Lab no.

2 Electrical measurement methods: bridge and compensation methods

Electrical measurement methods: bridge and compensation methods

1. Introduction
The experimental measurement process includes operations that have to be executed in
order to obtain the measurement result. No matter which kind of measurement is performed,
there are 4 main elements:
- The measurand;
- The measurement method;
- The measuring instrument;
- The standard.
The measurement method establishes the measurement function through which there are
assigned real numbers to each measured quantity, resulting the measured value. The
measuring instrument realizes this correspondence. The standards are instruments that
preserve, conserve and transmit the measurement units.
The classification methods of measurement methods can be done by several criteria, the
most important being:
I. The mode the measurand and the standard are compared by the instrument:
simultaneously or successively. The comparison methods involve the simultaneous presence
of both quantities, Fig. 1.a and the successive comparison involves two steps: the calibration,
fig. 1.b, and the measurement itself, fig. 1.c.

Standard Measurand Standard Measurand

Measuring Measuring Measuring

instrument instrument instrument

Result Calibration Result

a) Simultaneous comparison  b) Successive comparison: b) Successive comparison:

calibration step measurement step

Fig. 1

Simultaneous comparison measurement methods can be divided into 1:1 methods and
1:n methods, taking into account the ratio between the measurand and the standard.
a. Simultaneous 1:1 comparisons are used when the quantities (measurand and
standard) have close values. The comparison is done directly and the instrument indicates
either the equality (null indicator) or the difference of the two quantities, or indirectly and the
instrument is called comparator.
- Direct simultaneous 1:1 comparison is used for measuring active quantities. The
measurement is either a differential or a null one.
The differential method measures the difference between the measurand and the
unknown quantity. The result x can be computed with:
Lab no. 2 Electrical measurement methods: bridge and compensation methods

x = x0 + 
while the error is:
x  x0   x0  
= +  + .
x x x x0 x 
where x0 is the value of the standard and  is the measured difference. The instrumental error
can be neglected if the x and x0 are close to each other.
The null method is a particular case of the previous one (=0) but the instrument is
totally different. This is a null detector with a high sensitivity (S). Its error does not explicitly
appear into the measurement error formula, but rewriting the measurement equation:
x  x0  S
shows the instrument contribution to the final error:
x  x0 S
 +
x x0 x
These methods are between the most precise because a single instrument is used and the
sensitivity is chosen so that the last term is negligible. As disadvantages these methods use a
large number of standards and can be used for a limited number of quantities (1st degree).
- Indirect simultaneous 1:1 comparison methods are simple comparison, substitution and
permutation.
The simple comparison method uses a 1:1 comparator and a differential or null measurement.
The comparator is a bridge with equal arms that can indicate the difference or can be balanced
with a fine adjustment around unity. The comparator error has a complete contribution to the
final error, thus the comparator has to be with high accuracy. This can be a disadvantage.
The substitution method eliminates the comparator construction error by performing a double
measurement: balancing the comparator with the measurand connected at its input and
balancing the comparator with the standard connected at the input. The comparator ratio is not
compulsory to be unitary, but it has to be stable and adjustable in narrow limits. The
measurement is described by:
 m- n
x = (K + m) xt (a) x 0 = (K + n) xt (b) x =  1 +  x0 (c) (m , n » K)
 K 
x  x 0 | m - n | K  x 0
= +2 
x x0 K K x0
The error will be lower when the two quantities will have closer values.
The permutation method is applied just when the ratio is 1. The two quantities are connected
to the inputs and the comparator is balanced and then they are interchanged and the
comparator is rebalanced. The measurement is described by:
 m-n
x = (K + m) x0 (a) x0 = (K + n)x (b) x =  1 +  x0 (c) (m , n »K)
 2K 
x  x 0 | m - n | K  x 0
= + 
x x0 K K x0
The methods above are usually used for measuring with high accuracy of resistance,
capacitance and inductance. The quantities x, x0 and xt must be constant during the
measurement.
b. Simultaneous 1:n comparison applies when the measurand and the standard have
different values. Usually the standards have unique value or discrete values (10n or 1,2,5
x10n). There are two options to achieve this: the addition if the quantity has this property or
the division.
Lab no. 2 Electrical measurement methods: bridge and compensation methods

