the rate
of the PCAP.
Connect the 2 VDD pins to +5 VDC. Connect the 2 VSS pins to –5 VDC.
VDDP, VSSP, PCAP & NCAP are no connect.
-15 79LO5 -5
-15 -5
-5V 5.1V
10.2V ZENER
ZENER
-12 -5
6.8V -15V
ZENER
Figure 6. Typical -5 Volt Circuits
Signal Connections
Example:
2. R2 is relative to R1
3. R2 required = 12K
Next relative value to calculate is R4 (Note that the data sheet listing is in alpha
order, not working order)
1. R4 (70.8K) is relative to R1
3. R4 required = 70.8K
EXTERNAL
REF
LO HI
R1 R2 RB CBW
VELOCITY
R3 R4 CBW/10 Rv
5 4 7 8
Notes:
INH EL
1) Resistors selected to limit Vref peak to between 1 V and 5 V. 1000pf 1000pf EM
2) External reference LO is grounded, then R3 and R4 are not .01μf .01μf
needed, and -R is connected to GND. A B
{
3) Refer to the RDC-19220 Data Sheet for the digital output.
4) See the thin-film resistor network DDC-55688-1 data sheet. +5v -5v RESOLUTION
5) Same for RDC-19224 but pin-outs are different. CONTROL
Direct input is not recommended for high noise environments or where the
resolver transducer is a long distance from the converter. In these cases the
Common Mode Rejection (CMR) of a differential input will reduce the effects
of noise.
EXTERNAL
REF
LO HI
R1 R2 RB CBW
VELOCITY
R3 R4 CBW/10 Rv
30kΩ 30kΩ
6 7 18 19 9 10
10kΩ -R +R RSR C VEL
S3 14
+S 8
16 -VCO
-S
15
SIN 29 to 44
12
S1 COS DIGITAL
13
RDC-19222 OUTPUT
-C 16
10k Ω 11
+C 26
CB
S2 27 BIT
S4 A GND
21
{
4) See the thin-film resistor network DDC-55688-1 data sheet.
RESOLUTION
CONTROL
EXTERNAL
REF
LO HI
R1 R2
R3 R4
See Note 3. RL RH
+S
S3 -S
SIN
Note: The external BW components
S1 COS as shown in Figure 3
-C are necessary for the R/D to
function.
See Note 3.
+C
S2
S4 A GND RD-19230/RD-19240
RESOLVER GND
Notes:
1) Resistors selected to limit Vref peak to between 1.5 V and 5 V.
2) External reference LO is grounded, then R3 and R4 are not
needed, and -R is connected to GND.
3) 10k ohms, 1% series current limit resistors are recommended for SIN, COS, and Ref inputs when no voltage adjustments are
needed.
When the resolver output is higher than 2 Vrms the signals need to be scaled
down using the following equations, and as shown in Figure 11 below.
R2 2
=
R1 + R2 VRESOLVER(rms)
R1 + R2 ratio errors will result in 2 cycle angular errors. For example, a 0.1%
ratio error equates to 0.029 peak error.
Figure 11. X-Volt resolver, Example using Direct Input to RDC-19220 (40-pin DDIP)
Differential Resolver
A differential input signal will provide better stability than a single-ended
input. The circuit in Figure 10 applies to resolvers that have a signal output of
2 Vrms or higher. For devices with a signal output lower than 2 Vrms, a pre-
amplifier circuit should be used to provide gain, rather than using resistors to
increase the internal gain (please refer to information regarding Inductosyns).
Resistors should not be used to increase the gain for an output signal lower
than 2 Vrms for two reasons: 1. Noise on the signal lines will be gained up
Preferred Mode and become part of the signal, causing accuracy errors, and 2. Setting the
of Operation gain larger than 1/3 will introduce errors due to internal offsets.
The resistors used to scale down the resolver signals need to be precision
resistors matched to each other. DDC offers thin-film resistor networks that
are matched to within 0.02%.
The DDC-49530 and DDC-57470 are set-up to scale 11.8 Vrms down to 2
Vrms. The DDC-49590 and DDC-57471 are set-up to scale 90 Vrms down to
2 Vrms single ended. The DDC-55688-1 and DDC-73089 are 2 Vrms to 2
Vrms. Refer to the Thin-film Resistor Network data sheet for specifications
(available at www.ddc-web.com).
Please Note:
The External -REF
Components in SIGNAL AC
Note: Input options affect DC offset gains and therefore affect carrier
frequency ripple and jitter. Offset gains associated with differential mode
(offset gain for differential configuration = 1+Rf/Ri) and direct mode (offset
gain for direct configuration = 1) show differential mode will always be higher.
Higher DC offsets cause higher carrier frequency ripple due to the
demodulation process. This carrier frequency ripple rides on top of the DC
error signal, causing jitter. A higher carrier frequency versus bandwidth ratio
will help to decrease ripple and jitter associated with offsets.
Synchro Input
A synchro produces a three-phase (S1, S2, and S3) Y configuration ac output
from the reference excitation. The three-wire configuration of a synchro, like a
differential resolver input, cannot interface directly to the RD/RDC series.
Figure 13 uses a resistor network to scale a synchro input on a RD-19220
example.
-REF
SIGNAL AC
+REF
R1 R2
R3 R4
-REF +REF
3
SIN -VSUM
Rf
1 Ri 2
RDC 19220 Cbw
-S
6 Ri SERIES
5 +S Cbw/10
+R Rb
-R S1
S3 Rf
4
VEL VELOCITY
A GND
SYNCHRO S2
14 COS
Ri Rv
16
Rf
7 Ri 8 3 15
15
-C
Ri/2 -VCO
9 10 +C
Rf
3
11
In comparing Figure 14, detail A & B, you can see that the LVDT is scaled so
that the lower output voltage limit is scaled to equal zero and the upper
output voltage limit is scaled to equal 2 V.
The upper and lower output voltage levels could apply to a limited linear
range on the LVDT. With the output of the LVDT rescaled, the RD/RDC
series will interpret the LVDT with bits 1 and 2 for over range conditions,
similar to the first quadrant (0° to 90°) in the resolver mode (see Figure 14,
detail C).
Please note:
the external
components are
necessary for
the converter
to operate.
1 1 a
b= = SIN = − 1V + (VA − VB)
VANULL VBNULL 2
2 a
a= COS = − 1V − (VA − VB)
(VA − VB)MAX 2
Notes:
1. R’ > 10k Ohms
2. Considerations for the value of R is LVDT loading.
3. RMS values are given.
1 1 1
b= = b= = 0.204
VA NULL VBNULL 4 .9
2 2
a= a= = 0. 8
(VA − VB )MAX (6.1 − 3.6)
At –FS
a
sin = −1V + (VA − VB )
2
0.8
sin = −1V + (6.1 − 3.6)
2
sin = −1V + 1
sin = 0V
a
cos = −1V − (VA − VB )
2
0.8
cos = −1V − (6.1 − 3.6)
2
cos = −1V − 1
cos = −2V (see note below)
Note: Per Figure 15, the +REF signal is set to –2V, so this calculated cos
voltage in the above example looks like a positive 2V.
+S -S SIN
AGND
-REF
+REF
+C -C COS
C1
SIN
C1 and C2 aR
should be chosen
2 WIRE LVDT -S
to minimize R
-
phase-shift R
REF IN +S
between +S R
and +C inputs. + FS = 2 V
C2 aR
COS
bR R
R
2R
R -C
-
2R
+C
R
+ 2V R
bR
+REF
-REF
aR = 2V
VOUTMAX
R
VINMAX bR = 2V
R
R = 10 kΩ
aR
(0.5) = 2 therefore, aR = 40kΩ
10kΩ
bR
(3.5) = 2 therefore, bR = 5.7kΩ
10kΩ
In LVDT mode:
bit 16 is the LSB for 14-bit resolution
bit 14 is the LSB for 12-bit resolution
bit 12 is the LSB for 10-bit resolution
bit 10 is the LSB for 8-bit resolution
Refer to Table 8 for the digital output codes for the RD/RDC Series Converter
in LVDT mode.
Note 1: If using the RD/RDC series to replace the DTC-19300, bit 3 should be inverted.
Note 3: Accuracy specifications below for (2-wire) LVDT mode, null to + full scale travel (45° degrees)
4 Min part = 0.15% + 1 LSB of full scale “resolution set”.
2 Min part = 0.07% + 1 LSB of full scale “resolution set”.
1 Min part = 0.035% + 1 LSB of full scale “resolution set”.
Accuracy specifications below for (3-wire) LVDT mode, null to + full scale travel (90° degrees)
4 Min part = 0.07% + 1 LSB of full scale “resolution set”.
2 Min part = 0.35% + 1 LSB of full scale “resolution set”.
1 Min part = 0.017% + 1 LSB of full scale “resolution set”.
Note that these accuracy specifications are for the converter and do not consider front end and
external resistance tolerances.
