Shunt Calibration for Dummies; a Reference Guide

by LaVar Clegg
Interface, Inc. Advanced Force Measurement Scottsdale, Arizona USA Presented at

Western Regional Strain Gage Committee Summer Test and Measurement Conference August 2, 2005
1

ABSTRACT
Shunt Calibration is a technique for simulating strain in a piezo-resistive strain gage Wheatstone bridge circuit by shunting one leg of the bridge. The bridge may be internal to a discreet transducer or composed of separately applied strain gages. The resulting bridge output is useful for calibrating or scaling instrumentation. Such instrumentation includes digital indicators, amplifiers, signal conditioners, A/D converters, PLCs, and data acquisition equipment. Care must be taken to understand the circuits and connections, including extension cables, in order to avoid measurement errors.
2

1. What is Shunt Cal ?

Shunt Calibration = Shunt Cal = SCAL = RCAL A means of calibrating or verifying instrumentation by
simulating a physical input.

The simulation is accomplished by shunting one leg

of a Wheatstone bridge circuit.

Not a means of calibrating a transducer. A poor mans mV/ V simulator.

3

2. A Good Simulator is Superior to Shunt Cal

A simulator is not strain sensitive. Resistor ratios are less temperature sensitive. Thermal emf errors are minimized by design. Symmetrical shunting produces less common mode error. Provides specific and convenient mV/ V values. Provides a true zero output. Has no toggle. Shunt resistor doesnt get lost. A calibration history can be maintained.
4

3. Shunt Cal has some advantages

Low cost. The bridge circuit is already there. Need to make and break cable connections can be
avoided.

Can be applied conveniently and at any time and

during test programs.

4. About Formulae in this Paper

The formulae herein are derived by the author and he is responsible for any errors. Rs = Value of shunt resistor. Rb = Bridge resistance represented by single value. Vs = Simulated output at signal leads in units of mV/ V. Vs is always net (the difference between the shunted

and unshunted readings or similarly the difference between the switch-closed and switch-open readings). = means mathematically exact. means close enough for practical purposes. Infinite input impedance of instrumentation is assumed.
6

4. About Formulae (continued)

In each circuit type, only one leg of the bridge is

shown shunted, for simplicity. But the formulae apply as well to the opposite leg if the resistor numbers are placed in the formula according to position. They also apply to the adjacent legs if resistor numbers are placed according to position and polarity of Vs is reversed.

5. Basic Bridge Circuit

+ Exc

R4 R1
- Sig +Sig

R2 Rs
- Exc

R3

1000 R 2 Vs = R1 R 2 Rs + + 2 + R1 + R 2 R 2 R1

Simplified Basic Bridge Circuit, R1=R2=Rb

+ Exc

R4 Rb
- Sig +Sig

Rb Rs
- Exc

R3

Vs =

250 Rs + 0 .5 Rb
9

Basic bridge circuit examples

Rb (ohm) 350 350 350 350 700 1000 Rs (ohm) 30,000 40,000 59,000 60,000 120,000 120,000 Vs (mV/V) 2.89975 2.17797 1.47866 1.45409 1.45409 2.07469

10

Basic Bridge Formula Inverted to solve for Rs

+ Exc

R4 Rb
- Sig +Sig

Rb Rs
- Exc

R3

250 0 .5 Rs = Rb Vs

11

Basic circuit inverted formula examples

Rb (ohm) 350 350 350 700 700 1000 Vs (mV/V) 1 2 3 2 3 3 Rs (ohm) 87,325 43,575 28,992 87,150 57,983 82,833

12

+ Exc

6. Bridge With Series Trim or Compensation Resistors

-Sig

R5

Rb Rb
+Sig

Rb Rb Rs R6
-Exc

Vs

250 Rs + 0.5 Rb

R 6 R5 R 6 + R5 1 1 + 4 Rs R6 + R5 + Rb
13

250 Rs + 0. 5 Rb

R 6 R5 R 6 + R5 1 1 + 4 Rs R6 + R5 + Rb

Nominal term

Summation term

Difference term

This formula for the series resistors case is interesting because it shows that if R5 = R6, Vs is nearly the same as if R5 = R6 = 0. This fact allows a batch of transducers with variations in natural loaded outputs to be trimmed with series resistors to a standard output and all will have a similar Vs.
14

