Selecting an LCR meter is governed by four numbers on the spec sheet: test frequency range, basic accuracy, test signal level, and the impedance range the bridge can resolve without losing phase lock. Get any one of these wrong and the displayed inductance or capacitance is, at best, decorative.
The instrument class question has hardened into a three-tier market. Keysight groups its benchtop LCR meters as the ZA2-class (1 MHz to 3 GHz maximum frequency), handheld capacitance and LCR meters as the ZA1-class (1 kHz to 100 kHz), and high-performance impedance analyzers as the ZA5-class (120 MHz to 3 GHz) [S4]. IET Labs brackets the precision-bench segment with the 1693 at 0.01% capacitance accuracy and a 0 to 300 V test voltage range, paired with 12 Hz to 200 kHz frequency coverage [S3]. Buyers who skip the class question and shop headline price end up re-buying within 18 months.
What an LCR Meter Actually Measures
An LCR meter applies a sinusoidal AC stimulus to the device under test and computes the complex impedance Z = R + jX, from which it derives the operator-selected primary value (Ls, Lp, Cs, Cp, R) and a loss term: dissipation factor D, quality factor Q = 1/D, or equivalent series resistance ESR = D / (2 pi f C) [S2]. Because reactance scales with frequency (XL = 2 pi f L, XC = 1 / 2 pi f C), a single component returns a different impedance at every test frequency, which is why a reading quoted without its test frequency is essentially undocumented [S2]. Capacitors are graded by D, inductors and resonators by Q, and aluminum-electrolytic and DC-link capacitors by ESR because ESR governs ripple-current heating in power-supply loops.
Three instruments look similar but are not interchangeable. A handheld component tester or the capacitance range on a digital multimeter applies a DC charge ramp and reports only capacitance with coarse accuracy, no frequency control, and no loss term. A micro-ohmmeter measures DC resistance only. The LCR meter is distinguished by its phase-sensitive AC excitation at a controlled frequency and level, which is what lets it resolve the reactive and resistive parts independently and report D, Q, and ESR instead of a single number [S2].
Instrument Classes and the Frequency Decision
The first selection filter is frequency, because the stimulus frequency determines both the impedance range and the measurement technique inside the box. Keysight's three classes sort by upper frequency: ZA1-handheld at 1 kHz to 100 kHz, ZA2-benchtop at 1 MHz to 3 GHz, and ZA5 impedance analyzer at 120 MHz to 3 GHz [S4]. Across the same vendor's wider portfolio the stimulus range stretches from 5 Hz to 3 GHz when impedance analyzers are included [S5].
Inside the bench, the technique matters as much as the spec. The auto-balancing bridge is the workhorse for low-frequency general-purpose testing, with a typical frequency span of 20 Hz to 120 MHz and the widest usable impedance range for its class [S5]. RF I-V technique takes over above roughly 100 MHz and is what makes 3 GHz coverage practical. For a buyer specifying discrete passives at 1 kHz, the auto-balancing bridge inside a ZA2-class bench unit is the right architecture; specifying the same DUT at 1 GHz requires a ZA5-class analyzer or a network analyzer, not a faster LCR meter. More on the IET bench side, the 1920 covers 10 Hz to 2 MHz at 0.05% basic LCR accuracy and reaches 120 measurements per second, which makes it a credible production test candidate alongside the 1693 [S3].
Accuracy, Test Signal, and DCR
Basic accuracy on an LCR meter spec sheet is a single percentage that hides three different things: impedance magnitude accuracy, D or Q accuracy (often quoted as 0.001 or 0.0001), and frequency accuracy. IET's quick-selection table shows the spread: 1910 at 0.1% LCR with 0.001 DQ, 1920 and 7660 at 0.05% LCR with 0.0005 or 0.0003 DQ, 1692 and 1689 at 0.02% LCR with 0.0001 DQ, and the 1693 at 0.01% C with 0.0001 D, the latter being the only model in the IET bench family that swings the test voltage up to 300 V for high-voltage capacitor and dielectric characterization [S3]. Keysight's ZA2 benchtop class quotes basic accuracy as low as 0.05% [S4].
Test signal level is the second filter. Most general-purpose LCR meters run 5 mV to 1 V or 20 mV to 5 V; the 1690 family at 20 mV to 1.0 V is a typical low-voltage bench unit, while the 1920 broadens to 20 mV to 5.0 V [S3]. For incoming inspection of power-electronics capacitors the ESR measurement at rated ripple frequency is usually more diagnostic than the capacitance value, which is why bench units that pair AC LCR with a DC-resistance (DCR) mode, like the IET 1910, are preferred on a production line [S3]. A short, citable spec rule: most vendor datasheets state that basic LCR accuracy is specified at 1 kHz with a 1 V test signal, and derates outside that envelope. Buyers should request the accuracy envelope at the actual test frequency and voltage they plan to use.
Connection Method and Test Fixtures
At component values above a few hundred kilohms and at test frequencies above a few hundred kilohertz, the test lead itself becomes part of the measurement. Bench LCR meters use 4-terminal Kelvin connections to cancel lead resistance and capacitance, and higher-frequency units (1 MHz and up) use 4-BNC Kelvin or specialized fixtures. The IET 1693 supports 2-, 3-, and 4-terminal connections, which is the practical minimum for switching between axial leaded parts, SMDs on a custom fixture, and in-circuit measurements [S3].