Note 4: Data output is Binary Coded in LVDT mode. The most negative stroke of the LVDT is represented
by ALL ZERO’s and the most positive stroke of the LVDT is represented by all ONES. The most
significant 2 bits (2 MSBs) may be used as over range indicators. Positive over-range is indicated by code
“01” and negative over-range is indicated by code “11”.
Inductosyn®
A linear or rotary Inductosyn® consists of a slider and a scale. As the slider
moves over the scale there is a low voltage electrical output proportional to
the distance moved. Inductosyns usually are excited by a 5 kHz to 20 kHz
frequency. The OSC-15801 is an oscillator that was designed with a high
frequency output and a 90° phase shift, which is needed for Inductosyn®-to-
digital conversion (see Figure 19).
INDUCTOSYN
INDUCTOSYN SLIDER
SCALE
DUAL
MATCHED
INDUCTOSYN RESOLVER
QUADRATURE 2 Vrms
TO
PREAMPS
POWER DIGITAL
OSCILLATOR CONVERTER DIGITAL
OSC-15801 OUTPUT
PA OUT
RH RL
Note: Figure 19 will convert each pitch to 360° of digital data. A means to track
See DDC’s counts will be needed for multiple pitches. Using an RD-19230 and A quad B
Synchro/Resolver zero index pulse for counting pitches is one possible solution.
Conversion
Handbook “Using
an R/D with an Note that inductosyns typically have a (+ or -) 90-degree phase shift from
inductosyn” for input to output. This requires an oscillator with a 90-degree phase-shifted
further information.
second output to be connected to the RD converter reference so that the RD
input and reference will be in phase. DDC has two oscillators to cover +90 or
–90 degree phase shifts needed. See the OSC-15801 for -90PS and the
OSC-15802 for +90PS.
C1 RA1
RA2 RA2
RA1 +15V
_ _
R3 RA1
TBD
U1 U1
RA1
+ +
GND
RA2
CHASSIS GROUND
CIRCUIT GROUND
RA2 C2 RA1 RA2
+
RA1
_ _
R4 RA1
U1 TBD U1
RA1
+ +
RA2
DC Inputs
As noted in the specifications tables of the appropriate data sheets, the RDC-
19220/2/4, RD-19230, and RD-19240 will accept DC inputs.
4. Set the COS and SIN inputs such that the maximum signal is equal to 1.8
VDC. For example, at 90° the SIN and 0° for COS the input should equal to
1.8 VDC. This will keep the BW hysteresis consistent with AC operation.
5. Input offsets will affect accuracy. Verify the COS and SIN inputs do not
have DC offsets. If offsets are present, a differential op amp configuration can
be used to minimize differential offset problems.
7. Choose the bandwidth value of the converter based on the rate of change
of the systems input amplitude variation. It should be large enough to
minimize it’s effect on the system dynamics. Note that if the bandwidth is too
high the system will be more susceptible to noise.
Reference Input
The reference that is used to excite the Synchro/Resolver is used as an input
to the RD/RDC series. The conversion technique utilizes the reference for
phase information in the demodulator section of the converter. This ensures
quadrature rejection of the Synchro/Resolver signals. For example, the
resistors R1 and R2 should be selected as follows:
X R1+ R2
=
2Vrms R2
Note: For circuits where no voltage adjustment is needed add a 10k ohm
series resistor for turn on current surge protection.
-REF
SIGNAL AC
+REF
R1 R2
R3 R4
-REF +REF
3
SIN -VSUM
Rf
1 Ri 2
RDC 19220 Cbw
-S
6 Ri SERIES
5 +S Cbw/10
+R Rb
-R S1
S3 Rf
4
VEL VELOCITY
A GND
SYNCHRO S2
14 COS
Ri Rv
16
Rf
7 Ri 8 3 15
15
-C
Ri/2 -VCO
9 10 +C
Rf
3
11
Synthesized Reference
The synthesized reference section of the RDC-19220/2S, RD-19230, and
RD-19240 eliminates errors caused by quadrature voltage which is due to a
phase shift between the reference and the signal lines. Quadrature voltages
in a resolver or synchro are by definition the resulting 90° fundamental signal
in the nulled out error voltage(e) in the converter. Due to the inductive nature
of synchros and resolvers, their signals typically lead the reference signal
(RH and RL) by about 6°.
Due to the inductive nature of resolvers, the output signals typically lead the
reference by 6°, and a 6° phase shift will cause problems for a 1 arc minute
accuracy converters. A synthesized reference will always be exactly in phase
with the signal input, therefore eliminating accuracy error caused by phase
shift up to 45°.
Let:
F. S. Signal = 11.8 V
α = 6°
Speed Voltage = 0V
⎛ ( Rotational Speed ) ⎞
(SpeedVoltage) = ⎜⎜ ⎟⎟ • (F .S .Signal )
⎝ (Carrier Frequency) ⎠
where:
Fundamental
Quadrature
Third Harmonic
OUTPUT
TIME
Second
Harmonic
LAG
R
For a ref signal
that leads the
inputs {+ REF
- REF
C
+ REF
- REF
LEAD
C
For a ref signal
that lags the
inputs {
+ REF
- REF
R
+ REF
- REF
Transformer Isolation
System requirements often include electrical isolation from the
synchro/resolver and reference signals. DDC has transformers available for
use with the RD/RDC Series Converters. These transformers reduce the
voltage to the required 2 Vrms for a direct connection. The reference
transformers provide isolation of the reference signal.
Table 9. Transformers
ANGLE
FREQ IN OUT LENGTH WIDTH HEIGHT Available
P/N TYPE ACCURACY
(HZ)* (VRMS)* (VRMS)** (IN) (IN) (IN) FROM
***
52034 S-R 400 11.8 2 1 0.81 0.61 0.3 Beta
52035 S-R 400 90 2 1 0.81 0.61 0.3 Beta
52036 R-R 400 11.8 2 1 0.81 0.61 0.3 Beta
52037 R-R 400 26 2 1 0.81 0.61 0.3 Beta
52038 R-R 400 90 2 1 0.81 0.61 0.3 Beta
B-426 Reference 400 115 3.4 N/A 0.81 0.61 0.32 Beta
52039-X Synchro 60 90 2 1 1.1 1.14 .42 DDC
24133-X Reference 60 115 3/6**** N/A 1.125 1.125 .42 DDC
The following transformers can be ordered directly from DDC, Tel (631) 567-5600:
P/N 52039-X, 24133-X
The following transformers can be ordered directly from Beta Transformer Technology Corporation
(BTTC), Tel (631) 244-7393:
P/N 52034, 52035, 52036, 52037, 52038, and B-426.
1 6
RH +R
EXTERNAL B-426
5 10
REFERENCE RL -R
LO HI
OR
RB CBW
RH +R 60 hz
24133
RL -R CBW/10 RV
+15 V -VS
+15V -15V
-S SIN -R +R -VSUM VEL
S1 -VCO
RH +S
RL 1 10
S3 TIA DIGITAL
3 6 OUTPUT
+C RDC-19220
S4 16
11 20 -C
TIB COS CB
S2
15 16 BIT
AGND
52036(11.8V) INH
OR EM
52037(26V) GND Rs Rc
OR EL
52038(90V)
A B +5V -5V 30K Ω 30K Ω
OR
}
SYNCHRO INPUT
RESOLUTION
CONTROL
S1 +S
RL RH 1 10
S3
3 TIA
5 6
+C
11 20
S2 TIB
15 AGND
16
GND
52034(11.8V)
OR
52035(90V)
2 - The maximum tracking rate per resolution and the maximum carrier
frequency per resolution are as follows:
Table 10. Maximum Tracking Rate And Carrier Frequency Versus Resolution
Resolution Resolution
RC RS
PART NUMBER Or Or 10 12 14 16 10 12 14 16
RSET RCLK MAX TRACKING RATE *** CARRIER
*** IN RPS FREQUENCY
MAX IN KHZ
RDC-19220/2S,
30 kΩ*
RD-19230,
or 30 kΩ 1152 288 72 18** 10 10 5 5**
RD-19240
Open
SERIES
RD-19220/2 30 kΩ*
SERIES or 30 kΩ 1152 288 72 18 20 11 7 5
Open
* the use of a high quality thin-film resistor will provide better temperature
stability than leaving open.
*** Extended tracking rates and carrier frequencies per resolution are
available. Use the component calculation software for extended dynamic
needs.
4- The TR to BW ratio must be less than or equal to the ratios listed in Table
11. If the TR & BW ratio is greater than the listed ratio, a spin-around
condition may occur. Spin-around is a condition when the converters digital
output continually counts, and never settles on an angle.
When using the
charge pump
Table 11. Tracking / BW Relationship
feature the
maximum negative RPS(Max)/BW Resolution
voltage on velocity 1 10
is –3.5V.
0.45 (*0.50) 12
0.25 14
0.125 16
* = RD-19230/RD-19240 ONLY
3.2 • FS ( Hz ) * * • 10 8
CBW ( pF ) = (2)
RV • ( FBW ) 2
where FBW = bandwidth frequency
0.9
R B (kΩ ) = (3)
CBW • FBW
where RB = bandwidth component
Resolution = 16-bits,
1st check: Ensure that the resolution will be acceptable for the carrier
frequency. Use the maximum carrier frequency lookup table. The example
below shows that the chosen resolution of 16-Bits will be acceptable for a
5 kHz carrier frequency.
