A supplier's statement that a harness is electrically tested leaves several purchasing questions unanswered. Were the intended connections checked? Were circuits that should remain isolated evaluated? Which conditions and acceptance limits produced the pass result?
Continuity, insulation resistance and dielectric withstand testing address different questions. A useful cable harness test report connects each required test to a released drawing, an approved program and the units being supplied. This guide helps buyers request that evidence without inventing a universal test voltage or treating a pass label as proof of every electrical property.
Separate the three test purposes
Continuity concerns the conductive paths that should exist. Insulation resistance, usually abbreviated IR, concerns resistance across insulation separating conductive parts. Dielectric withstand, abbreviated DWV and commonly called hipot, checks whether the specified insulation withstands a defined electrical stress for a defined period. Cirris distinguishes these purposes in its cable high-voltage testing guide.
Terminology needs one clarification in the RFQ: Cirris also uses hipot as an umbrella term that includes both DWV and IR. Therefore, write the required test names explicitly and ask the supplier to identify the corresponding report fields. Its guide to assemblies containing components describes DWV as an AC or DC test and IR as a DC test.
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| Test named in the RFQ | Procurement question | Evidence to request |
|---|---|---|
| Continuity and connection verification | Does each intended path meet its limit, and does the programmed wiring pattern match the drawing? | Drawing/netlist revision, endpoints, connection limits, actual results or agreed result format, and explicit unexpected-connection coverage |
| Insulation resistance | Does each required isolation boundary meet its minimum resistance under the approved conditions? | Tested nets or groups, DC voltage, measurement timing, minimum limit, resistance result or clearly identified instrument bound |
| Dielectric withstand | Does the required insulation withstand the approved stress and duration without the specified failure conditions? | AC/DC method, voltage, applicable frequency, ramp and dwell, current/arc criteria where applicable, result and failure location |
This table is a purchasing comparison, not a test specification. The product's responsible engineer must select the applicable requirements, conditions and acceptance rules.
Define the wiring coverage before comparing results
A net is a group of points intended to be electrically connected. The buyer should supply the released connection drawing and ask the supplier to map connector names, cavity numbers, shields, shells and unused contacts into the test program. Give intentional joins and intentional isolation equal attention.
For continuity, ask whether the program checks only listed paths or also checks for unintended connections. A statement that every listed wire passed does not explain what happened to an omitted cavity. Likewise, a high-voltage pass between two grouped circuits does not establish that every possible individual pair was tested.
Harnesses containing diodes, capacitors or other components need particular care. Cirris explains that component-aware grouping and a verified program are necessary to avoid inappropriate electrical stress across embedded parts. Its implementation performs continuity checks before high-voltage testing; that sequence should not be assumed for an unspecified tester. Cirris component-testing guidance.
The purchasing action is to request an approved coverage matrix showing included boundaries, grouped nets and exclusions with their reasons. Have the responsible engineer resolve exclusions before sample acceptance. This article does not provide high-voltage connection instructions.
Make test conditions part of the reported result
An IR value without its conditions is difficult to compare across quotations. Ask for the specified voltage, measurement timing, conditioning requirements and relevant ambient conditions. Also distinguish an actual measured value from a display meaning that resistance exceeded the instrument's reporting range. A greater-than result can be useful evidence, but it is not an exact measurement. If its stated lower bound is below the approved minimum resistance, the report alone cannot demonstrate compliance. Request evidence under the approved conditions; an inconclusive record is not proof that the product failed.
For DWV, record whether the method uses AC or DC, together with the approved settings and result criteria. Chroma explains that AC testing of capacitive devices involves reactive current, while DC voltage changes produce charging current. Consequently, a current reading or trip cannot be interpreted independently of the method and timing. Do not convert an AC hipot current into a claimed DC insulation resistance or substitute one method simply because the numerical voltage looks comparable. Chroma, A Practical Guide to Dielectric Testing.