Series resistor circuit examples

Rb (ohm) 350 350 350 350 350 700 700 Rs (ohm) 60,000 60,000 60,000 60,000 60,000 60,000 60,000 R5 (ohm) 0 50 0 50 175 0 100 R6 (ohm) 0 50 50 0 175 100 200 Vs (mV/V) 1.45409 1.45349 1.63551 1.27207 1.45198 3.26086 3.18574 15

7. Bridge With Parallel Trim or Compensation Resistor

+Exc

Rb Rb R7
- Sig +Sig

Rb Rs
-Exc

Rb

Vs =

250 Rb Rs Rs + 0.5 1 + + 2 R7 2 R7 Rb
16

Parallel resistor circuit examples

Rb (ohm) 350 350 350 350 350 350 R7 (ohm) Rs (ohm) 60,000 60,000 60,000 60,000 60,000 30,000 Vs (mV/V) 1.45409 1.44903 1.40499 1.07751 0.72758 1.45198

4
100,000 10,000 1,000 350 350

17

8. How Does Tolerance of Rb and of Rs affect Vs ?

To a close approximation for all of the circuits above with reasonable values,

Vs is proportional to the value of Rb and Vs is inversely proportional to the value of Rs.

18

9. What Errors are Contributed by Extension Cables ?

It often happens that initial shunt calibration data is recorded on a particular bridge and then in subsequent tests an extension cable has been added between the bridge and the instrument with Rs located at the instrument. It is useful to know the resulting error. In the following discussion Rc represents the resistance of one conductor of a cable made of multiple similar conductors.
19

Error contributed by a 4-conductor extension cable

+Exc

Rc

Rb
-Sig

Rc

Rb Rb Rs Rb

Rc
+Sig

Error is due primarily to current flow in Sig lead. Vs error = 500 Rc / Rs (in units of mV/V). Error is always same polarity as Vs.

Rc
-Exc

20

4-Conductor Error Examples

Based on 10 ft cable length
Conductor

Gauge 22 22 28 30 30 30

Rc (ohm) 0.16 0.16 0.65 1.03 1.03 1.03

Rb (ohm) 350 350 350 350 700 700

Rs (ohm) 30,000 60,000 60,000 60,000 60,000 120,000

Nom Vs (mV/V) 2.8998 1.4541 1.4541 1.4541 2.8998 1.4541

Vs Error (mV/V) 0.0027 0.0013 0.0054 0.0086 0.0086 0.0043

Vs Error (%Vs) +0.09 +0.09 +0.37 +0.59 +0.30 +0.30

21

Caution !
If Vs is being converted to physical units, remember that the 4-conductor extension cable changes output of the circuit by the factor
Rb Rb + 2 Rc

If Rc is not accounted for in the conversion, the error compounds the Vs error, doubling it as a % of reading !
22

10. What Error is Contributed by a 6-Conductor Extension Cable ?

-Sig

+Sense

Rc Rc Rb

+Exc

Rc

Rb

Rc
+Sig

Rb Rs Rc
-Sense

Rb

Rc

-Exc

Same error in Vs as for 4-conductor extension cable. No error in physical load output due to Rc in Exc leads. Remote sensing of excitation is the benefit of a
6-conductor cable.
23

11. What Error is Contributed by a 7-Conductor Extension Cable ?

-Sig

+Sense

+Exc

Rc

Rc Rb

Rc

Rb

Rc
+Sig

Rc Rs

Rb Rb Rc

-Sense

Rc

-Exc

Only error is Rc adding to Rs for total shunt =Rs+Rc. Error is negligible for all practical purposes.
24

12. Why are 8-Conductor Cables Sometimes Used ?

To allow shunt connection to either Sig or + Sig and

have negligible error.

7-Conductor and 8-Conductor cables solve the

extension cable error problem.