Fixture choice is not optional. SMD components with sub-millimeter terminations cannot be measured reliably with clip leads; vendor-supplied SMD tweezers or test fixtures with defined geometries are required to keep the parasitics reproducible. For high-voltage or safety-rated testing, the fixture must also carry the relevant working voltage and creepage clearance, which is where 0 to 300 V test instruments such as the 1693 are paired with vendor-approved dielectric test cells [S3]. A recurring line-side failure mode is using a handheld's alligator clips for an SMD and blaming the meter for a 20% error; the instrument is fine, the fixture is wrong.
Comparison: Three Realistic Choices for Three Duty Profiles
Three duty profiles cover most procurement cases, and the comparison below is a quick filter, not a recommendation. Accuracy figures are basic LCR accuracy as quoted by the vendor at 1 kHz / 1 V. [S2]
Duty 1, field service and quick component sorting: a handheld ZA1-class unit, 1 kHz to 100 kHz, basic accuracy around 0.5% to 1%, optional DCR, battery powered. Best for troubleshooting on the bench, not for QA sign-off. Duty 2, incoming inspection and production test of discrete passives: a ZA2-class benchtop or a precision bench unit like the IET 1920, 10 Hz to 2 MHz, 0.05% basic LCR accuracy, 4-BNC Kelvin, 120 measurements per second, handler and RS-232 / GPIB / LAN options for the ATE rack [S3][S4]. Duty 3, high-voltage capacitor and dielectric characterization: a precision bench unit rated to 0.01% C with 0 to 300 V test signal, such as the IET 1693, 12 Hz to 200 kHz, 4-BNC Kelvin, GUI and LabVIEW driver [S3].
Series vs Parallel Models and Frequency-Dependent Behaviour
Every LCR meter offers the operator a choice of equivalent circuit: series (Ls, Cs, Rs) or parallel (Lp, Cp, Rp). The choice is not aesthetic. Low-impedance parts (large capacitors, small inductors) read more accurately in the series model because their parasitic resistance is in series with the ideal element; high-impedance parts (small capacitors, large inductors, high-value resistors) read more accurately in the parallel model because the parasitic conductance is across the element [S2]. Picking the wrong model gives a reading that is numerically plausible but physically wrong, and disputes over a 5% to 10% spread between supplier and incoming-inspection data are usually traced to this single switch.
Frequency dependence is the other trap. Buyers who spec a single capacitance value without a paired frequency create disputes that the instrument cannot resolve. Reference [S2] grounds the parameter conventions in the Keysight Impedance Measurement Handbook, IEC 60384-1 for fixed capacitors, and IEC 62391-1 for electric double-layer capacitors, which is the standards stack to anchor any incoming-inspection procedure in.
Who Should Not Pick the Mainstream Bench Unit
Three profiles should bypass the ZA2-class mainstream bench entirely. First, buyers characterizing RF materials (ferrite permeability, dielectric permittivity) above 100 MHz: they need a ZA5-class impedance analyzer with continuous sweep and equivalent-circuit analysis, not a spot-frequency LCR meter [S4][S5]. Second, buyers testing small-signal SMD MLCCs at production speed: they need a production-grade LCR or impedance analyzer with a handler interface, comparator, and at least 40 measurements per second; the IET 1920 at 120 measurements per second is the lower bound, slower units will bottleneck the line [S3].
Standards, Sourcing, and Shortlist Logic
Parameter conventions in any LCR procedure should be tied to the Keysight Impedance Measurement Handbook, with IEC 60384-1 governing fixed-capacitor test methods and IEC 62391-1 covering electric double-layer capacitors, as cited in the reference LCR guide [S2]. The seven selection FAQs in the same reference cover measurement method, instrument class, test frequency and signal, series-vs-parallel models, and spec-sheet decoding, which is the minimum reading set before any quote goes out [S2]. For a working shortlist on the bench, the practical filter is: confirm the test frequency span covers at least 100 Hz to 100 kHz for general passive work, confirm basic LCR accuracy is 0.1% or better at 1 kHz, confirm 4-terminal Kelvin as standard, and confirm the vendor publishes accuracy envelopes at the test frequency and voltage you actually plan to use, not only at the 1 kHz / 1 V sweet spot.
Trackable signals for the next procurement cycle: vendor releases adding higher test frequencies to handheld ZA1-class units, vendor updates to LabVIEW driver and handler interface firmware for the 1920/1693 family [S3], and any new IEC or manufacturer-published revisions to IEC 60384-1 or IEC 62391-1 that move test-condition requirements. For a deeper walk-through of the same selection logic anchored to spec-sheet lines, the in-depth guide How to Choose an LCR Meter: Spec-Anchored Selection Guide covers frequency-class fit and accuracy envelopes in more detail. For the broader test-and-measurement context, the LCR meter encyclopedia entry consolidates measurement methods, instrument classes, and the series-versus-parallel decision in a single reference.
Spec-level background on the components involved: linear guide, and crossed roller guide.