2nd check: Ensure that the resolution will not exceed the rps maximum
tracking rate. Use the maximum tracking rate lookup table. The example
below shows that the chosen resolution of 16-Bits will be acceptable for
15rps.
3rd check: Now for a 5 kHz frequency being used at 16-bit resolution, we
need to select the bandwidth. Per the specific component data sheets
dynamic characteristics table, locate the maximum BW selectable for 16-bit
mode. For this example, 300 Hz is the maximum. Per BW component general
guidelines, the BW to Frequency rule is no greater than ¼. In this case the
dynamic characteristics table will take precedence since 300 is much less
than ¼ of 5 kHz. So, for this example we will pick an arbitrary BW of 130 Hz
for noise rejection.
Dynamic Characteristics
Resolution Bits 10 12 14 16
Tracking Rate (min) rps 1152 288 72 18
Bandwidth (Closed Loop) Hz 1200 1200 600 300
(at maximum bandwidth)
After verifying that the four dynamic parameters will operate in your system,
use these parameters in the component selection software to calculate
resistor and capacitor values.
Note: If resistor values are calculated as high Meg ohm resistors, recheck
entered dynamic parameters. If a parameter is entered much lower than the
maximum allowed value, then relaxing the parameter will impact resistor
values. As an example, if the maximum rps allowed entry is 15 and 5 was
entered, raise the rps entry value to 15 and recalculate components.
At 10-bits of resolution, the tracking rate can never exceed the bandwidth, but
may be equal to the bandwidth, as there is a 1:1 ratio. Maximum tracking
rates are maximum converter tracking rates and not necessarily maximum
Resolver/Synchro tracking rates. Maximum tracking rates of the converter are
absolute maximum tracking rates achievable and not intended for standard
operation. If changing resolution, divide the 10 bit resolution maximum
tracking rate by 4 for every resolution increase desired (Example: For 12 bit
resolution the maximum tracking rate is 288 RPS. For 14 bit resolution the
maximum tracking rate is 1152/4/4 = 72 RPS, For 16 bit resolution the
maximum tracking rate is 1152/4/4/4 = 18 RPS).
Notes:
The converter can track four times faster for each step down in resolution
(i.e., a step from 16 bits to 14 bits). The velocity output will be scaled down by
a factor of four with each step down in resolution. For example, if the velocity
output is scaled such that 4 Volts = 10 RPS in 16 bit resolution, then the
same converter will output 1 Volt for 10 RPS in 14 bit resolution. To avoid
glitches in the velocity output, the second set of bandwidth components can
be pre-charged to the expected voltage, and switched in using the SHIFT
input at the same time the resolution is changed. This will allow for a smooth
velocity transition, resulting in reduced errors and minimal settling time after
the change.
2) ERROR: this is the analog representation of the error between the input
and the output of the converter
When this system uses the switch resolution on the fly implementation,
the velocity signal immediately assumes the pre-charged level of the
second set of components, resulting in small errors and reduced settling
times. Notice that the BIT output, in Figure 26, does not indicate a fault
condition.
When this system type does not use the switch resolution on the fly
implementation, large errors and increased settling times result. The
errors exceed 100 LSBs causing the BIT to flag for a fault condition.
GND
D1
RD-19230/RD-19240
Control D0
SHIFT
UP/DN
Figure 25. Input Wiring – Switch On The Fly Between 12 and 14 Bit Resolution
Dual Bandwidths
With the second set of BW component pins, the user can set two bandwidths
for the converter and choose between them. To use two bandwidths, proceed
as follows:
2) Choose the two bandwidths following the guidelines in the General Setup
Considerations; the RV resistor must be the same value for both bandwidths.
3) Use the SHIFT pin to choose between bandwidths. A logic 1 selects the
VEL1 components and a logic 0 selects the VEL2 components.
0V
VEL
-5V
ERROR 0˚
5V
D0
0V
5V
BIT
0V
0V
VEL
-5V
ERROR 0˚
5V
D0
0V
5V
BIT
0V
Note: Depending on the bandwidth, step error may be greater. Less velocity
or movement will lessen the error glitch shown in the graph. The graph shows
a worst-case condition based upon one bandwidth and tracking rate setup.
Worst case is when the overshoot on the velocity hits the saturation point as
per the graph.
VELOCITY
OUT
ERROR PROCESSOR
VCO
CT A1 S + 1
RESOLVER + A2 DIGITAL
e B
INPUT S POSITION
(θ) S S +1
- 10B OUT (φ)
H=1
(CRITICALLY DAMPED)
ct
GAIN = 4
2A
OPEN LOOP ω (rad/sec)
B A -6
db 10B
(B = A/2) /oc
t
GAIN = 0.4
f BW = BW (Hz) = 2A
π
2A 2 2A
CLOSED LOOP ω (rad/sec)
⎛S ⎞
A 2 ⎜ + 1⎟
Open loop transfer function= ⎝B ⎠
⎛ S ⎞
S2 ⎜⎜ + 1⎟⎟
⎝ 10 B ⎠
2
Where A is the gain coefficient and A = A1A2
and B is the frequency of lead compensation
The components of gain coefficient are error gradient, integrator gain, and
VCO gain. These can be broken down as follows (see Figure 2, Figure 3,
and Figure 4):
- error gradient = 0.011 volts per LSB (ct + error amp + demod with
2 Vrms input)
CSFS
- integrator gain = volts per second per volt
1.1CBW
1
- VCO gain = LSBs per second per volt
1.25RVCVCO
where: CS = 10 pF
FS = 67 kHz when Rs = 30 kΩ (RDC-19220/2/4 Series)
FS = 100 kHz when Rs = 20 kΩ ( RDC-19220/2/4 Series)
FS = 134 kHz when Rs = 15 kΩ ( RDC-19220/2/4 Series)
CVCO = 50 pF (RDC-19220/2 Series)
RV, RB, and CBW are selected by the user to set velocity scaling and
bandwidth.
Once the external component values are known, the A1 and A2 can be
calculated as follows:
1
A1 = 0.011•
Ri • CBW • 1.1
1
A2 =
RV • CCVO • 1.25
1
Ri =
CS • FS
where:
CS = 100pF
A 2 = A1A 2
A
and B =
2
Example:
Resolution: 16 Bits
Bandwidth: 300 Hz
Rv = 54 kΩ
CBW = 4390 pF
CBW/10 = 439 pF
RB = 683 kΩ
RC = 30 kΩ
RS = 30 kΩ
1
Ri =
(100pF)(67kHz)
1
A1 = 0.011•
(1.5 • 10 ) • (4390 • 10−12 ) • (1.1)
6
A1 = 1.52
1
A2 =
(54 • 10 ) • (50 • 10−12 ) • (1.25)
3
A 2 = 296,000
To interpret the digital output for a LVDT input refer to Table 7 & Table 8.
Binary-To-Decimal Conversion
Conversion from binary to decimal angular data is simply a matter of adding
all the decimal angular weights for all of the binary bits which are set to logic
1.
For example, if the converter is set for 16-bit resolution and the binary
measurement is:
+(0 • All Other Bit Weights) All remaining Bits are logic low
Total = 151.875°
If the 16-bit binary value is treated as a decimal integer (in this example, it
has the hex value of 6C00H and the decimal value of 27648), the decimal
angle, in degrees, can be calculated by the following formula:
65536
R2 U2D
2k 74AC86
13
C2 11
RDC-19220 220 pF A
12
U2A
74AC86 R1 U2B
LSB +1 2 2k 74AC86
3 4
1 C1 6 B
LSB 220 pF 5
R3 U2C
2k 74AC86
9
CB/NRP
C3 8
120 pF 10 NRP
EL
D1
1N4148
-5 V
NOTE: CMOS LOGIC IS RECOMMENDED. TTL AND TTL
COMPATIBLE LOGIC WILL SKEW THE DELAYS.
Tying EL to -5v changes the function of the converter busy (CB) output of the
converter to a Zero Index (ZI) or North Reference Pole (NRP) and the LSBs
are enabled. The Inhibit INH input can not be used to freeze the data on the
output lines because the A quad B (or incremental encoder signals) will be
lost.
360°
t=
(Counts) (Velocity{ °/s})
360°
T=
Velocity (°/s)
A(LSB+1)
2t
B(X-OR
LSB & LSB+1)
ZI(CB) t
0
0 T 359.95
0
Note: The encoder resolution must be less than or equal to the resolution of
the parallel data outputs. Refer to Figure 31.