25

13. What Error in Shunt Cal is Caused by +Exc a Physical Load ?

Rb+r Rb-r

-Sig

+Sig

Rb-r
Rs -Exc

Rb+r

Analyze by assuming a fully active basic bridge circuit. r = change in gage resistance due to strain. Error in Vs is proportional to r / Rb.
26

Error due to Physical Loads, Examples

(Assuming all legs equally active)
Rb (ohm) 350 350 350 350 350 700 Rs (ohm) 60,000 60,000 60,000 60,000 60,000 60,000 Vs without load (mV/V) 1.45409 1.45409 1.45409 1.45409 1.45409 2.89975 Physical Load (mV/V) 0 +1 +2 -2 -1 -1 r (ohm) 0 0.35 0.70 -0.70 -0.35 -0.70 Vs with load (mV/V) 1.45409 1.45264 1.45118 1.45699 1.45554 2.90265 Error (mV/V) 0 -0.00146 -0.00291 +0.00290 +0.00145 +0.00290 % 0 -0.10 -0.20 +0.20 +0.10 +0.10

Generalized Rule: For any Rb and Rs, Error in % = - 0.1 X Physical Load in mV/V
27

14. What is the Effect on Shunt Cal of a Permanent Zero Balance Shift ?

If all legs are equally active and somewhat equally

shifted in resistance, as is often the case with a mild physical overload, error in Vs follows the same analysis as in paragraph 13 above.

It is wise to assume that any significant permanent

zero shift invalidates a prior Shunt Cal.

28

15. What is the Effect on Shunt Cal of Transducer Toggle ?

Toggle is a reversible change in bridge zero resulting

from the most recent loading changing from tension to compression or vice versa. The error in Vs follows the same analysis as in paragraph 13 above. Toggle seldom exceeds 0.01 mV/V even for load excursions as high as +/- 4 mV/V. Therefore error in Vs seldom exceeds 0.001% for any Rb or Rs. The error is normally negligible.
29

16. Is There Reason to Prefer Any Particular Leg of the Bridge to Shunt ?

Probably not. In terms of stability and repeatability, all legs are

contributing equally to a shunt measurement.

R5 and R6 contribute equally to Vs for shunt

measurement on any leg regardless of their values. Remember paragraph 6 !

R 6 R5 250 R6 + R5 Vs 1 1 + Rs 4 Rs R6 + R5 + Rb + 0 .5 Rb
30

17. Can a Value for Vs be Calculated for any Rs, Knowing Only Rin and Rout of the Bridge ?

Only approximately because Rin and Rout only

approximate the value of R1, R2, R3, or R4 in the basic circuit. With tolerances typical of strain gages and bonding processes in transducer production, 0.25% is about the uncertainty of a calculation for Vs with the basic circuit. It gets worse if R5, R6, or R7 are present.
31

18. Where May the Shunt Resistor Be Located ? a. Internal to a transducer or permanently wired to a bridge circuit.

Resistor tolerance not important. Resistor should have low temperature coefficient of
resistance (TCR).

32

b. Internal to an Instrument.

Low resistor tolerance is important unless a bridge and specific instrument are always used together. TCR should be low.

HRBSC

9840

These high end instruments have 0.01% Low TCR Internal Shunt Cal Resistors in two different values. Instruments may be substituted without significant error. 33

9830

9820

These lower cost instruments have 1% tolerance Shunt Cal Resistors. For good Vs measurements, instruments should not be substituted. 34

500

DMA

DCA

These lower cost signal conditioning modules have 1% tolerance Shunt Cal Resistors and a manual switch permanently installed. For good Vs measurements, instruments should not be substituted. 35

c. External resistor, portable.

Substitutability depends on tolerance. Potential for high accuracy. Potential to get lost or mixed.
36

d. External, laboratory instrument.

0.01% tolerance available. Good substitutability.

Special 3-bank decade resistor, tests up to 3 bridges simultaneously. 1 ohm to 1111 Kohm, 0.01 % tolerance. 37

19. What Shunt Cal Repeatability Can Be Expected From Modern Transducers ?

Procedure for a repeatability test performed 100 Klbf Load Cell specimen loaded in compression. 12 test cycles of 4 mV/V hydraulically applied physical load and 1 mV/V Shunt Cal on two bridge legs. Rb = 350 ohm, Rs = 88750 ohm, 20 ppm/C, internal to load cell. Measurements over 3 days. Interface Gold Standard HRBSC instrumentation. Results of test Std Dev of physical load measurement: 0.004%. Std Dev of Shunt Cal: 0.001% pos, 0.001% neg. 38

