26.

Impedance Measurement Instruments

26.1. Introduction to Impedance Measurements
Impedance (admittance for its reciprocal value) is a fundamental parameter associated with electronic materials, components,
and circuits (these are inclusively called devices in the later sections). When a current flows through a device, the device gives
an opposition to the current flow; the degree of opposition is called resistance if the current is a direct current (dc), and
impedance if it is an alternating current (ac). Ohm’s law governs the relationship between the resistance, the current, and the
resultant voltage drop across the device under test (DUT) for dc cases, and can be mathematically extended to deal with ac
cases. The DUT impedance can, therefore, be expressed by the ratio of the voltage across the DUT to the current flowing
through it. However, impedance is a more fundamental property inherent to a device; it can be accurately determined and
maintained.

The many impedance- or admittance-measuring methods vary in DUT type, cost, accuracy, and test conditions. For many
years, a variety of bridges, most of which were manually operated, were used for medium- to high-accuracy measurements.
Nowadays, however, digital impedance meters (names vary widely), which automatically perform impedance measurements,
have also been widely accepted, owing to their ease of use (less susceptible to human error), relatively high accuracy, and
convenience for systemization.

26.2. Basic Definitions

When a direct current is applied to a DUT, the relationship between parameters is expressed by Ohm’s law:

(26.1)

where the resistance (more specifically, dc resistance), R, is the degree of opposition that the DUT offers to the current, I, and
E is the resultant voltage drop across the DUT.

The same description holds true if an alternating current flows through the DUT. In this case, however, an extended version of
Ohm’s law should be applied because ac voltage or current consists of two independent elements, magnitude and phase.
Alternating current can be conveniently dealt with if voltage, current and impedance are each expressed as complex numbers.
Then, DUT impedance Z is derived with the following equation, where each value is a complex number:

(26.2)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
A complex impedance (Z) can be further expressed either in the rectangular coordinate formZ=R+jX or in the polar form
Z=|Z|e jθ. Similarly, admittance Y(=1/Z) can be expressed as either Y=G+jB or Y=|Y|e jθ= | 1≠ Z≠e −jθ . These expressions are
summarized in Table 26.1 and illustrated in Fig. 26.1. In the figure, an impedance and its associated admittance are shown
together in the same complex plane. The unit for impedance, resistance, and reactance is the ohm (Ω), and the unit for
admittance, conductance, and susceptance is the siemens (S). Generally, impedance is convenient for calculations when
devices are connected in series, whereas admittance is convenient when they are connected in parallel. If alternating current I
with frequency f is applied to an inductor of inductance L (a pure or lossless inductor is assumed here), its degree of
opposition to the current is proportional to L and to the angular frequency ω(=2πf) of the current. The phase of the resultant
voltage (E) will lead by 90° (or π/2 radian) relative to I. This can be denoted as follows because a 90° phase lead is
mathematically expressed by multiplying by the imaginary unit j:

Table 26.1 Definitions of Impedance and Admittance

Figure 26.1 An impedance vector and its associated admittance vector on the complex plane.

(26.3)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
If E in Eq. 26.2 is substituted by Eq. 26.3, the impedance Z of the inductor becomes:

(26.4)

By comparing Eq. 26.4 with the general form Z=R+jX, it can be seen that ωL is the reactance given by inductance L at angular
frequency ω. The reactance provided by an inductance is called the inductive reactance, X L.

Likewise, a pure capacitance (C) provides the capacitive reactance (X c) of 1/ωC, and the phase of the voltage across the
capacitor lags by 90°, relative to the applied current. Thus, impedance Z of the capacitor becomes:

(26.5)

The quality factor (Q) is defined for a device of impedance Z, as follows. Here,

a resistance element is involved, in addition to a reactance element. That is, the device impedance is no longer pure:

(26.6)

Because a reactance element stores energy, but a resistance element dissipates it, a large Q of a device implies it can store
energy with little loss. Also, a large Q for an inductor or capacitor implies that the degree of its “purity” is high.

The dissipation factor (D) is also defined. It is the reciprocal value of Q; thus, it indicates the size of the energy loss, when
compared with the stored energy in a device:

(26.7)

A similar description can be applied to admittance, and the related expressions are listed in Table 26.2.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Table 26.2 Impedance, Admittance, and Their Related Definitions for RL and RC Combinations

26.3. Characteristics of Electronic Components

26.3.1. Equivalent circuits
No actual device consists of a single element of inductance, capacitance, or resistance in a strict sense. Even an air dielectric
capacitor, which is usually considered to be a very pure capacitor, has the lead inductance, and losses caused by the lead
resistance, electrode surface finish, contamination, eddy current, etc. For actual applications, however, it can be treated as a
pure capacitor if its capacitance value is less than a few hundred pF and the operating frequency is less than several kHz.

