A finished wire harness is tested three ways, because a harness can fail three ways. Continuity testing proves the electrical map is right — every net connects where it should and nowhere it should not. HiPot and insulation resistance prove the dielectric holds — the insulation will not break down under voltage. Pull-force testing proves the mechanical joints hold — a crimp will not back out under load. A harness that passes all three is not "probably fine"; it is verified against the failure modes that actually put a build on the line. Here is what each method catches, which run on every unit versus a sample, and how the acceptance class sets how far testing goes.
These are different questions, and passing one says nothing about the others. A perfectly crimped harness can still be miswired; a correctly wired one can still have a nicked insulation that only fails at voltage. That is why a serious test plan runs all three families, not whichever is quickest.
1. Continuity testing — the electrical map
Continuity is the first and most fundamental test: does current flow along every intended path, and only along those paths. A continuity tester energizes each net in turn and confirms three things — that the connection is present (no open), that it goes only where the drawing says (no short or crossed wire), and that its resistance is low enough (no high-resistance partial from a poor crimp). On our floor this is a 100% test — every net on every harness through a Cirris bench, with a pass threshold under 10 milliohms, run across eight benches so it never becomes the bottleneck.
Continuity catches the errors that are invisible to the eye: a terminal seated in the wrong cavity, two nets swapped at a connector, a strand count that passed a visual but reads high resistance. It is cheap, fast, and non-destructive, which is why it is the one test that should never be sampled — there is no reason to ship a harness whose wiring has not been confirmed. What continuity does not prove is that the insulation will survive voltage, or that the crimp will survive being pulled. Those need the next two families.
2. HiPot and insulation resistance — the dielectric
The second question is whether the insulation between conductors — and between a conductor and any shield or chassis — will hold under stress. Two related tests answer it.
HiPot (high-potential, or dielectric withstand) applies a voltage far above the working voltage across the insulation for a fixed time and watches the leakage current. If the insulation is nicked, thin, contaminated, or pinched, it breaks down and the leakage spikes past the limit — a failure. Our HiPot runs at 1500 V AC for 60 seconds with a 0.5 mA leakage limit per IPC-9252; a failed unit is locked in the fixture and quarantined for engineering review rather than re-tested until it passes. HiPot is a proof stress, so it is applied where the application warrants it — multi-conductor, shielded, or higher-voltage harnesses — and its parameters come from the product spec, not a default.
Insulation resistance (IR) is the gentler companion: a DC voltage measures how many megohms of resistance the insulation offers. It trends the health of the dielectric rather than proving a single withstand point. We measure IR at 500 V DC with a pass floor of 100 MΩ per UL 758. Together, HiPot catches the outright breakdown and IR catches the slow degradation — a harness that reads a falling IR is warning you before it ever fails a HiPot. Neither says anything about the crimp, which is the third family.
3. Pull-force testing — the mechanical joint
A crimp is a gas-tight mechanical and electrical joint formed by deforming a terminal around a conductor. It can read perfect on continuity and still be under-crimped — holding now, but ready to back out under vibration or a harness pull. Pull-force (crimp tensile) testing is how that is caught: the terminated wire is pulled in a gauge until it separates, and the force at separation is compared against a minimum for that conductor size. The minimums are not arbitrary — IPC/WHMA-A-620 and USCAR-2 publish pull-force tables by wire gauge, so a 20 AWG crimp and a 12 AWG crimp are held to different, defined thresholds.
Because the test destroys the sample, pull-force is sampled, not 100% — a defined number of terminations per setup, per shift, or per crimp-die change, recorded on a control chart. That sampling is backed by 100% non-destructive checks: crimp-height measurement and a 10× microscope inspection of every crimp transition per IPC-A-620 §10, run by a different operator than the one at the bench so the check is independent. The pull test proves the process window is right; the 100% checks prove each unit sits inside it. The crimp setup and control detail is in our crimping capability page.
How much testing? The acceptance class decides
The same harness can carry very different test scope, and the difference is the acceptance class, not the BOM. IPC/WHMA-A-620 defines three classes, and each raises the bar on documentation and sampling:
- Class 1 — general use. Basic continuity, minimal records.
- Class 2 — dedicated service. 100% continuity plus workmanship inspection; the common commercial default.
- Class 3 — high-reliability (aerospace-grade process rigor, medical, safety-critical). Adds pull-force sampling with records, crimp-height logging, first-article documentation, and tighter workmanship limits.
Stating the class up front is what lets a quote and a test plan match the real requirement instead of guessing. The full walk-through of what Class 3 changes — and when you actually need it versus when Class 2 is enough — is in our IPC/WHMA-A-620 Class 3 article. Which gauge and insulation you specified also feeds the test limits; that decision is covered in the wire gauge selection guide.
What a real test floor looks like
Methods only matter if they are actually run and recorded. Our line pairs 100% electrical test with sampled mechanical test and independent inspection: eight Cirris continuity benches, HiPot and IR per the specs above, crimp pull-force sampling against A-620/USCAR-2 minimums, and FQC microscope inspection by a separate operator. For automotive programs, continuity resistance is run as a capability study (Cpk targets) to support a customer's PPAP submission — the test data the customer's own quality system needs. The equipment list, environmental chambers, and a sample test report are on the testing and inspection page, and the broader process sits under manufacturing capabilities.
Representative project (anonymized to industry and region — no client name). An industrial-equipment OEM moving a harness program to a Cavite build was not worried about price; it was worried about intermittent field faults on the harness it was replacing. The qualification plan led with test evidence: 100% continuity records, HiPot results, and pull-force sample charts on the first article, plus crimp-height logs. Seeing the actual data — not a certificate, the measurements — was what let the OEM's engineers sign off and release the program. The lesson: for a reliability-driven buyer, the test report is the sales document.
FAQ
What is the difference between a continuity test and a HiPot test?
Continuity confirms the wiring is correct and low-resistance — every net connects where it should. HiPot confirms the insulation withstands a voltage well above the working voltage without breaking down. One checks the conductors, the other checks what separates them; a harness needs both, because passing one says nothing about the other.
Is every harness pull-force tested?
No — pull-force is a destructive test, so it is sampled by conductor size and crimp setup and recorded on a control chart, not run on every unit. Every unit instead gets 100% non-destructive checks: continuity, crimp-height measurement, and microscope inspection of the crimp. The sampling proves the crimp process is in spec; the 100% checks prove each harness sits inside that proven window.
Do I need Class 3 testing?
Only if the application demands it. Class 3 adds pull-force records, crimp-height logging, and first-article documentation — real value for safety-critical, medical, or high-vibration programs, and unnecessary cost for a general commercial harness that Class 2 covers. State the class in your RFQ so the test plan and price match the requirement. Note that we build to standard-defined test methods and support customer PPAP with capability data; we do not test aerospace, military, or defense harnesses.
What does a continuity failure usually mean?
Most often a terminal in the wrong cavity, two nets crossed at a connector, or a high-resistance crimp. Continuity localizes it to the net, so the harness is corrected and re-tested rather than scrapped — which is exactly why the test runs before anything ships.
Have a harness with a defined test and acceptance class? Send the drawing, BOM, and test scope to a drawing-driven quote, and for a Philippines-origin program we will confirm the test plan alongside the build. The standards behind the numbers are on our standards and certifications page.