The timing of the A, B and ZIP (or North Reference Pole [NRP]) output is
dependent on the rate of change of the synchro/resolver position (rps or
degrees per second) and the encoder resolution latched into the RD-19230
and RD-19240 (refer to Figure 32). The calculations for the timing are:
t = 1 / ( 2n* Velocity(RPS))
T = 1 / ( Velocity(RPS))
Note: U indicates the direction of the counter. It stands for “UP”. If the RD-
19230/RD-19240 is at a static angle awaiting a new angle θ, U indicates the
direction the counter was going to get to the current angle φ. As the error is
approaching zero, the internal analog circuitry voltage may over shoot before
settling - which would then indicate an incorrect direction. Because of this
over shoot, the U output should not be relied on after settling to a static state.
Only during active resolver movement will the U output state be reliable. U is
a logic 1 when going in the positive direction (increasing angle). It is a logic 0
when going in the negative direction (decreasing angle). This is the same as
it is in the RDC-19220.
ZIP _ EN chooses between the CB and Zero Index pulse outputs and is
independent of encoder emulation mode. A logic 1 enables the CB pulse, a
logic 0 enables the Zero Index pulse (see Table 19).
Note: When the RD-19230 is set for 16-bit mode, the LSB is bit 16. When the
RD-19230 is set for 14-bit mode, the LSB is bit 14 and bits 15 and 16 are set
to logic “0”. This is also the same for the RD-19240 except it will only operate
in 10-, 12-, 14-bit modes.
RD-19230
1 MSB
2
3
4
5
6
7
8
9
10
11
12
13
14
15
BIT 16 LSB
1 1 1 1
0 2 4 6
A
A (LSB+1)
ZIP (NRP)
t
0 T 359.95
DATA
D0/D1 VALID
50 nsec
A QUAD B
Velocity Output
A benefit of using a type II tracking Synchro/Resolver-to-Digital converter is
the analog velocity output. The velocity on the RD/RDC series is scaled by
the mandatory external component RV. If the application is using a +5 V and
a -5 V power supply the maximum tracking rate will be equal to ±4 V (±
depending on the direction of rotation). If the converter is used in +5 V only
mode, the maximum tracking rate should be scaled to ±3.5 V (see Figure 34).
The velocity characteristics have been broken down into four specifications:
1. scale factor error
2. reversal error
3. linearity
4. zero offset
Scale factor error is the gain error for the positive rotation and the gain error
for the negative rotation averaged together. Gain error for either positive or
negative rotation is the average of the percent error between the actual
measured velocity reading and the theoretical velocity output at certain
speeds in that direction. For example, if the percent error at 10 RPS is 5%
and the percent error at 5 RPS is 4% the gain error is 4.5%.
Reversal error, or bipolarity, is half the difference between the positive gain
error and the negative gain error.
Zero offset is the measured velocity at a static position (speed of the resolver
= 0 RPS).
When measuring velocity, the return should be with respect to AGND. The
AGND provides a reference for analog signals. (No current drop)
Velocity Specifications
Velocity
Voltage
Actual Avg
Max Linearity
Actual
Theoretical
Scale Factor
Max RPS
(a)
Velocity Specifications
+V
+ Actual
Theoretical
+ Slope
Offset
-RPS +RPS
Reversal Error ( Bipolarity )
= Difference in + Slope to
- Slope - Slope
- Actual
-V
(b)
Figure 34. Velocity Specifications
Velocity Ripple
The velocity signal is a DC signal with an ac ripple. This is a by-product of the
conversion technique. The main components of the ripple are carrier
frequency and twice the carrier frequency, which can be filtered as seen in
Figure 35. Set the RC to cut-off below the carrier frequency.
R
VEL +
-
C
(
τ = RC = 1/A
3. Excessive error - if the digital output is more than 100 LSBs different from
the input angle, a BIT is generated. This can happen normally during large
step changes of the input and high accelerations which cause an acceleration
lag. In a static mode this should not be a problem. Since the R/D is an analog
system, the power supply quality is a concern. If the power supply dips, it
may cause the signal level to drop at the output of the input buffer amplifiers.
This may cause the signal level to drop enough to indicate an LOS. Also, the
distortion of the input signal may cause significant error in the error channel.
A 500µs dynamic delay occurs before the excessive error BIT becomes
active. This dynamic delay is responsive to the active filter loop.
Timing
CB
*
50 ns
DATA DATA
DATA VALID VALID
Note: Refer to the specific converter data sheet for timing specifications.
EL
EM
INH
CB
Enable
INH
Timing
For an 8-bit system, the MSB and LSB bytes are transferred sequentially. For
a 16-bit system all bits are transferred simultaneously.
When EM and EL are in a valid logic high state, the digital output is in a high
impedance tri-state mode.
Note: The converter INH line may be applied regardless of the CB signal
state, as the INH line will not latch until CB is not busy.
1. The converter Inhibit ( INH ) control is applied and must remain low for a
minimum of 300 nsec before valid data is available.
2. The Enable Most significant byte ( EM ) is set to a low state (logic 0) and
must remain low for a minimum of 150 nsec before the MSB data (bits 1
through 8) is valid and transferred.
3. After EM is set to a high state (logic 1), the Enable Least significant byte
( EL ) is brought low 150 ns minimum before the LSB data (bits 9 through 16)
is valid and transferred.
4. EL should go high (to logic 1) at least 100 nsec maximum before another
device uses the bus.
5. INH is set high and the data transfer is done. The data refresh cycle can
begin.
INH
INH
300 ns MAX
;;;; ;;;
EM 150 ns MIN
;;;;;;;
DATA 1-8
100 ns MAX
VALID
;;;;;; ;
EL 150 ns MAX
;;;;;;;
DATA 9-16
VALID
100 ns MAX
Referring to Figure 41(b), since INH has already been asserted low, the
output of the RD/RDC series transparent latch settles to a valid value within a
maximum of 150 ns after the internal counter value is stable. Therefore, the
overall worst-case delay from INH low to RD/RDC series output data valid =
375 + 50 + 150 = 575 ns. This allows the user to read valid data without
having to monitor the CB pulse.
CB
50 ns
DATA DATA
DATA VALID VALID
(a)
INHIBIT
150 ns max
(b)
3. Address line A0 sets and resets the converter INHIBIT line. When A0 is
low, the INHIBIT command INH is invoked.
5. Since the output data is not valid until 0.5 µs after the INHIBIT command is
invoked, the I/O READY line is held low for this period of time. When I/O
READY returns to the high level, the data on the bus reads on the next
negative clock edge.
EM EL
BIT 16(LSB)
BIT 15
BIT 14
BIT 13
BIT 12
BIT 11
IBM
BUS
Y1 Y2 Y3 Y4
ALE
A1 A2 A3 A4
G P=Q 3 STATE
A9 P0 Q0 BUFFER
A8 P1 Q1
A7 P2P Q2
ADDRESS ADDRESS
A6 P3 DECODER Q3 SELECTION
74LS688 JUMPERS
A5 P4 Q4
A4 P5 Q5
A3 P6 Q6
A2 P7 Q7
A1
A0
I/O READY
RV = 100kΩ
CBW = 0.022μF
CBW
= 2200pF
10
RB = 410kΩ
RS = 30kΩ
RC = 30kΩ
RB = CBW =
410k Ω 0.022µf
VELOCITY
CBW =
10 Rv =
2200pf
100k Ω
RS = RC=
30kΩ 30kΩ
18 19 9 10
-VSUM VEL
8
-VCO
RDC-19222 29 to 44
DIGITAL
OUTPUT
16
Figure 45. Mandatory External Component Values and Connections
xVRMS R1 + R 2
Using the formula =
2Vrms R2
R2 is given as 10 kOhm
Note: For circuits where no voltage adjustment is required add a 10k ohm
series resistor for oscillator turn on protection.
– SIGNAL REF
26 VAC
+SIGNAL REF
R1 = 130k Ω R2 = 10kΩ
R3 = 130k Ω R4 = 10kΩ
7 6
-REF +REF
15
SIN
16
-S
RDC-19222
14 +S
Note : For circuits where no voltage adjustment is required add a 10k ohm 1%
tolerance resistor to the SIN and COS line for circuit turn on protection.
57470
3 15
SIN
Rf
1 Ri
S1 16
2 -S
+R Ri
-R S1 S3 6 14 +S
S3
S4 Rf
RESOLVER S2 RDC-19222
A GND 4
13
Rf 12 COS
16 Ri
S4 15
7 Ri 13 -C
S2
8 10 11 +C
Rf
12
29 to 44
DIGITAL
OUTPUT
16
28
CB
RDC-19222 27 BIT
-CAP 23
10 µf
25 +
+CAP
(-5c)-5v(+5c) +5v
22 17 26 2 POWER
3 4 5 1 20 SUPPLY
+5VDC
47u f
0.01uf +
0.01uf
INH EM +
EL 47uf
A B
{
RESOLUTION
CONTROL
Figure 49 illustrates the example system circuitry with all five steps combined.