For quantitative evaluation of a device, it is convenient if its characteristics can be approximated by a combination of ideal
elements near the operating condition concerned. The resultant combination is called an equivalent circuit. Such an equivalent
circuit does not necessarily reveal the physical structure of an actual device. For the air capacitor example covered earlier, its
equivalent circuit might be a pure capacitor in the low-frequency region, and even a series combination of a resistor and an
inductor at a few hundred MHz, above the series resonant frequency of the capacitor. In general, a complex impedance at a
given frequency is expressed by two quantities, a real part and an imaginary part, and can be synthesized by a combination of
a resistance (or conductance) and either an inductance or a capacitance. For example, if a device has an impedance of
1000+j1000 Ω at 1 kHz, its equivalent circuit can be that of eitherFig. 26.2A or 26.2B, and the two circuits exhibit exactly the
same characteristics at this frequency. If the device, however, provides an impedance of 1000+j2000 Ω at 2 kHz, it can be said
that the equivalent circuit better represents the actual characteristics of the device. If the configuration of an equivalent
circuit is close enough to the actual structure of a device (often with many elements), it can exhibit a good match with actual
characteristics of the device. In reality, however, equivalent circuits with only a few elements are frequently used for simplicity
and convenience reasons.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.2 Series (a) and parallel equivalent circuits of an inductor (b).

Table 26.3 Parameter Conversion between Series and Parallel Equivalent Circuits

For equivalent circuits with two elements, parameter conversion equations are listed in Table 26.3.

26.3.2. Parameter dependence

With actual devices, their inductance, capacitance, and resistance elements are subject to change, depending on operating
conditions, such as frequency, temperature, applied voltage and current, and time passage, for example. Therefore, it is
important to remember that a test result of a device for a given test condition might not hold true at another condition. In other
words, it is important to measure the characteristics of a device with all parameters as close as possible to the actual
operating condition.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.3 An equivalent circuit of a ceramic capacitor.

Figure 26.3 illustrates an equivalent circuit of a ceramic dielectric capacitor with major parasitic elements included, where,C
0 =nominal capacitance at low frequencies, L s and R s=series inductance and resistance mainly caused by leads and
electrodes, R i=insulation resistance, R d=dielectric loss of ceramic material, and C a and R a=dielectric absorption capacitance
and resistance.[*] Owing to the series inductance, the equivalent series capacitance becomes larger than the nominal
capacitance value (C 0) near the series resonance frequency. The dielectric material is frequency-dependent, too, and its loss
becomes larger as the operating frequency increases. Usually, a material with a high dielectric constant, useful to make a
capacitor size smaller, has a large dependence on temperature (i.e., its capacitance value varies with change in ambient
temperature), and its temperature coefficient might not be constant over temperature. It usually has a large dependence on
the applied signal level, too.

A similar description can be applied to an inductor with a magnetic material core, shown in Fig. 26.4, where, L 0=nominal
inductance at low frequencies, R s=series resistance, C i and R i=parasitic capacitance and associated dielectric loss, and R
p=hysteresis and eddy current loss for core material.

Figure 26.4 An equivalent circuit of an inductor with a core.

Because an inductor is usually realized by several layers of wire windings around a magnetic core material, parasitic
capacitance between layers and windings tends to be introduced. Frequency dependence of an inductor is mainly determined
by the parasitic capacitance and the magnetic material property. Wire resistance, skin effect, and losses of magnetic material
are the major sources of the loss of an inductor, and each of them has frequency and temperature dependence, too. The
property of the magnetic material is also sensitive to the magnetic field strength. Therefore, the inductance of an inductor
with a magnetic core varies, depending on the applied signal level, which is converted into the magnetic field strength.

26.4. Impedance Measuring Techniques

A wide variety of techniques are used to measure the impedance of a device, and a suitable technique should be chosen.
Although it is impossible to list all the techniques currently available, the representative techniques often used are explained
here. Each technique also has variations, depending on actual applications.

26.4.1. The voltmeter-ammeter method

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
26.4.1. The voltmeter-ammeter method
Figure 26.5 shows the basic connection for measuring the dc resistance of a device with the voltmeter-ammeter method. The
resistance can be derived by directly applying Ohm’s law:

(26.8)

where E (the reading of the voltmeter, V) is the voltage across the device and I (the reading of the ammeter, A) is the current
flowing through the device.

The dc supply (V s) can be either a voltage or a current source. If an ac supply and appropriate ac voltage and current meters
are used, the magnitude of the device impedance can be obtained. Although this method is easy to understand and to
implement, errors associated with the meters (nonlinearity, internal resistance, etc.) directly affect the measurement results.
Therefore, this is mainly used for low- to medium-accuracy applications (no better than 1 percent).

Figure 26.5 A dc resistance measurement with the voltmeter-ammeter method.

Figure 26.6 The Wheatstone bridge.