-The addition method uses many standards with discrete values in order to get a standard with
a value close to the measurand value and to make a 1:1 comparison. The total accuracy of the
standard is calculated based on the individual error of each standard and on the error
composition mode calculated from the measurement equation. Usually, they are laborious
methods and they are applied in metrological labs for comparing the standards.
- The ratiometric method is a null method that compares a quantity with a fraction of the other
quantity obtained with a dividing device. The measured value is calculated based on the
equation x=K*x0 where K is a device parameter that can be varied in large limits. The total
error is
x  x0 K
= +
x x0 K
The most known ratiometric methods are the bridge method and the compensation method.
c) Successive comparison methods are used in all instruments with analogic or
digital display. These instruments realize one or more conversions of the measured quantity in
order to get the useful information for the operator. The calibration information is stored into
the instrument in a memory (mechanic memory for electromechanical instruments or
electronic memory for electronic and digital instruments). Because the stored information
during calibration changes in time, from time to time these instruments have to be verified at
certain time intervals established by legal regulations.
II. The type of the metrological signal: analogic, digital and mixt. The methods can be
divided into analogic methods, digital methods, mixt methods and computerized methods.
Most of the measured quantities have continuous variation in time, but analogic
methods do not lead to high accuracy results. The digital methods, using discrete signal
processing, give better results and facilitate the processing, display and transmission of data.
While the computers were developed (computing speed, minimization, reduced costs, etc)
they became usual components of measuring systems and by applying new computing
methods they dramatically increased the performances.
III. They way the result is obtained: direct methods (measuring instruments) and indirect
methods (measuring facilities or systems). In case of direct measurements the instrument is
sensitive to the measured quantity and it is described by a simple law, usually linear. Indirect
methods need to compute the result using many measured values by direct methods.
IV. The desired accuracy: usual methods (current measurements), industrial (regulation
and control for industrial processes), laboratory (verification, calibration, standardization and
constant determination).
V. The domain in which is performed the measurement: time domain and frequency
domain. Most frequent methods works in the time domain and works for constant and variable
(stationary or random) quantities. Frequency domain methods do not have the same
performances as time domain methods but are necessary in certain applications.

2. Bridge measuring methods

The bridge is a circuit that looks like a square that has the arms as impedances, a
diagonal for power supply and the other for measurement. The indicating instrument is a null
indicator for balanced bridges or a simple indicator for unbalanced bridges. The power supply
is a voltage source (low or high voltage) or current source (DC, AC – audio or RF). The
bridge can measure a resistance (DC bridge) or a capacitance or inductance (AC bridge). They
are used for high accuracy measurements. In order to exemplify the method, the DC bridge is
a suggestive method and there are 3 bridge types:
a) bridges for usual resistance values (1 to 106 Ω) – Wheatstone bridge – the measured
device is dipolar;
Lab no. 2 Electrical measurement methods: bridge and compensation methods

b) bridges for small resistance values (10-6 - 10 Ω)- Thomson bridge - the measured device
is quadripolar;
c) bridges for high resistance values (106 – 1010 Ω)- Megohm bridge - the measured device
is tripolar;
The quadripolar and tripolar connections are needed in order to eliminate the parasitic
series elements or insulation respectively.

2.1.Resistance measurement using the Wheatstone bridge

In the following figure it is shown the DC Wheatstone bridge basic configuration. It can
be used either in balanced or unbalanced mode. The unbalanced type is used for non-electric
quantities that use a resistive transducer.
In order to balance the bridge, the bridge ratio (Ra/Rb) is adjustable in decades (10-3-103)
and the resistor R is finely adjustable (4 decades with 0.1% accuracy). The direct
measurement is done when the bridge is balanced (VB=VD):
V B = V D   U = U BD = V B - V D = 0
Rx = R  Ra = n  R
U BC = U DC  R x = R 10
Ra Rb Rb

Ra A
E
Sa Sb

D
Rb
B
IN
R
RX
C

The balancing process is the following:

