Electrical testing decides whether a material will isolate, conduct or fail under field, and it splits into distinct measurement crafts that share almost nothing at the bench. Resistivity and conductivity measurements quantify current paths, dielectric testing and breakdown voltage testing push insulation to failure, impedance spectroscopy reads frequency response, and electrical reliability testing accelerates years of service into weeks. The discipline has a new front end in power electronics: JEDEC's JC-70 committee, formed in 2017 and grown past sixty member companies, published JEP180 because silicon-era reliability tests do not validate gallium nitride devices under real switching duty . Every converter and inverter riding those devices inherits the same measurement problem, and so does every laboratory staffed to run it.
Challenges in Electrical Testing Recruiting
Electrical reliability testing follows GaN and SiC into switching duty
Wide-bandgap power has rewritten what an electrical reliability engineer does. JEP180 exists because tests written for silicon power transistors do not necessarily validate operation under the actual-use conditions of power conversion equipment and may simply not apply to GaN . The document specifies switching accelerated life and dynamic high-temperature operating-life tests for enhancement-mode, depletion-mode, cascode and integrated GaN switches, which is to say the discipline now demands people who understand dynamic on-resistance, continuous-switching stress circuits and the junction-temperature measurement behind them . Stress conditions, soak temperatures and the failure criteria that separate device wear-out from test artifacts all had to be written down before vendors could compare anything. That population does not exist in the old silicon reliability benches. The committees drawing these standards mix device makers, users and test-instrument vendors precisely because no single group holds the whole measurement chain, and employers hiring against the same skills are fishing in that same mixed pond. The posting that asks for generic electrical testing experience will collect silicon-era resumes and miss the GaN bench entirely.
Breakdown voltage testing splits short-time from step-by-step protocols
ASTM D149 is the workhorse for dielectric strength, and its internal structure is the hiring lesson. The standard defines three voltage-application methods: Method A, the short-time test used for quality control; Method B, the step-by-step test; and Method C, the slow rate-of-rise test, with B and C usually giving lower but more meaningful results when materials are compared . The same standard draws the line between puncture through the specimen and flashover along its surface, two failure modes the record must separate . An operator who runs Method A against a specification produces numbers; a method owner knows which ramp regime their application needs, whether oil or air was the medium, and what a thickness dependence should look like. Breakdown voltage testing is a cheap phrase and an expensive misunderstanding, because the number depends on the protocol that produced it. Wire and cable makers, laminate suppliers and high-voltage component builders all run variants of the same test under different product specifications, and their staffs do not swap freely.
Resistivity measurements hide the electrification time inside the number
Insulation resistance looks like one measurement and is actually a time-dependent one. ASTM D257 notes that all dielectric resistances depend on the length of time of electrification and on the applied voltage, and that these conditions must be known and reported for the value to mean anything; surface resistance also swings rapidly with humidity while volume resistance moves more slowly . The IEC successor to IEC 60093, IEC 62631-3-1, makes the consequence explicit: for materials with volume resistivity below roughly ten to the tenth ohm-meters the steady state arrives within a minute, but for higher-resistivity materials the current can keep decaying for hours, days or weeks . The same guarding discipline returns in impedance spectroscopy of insulating systems, where a guarded three-terminal cell decides what the analyzer actually reads. A resistivity number with no electrification time attached is a reading, not a property, and the CV never carries the difference. That is why insulation testing laboratories screen for people who write down their measurement conditions as reflex, and why the ones who do not stay invisible in a pile of identical-looking certificates.
Conductivity measurements hand the four-point probe to people who never built one
On the conducting side, the craft is geometry. The four-point, or Kelvin, technique sources current through two probes and measures voltage on two others, which eliminates the probe resistance, spreading resistance and contact resistance that would otherwise bury the answer; the collinear four-point probe handles wafers and films, while van der Pauw works on small samples of arbitrary shape where probe spacing stops mattering . The instrument details are where competence shows: sub-nanoamp current sourcing, settling time on high-resistance samples, and reversal of current polarity to cancel offsets . Semiconductor materials laboratories and conductive-coating groups both advertise conductivity measurements, but a technician who has only ever read a benchtop meter has never built the measurement the number depends on. The correction factors for probe-to-wafer ratios, the thickness entry that turns sheet resistance into volume resistivity, and the polarity reversal that cancels offsets are the difference between a characterization and a casual reading .
Dielectric testing moves from flashover-guarded benches to partial discharge diagnosis
Beyond breakdown, dielectric testing extends into partial discharge territory, where the questions change from does it fail to how is it degrading. IEC 60270 defines the measurement of partial discharges in apparatus tested with alternating or direct voltage, and its centerpiece is the apparent charge, the injected charge that would reproduce the instrument reading, expressed in picocoulombs . That definition exists because the real charge at the discharge site cannot be measured directly, which is precisely why the method is calibration-heavy: coupling devices, pulse-train response and the discrimination of test-object discharge from external interference all sit inside the standard . Measuring partial discharge without a proper calibrator run is guesswork with a pretty screen. High-voltage apparatus qualification keeps hiring this skill, and it does not come from material-coupon laboratories. A candidate who can defend an apparent-charge measurement with its calibrator history is scarce precisely because so few benches teach it. Utilities, rotating-machine builders and cable manufacturers recruit from the same small group of partial discharge practitioners, most of whom learned on commissioning sites rather than in classrooms.
Electrical characterization claims collapse at the guard electrode question
Verification in this discipline runs through one piece of hardware: the guard. IEC 62631-3-1 devotes its equipment clauses to guarding because stray currents, spurious external voltages and surface leakage an order of magnitude below the volume resistance would otherwise distort every high-resistance measurement; guard conductors intercept those currents and leave the specimen as the only direct path between the measuring terminals . Ask a candidate to draw their electrode system. Ask what happens to a surface-resistivity reading in humid air, what electrification time they report, and what a voltage ramp did to the numbers. Ask which dielectric testing records they personally signed and which they only witnessed. In electrical characterization the vocabulary is uniform and the practice is not, and the difference between a guarded measurement and an unguarded one is often the difference between a spec and a rumor . A mis-hire writes those rumors into design margins, where they surface later as field failures nobody traces back to the bench.
References
- JEDEC Wide Bandgap Power Semiconductor Committee Publishes Guideline for Switching Reliability Evaluation Procedures for GaN Devices — JEDEC Solid State Technology Association. (accessed 2026-09-28)
- ASTM D149: Standard Test Method for Dielectric Breakdown Voltage and Dielectric Strength of Solid Electrical Insulating Materials at Commercial Power Frequencies — ASTM International. (accessed 2026-09-28)
- ASTM D257: Standard Test Methods for DC Resistance or Conductance of Insulating Materials — ASTM International. (accessed 2026-09-28)
- IEC 62631-3-1:2016: Dielectric and resistive properties of solid insulating materials — Determination of resistive properties (DC methods) — Volume resistance and volume resistivity — International Electrotechnical Commission (IEC). (accessed 2026-09-28)
- Four-Probe Resistivity and Hall Voltage Measurements with the Model 4200-SCS — Tektronix (Keithley Instruments). (accessed 2026-09-28)
- IEC 60270:2000: High-voltage test techniques — Partial discharge measurements — International Electrotechnical Commission (IEC). (accessed 2026-09-28)