26.4.2. The bridge method

A variety of bridges have been implemented depending on the applications. The word bridge was originally used for a null
detector bridging the two points of balancing arms. But it is now widely used to indicate impedance-measuring circuits
employing null detection and balancing arms. Generally, a bridge can yield relatively good measurement accuracy because it
is primarily determined by the stable passive components forming the balancing arms. Figure 26.6 is the Wheatstone bridge
for dc resistance measurement with a dc voltage source, V s. Assuming that E=0 (the reading of null detector D is 0),
resistance R x is derived as:

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 (26.9)

Neither V s nor E explicitly appears in this equation; that is, neither the variation of the power supply nor the nonlinearity of the
voltmeter becomes the primary source of the measurement error. The same description can be applied to the bridges used for
ac impedance measurements. For ac bridges, the balancing arms always consist of one or more reactance elements, in
addition to resistance elements.

26.4.3. The resonance method

A combination of inductance and capacitance resonates at a certain frequency. If one of the values of the combination and
the resonance frequency are known, the other value can be calculated. This is the principle of the resonance method. The Q
meter is a direct implementation of this method, and Fig. 26.7 gives its basic circuit diagram. In addition to the manually
operated Q meter, the automatic Q meter, with which the tuning process is automatically performed is available.

Figure 26.7 The principle of the Q meter.

The Q meter typically has a built-in variable capacitor, and an inductor is externally connected. The resonance frequency(f) of
the combination circuit can be calculated as:

So, if f and C j are known, L x is calculated as:

(26.10)

The Q value of the inductor is also obtained as:

(26.11)

26.4.4. The automatic bridge method

Many bridges, such as the Wheatstone bridge described in Section 26.4.2, are manually operated (i.e., iterative balancing
manipulation is needed to reach the final null, balance, condition). In recent years, however, many automatic bridges have
been developed to eliminate the time-consuming manual operation and associated measurement errors.

Figure 26.8 is an example of the automatic impedance bridge. In this block diagram, the reference resistor (R r) is connected

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
to the null point (P) and the residual voltage there is amplified to be fed back to the other end of the resistor to establish the
null condition. Once the null condition is achieved, the DUT impedance (Z x) is calculated as follows:

Figure 26.8 An example of the automatic impedance bridge.

Then:

(26.12)

One or more microprocessors are used to control the internal functional blocks and converting the measured data into the
desired final outputs. Modern automatic bridges are also equipped with communication paths with external devices, such as a
computer and a device handler for systemization.

26.4.5. Specifications of impedance measuring instruments

Fundamental items found in the specifications of impedance measuring instruments are listed here.

Impedance range. The impedance range specifies the limits of DUT impedance, within which an instrument can make
measurements. The range of R, C, L, and other parameters is often specified as well. Measurement performance of an
instrument is usually degraded near the boundary of its specified range.

Measurement accuracy. The measurement accuracy specifies how close the measured impedance value is to the true value.
To express the measurement accuracy, the uncertainty of the measured value is usually indicated in the basic form of percent
of measured value plus offset value. The percent term becomes significant if the measured value is near the full scale of a
measuring instrument; the offset term becomes significant if it is close to zero. Because the accuracy of an instrument is
determined by many parameters, such as DUT impedance itself, test frequency, and test signal level, it is sometimes
expressed with a complicated formula, which is the function of these parameters.

Test signal level. This specifies the test signal level applied to a DUT at measurement. This specification becomes important
for DUTs, which have impedance that varies with signal level. Also, in some situations the test signal level is specified by the
industry standard for commercial transaction purposes.

Measurement time. The measurement time for an automatic impedance measuring instrument is a period of time for which a
DUT is being measured. This specification is important at the production line environment of component manufacturers, for
example, where the test throughput is of primary concern.

A simple measurement time specification often specifies only the period of each consecutive measurement with a DUT
continuously connected to the instrument. The measurement time, however, becomes longer because of additional

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
rebalancing (settling) time if the DUT is disconnected from the instrument each time after measurement. This is usually the
case at the production line, and the measurement time specification should, therefore, be carefully examined for actual
applications.

Operating temperature range. The operating temperature range is the range of ambient temperature within which an
instrument can be used. Because every instrument has temperature dependence, its accuracy is somewhat degraded if the
operating temperature is different from its specified (calibration) temperature, usually 23°C. The degree of accuracy
degradation is usually specified, too.

26.5. Connection and Guarding

26.5.1. Multiterminal connections
The two-terminal connection. As covered in the previous sections, impedance is usually defined between two terminals of a
device. Therefore, it is natural to connect it to an impedance measuring instrument with two test leads, as shown in Fig. 26.9.
Although this connection is simple to implement and easy to understand, it has some potential measurement errors. First, a
stray or residual impedance (Z s) might exist between the two test leads. If a stray impedance exists in parallel with a device,
the measured impedance becomes their parallel combination value, which is not the desired result. Second, a contact
resistance (R c) appears between the device terminal and test lead at each of the contact points, and it is placed in series with
the device. The contact resistance varies from 10 Ω to as large as several ohms, and is unstable. As a result, it introduces a
significant amount of error if the device impedance is relatively small, such as less than several hundred ohms.