57470 7 6 18 19 9 10
-REF +REF RS RC -VSUM VEL
3 15
SIN 8
Rf -VCO
1 Ri
S1 16
2 -S
+R
Ri
-R S1 S3 6 14 +S 29 to 44
DIGITAL
S3 OUTPUT
S4 Rf 16
RESOLVER S2 A GND 4
RDC-19222
28
CB
13 27
Rf 12 COS BIT
16 Ri
S4 15
7 Ri 13 -C
S2
8 10 11 +C
Rf
AGND GND -CAP +CAP (-5c)-5v(+5c) +5v
12
23 25 22 17 26 2 POWER
21 24 3 4 5 1 20 SUPPLY
+5v
10uf 0.01uf 0.01uf
INH EM
EL
A B
1000pf
{
RESOLUTION
RESISTORS SELECTED TO LIMIT V REF TO BETWEEN 2V AND 4V PEAK 47u f
CONTROL
Grounding Tips
When designing with the RD/RDC Series Converter’s, special consideration
should be given to the two ground connections, Ground (GND) and Analog
Ground (AGND). These two internal ground planes help to reduce noise in the
analog input signal from the digital ground currents. The digital inputs and
outputs are referenced to GND, and the resolver inputs and velocity output
are referenced to AGND. To minimize ground loop noise, and therefore
unstable output results, both grounds must be tied together as close to the
converter package as possible.
When using a single-ended resolver, the resolver returns (S1 and S4) should
be connected to AGND. The velocity output can then be measured with
respect to AGND. Additionally, AGND provides a reference for internal analog
signals, with no current drop.
Layout Considerations
As in any analog and digital design, special attention should be given to
routing of digital signals near analog lines. This converter has several high
impedance nodes that should be considered (+S, -S, +C, -C, -VSUM and
–VCO). Mount components as close to these pins as possible and avoid
running digital or other signals near these pins.
Transient Protection
The inputs of the RD/RDC Series are susceptible to overvoltage damage
dependent on the magnitude of the current that is injected into the input. The
failure mode is one where the unit latches, and may damage or destroy the
device. If the current is not excessive (<100 mA) the unit may function
normally after power cycling.
The two summing junctions -C and -S are the most susceptible and will latch
at ±6 mA. The inputs RH, RL, +C and +S show no problem up to 25 mA.
Each power supply input to the RD/RDC Series [+5 V (VDD) and -5 V (VSS)]
Power supply has an internal diode clamp. If there is a possibility that the amount of current
protection.
could exceed the limit, a reversed-biased diode should be installed at the
input to VDD and VSS. If using a single-ended input, a 10 kΩ series resistor is
recommended.
When using a thin-film resistor network for the signal input and additional
resistors to scale the reference voltage, the solid-state signal and reference
inputs are true differential inputs with high ac and DC common rejection. Input
impedance is maintained with power off, therefore, most applications will not
require units with isolation transformers. The recurrent ac peak DC + common
mode voltage should be limited.
Direct When direct connect is used (No voltage divider used) for either the input or
Connect ref then (1%) 10k ohm series resistors should be used for turn on protection.
Applications Only one resistor per SIN, COS, REF line is required.
+S RH
THIN-FILM
S1 CR3 RESISTOR
NETWORK -S
RD/RDC
CR4
CR2 S2 COS
S2
+S
RL
S1 -S
S1 S1 SIN
S2 S2 +S RH
CR4 CR5 THIN-FILM
90 V L-L RESISTOR -S
RESOLVER NETWORK RD/RDC CR6
INPUT COS
S3 S3
+S
S4 RL
S4 -S
Bandwidth
increasing the hysteresis in the converter (there will be missing digital output
codes). An increase in amplitude will increase the BW & decrease the amount
of hysteresis thereby causing jitter on the output digital code. For example, a
50% reduction in signal will lead to a 50% reduction in bandwidth. This is due
to the conversion technique used. An internal voltage-controlled oscillator
(VCO) expects to see a certain voltage per LSB. If this voltage is not
maintained, the amount of hysteresis in the converter will change affecting
normal operation.
Bandwidth Optimization
When using a low-cost monolithic converter for position and velocity feedback,
it is important to understand the dynamic response for a changing input.
When considering a bandwidth value to set to your converter, several
parameters have to be taken into consideration. The ability to track step
responses and accelerations will determine what bandwidth to select. The
lower the bandwidth of the resolver-to-digital converter the greater the noise
immunity; high frequency noise will be rejected. For a small step input, the
bandwidth determines the converter settling time. For a large step input the
maximum velocity, slew rate and bandwidth determine the settling time.
Acceleration Lag
The RD/RDC Series Converters use a Type II tracking loop. As the
acceleration increases the converter output signal will lag the input signal by a
In a Type II loop,
constant value, the acceleration constant (Ka). Ka is dependent upon the
acceleration lag is
a constant.
bandwidth and maximum tracking rate as shown in the following example.
EXAMPLE: Find the Acceleration Constant (Ka) and the Acceleration Lag for
a system with the following parameters:
Resolution: 16-Bit
Bandwidth: 100 Hz
Reference: 1000 Hz
Maximum Tracking:10 RPS
Equation 1 : BW =
( 2 )(A ) and Equation 2 : Ka = A 2
π
2
(BW )(π) ⎡ (BW )(π ) ⎤
therefore Equation 3 = A = and Equation 4 = Ka = ⎢ ⎥
2 ⎣ 2 ⎦
From Equation 4 :
2
⎡ (BW )(π ) ⎤
Ka = ⎢ ⎥
⎣ 2 ⎦
2
⎡ (100Hz )(π ) ⎤
Ka = ⎢ ⎥
⎣ 2 ⎦
49,298
Ka =
sec 2
Given :
Acceleration Rate
Equation 5 = ∗ Acceleration Lag =
Ka
* Acceleration lag is in degree’s.
From equation 5 :
Acceleration Rate ( for 1 LSB error ,16 − bit mod e) = (Ka )( Lag for 1LSB error )
Acceleration Rate ( for 1 LSB error ,16 − bit mod e) =
(49,284)(0.0055°)
sec 2
270°
Acceleration Rate ( for 1 LSB error ,16 − bit mod e) =
sec 2
Acceleration rate is the max rate at which the converter will track within 1LSB
of error. In the example, an acceleration of 270°/sec2 will cause 1 LSB of lag.
2
Also written as 270°/S .
Example: Solve for the large step (179°) settling time for a 16-bit mode
system with the following parameters:
Bandwidth: 100 Hz
Maximum Tracking Rate: 10 RPS
360° 3600°
Maximum Tracking Rate= x 10 RPS =
1revolution sec
1
Time Cons tan t =
A
A=
(BW )(π)
2
A=
(100 Hz )(π)
2
A = 222 Hz
1
= 4.5 msec
A
Large Step Settling Time = tTR + [(time constant)(tc from Table 23)].
Therefore, the settling time for a large step in a 16-bit application of the
converter with a 100 Hz bandwidth would be 98.5 msec (This is an
approximation).
Table 23 lists the number of counts per rotation and time constants for various
resolutions.
Table 23. Counts per Rotation and Time Constraints For Various Resolutions
Resolution Number of Counts per Rotation Number of Time
(bits) (cpr) Constants (tc)
10 1024 7
12 4096 8
14 16384 10
16 65536 11
ANGLE
179˚ 1 rps
12 bit resolution
100 Hz BW
0˚ time (ms)
500 36
12 bits = 8 time constants
time constant = 1/A
179˚
(RPS)˚/s
ANGLE
179˚ 10 rps
12 bit resolution
100 Hz BW
0˚ time (ms)
50 36
12 bits = 8 time constants
time constant = 1/A
179˚
(RPS)˚/s
1
*If you decrease the BW you increase the settling time constant ( ): BW = 2 A
A
π
Let Rs = 30 kΩ, therefore the internal clock rate is 1,333,333 Hz (or counts-
per-second).
Note:
If Rs = 20 kΩ then the internal clock is nominally 2,000,000 Hz
If Rs = 15 kΩ then the internal clock is nominally 2,666,666 Hz
Many applications for R/D converters require an output with high resolution,
improved angular position accuracy and improved resolution of the calculated
velocity. The RDC-19220/2/4 & S Series, like most R/D converters, is based
on a type II tracking loop. This loop uses a VCO (voltage-controlled oscillator)
to increment the output code. As the resolution of the output is increased, the
number of counts per rotation is increased. The VCO will be required to
operate at a very high counting rate as the input is changing at a fast speed.
As shown in the maximum tracking rate section of this manual, the VCO limits
the maximum tracking rate of the converter.
To alleviate this limitation, the RDC-19220/2/4 & S Series Converter has the
For some
applications,
ability to change the resolution on the fly (sometimes called gearshifting). This
a velocity feature allows the user to increase the resolution when the input is changing
signal at slow speeds, or decrease the resolution when high tracking rates are
transient required. This change is accomplished by simply changing the digital
caused by resolution controls. The change in resolution will not change the bandwidth of
changing the the converter. However, the change in resolution will introduce a transient in
resolution on the output code as well as the velocity. This is due to internal gain changes
the fly may when the resolution is switched. In applications where this transient is a
not be a problem, the RDC-19220/2/4 & S Series Converter will require additional
problem. external components to make the transition from one resolution to the next
resolution smooth (greatly reducing the transient). Refer to Figure 53 (Note
the RD19230 & RD-19240 has this feature built in).