Request the source of the acceptance requirements: the customer specification, applicable product requirement or approved engineering test plan, including its revision. Separate qualification conditions from routine production conditions if both apply. A supplier's tester capability range is not an acceptance limit, and a higher setting is not automatically a better purchase specification.
Ask what “100% tested” actually covers
Define the denominator. Does 100% mean every finished unit or every required circuit? Those statements answer different questions. If only a sample was tested, report the tested quantity and lot quantity explicitly; do not describe the whole lot as 100% tested.
A practical report agreement should identify:
- Units and stage: part number, drawing revision, quantity, serial or lot identity, and whether testing occurred before or after the final operation relevant to the requirement.
- Program and fixture: controlled program revision, fixture identity and the approved mapping to the product drawing.
- Equipment: tester identity and calibration status sufficient to establish its validity at the recorded test date.
- Results: required limits, units, result format, covered boundaries, failures and any instrument-range qualifiers.
- Disposition: repair or rework history, retest authorization and the relationship between the original failure and final acceptance.
These are proposed purchasing fields. Agree their detail and retention period before ordering rather than assuming every standard report contains them.
For example, a report linked only to a batch can support batch-level traceability, but it cannot by itself identify the result for a particular serial-numbered harness. Choose the required level according to the project's acceptance needs.
Compare two hypothetical supplier replies
Assume a buyer is reviewing a hypothetical order for 100 passive harnesses under an already approved test plan. No test voltage, current threshold or resistance limit is proposed here.
Supplier A replies: “All 100 pieces passed continuity and hipot.”
Supplier B replies that all 100 finished units were checked against the released drawing and named program revision. It supplies the isolation coverage matrix, distinguishes IR from DWV, identifies the approved conditions, and provides the agreed unit-linked results. Its report also preserves one repaired unit's initial failure and authorized retest.
As a further hypothetical reconciliation, assume each complete program run creates one event: 99 units pass once, while one unit fails and then passes its authorized retest. That produces 101 events for 100 unique units. Count the accepted unit identities and their dispositions, not merely the number of test events.
Supplier B's reply is more reviewable because the buyer can reconcile the claimed scope with the evidence. It is not proof of superior product performance: the drawing, program, actual results and disposition still need review. Supplier A may have performed suitable testing but has not yet documented enough to make the comparison.
Ask Supplier A for the missing fields before ranking quotations. Ask Supplier B to explain any exclusions or departures. Compare both against the same acceptance plan, rather than rewarding a more impressive-looking test name.
Keep electrical acceptance connected to the wider RFQ
These reports do not replace every other acceptance activity. For precision measurement of a conductive path, use the separate Kelvin resistance testing RFQ guide. For drawing release, sample approval and later manufacturing changes, use the first-article and change-control guide.
When preparing an OEM/ODM cable assembly inquiry, include the drawing, embedded-component information, application, acceptance-plan reference, required report format and order quantities. Identify unresolved test requirements clearly so they can be reviewed before a quotation is finalized.
Buyer FAQ
Does passing continuity prove the insulation is acceptable?
No. Continuity addresses intended conductive paths. Request the separately specified isolation tests and their conditions when the acceptance plan requires them. A continuity pass also needs a clear explanation of connection coverage.
Can a supplier replace insulation resistance with a hipot pass?
Do not assume equivalence. First establish whether the supplier uses hipot to mean DWV alone or a sequence including IR. Then compare the reported tests with the approved acceptance requirements and resolve any proposed substitution with the responsible engineer.
Is a pass/fail report enough, or should we request measurements?
Specify the evidence needed for acceptance. Pass/fail records tied to the correct program may meet an agreed requirement; numeric results can support comparison or investigation. Ask whether exported values are measured readings, instrument bounds or only thresholds, and agree this before production.
Should every repaired harness receive the same retest?
Follow the approved rework and retest plan. Ask which changes trigger which checks and how the original failure remains traceable. A later pass should be linked to its disposition rather than presented as if no earlier failure occurred.