Figure 26.9 The two-terminal connection.

The three-terminal connection. To solve the first problem, the three-terminal connection is used. The third terminal and its
associated structure, called the guard terminal, is used to prevent the stray impedance from causing error (Fig. 26.10). The
stray impedance is now placed in parallel with the signal source (Z a) and the ammeter (Z b). The leakage current caused by Z a
does not flow through the ammeter and the leakage current caused by Z b becomes small because the internal impedance of
the ammeter is small and the resultant voltage drop across it becomes small, too. Thus, only the current flowing through the
DUT is measured by the ammeter. Usually, the stray impedance is either capacitive, resistive, or a combination of both, and
such a guard terminal is effective when DUT impedance is relatively high, above several kΩ.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.10 The three-terminal connection.

Figure 26.11 The four-terminal connection.

The four-terminal connection. To deal with the second problem, namely, to eliminate the measurement ambiguity caused by
the contact resistance, the four-terminal connection is used. The current supply leads and the voltage sensing leads are
independently provided, as shown in Fig. 26.11. With this connection, DUT impedance is defined across the voltage sensing
points (A-B) because of no voltage drop through the voltage-sensing leads, and the effect of the contact resistance can be
completely eliminated. Note that the input impedance of the voltmeter should be much larger than the device impedance to
be measured so that it can be neglected.

The five-terminal connection. The three-terminal and the four-terminal connections can be combined, resulting in the five-
terminal connection. With this connection, benefits from both connections can be obtained with a drawback of having a
somewhat-complicated configuration.

26.5.2. The four-terminal-pair connection

When an alternating current flows through a conductor, ac magnetic flux is generated around it. An ac voltage is induced at
another conductor if the magnetic flux is coupling with it. With the four-terminal connection covered in Section 26.5.1, there is
a possibility of an error voltage being induced at the voltage-sensing leads with this mechanism (Fig. 26.12). The higher the
test frequency and the larger the current in the test leads, the larger the error voltage that tends to be introduced. One solution
for this undesirable voltage is to twist the two current supply leads together and also do the same for the voltage-sensing
leads. If the current supply leads are twisted together, the current flowing in each lead is of opposite direction and the
generated magnetic flux by each current cancels out. Also, each of the generated error voltages at the twisted voltage-sensing
leads has opposite polarity and again

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.12 Error voltage induction with magnetic coupling.

Figure 26.13 The four-terminal-pair connection.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.14 Examples of lower-frequency test fixtures: Leaded device test fixture (a), SMD/chip
device fixture (b), SMD/chip device tweezer fixture (c), dielectric material test fixture (d).

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.14 (Continued)

cancels out. Twisting the test leads, however, is not always practical to implement and the degree of cancellation becomes
worse as the test frequency becomes higher. In such a situation, the four-terminal-pair connection provides a better solution.

Figure 26.13 illustrates the basic configuration of the four-terminal-pair connection, where coaxial cable is used for the test
leads and the outer conductor of each test lead is tied together at the DUT end. With this structure, the current flowing through
the inner conductor returns via the outer conductor of the same lead, and the magnetic flux in each conductor is canceled out
and not radiated outside the coaxial structure. Thus, unwanted error voltage is not introduced. Needless to say, the four-
terminal nature of this configuration prevents the contact resistance from causing error. Another benefit of this structure is a
smaller stray inductance introduced by the test leads in series with the DUT because it is suppressed by the mutual magnetic
coupling between the inner and outer conductors. Therefore, this connection is preferably used with modern impedance-
measuring instruments.

26.5.3. Test fixtures

A test fixture is a crucial element for impedance measurement to interface a device to an instrument. Having easy mounting of
a device to a test fixture often conflicts with minimizing the amount of induced error. So, various types of test fixtures have
been designed, depending on applications. A variety of electrode shapes are also available for accepting various devices, such
as axial-lead, radial-lead, and surface-mount devices. A test fixture to measure the volume and surface resistivity of materials
is available, too. It is advisable to select an appropriate test fixture for each application with the supplier’s consultation. Some
examples of test fixture are shown in Fig. 26.14.

26.6. Accuracy Considerations

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
26.6. Accuracy Considerations
26.6.1. Sources of errors
If an impedance measuring instrument is very strictly examined, it might have potential sources of errors, such as dependence
on temperature, humidity, or line voltage; internal system noise; and nonlinearity with test signal level and DUT impedance.
Even if such errors that are intrinsic to an instrument are small enough for actual measurement applications, other error
sources related to DUT connection still exist and care must be taken to eliminate them.