- the bridge ratio is switched to 1 (Ra/Rb=1);
- the resistor R is fixed to zero and the switch K is closed is order to power the bridge. If
the null indicator does not change its polarity it means that the bridge ratio is too small and
must be increased with one decade. If the polarity changes without using the most significant
decade of R it means that the bridge ratio is too big and must be decreased with one decade;
- when all the decades of R are used, the sensitivity of the null instrument must be
increased while the balance is done using the least significant decades until the maximum
sensitivity is reached and the minimum indication is obtained.
- the value Rx is computed with last equation.
- the measurement error is estimated using the following equation:
 R x R Ra Rb  R 
= + + + 
Rx R Ra Rb  R sensitivity
 R  
  =
 R sensitivity S
where  is the variation of the null instrument indication and S is the bridge sensitivity.
A Wheatstone bridge is appreciated by several metrological parameters:
Lab no. 2 Electrical measurement methods: bridge and compensation methods

- the measuring interval: it is calculated from the balance equation using the upper and
lower limit for the bridge ratio:
Ra min 1
Rx min = R min  = 0,1  3 = 10 -4 
Rbmax 10
3
Ra max 10
Rx max = R max  = 104  = 107 
Rbmin 1
This interval is limited by the parasitic resistances (contact and insulation).
- the bridge sensitivity: it is the ratio between the variation of the output versus the
relative variation of the input quantity:
 
 d d
S = lim  =  = R  [div]
R 0 R dR dR
R echil. R echil.
It is calculated starting from α = SuU written around the equilibrium.
R Ra dR A dR Ra A
dU BD = E    = 2
E ; A= ; S = Su  E 
R + Rb Rx + Ra R (1 + A ) R Rb (1 + A )2
From the above equation it is observed that the sensitivity is direct proportional with the supply
voltage and the null instrument sensitivity and depends nonlinearly on the bridge ratio (A=
Ra/Rb). The term A/(1+A)2=F is the bridge factor. When A varies from 10-3 to 103 it results a
sensitivity variation of about 1000 like in the following figure.

1000 
A

The bridge factor is maximum when the bridge ratio is 1 and it does not exceed 0.25. For
applications with resistive transducers this is a major problem due to the small variations of
transducer resistance.
- the measuring accuracy: is the limit error containing 2 components – the intrinsic
error and the supplementary error. The intrinsic error is limit error in normal conditions
(20C, 40% RH) and can be computed with:
Rx  
=  c  1+ k  RN 
Rx  Rx 
Lab no. 2 Electrical measurement methods: bridge and compensation methods

where k is a constant given by the producer, RN =10n [] is the maximum value of a resistor
in the measuring interval for a given bridge ratio and c is the bridge accuracy class.
The supplementary error is the error generated by the variation of a single influence
factor (temperature, humidity) outside the reference domain and it is expressed as a percent
from the intrinsic error. While the influence quantities are in the reference domain the error is
the intrinsic error.
Observations:
- the Wheatstone bridge is an example of indirect simultaneous 1:1 comparison (A=1)
or 1:n for A1.
- the measurement is a null method if the bridge is balanced or differential if the bridge
is unbalanced.
The substitution method is used to increase the accuracy of the measurement. The
measurement has 2 stages: during the first stage the unknown resistance is connected at the
input and the bridge is balanced and during the second stage the unknown resistance is
replaced with a decade standard resistor and the bridge is rebalance from the decade resistor.
The measurement errors include two sensitivity errors:
Rx = K  Rt Ro = K  Rt Rx = Ro
Rx Ro Rt Ro
= + +
Rx Ro Rt Ro
where δR/R is the sensitivity error. If they are small enough comparing with the construction
error, they can be neglected and the final error is the best possible for a given standard.
The permutation method is usable when the bridge has the ratio 1 and 2 of its resistors
can be interchanged. The measurement consists of two stages resulting two values for the
balance R1 and R2:
Ra Rb
Rx = R1  Rx = R 2  Rx = R1  R 2
Rb Ra
Rx 1   R1  R 2  R R
=  +  = +
Rx 2  R1 R2  R R

3. Compensation measuring methods

The compensation measuring methods are considered 1:n indirect comparison methods using
ratiometric methods for active quantities (voltage, current). There are two types of
compensations: complete – the null method and incomplete – the differential method. The
complete compensation method uses the so called compensators and the incomplete
compensation method uses differential voltmeters (next figure).