In a typical type II servo loop the velocity voltage is charged onto the
capacitors used to control the bandwidth (CBW and CBW/10). When the
resolution is increased (i.e., 14-bits to 16-bits) the velocity voltage that is
presented to the VCO must be 4 times larger to achieve the same tracking
rate of 214 counts per revolution vs 216 counts per revolution. The time to
change this voltage is controlled by the RC time constants of the bandwidth
components. This is the time for the velocity voltage to re-scale itself. A
tracking error will result due to this re-scaling period.
For applications requiring switch on the fly, the RD-19230 & RD-19240 series
converters are recommended.
Details on setting up switch on the fly operation are available in the “Basics”
section of this manual, the RD-19230FX & RD-19240FX data sheets and the
application note AN/MFT-3.
Error Minimizing for the RDC-19220/2/4 & S Series during Resolution Switching
By using the external circuitry shown in Figure 53, the error will be reduced.
While one set of bandwidth components are being used in the loop, a second
set of components are being charged to 4 times the velocity voltage and a
third set to 1/4 of the velocity voltage. When the resolution is either increased
or decreased one step (i.e., 14- to 16-bit or 14- to 12-bit) the precharged
bandwidth components are switched into the loop. Since the capacitors are
already charged to the correct velocity voltage this circuit greatly reduces the
error caused by changing resolution.
The circuit shown in Figure 53 allows switching from any of the four possible
resolutions, up or down one resolution step, without generating a significant
error in the output, provided the maximum tracking speed is not exceeded
when switched to a resolution (i.e., 16 bits when input is at 20 RPS). If the
resolution needs to be changed a second step in the same direction (i.e., 12-
bit to 14-bit to 16-bit) sufficient time must be allowed between steps to
precharge the proper bandwidth components, or an error may be generated.
DDC’s RD-19230 and RD-19240 converters provide another way to solve this
problem using a on chip switch for a second set of bandwidth components
detailed in the Optional Bandwidth Components section on page Error!
Bookmark not defined. of this manual.
R4
30K
R3
RC4741 - 13
U3b 12
14 11 + 10K
-5v
+5v
2 R1
1 - 4
U3a 3 10K
RC4741 +
C1 R2
+5v 100pF 30K
U2
12 VDD 16 GND
14 X0
15 X1 X 13
11 X2
X3
1 Cbw3 Rb3
Y0
5 Y1
Resolution 2 Y 3
Controls 4 Y2
Y3 VEE 7
10 INH 6
-5v
A Cbw3/10
B 9 A
B VSS 8
4052
U1 +5v
12 X0 VDD 16 Cbw2 Rb2
14 X1
15 13
X2 X
11 X3 Cbw2/10
1 Cbw1 Rb1
Y0
5 Y1
2 Y2 Y 3
4 Y3 VEE 7 -5v
10 A INH 6 Cbw1/10
9 B VSS 8
4052
VEL
RS Rv
CT
RESOLVER R1 16 BIT
GAIN DEMOD VCO UP/DOWN
INPUT COUNTER
+1.25V
-
THRESHOLD
DIGITAL
H=1
OUTPUT
RDC-19220/2 & RDC-19220/2S GND A GND
Two-Speed Resolution
LogX
Resolution derived from 2-speed = ( − 1 ) + (Converter Resolution)
Log2
X = The ratio of the fine shaft compared to the binary look-up chart (Table 24).
For the value of X, use the next highest number listed that is greater than the
fine shaft ratio.
(Note : For higher values of X, simply continue to double last listed value)
Example:
In the single-speed case (CT = the transducer used) the system will,
essentially, have an accuracy dependent upon the accuracy of the CT (we will
assume a perfect transducer) and the positional resolution of the servo loop.
Assume we can position the CT with a certain accuracy, say within ± 0.1°. If
we now gear this CT to another with a gear ratio of 1:n (called a “coarse” CT)
where the CT we are positioning rotates n turns for each single turn of the
other, then 0.1° of rotation of the fast CT (called the “fine” CT) will turn the
slow CT (called the “coarse” CT) 0.1 ÷ n, so that an inaccuracy in the fine CT
position is effectively divided by the gear ratio (often called the speed ratio).
For greater detail on 2 speed theory review the synchro resolver handbook
which can be downloaded from www.ddc-web.com.
DDC provides 2 speed support in the PCI card family. The SB-36200, SB-
36210, SB-36220 cards contain 2 speed functions.
User-Written Software
The following software example is a concept of one of many possible software
algorithms:
If the difference between the calculated and actual fine angles is greater than
180° an additional turn of the coarse angle is added to the resultant angle,
since the Actual Fine Angle indicated an additional coarse turn is necessary.
The result is calculated by taking the number of turns calculated from the
Coarse Angle, multiplying by 360.0° to get a rough calculated rotation, then
adding the Actual Fine Angle and any necessary correction value to get an
exact calculated rotation. This calculated rotation, when divided by the speed
ratio, gives the final two-speed output. An example program written to
interpret output data is illustrated below:
IN->COARSE_ANGLE
IN->FINE_ANGLE
IN->SPEED
CORRECTION =0
ERROR_FLAG = CLEAR
IF (ERROR>0)
IF (ERROR>135) AND (ERROR<225) ERROR_FLAG = TWOSPEED_ERROR
IF (ERROR>180) CORRECTION = 360.0
ELSE
IF (ERROR<-135) AND (ERROR>-225) ERROR_FLAG = TWOSPEED_ERROR
IF (ERROR<-180) CORRECTION = 360.0
Synchro/Resolver-to-DC Conversion
A technique used to convert Synchro/Resolver information to a DC voltage
that represents an angle is shown in Figure 56. This technique can be
accomplished by connecting the RD/RDC or similar series product as
discussed in earlier sections and interfacing the digital output to a digital-to-
analog converter (DAC). The positive full-scale voltage will represent +180°,
and the negative full-scale voltage will represent -180°. The use of an
RD/RDC has the advantage of eliminating the effects of quadrature signals
and decreasing the noise sensitivity of the Synchro/Resolver-to-DC
conversion.
+5 V +15 V
0.1 µF 0.1 µF
RH RH
BIT 16 (LSB) BIT 16 (LSB)
RL RL
A +5 V
B
-VSUM
CBW CBW
ENM BURR BROWN
10 RBW DAC 703
ENL
VEL
RV
-VCO
GND AGND -VCC VDD COMMON
0.1 µF
-15 V +5 V
-5 V
NOTES: 1. CONCEPT DRAWING. SEE MANUFACTURERS' DATA SHEETS FOR COMPLETE INFORMATION.
2. RH AND RL ARE SWAPPED TO PROVIDE 2'S COMPLEMENT OPERATION. NORMAL WIRING OF
RH AND RL PROVIDES OFFSET BINARY OPERATION.
SEU/SEL Solutions
Preface
The RD/RDC series are versatile 16-bit state-of-the-art monolithic synchro,
resolver, or LVDT to digital converters. These single chip monolithics are
CMOS components which need protection from the effects of SEU/SEL or
latchup. The following summary outlines the most common approach used to
protect the converters.
For a quick, but limited functional test, the reference excitation or a single
oscillator source can be used to test the converter in the system. This test is
achieved by shorting various combinations of the input signals and connecting
these to the reference inputs. The amplitude of the reference or the oscillator
source should be a factor of the intended line-to-line input, as shown in the
reference multiplier column of Table 25. For example, a resolver-based
system with an intended 11.8 VL-L input should have the oscillator set at either
11.8 Vrms or 8.26 Vrms, depending on the input signal connection
combination. Connect shorted inputs to reference inputs as specified in Table
25.
*Single ended mode can only be tested in the 1st quadrant (0° to 90°) where SIN, COS, and REF are all in phase.
Boards that are designed with an LVDT input can also use a
Synchro/Resolver simulator for testing purposes. To test an LVDT application
using a Synchro/Resolver simulator, the angular bit weights of the R/D
converter, in linear mode, are modified as follows (also see Table 26):
Note: For LVDT Simulation Guidelines see the DSC-11524 Converter Data Sheet.
Troubleshooting
The RD/RDC Series converters present a challenge for troubleshooters when
things go wrong in a system. This section includes a Checklist and a detailed
symptoms and solutions table.
Troubleshooting Checklist
The following checklist offers a starting point for the troubleshooter to follow.
Perform these tests as the first step in problem solving.
2. Check the converter COS pin. It should be a sampled sine wave as shown
in Figure 57 (w.r.t. AGND) for converters without a synthesized reference
such as the RDC-19220/2/4. For converters with a synthesized reference,
such as the RDC-19220/2S, RD-19230 and RD-19240, the sine wave will not
be sampled at those pins.
3. Check the converter SIN pin. It should be a sampled sine wave as shown in
Figure 57 (w.r.t. AGND) for converters without a synthesized reference such
as the RDC-19220/2/4. For converters with a synthesized reference, such as
the RDC-19220/2S, RD-19230 and RD-19240 the sine wave will not be
sampled at those pins.