When a test fixture is used with an instrument, it usually has a certain amount of stray impedance or admittance—even without
a device inserted. Its form might be stray capacitance and conductance in parallel with the DUT, and/or stray inductance and
resistance in series. Such stray components become on offset error and can be mathematically subtracted from the
measurement results on which they are superimposed. Most instruments are equipped with stray impedance compensation
(cancellation) capability.

Another error source is associated with test-lead extension at ac measurements. When an ac signal travels through a test
lead, its magnitude and phase vary as transmission-line theory indicates. Errors caused by this effect become larger as the
lead length becomes longer and the test frequency becomes higher. Compensation of this type of error is more difficult than
compensation for stray impedance because it is mathematically equivalent to multiplying the measured values by complex
compensation coefficients. Again, some modern instruments have built-in compensation capability for extended test leads.

26.6.2. Compensation and calibration

As shown in the previous section, compensation usually indicates the means to eliminate error sources given by the external
conditions. On the other hand, calibration usually indicates the procedure used to adjust an instrument using appropriate
impedance standards so that it can measure devices with specified accuracy. Therefore, compensation and calibration are
sometimes dependent. However, the sources of measurement errors should be always clearly identified and each of them
must be appropriately implemented to obtain reliable measurement results.

26.6.3. Impedance standards and traceability

Many impedance standards are available to calibrate an impedance-measuring instrument. Their primary parameter is
resistance, capacitance, or inductance, and sometimes a secondary parameter (such as D, Q, or time constant is indicated
together). “Open” and “short” standards are often used in a calibration process, too. Although the resistance standard
(including open and short) is the only possible standard for dc resistance calibration, capacitance and inductance standards
are also used for ac impedance calibration. Because stable and low-loss inductors are difficult to realize, capacitance
standards are preferred for precision applications.

Because an impedance-measuring instrument is calibrated with respect to the values of impedance standards, it is essential
to provide a reliable value to a standard. This is achieved with the concept of traceability: A value given to a standard can be
traced back to upper-level standards, and eventually to the national standards. It is apparent that a lower-level standard has
larger uncertainty. When an impedance value cannot be directly traced back to the same type of impedance standard, it can be
substituted by a combination of other traceable quantities and conversion theory. One example is a coaxial line with air
dielectric material. A combination of physical dimension measurements and microwave transmission-line theory allows a
coaxial line to be used as an impedance standard.

26.7. Impedance Measuring Instruments

To become more familiar with the popular impedance-measuring instruments, internal structure and features of typical

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
instruments will be discussed. Although the guard structure is almost always provided with an impedance-measuring
instrument, it is not explicitly shown, unless otherwise indicated to simply illustrate the basic operation principle of each
instrument.

26.7.1. Multimeters
Dc resistance-measuring capability is usually provided as one of the functions of a multimeter; a variety of which are currently
available, ranging from simple analog circuit testers to precision digital multimeters. Although the dc resistance measurement
circuits in a multimeter have some variation, they are all essentially based on the voltmeter-ammeter method. Figure 26.15A
shows a circuit used in an analog circuit tester with measurement accuracy of 3 percent of the meter scale, and Fig. 26.15B
shows a dc resistance measurement block diagram for a digital multimeter with a basic accuracy of 50 ppm of the reading.
Note that some digital multimeters have the ability to select between the two-terminal and the four-terminal connection.

26.7.2. The universal bridge

An impedance-measuring tool containing several types of manual bridges, sharing the key bridge elements, is available and is
sometimes called the universal bridge. Typically, it consists of resistance-, capacitance-, and induc-

Figure 26.15 Dc resistance-measurement circuits: A part of an analog multimeter (a) and a part of a
digital multimeter (b).

Figure 26.16 A Wheatstone bridge implementation.

tance-measuring bridges, one of which can be chosen and configured, depending on the DUT type.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
For a resistance measurement, the Wheatstone bridge configuration is used and its operation principle is explained in Section
26.4.2. Strictly speaking, because it consists of only resistors for the balancing arms, it can only be used for dc resistance
measurement. However, if the measurement frequency is low enough, such as less than a few kHz, the reactance elements
associated with the circuit are negligibly small, so the ac resistance of a DUT can be measured, too. This is convenient in such
cases when electrochemical reaction occurs if a dc voltage is applied to a device, or when a thermal electromotive force
prevents the bridge from making accurate dc measurements.

To expand the measurement range, the Wheatstone bridge in actual use is modified from its original form, as shown in Fig.
26.16, for example. In this configuration, R b is a step-variable resistor whose value (reading of the step dial) is in proportion to
the value of R x, and R c is a range multiplier to extend the measurement range (four decades, in this case). This modification
is also applicable to the capacitance- and inductance-measuring bridge configuration.