3.1.Complete compensation method

This is a null method. The balance (equality) is achieved between the unknown
quantity (voltage) and a standard voltage generated with a constant current through a standard
Lab no. 2 Electrical measurement methods: bridge and compensation methods

adjustable resistor – constant current (Poggendorf, Feussner, Kelvin-Varley) or with a

adjustable current through a fixed standard resistor (usually quadripolar) – constant resistance
(Lindeck-Rothe).
a) Constant current compensation methods
The Poggendorf compensator is presented in the following figure. There are two separate
circuits, one for calibration and the other for measurement. This has larger errors than the
basic schematic above but allows consecutive measurements without recalibration. The
constant current Ia is chosen as 10-imA (i=0 or 1) adjusted with the resistor Ra. The values r0
and r results according with the voltages Er and Ex. The measurement has 2 stages:
- The calibration stage while comparing the voltage drop over R0 with the standard
voltage Er (switches on position 1)
Er = r o Ia
- The measuring stage while comparing the voltage drop over resistor R with the
unknown voltage Ex (the switches on position 2)
Ex = r  Ia
The result of the measurement is:
r
Ex = Er 
ro
with the error
Ex Er R Ro  U IN  Er 
= + + +  1+ 
Ex Er R Ro Er  Ex 
The last equation shows the construction errors (R/R, R0/R0, Er/Er) and the sensitivity
errors (Uin/Er) generated by the limited resolution of the adjustable elements. The error
increases when the unknown voltage is smaller than the standard voltage. The theory above
accepts that at the equilibrium there is no current flow through the null indicator. Another
advantage is that it allows to measure voltages in large domain. Constructively it uses decade
resistors (4-6 decades) and Weston standard voltage: the nominal voltage is 1V and the
accuracy class is 0.05.
The Feussner compensator (double decades) is coming from the Poggendorf compensator.
As a difference, the resistor R0 has a constant value (10,180k) and the resistor R has double
decades connected in order to keep constant the resistance in the auxiliary circuit. The voltage
drop that compensates the unknown voltage is adjustable with a resolution of 4-7 digits.

In the figure above we can see the existence of adapting adjustments (Ro' and Ro")
and the mode of obtaining the most significant digit and the least significant one. The
measurement has 3 stages:
- R0 is adjusted according with the last significant digits of the standard voltage
Lab no. 2 Electrical measurement methods: bridge and compensation methods

- S2 is switched on position 1 and on proceed to calibrate the compensator by adjusting

the resistor Ra (Ia=0.1mA)
Ea Er
Ia = = = 0,1 mA
Ro + R + Ra 10180 + m  1 + n  0,1
- S2 is switched on position 2 and the compensator is balanced using the resistor R,
starting from the most significant decade to the least one.
Ex = Ia  (A  1000 +  100 + C  10 + D  1 + E  0,1) = 0, ABCDE [V]
This kind of compensator has the advantage that the measured value is direct on decimal
format, but it has also two disadvantages – there is a large number of contacts in series and the
resistor R has double decades that must be adjusted with accuracy. The measurement errors
are:
Ex c  Un 
=  1+ k  
Ex 100  Ex 
where k is the compensator factor (0.1 or 0.01) given by the producer and Un is the nominal
voltage (0.2 or 2 V). The achievable resolution is 10µV and the best accuracy class is 0.01.
The Kelcin-Varley compensator is another solution for obtaining an adjustable voltage using
a constant current. As it is shown in the following figure, there are used successive decades
made up from 11 equal resistors (exception the last that have 10) and in the ratio of 1:5 with
respect to the previous one. Each decade shorts two resistors from the previous. The voltage
drop on each decade is 1:10 compared with the preceding, resulting decade adjustable
voltages with the resolution given by the number of decades. At the equilibrium we have:
Ui Ui Ui Ui
Uo = Ua + Ub + Uc + Ud =  A + 2  B + 3  C + 4  D = Ui  0, ABCD
10 10 10 10
while the error is:
Uo Ui k
= +
Uo Ui k
where A, B,… are the numbers corresponding to the selected position of each decade and k is the
resulting division factor.