6. Ensure that there is no digital routing near the ±S, ±C, -VCO and -VSUM
pins. These are amplifier summing junction nodes (high impedance) and are
very sensitive to coupling.
7. Check the -5 V signal on pin 17 (if using the devices’ internal -5 volt
feature). It should measure between -4 VDC and -5 VDC.
8. Check the built-in test (BIT) pin. It should be high during normal operation.
9. Check the +5 VDC and –5 VDC power supply lines for excessive noise or
ripple.
10. Verify that AGND & GND are tied together close to the converter chip.
11. Verify that no DC-offsets are present at the converter SIN and COS
inputs.
12. Verify that the bandwidth components are correctly calculated, and
properly installed.
Further troubleshooting guidelines are available, ask for the following FAQ’s:
FAQ – gcnvtr-016
FAQ – gcnvtr-017
BIT (Built-in Test) fault (cont): reference and the input signals (see Reference
Input section)
4. For RDC-1922XS and RD-19230, RD-19240
versions verify there is no loss of the reference
signal (LOR) ref input less than 500mv.
5. For intermittent BIT faults verify correct input
voltage configuration. An indication of this problem
would be if the BIT faults occur when the SIN and
COS input absolute values are equal. For example,
at 45° the SIN and COS inputs should both be
1.41Vrms ± 10% measured at the converter pins.
Jitter:
• Least significant bits (LSBs) continuously 1. Verify that the correct input voltages at the SIN and
changing, or jittering. COS pins are sampled 2 Vrms waveform.
• BIT logic level is high, indicating normal (See Figure 57 ). SIN is max voltage at 90°, COS is
operation. max voltage at 0°.
2. Verify that there is no excessive phase shifting
between the reference and the input signals, or the
input signals from each other (see Reference Input
section).
3. Ensure that there are no traces near the mandatory
external component connections (specifically +S, -S,
+C, -C, -Vsum and -VCO). This may cause coupling
into the converter.
4. Verify that the carrier frequency is at least 3.5 times
the bandwidth (see Mandatory External Components
section).
5. Check for excessive ripple or noise on the power
supplies.
6. Tie the GND and AGND pins together close to the
converter (see Grounding Tips section).
7. If using a differential input, check resolver for
undesirable coupling of the sampled sine wave
signal. To verify this, ground +S and +C. One
suggestion to alleviate this problem is to use a
single-ended input to buffer the input.
Digital output not responding:
Note: All digital inputs have an internal pull up.
• Digital output does not respond to changes in
input from resolver.
1. Ensure that the EM and EL pins are set to logic 0
or tied to ground (see Timing section).
2. Ensure that INH is allowed to pull up to logic 1, or
is set for a logic 1 after the first position is read.
Additionally, set INH to a logic 0 for a minimum of
300 ns before taking the next reading (see Timing
section).
Large error in digital output: 1. Verify that the synchro or resolver lines are
connected correctly.
• Error larger than the rated accuracy of the If the synchro is ARINC 407 compliant, X, Y, and Z
converter, and larger than 1° (or > 60 arc should be interpreted as follows:
minutes). X = S1, Y = S3, Z = S2.
(The Z may be grounded).
2. Verify that the reference input polarities are correct.
The high side of the reference should be connected
to RH and the low side should be connected to RL.
If the reference inputs are swapped, the output
error would be approximately 180° (or 10,800 arc
minutes).
3. Verify that the input resistors are correctly scaled to
provide a 2 Vrms signal to the converter input.
4. Ensure that the -5 V supply is between -4 V and
-5 V.
5. Verify that no external circuits are added to the
velocity output of the converter. See the “basic”
section of the manual on velocity.
Table 27 and Table 28 list the output of the RD/RDC series converters when
the Synchro and Resolver input signal lines are swapped.
For further troubleshooting hints contact DDC Applications and request FAQ-gcnvtr-016 and FAQ-gcnvtr-
017.
Introduction
The requirement for velocity and position feedback play an important role in
today’s motion control systems. With the development of low-cost monolithic,
resolver-to-digital converters, a resolver-based system provides design
engineers with the building blocks to handle a wide variety of applications. A
resolver’s small size, rugged design and the ability to provide a very high
degree of angular accuracy under severe conditions, make it an ideal
transducer for absolute position sensing. These devices are also well suited
for use in extremely hostile environments such as continuous mechanical
shock and vibration, humidity oil mist, coolants and solvents. Absolute position
sensing vs. incremental position sensing is a necessity when working in an
environment where there is the possibility of power loss. Whenever power is
supplied to an absolute system, it is capable of reading the current position
immediately, eliminating the need for a go home or reference starting point.
Figure 58 illustrates an example of when DDC components would be used in
a two-axis antenna control application.
Applications
Specific applications require unique profiles to control the speed and acceleration of
the motor to perform the task at hand. By reducing the accelerations and
decelerations which occur during each operation it is possible to lower the cost and
use more efficient motors. Industrial applications include the following:
• Stability
Control
Resolver-to-Digital Converters
For proper operation the RD/RDC Series Resolver-to-Digital converter
requires only five external passive components. These components are used
to set the bandwidth, which controls the speed at which the converter will
react to changes in the inputs, and the maximum tracking rate. Depending on
converter selection resolution can also be programmed to provide 10-, 12-,
14- or 16-bits of parallel data.
multiplier is cos θ, and the digitally encoded “word” presented to the sine
multiplier is φ, then the output code is cos θ sin φ.
ERROR ERROR
POSITION
COMMAND
CONTROLLER AMPLIFIER MOTOR
PLC
RESOLVER
RESOLVER-TO-DIGITAL
CONVERTER
POSITION SENSOR
( BIT ) flag. When the AC error signal exceeds the equivalent ac value of 100
LSBs the BIT flag will indicate a tracking error. This angular error signal is
then fed into the error processor and VCO. This circuit consists essentially of
an analog integrator whose output (the time integral to the error) controls the
frequency of a voltage-controlled oscillator (VCO). The VCO produces clock
pulses that are counted by the up-down counter. The “sense“ of the error (φ
too high or φ too low) is determined by the polarity of (φ), and is used to
generate a control counter signal (U), which determines whether the counter
increments upward or downward. Finally, note that the up-down counter, like
any counter, is functionally an incremental integrator, therefore the tracking
converter constitutes in itself a closed-loop servomechanism (continuously
attempting to null the error to zero) with two integrators in series.
+REF -REF BIT
SIN
-S C BW
S1 - C BW
S2 +S 10 RB
SIGNAL + D R1
S3 CONDITIONER COS CONTROL e
TRANSFORMER GAIN DEMODULATOR VEL
S4 -C
-
+C INTEGRATOR
+ RV
HYSTERESIS
A B
16 BIT VCO
-5 V
UP/DOWN & RS
INVERTER
COUNTER TIMING
E RC
A GND
+5 V DATA
GND LATCH
-5 V
INH EM BIT 1 EL A B CB
THRU
BIT 16
This is called a “Type II“ servo loop, which has decided advantages over Type
1 or Type 0 loops. In order to appreciate the value of a Type II servo behavior
of this tracking converter, consider first that the shaft is not moving. Ignoring
inaccuracies, drifts, and the inevitable quantizing error, the error should be
zero (θ = φ), and the digital output represents the true shaft angle of the
resolver. Now, start the resolver shaft moving, and allow it to accelerate
uniformly, from dθ/dt = 0 to dθ/dt =V. During the acceleration, an error will
develop, because the converter cannot instantaneously respond to the
change of angular velocity. However, since the VCO is controlled by an
integrator, the output of which is the integral of the error, the greater the lag
(between θ and φ), the faster the counter will be called upon to catch up.
When the velocity becomes constant at V, the VCO will have settled to a rate
of counting that exactly corresponds to the rate of change in θ per unit time
and instantaneously θ = φ. Therefore, dφ/dt will always track dθ/dt without a
velocity or position error. The only error will be momentary (transient) error,
during acceleration or deceleration. Furthermore, the information produced by
the tracking converter is always “fresh“ being continually updated, and always
available at the output of the counter. Since dθ/dt tracks the input velocity it
can be brought out as velocity, a DC voltage proportional to the rate of
Velocity (Speed)
Motor speed is monitored by using the velocity output signal generated by the
resolver-to-digital converter. This signal is a DC voltage proportional to the
speed, positive for increasing angles and negative for decreasing angles, with
a linearity specification of .25% typical and a reversal error of .75% typical.
The error processing is performed using the industry standard technique for
type II tracking, resolver-to-digital converter (see Figure 60).