For capacitance measurement, two types of capacitance bridges are provided and one of them is chosen to measure either
the equivalent series capacitance or equivalent parallel capacitance. They are called the series and parallel resistance bridges,
shown in Fig. 26.17. If the dissipation factor (D) of a capacitor under test is small enough (less than 0.03, for example), the
difference between the capacitance values, depending on the choice of equivalent circuit is small, but becomes significant as
D increases (see Table 26.3). Therefore, the selection of bridge circuit, based on some insight about the device structure,
becomes important for large-loss capacitor measurements.

Figure 26.17 Capacitance-measuring circuits for the universal bridge: The series resistance bridge
(a) and the parallel resistance bridge (b).

The balance condition for the series resistance bridge is:

(26.13)

(26.14)

(26.15)

For the parallel resistance bridge, the same equations can be used for C x and R x, but the dissipation factor is:

(26.16)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
To measure inductance, the Maxwell-Wien bridge and the Hay bridge configurations are used. The Maxwell-Wien bridge is
chosen for an equivalent series inductance measurement and the Hay bridge for an equivalent parallel inductance
measurement. The basic circuits of these bridges are shown in Fig. 26.18.

The balance condition of the Maxwell-Wien bridge is:

(26.17)

Figure 26.18 Inductance-measuring circuits for the universal bridge: The Maxwell-Wien bridge (a)
and the Hay bridge (b).

(26.18)

(26.19)

For the Hay bridge, the same equations can be used for L x and R x, but the Q factor is:

(26.20)

An example of a commercially available universal bridge with an internal 1-kHz signal source has a basic measuring accuracy
of 0.1 percent of the reading with four significant figures.

26.7.3. The Kelvin bridge

The Kelvin bridge is a four-terminal connection bridge that is frequently used to measure dc resistances as low as 100μΩ with
accuracy of better than 1 percent. Its circuit schematic diagram is shown in Fig. 26.19. The Kelvin bridge is usually designed in
such a way that R a and R u, and R b and Rv are ganged (i.e., R b/R a≈R v/R u). If each contact and lead resistance is small
enough, compared with each resistor value, the balancing condition becomes:

(26.21)

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.19 The Kelvin bridge.

If R w is reasonably small, the second term of Eq. 26.23 becomes small, compared with the first term—even if the error of
ganging is in the order of a few percent. Then it can be simplified to:

(26.22)

Because this bridge has the second ratio arms, R u and R v, it is sometimes called the Kelvin’s double bridge.

26.7.4. Milliohmmeters
A milliohmmeter with the four-terminal connection is designed so that a small resistance can be easily measured in various
applications. Examples of a small resistance include the contact resistance of a connector and a switch, the internal
resistance of a battery, and the series resistance of a fuse. Because the contact resistance of a delicate contact sometimes
has dependence on the test-signal level, some milliohmmeters are equipped with a test-signal limiting circuit. Without this
circuit, a large signal might appear across the test leads and “clean up” the contact contamination of a delicate device upon
connection, resulting in a measured value of the contact resistance that is smaller than it should be. Both dc and ac
milliohmmeters are commercially available. A dc meter is less expensive, owing to its simpler structure, but an ac meter can
make measurements under the presence of a dc error signal in such cases as a battery measurement and a contact resistance
measurement with thermal electromotive force presence.

Figure 26.20 is an example of the ac digital milliohmmeter. A phase-sensitive detector (PSD) is used to extract only the
resistance element of a DUT—even if a large reactance element exists. The series resistance measurement of transformer
windings is an example. A constant-current ac source is pro vided; thus, the output voltage from the PSD is proportional to the
resistance element (R x) of the DUT. With the test-signal limiting circuit, the signal voltage across the test leads is limited at
the level of not more than 20 mV. Such a meter can offer a basic measurement accuracy of 0.4 percent of the reading.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.20 The ac milliohmmeter.

26.7.5. High-resistance meters

An example of the high-resistance meter is shown in Fig. 26.21. A high-resistance meter typically consists of a high-voltage dc
source and a current-voltage converter with the usual guard structure. Also, an electrode contact check mechanism is usually
incorporated to distinguish a high DUT resistance from open or insufficient contact. Because the current to be detected is very
small (sometimes in the order of nanoamperes or less) and noisy, enhanced noise-rejection capability is provided inside of the
meter to obtain stable readings at the display. In this example, the ammeter section is completely isolated using a separate
power supply, and the measured data and the control signal are transferred through an optical fiber cable between the
ammeter section and the digital section. A typical high-resistance meter can offer a basic measurement accuracy of 0.6
percent of the reading up to 10 16 Ω.

Some meters also have a timer function, with which the resistance of a device can be measured at a specified time after
applying a test voltage to a DUT. This is convenient because many high-resistance and insulation materials have a large time
constant and absorption current, and measured resistance values have time dependence.