These compensators have the advantage that the effect of the contact resistances is
reduced because of the reduced number of contacts and of the smaller current through them,
for the same resolution. The resulting accuracy is maximum 10ppm and the best achievable
resolution is 1µV.
b) Constant resistance compensation methods
These methods use a cuadripolar resistor traveled by an adjustable current in order to generate
the compensation voltage. The auxiliary source is connected at the resistor’s current contacts
and the measuring circuit is connected at the voltage contacts (following figure). The method
is described by:
Lab no. 2 Electrical measurement methods: bridge and compensation methods

 Ex  R  I
Ex = R  Ia = +
Ix R I

Because the resistor R can have high accuracy, the overall accuracy will be given by the
current accuracy. For knowing the current with high accuracy the current must be measured like
in the schematic on the right. The best achievable accuracy is 0.1ppm (the inductive current
compensator) and a resolution of 0.1 µV.
The metrological characteristics of compensators:
- the nominal voltage is the maximum value that is accepted at the inputs without
changing the configuration. It is between 20 – 200mV for Lindeck-Rothe and 0.2-
2V for the other types.
- the resolution is the smallest variation of the compensation voltage that can be
achieved while acting on the adjustment elements. It is between 10-4 – 10-7 for
constant current compensators and between 10-2 – 10-7 for constant resistance ones.
- The accuracy is given by the limit error through the accuracy class. Like in the case
of bridges, the limit error includes the intrinsic error and the supplementary error
(20% to 100% from intrinsic error) when the influence quantities vary outside the
nominal domain. The intrinsic error is:
Ux c  Un 
=  1 + 0 ,1  
Ux 100  Ux 
while the sensitivity errors were neglected.
Lab no. 2 Electrical measurement methods: bridge and compensation methods

Group Name Date

Lab. no. 2 Electrical measurement methods: bridge and compensation

methods
Abstract:
………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
………………………………………………………………
Experimental work

1. Using the Wheatstone bridge, measure 3 resistors having values that require unitary
ratio, subunitary ratio and greater than one ratio. Use the following equation to compute the
error of the measurements:

Rx  
=  c  1 + k  R N 
Rx  Rx 
where c is the bridge accuracy class (marked on the bridge front panel) and k=0.1 for the given
bridge.

Ra
A
E
Sa Sb

D
Rb
B
IN
R
RX
C

Table 1
Crt. Bridge Bridge k c R Rx Scomp. Sexp. ΔRx/Rx Obs.
No. ratio: A factor: F [Ω] [Ω] [%]
1
2
3
Lab no. 2 Electrical measurement methods: bridge and compensation methods

2. For the same resistors apply a method for accuracy increasing and calculate the new
values and the measurement errors.
Nr. Rx ΔRx/Rx sensb. 1 Rsubst ΔRsb/Rsb sensb. 2 Rx ΔRx/Rx Obs.
crt [Ω] [%] [Ω] [%] [Ω] [%]
1
2

3. Use the constant current compensator to measure a voltage source and calculate the
errors:
Ra Ea S1

R' o Ro R"o (10-D) (10-C) (10-B)

10x1 10180 10x0,1 Ax1000 Bx100 Cx10 Dx1 Ex0,1 x1 x10 x100

S 21
1 2

Eo IN Ex
1 2

S 22

Ux  Un 
= c  1 + 0,1   [%]
Ux  Ux 

Table 2
Crt. Compensator type c Un Eref Ia Ex ΔEx/Ex Obs.
No. [V] [V] [mA] [V] [%]
1
2

Conclusions:
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
…………………………………………………………………………………………………
……………………………………………………………………………..
Lab no. 2 Electrical measurement methods: bridge and compensation methods

Wheatstone bridge

DC constant current compensator

OBSERVATION: WRITE ON THE SCHEMATIC DIAGRAMS THE VALUES OF

THE COMPONENTS USED IN THE EXPERIMENTS.