PIN NUMBERS
FOR REF ONLY
0.125 ±0.020
(3.18 ±0.508)
0.590 ±0.010 0.115 ±0.010
(14.99 ±0.25) (2.92 ±0.25) 0.050 ±0.010
(1.27 ±0.25)
1 40
20 21
0.075 ±0.010
(1.91 ±0.25) 0.040 x 45˚
CHAMFER
0.500 ±0.010 (1.02)
0.020 x 45˚ (12.70 ±0.25) (3 PLACES)
0.143 ±0.010
(0.51) (3.63)
CHAMFER 0.050 TYP
(ORIENTATION (1.27) 0.095 ±0.007
MARK) (2.41 ±0.18)
6 1 40
7 39
0.017 TYP
(0.43)
17 29
18 28 PIN NUMBERS
FOR REF ONLY
0.075 ±0.010
(1.91 ±0.25) 0.650 SQ ±0.010
(16.51 ±0.25)
0.078+0.004
-0.002
(2.00+0.10
-0.05 )
32 17
0.096MAX
(2.45MAX)
33 16
0.0098MIN,0.0197MAX
(0.25MIN,0.50MAX)
0.0197
0.520±0.010 (0.50)
(13.2±0.25)
RD-19230FX
-XXX 0.394±0.004
(10.00±0.10)
Date Code
48 pin1
0.0098MIN,0.0197MAX
0.096MAX (0.25MIN,0.50MAX)
49 64
(2.45MAX)
*
0.394±0.004
(10.00±0.10) 0.007MAX
(0.17MAX) 0.008
(0.22)
0.520±0.010 0.035+0.006
(13.2±0.25) -0.004
( 0.88+0.15
-0.10 )
12 Eq Sp @
0.026 = 0.312
(0.65 = 7.800)
0.035 +0.006
-0.004
-XXX 0.394±0.004
(10.00 ±0.10)
Date Code
pin1
Notes: 1. When using the built-in –5V inverter: connect pad 33 to 58, pad 16 to 17, and a 10 mf/10 VDC
capacitor from pad 24 (negative terminal) to pad 26 (positive terminal). Connect a 47 mf/10
VDC capacitior from –5V to GND. The current drain from the +5V supply doubles. No external
–5V supply is needed.
Transformers
1 2 3 5
0.600 0.81 MAX
T1A (15.24) (20.57)
10 9 8 7 6
0.105 (2.66)
SIDE VIEW 0.100 (2.54) TYP
TOL NON CUM
TERMINALS
0.025 ±0.001 (6.35 ±0.03) DIAM BOTTOM VIEW
0.125 (3.18) MIN LENGTH
SOLDER-PLATED BRASS
1 6
INPUT OUTPUT
5 10
(SEE Table 9)
0.61 MAX
(15.49)
0.61 MAX
(15.49) 0.15 MAX 0.09 MAX
(3.81) (2.29)
0.30 MAX 0.09 MAX 0.15 MAX
(7.62) (2.29) (3.81)
1 3 4 5 11 12 14 15
0.81 MAX
(20.57)
T1A T1B 0.600
(15.24)
10 9 8 7 6 20 19 18 17 16
0.115 MAX
(2.92)
SIDE VIEW 0.100 (2.54) TYP BOTTOM VIEW
TOL NON CUM
TERMINALS
0.025 –0.001 (6.35 –0.03) DIAM BOTTOM VIEW PIN NUMBERS FOR REF. ONLY
0.125 (3.18) MIN LENGTH
SOLDER PLATED BRASS
T1A
S1 1 6 -SIN
5
3 10
S3 +SIN
SYNCHRO
RESOLVER
INPUT
T1B OUTPUT
11 16 -COS
S2 15 20 +COS
0.61 MAX
(15.49)
0.61 MAX
(15.49) 0.15 MAX 0.09 MAX
(3.81) (2.29)
0.30 MAX 0.09 MAX 0.15 MAX
(7.62) (2.29) (3.81)
1 3 4 5 11 12 14 15
0.81 MAX
(20.57)
T1A T1B 0.600
(15.24)
10 9 8 7 6 20 19 18 17 16
0.115 MAX
(2.92)
SIDE VIEW 0.100 (2.54) TYP BOTTOM VIEW
TOL NON CUM
TERMINALS
0.025 –0.001 (6.35 –0.03) DIAM BOTTOM VIEW PIN NUMBERS FOR REF. ONLY
0.125 (3.18) MIN LENGTH
SOLDER PLATED BRASS
Dimensions are shown in inches (mm).
T1A
S1 1 6 -SIN
3 10
S3 +SIN
RESOLVER
RESOLVER
INPUT
T1B OUTPUT
11 16 -COS
S4
S2 15 20 +COS
+15 V +15 V
V -Vs V -Vs
(Analog (-15 V) (Analog (-15 V)
Gnd) Gnd)
Figure 72. 60 Hz Transformer Mechanical Outline – (Syn Input – 52039, Ref Input - 24133)
BW (Bandwidth) .........................The frequency band in which a system can process information. Also known as the
3dB point.
Bipolarity ....................................The difference between a measurement in one direction and a measurement in the
BIT ( Built-in test) .......................Digital output that will indicate if a fault condition exists.
Charge Pump..............................A voltage-doubler built into the converter to provide a -5 V supply to the chip.
Coarse Angle..............................In a two-speed system, the reading from the resolver with a 1:1 ratio.
CB (Converter Busy)..................A pulse that is generated for each LSB of change in the converter.
Differential ..................................Input signal that is with respect to another signal (S3-S1), but not to ground.
Differential Linearity…………… Measurement (in LSBs) of the deviation of the digital output from equal step-per-bit
behavior. The worst-case deviation of the smallest or largest step (bit) from the
theoretical (LSB) size, expressed as a percentage of the theoretical (LSB) size, is
the differential non-linearity.
Encoder.......................................A device used to translate angular position data into a series of digital pulses.
Fine Angle...................................In a two-speed system, the reading from the resolver with a 1:N ratio, where N >1.
GND.............................................Digital Ground.
HOLE...........................................Change data by 1 LSB and converter does not respond with any change. This is
referred to as a hole in data.
Inductosyn® ...............................A device that translates linear or rotational position data into analog signals.
LSB..............................................Least significant bit. In 10-bit resolution this would be bit 10, in 16-bit resolution this
would be bit 16.
MSB ............................................ Most significant bit. For any bit resolution this would be bit 1.
Over Voltage .............................. A voltage amplitude that is too large for the circuit to operate properly.
Quadrature ................................. A 90° out of phase portion of a signal, with respect to reference.
Resolver ..................................... Type of transducer whose four-wire output has a direct relationship to the anglular
position of the shaft.. A resolver contains two output windings 90° out of phase.
Reversal Error............................ The difference between a measurement in one direction and a measurement in the
RPS
(Revolutions per second) ......... A unit of measure for velocity or tracking rate.
Scale Factor ............................... Accuracy measurement for the velocity output of the RDC-19220 series. Average of
the gain errors or the best fit straight line, for positive and negative rotations.
Synchro ...................................... Type of transducer with a three-wire output having a direct relationship to the
angular position of the shaft. A synchro contains three output windings 120° out of
phase.
TR (Tracking rate)...................... A velocity or speed at which the converter can monitor the changes of the signal
output.
A G
A quad B.......................................................... 57 gearshifting .. See resolution: changing on the fly
acceleration lag................................................ 64
I
angular data ..................................................... 55
B
x
bandwidth Icons
and carrier frequency................................... 80 Caution......................................................... ix
and external component values.. 36, 50, 87, 88 Disk.............................................................. ix
and input signal amplitude........................... 81 Idea/Tip ........................................................ ix
and resolution .............................................. 80 Note.............................................................. ix
optimization................................................. 81 Reference ..................................................... ix
bipolarity ......................................................... 61 Warning........................................................ ix
x x
x imbalance matching .........................................10
built-in test................................................... 5, 64 incremental encoder .........................................57
incremental encoder output ..............................56
C
incremental encoder signals .......................56, 57
calculation of external components ........... 36, 73 Inductosyn input.........................................26, 27
x Inductosyn-to-digital conversion .....................26
CBW ........................................................ 4, 50, 51 inhibit .........................................................57, 66
CBW/10 ........................................................... 4, 51 input configurations
4, 38 DC sin/cos input...........................................28
converter busy (CB) ........................................ 57 differential resolver input.............................16
x direct input .............................................13, 15
direct sin/cos input .......................................13
D Inductosyn input...........................................26
DC sin/cos input .............................................. 28 LVDT input..................................................19
differential resolver input ................................ 16 reference input..............................................29
digital output data ............................................ 66 single-ended input ........................................13
digital position data ......................................... 52 synchro input................................................18
direct input........................................... 13, 14, 15
J
direct sin/cos input........................................... 13
x J-lead..............................................................113
E L
electrical static discharge................................... x layout considerations........................................77
enable .............................................................. 66 linearity ............................................................61
encoder emulation ........................................... 56 LVDT
Encoder Emulation .......................................... 56 application testing ........................................97
error digital output code........................................25
accuracy....................................... 16, 100, 102 input .......................................................18, 20
phase.................................................... 64, 100 input scaling ...............................20, 21, 23, 24
30, 31, 32, 100, 102 resolution................................................19, 24
ESD ................................................................... x
M
external components.................. 3, 36, 51, 56, 61
maximum tracking rate ........................36, 38, 61