Because a high dc test voltage is applied to a DUT, safety considerations should be given to the measurement system (high-
resistance meter and test fixture). Provision of some interlock mechanism is advisable so that the operator can be kept from
touching places where high voltage is exposed during operation.

Figure 26.21 The high-resistance meter with contact check function.

26.7.6. Transformer-ratio-arm bridges

With a well-constructed transformer, an ac voltage or current ratio can be precisely determined by the ratio of wire windings;

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
thus, accurate DUT values can be read by counting the number of windings of the transformers. The transformer-ratio-arm
bridge implementations have many variations. An example is shown in Fig. 26.22, where the combination of the transformer
taps, several internal reference capacitors, and resistors are used to cover a wide range of DUT values. Because a transformer
cannot be considered ideal at higher frequencies, a typical transformer-ratio-arm bridge accuracy is specified at 1 kHz, and
error terms in proportion to the measurement frequency squared are added above 1 kHz. A typical transformer-ratio-arm
capacitance bridge has basic measurement accuracy of 0.01 percent of the reading, which is mainly limited by the uncertainty
of the internal standard capacitors.

A microprocessor-controlled transformer-ratio-arm capacitance bridge is available for automated applications. Its test
frequency can be chosen from 50 Hz up to 20 kHz and automatic balancing is realized with all of the previously mentioned
features preserved. Its basic accuracy is specified as 5 ppm of the reading at 1 kHz after calibration.

Figure 26.22 A transformer-ratio-arm bridge.

Because of their inherent high precision and long-term stability, transformer-ratio-arm bridges are mostly used in the standard
lab environments.

26.7.7. Capacitance meters

In addition to the transformer-ratio-arm bridges previously described, many digital capacitance meters are commercially
available. Figure 26.23 is an example of 1-MHz digital capacitance meter. In this meter, 1:1 ratio transformers are used to
implement an octave ranging sequence, instead of the popular decade-ranging sequence to realize good linearity and to
guarantee long-term stability. Also, the imaginary component signal-canceling circuit is provided to utilize the full dynamic
range of the analog-to-digital converters (ADC). This meter can offer a basic measurement accuracy of 0.05 percent of the
reading after performing the proper calibration procedure.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.23 The 1-MHz digital capacitance meter.

Figure 26.24 The high-capacitance meter with ALC.

Another example of the digital capacitance meter is shown in Fig. 26.24. It is designed to measure relatively high values of
capacitors. With a high-capacitance DUT, the measured impedance becomes small and the signal level across it, therefore,
tends to become small. Because the industry standard often requires a certain level of test signal across DUT (even under
such a condition), 1 V for 10 μF at 1 kHz, for example, a powerful test-signal source and anautomatic signal-level control
circuit (ALC) are provided. Its accuracy is specified as 0.3 percent of the reading at a typical measurement condition.

26.7.8. LCR meters

Many digital LCR meters are commercially available these days. They usually measure DUT impedance, and the internal
microprocessor converts the measured impedance values into user-specified parameter values such as L, C, R, D, and Q.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
Unlike the dedicated meters, such as the digital capacitance meter covered in the previous section, a typical digital LCR meter
can cover a variety of DUT types with somewhat moderate measurement impedance range. A typical digital LCR meter front
panel is shown in Fig. 26.25. A test fixture is attached to the LCR meter via the four BNC connectors located at the lower left
area of the front panel.

Figure 26.25 Typical digital LCR meter front panel.

A schematic diagram example of inductance measurement with a digital LCR meter is shown in Fig. 26.26. It can supply a
constant current/voltage ac test signal of a specified level to the inductor under test. It can also supply a dc bias current up to
40 A with the bias current source unit. Because the permeability of a magnetic core material varies according to the applied
magnetic flux density, it is necessary to evaluate the inductor near the actual operating condition. When an inductor is used as
a choke inductor in a switching power supply, for example, its inductance significantly decreases under a large signal level and
even saturates (inductance becomes almost zero) in extreme cases.

Recently, many surface-mount electronic devices (SMDs) are popularly used to reduce the size of electronic products. As the
actual operating frequency of such devices becomes higher, with the cellular phones and BS tuners, for example, it is
necessary to evaluate them at high frequencies. A commercially available digital LCR meter can measure devices up to 1 GHz.
To make measurements at high frequencies, a proper test fixture should be chosen and an appropriate calibration procedure
should be performed before connecting a DUT to minimize the measurement uncertainly introduced by the unwanted parasitic
elements associated with the test fixture.

Because it is difficult to obtain accurate inductance standards, stable resistors and capacitors are usually used as the
standards for calibrating the digital LCR meter. Capacitance is treated as negative inductance and its polarity is converted to
positive by the microprocessor in the inductance meter to guarantee the accuracy—even for inductance measurements. Thus,
a typical 1-MHz digital LCR meter can offer a basic measurement accuracy of 0.05 percent of the reading, whereas the 1-GHz
LCR meter provides a basic measurement accuracy of 1 percent of the reading after proper calibration.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.26 Inductance measurement with the digital LCR meter and the dc bias current source.