Shunt Calibration Repeatability Data

Model: 1232BKN-100K S/N: 103014 Bridge A Standards usedSTD-16,BRD4,BRD1 Sequence Date Time Temp F Humidity % Load mode Units LOAD (Klbf): 0 20 40 60 80 100 40 0 Span (mV/'V) Raw zero Raw Neg SCAL Net Neg SCAL % Dev from avg -0.01424 -0.81384 -1.61358 -2.41373 -3.21476 -4.01617 -1.61607 -0.01413 -4.00193 -0.01429 -0.99896 -0.98467 -0.001 -0.01429 -0.81388 -1.61367 -2.41383 -3.21483 -4.01619 -1.61617 -0.01416 -4.00190 -0.01426 -0.99894 -0.98468 0.000 -0.01429 -0.81394 -1.61370 -2.41380 -3.21484 -4.01616 -1.61616 -0.01415 -4.00187 -0.01428 -0.99897 -0.98469 0.001 -0.01432 -0.81402 -1.61382 -2.41390 -3.21486 -4.01620 -1.61619 -0.01418 -4.00188 -0.01429 -0.99898 -0.98469 0.001 -0.01433 -0.81406 -1.61385 -2.41392 -3.21486 -4.01613 -1.61613 -0.01421 -4.00180 -0.01434 -0.99901 -0.98467 -0.001 -0.01436 -0.81403 -1.61385 -2.41390 -3.21483 -4.01609 -1.61603 -0.01421 -4.00173 -0.01434 -0.99904 -0.98470 0.002 -0.01340 -0.81302 -1.61283 -2.41297 -3.21399 -4.01558 -1.61555 -0.01332 -4.00218 -0.01340 -0.99810 -0.98470 0.002 -0.01343 -0.81309 -1.61283 -2.41294 -3.21403 -4.01549 -1.61555 -0.01334 -4.00206 -0.01344 -0.99812 -0.98468 0.000 -0.01333 -0.81303 -1.61280 -2.41284 -3.21392 -4.01537 -1.61542 -0.01329 -4.00204 -0.01346 -0.99814 -0.98468 0.000 -0.01341 -0.81305 -1.61286 -2.41296 -3.21395 -4.01530 -1.61540 -0.01332 -4.00189 -0.01345 -0.99812 -0.98467 -0.001 -0.01331 -0.81293 -1.61271 -2.41271 -3.21365 -4.01510 -1.61520 -0.01327 -4.00179 -0.01339 -0.99807 -0.98468 0.000 -0.01342 -0.81306 -1.61287 -2.41284 -3.21373 -4.01505 -1.61520 -0.01332 -4.00163 -0.01343 -0.99810 -0.98467 -0.001 16:44:11 75 32 COMPR Klbf 1 16:51:29 75 32 COMPR Klbf 2 16:56:25 75 32 COMPR Klbf Internal Shunt Resistor: 88.75 Kohm

"

3 17:01:10 75 32 COMPR Klbf

4 17:11:22 75 32 COMPR Klbf

5 17:17:24 75 32 COMPR Klbf

6 18:03:14 75 32 COMPR Klbf

7 18:08:02 75 32 COMPR Klbf

8 20:33:34 75 32 COMPR Klbf

10 20:37:16 75 32 COMPR Klbf

11 07:28:39 75 32 COMPR Klbf

12 07:33:36 75 32 COMPR Klbf

Std Dev

3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-26-1999, 3-29-1999, 3-29-1999,

SHUNT CAL (mV/V)

0.00014 mV/V -0.004 %

0.00001 mV/V -0.001 %

Raw zero Raw Pos SCAL Net Pos SCAL % Dev from avg

-0.01428 0.97123 0.98551 -0.001

-0.01428 0.97125 0.98553 0.001

-0.01430 0.97123 0.98553 0.001

-0.01432 0.97122 0.98554 0.002

-0.01433 0.97118 0.98551 -0.001

-0.01434 0.97118 0.98552 0.000

-0.01343 0.97210 0.98553 0.001

-0.01344 0.97206 0.98550 -0.002

-0.01346 0.97207 0.98553 0.001

-0.01347 0.97204 0.98551 -0.001

-0.01341 0.97210 0.98551 -0.001

-0.01347 0.97205 0.98552 0.000

0.00001 mV/V 0.001 %

39