26.7.9. Impedance analyzers

In some cases, the impedance value of a device is more important than itsL, C, or R value, and the impedance variation over
frequency or other parameters is of primary concern. Examples include evaluation of a bypass capacitor, an electromagnetic
interference (EMI) choke inductor, and the input-output impedance of a filter. A dielectric material evaluation over frequency
and temperature variation is extremely important to design a stable dielectric resonator and filter. An impedance analyzer is
usually equipped with a sweep function of frequency and other parameters, so the impedance of a device can be evaluated
over a specified range of a parameter.

An example of the impedance analyzer covers the test-frequency range from 100 Hz to 40 MHz with the four-terminal-pair
connection and up to 100

Figure 26.27 The 1.8-GHz impedance analyzer with two test head configurations for wide range of
DUT impedance.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 Figure 26.28 An example of impedance analyzer front panel.

Figure 26.29 Examples of high-frequency test fixtures: SMD test fixture (A) and dielectric material
test fixture (B).

MHz with an impedance probe; the basic measurement accuracy is 0.17 percent of the reading with the former configuration
and 1.5 percent with the latter. Its measurement principle is based on the automatic bridge method described in Section
26.4.4. As an example of device analysis, the variation of magnitude and/or phase of the device impedance relative to the
test-frequency change can be graphically displayed. The internal microprocessor allows the impedance values to be
converted into L, C, R, and other values, too. For simple devices (such as a capacitor, an inductor, or a resonator), the
equivalent circuit analysis function provides an estimation of the device’s equivalent circuit element values, based on the
frequency characteristics. This function is useful to evaluate the parasitic elements associated with a device.

Figure 26.27 illustrates another example of the impedance analyzer up to 1.8 GHz with basic accuracy of 0.8 percent of the
reading. Two types of test heads are provided and appropriately selected, depending on DUT impedance. Dc bias can be
supplied to the device via the test head. The analyzer has the capability of controlling an external oven for evaluation of DUT
material property over temperature variation up to 200°C. The analysis results are graphically displayed on the screen, as
shown in Fig. 26.28.

Figure 26.29 shows two examples of test fixtures used with the analyzer to evaluate devices at high frequencies. At high
frequencies, such as 1.8 GHz, a significant amount of error caused by inappropriate DUT connection would be introduced, and
proper use of a test fixture with consultation by its supplier is strongly recommended.

26.7.9. Bibliography

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.
 1. 1. HP 42851A Precision Q Adaptor Operation Manual (Second Edition), Hewlett-Packard Co., December 1995.
2. 2. Cutkosky, R.D., “Techniques for Comparing Four-Terminal-Pair Admittance Standards,” Journal of Research of the
N.B.S., Vol. 74C, Nos. 3 and 4. July–December 1970.
3. 3. HP 16380A Standard Air Capacitor Set Operation and Service Manual, Hewlett-Packard Co., 1990.
4. 4. HP 4338B Milliohm Meter Operation Manual (Third Edition), Hewlett-Packard Co., June 1998.
5. 5. HP 4339B High Resistance Meter Operation Manual (Third Edition), Hewlett-Packard Co., June 1998.
6. 6. Type 1615 Capacitance Bridge-Type 1620 Capacitance Measuring Assembly, QuadTech Inc., December 1992.
7. 7. Model 2700A Multi-frequency Ultra-precision Capacitance/Loss Bridge Preliminary Specifications, Andeen-Hagerling
Inc., 1998.
8. 8. HP 4278A 1kHz/1MHz Capacitance Meter Operation Manual, Hewlett-Packard Co., December 1996.
9. 9. HP 4268A High Capacitance Meter Operation Manual, Hewlett-Packard Co., December 1998.
10. 10. HP 4284A Precision LCR Meter Operation Manual (Fifth Edition), Hewlett-Packard Co., December 1996.
11. 11. HP 42841A Bias Current Source Operation Manual, Hewlett-Packard Co., December 1996.
12. 12. HP4286A RF LCR Meter Service Manual, Hewlett-Packard Co., July, 1995.
13. 13. HP 4194A Impedance/Gain-Phase Analyzer Operation Manual, Hewlett-Packard Co., December 1996.
14. 14. HP 4291B RF Impedance/Material Analyzer Service Manual, Hewlett-Packard Co., May 1998.

[*] The dielectric absorption is explained by the slow diffusion of electric charges into the dielectric material, and it sometimes has a
very large time constant of the order of hours or more.

© McGraw-Hill Education. All rights reserved. Any use is subject to the Terms of Use, Privacy Notice and copyright information.