Two suppliers can report different resistance values for equivalent battery cables because they measured different spans, included different fixture contacts or tested at different temperatures. A smaller number alone does not establish a better assembly.
For B2B procurement, request a defined measurement boundary and reproducible conditions before comparing results. This guide addresses isolated, de-energized passive cable assemblies tested under an approved laboratory procedure.
Define exactly what the resistance includes
Label the two voltage-sensing locations on a drawing. “Cable resistance” could describe the conductor alone, a finished assembly between lug palms, or an installed path that includes bolted interfaces. These are different acceptance characteristics.
For a finished assembly, identify the conductor length, both terminations and the terminal material between the sense points. State whether mating plates, fasteners or other fixture parts belong inside the measurement boundary. A test across an installed joint deliberately includes that interface; a cable-only acceptance test may exclude it.
A crimp-specific result needs separately defined access points and any approved method for accounting for adjacent conductor resistance. Do not relabel a complete lug-to-lug result as the resistance of one crimp.
Establish conductor construction and length first using the battery cable sizing RFQ guide.
Separate current contacts from voltage-sensing contacts
Four-wire, or Kelvin, measurement uses one pair of leads to carry test current and a second pair to sense voltage. Resistance is calculated from the measured voltage and current. This arrangement reduces the influence of test-lead resistance on low-resistance results. Keysight explains the four-wire measurement principle.
The fixture still matters. Show force and sense locations independently, including their position on each lug. Joining the connections remotely from the intended measurement points leaves the intervening material inside the measured span. Tektronix recommends locating voltage-sensing connections close to the resistance being evaluated. Tektronix low-resistance measurement white paper.
Request a fixture photograph and controlled contact locations. Include a repeat-clamping check to assess whether placement changes the result. Confirm that the current source reaches the specified current without reaching its output limit.
Identify how thermal voltage offsets are handled
A tiny measured voltage may contain both the resistive voltage drop and an unwanted thermoelectric voltage. Temperature differences and junctions between dissimilar conductors can generate these offsets.
For equal-magnitude positive and negative test currents, with the voltage-sense orientation unchanged:
V+ = I × R + EV− = −I × R + ER = (V+ − V−) / (2 × I)
Here, E is the offset assumed stable during the pair of measurements. Current reversal cancels that stable offset; changing thermal conditions can leave residual error. Offset-compensated ohms instead compares current-on and zero-current readings. Delta measurement uses three readings to address approximately linear offset drift over a sufficiently short interval. Tektronix compares current reversal, offset compensation and delta methods.
The RFQ should identify the selected method, timing and retained raw readings. A report stating only “offset corrected” leaves the measurement difficult to reproduce.
Worked example: a stable offset changes a small resistance result
The numbers below are invented solely to demonstrate the equation. They are not SINAWATTS measurements, a recommended test current or an acceptance limit. Assume equal-magnitude currents of +10 A and −10 A, unchanged sense polarity, settled readings and an offset that stays constant during the pair.
- At +10 A, suppose the sensed voltage is +2.2 mV.
- At −10 A, suppose it is −1.8 mV.
- The difference is 4.0 mV, so resistance is 4.0 mV / 20 A = 0.20 mΩ, equivalent to 200 µΩ.
- Using only the positive reading would give 2.2 mV / 10 A = 0.22 mΩ. The average of the two signed voltage readings is +0.2 mV, representing the stable offset in this simplified example.
This demonstrates the current-reversal principle described by Tektronix. It does not correct a changing offset, mismatched current magnitudes, self-heating, a different sense span or an unqualified fixture. Retain signed readings, actual currents, timing and units so the calculation can be checked.
For a supplier comparison, first align the measurement spans and conditions. If one report uses conductor-only points and another includes both lug terminations, their final numbers are different characteristics even if both use Kelvin leads. Request comparable reports before ranking the assemblies.
Control current, settling and temperature together
Test current must generate a measurable voltage while limiting self-heating. Source settling and voltage acquisition must also be coordinated. Keysight identifies sufficient signal, minimized power dissipation and synchronized measurements as separate requirements for precise low-resistance testing. Keysight precision low-resistance application note overview.
Specify current magnitude, waveform, measurement delay, acquisition duration and recovery time where relevant. There is no single test current appropriate for every cable geometry, instrument and test purpose.
Record sample temperature and how it was determined, alongside ambient temperature. Agree on a stabilization criterion rather than relying solely on elapsed waiting time. If results are corrected to a reference temperature, require the correction method and its applicability to the measured assembly.
Treat electrical settling, sample temperature stabilization and thermoelectric drift as separate controls. Increasing averaging time does not establish that all three are satisfactory.
Put the measurement requirements into the RFQ
Use the following table to request comparable evidence. Let the responsible engineer approve the completed method and acceptance criteria before production testing.
On small screens, swipe or scroll sideways to read every column. Keyboard users can focus the table and use the arrow keys.
| RFQ item | Buyer definition | Supplier evidence to request |
|---|---|---|
| Assembly identity | Drawing revision, conductor, length and terminal parts | Sample and material-lot identifiers |
| Measurement span | Marked sense datums and included interfaces | Annotated drawing and fixture photograph |
| Instrumentation | Required measurement capability | Instrument identity, range and calibration status |
| Excitation | Current, waveform and permitted limits | Actual settings and achieved current |
| Offset control | Reversal, current-on/off or another approved method | Raw readings, timing and calculation |
| Temperature | Measurement location and stabilization criterion | Sample temperature, ambient and timing |
| Repeatability | Defined reconnection and repeat-reading checks | Individual results and observed spread |
| Acceptance | Approved limit, units and decision rule | Actual result, uncertainty information and disposition |
Ask bidders to distinguish fixture development, method verification, sample testing and recurring production inspection in their quotation. Confirm test availability, responsibility and reporting scope in writing.
Approve the limit separately from the measurement method
A Kelvin instrument explains how a resistance value was obtained; it does not supply the assembly’s acceptance limit. That limit should trace to the released design, applicable requirements and the validated measurement boundary.
Do not convert a conductor resistance-per-length requirement into a finished-assembly limit without accounting for the included terminations. Likewise, a reference sample supports comparison but needs an approved rationale before becoming a pass/fail threshold.
Connect these records to the battery cable first-article and change-control plan. A resistance check also remains separate from installation-specific joint and temperature-rise validation.
Turn a resistance report into a purchasing decision
Separate three outcomes in the review:
- Comparable evidence: the configuration, sense points, units, temperature, excitation and offset method match the agreed procedure. The report can proceed to comparison with the approved limit.
- Incomplete or non-comparable evidence: a span, current, temperature or unit is missing or different. Ask for clarification or a measurement under the agreed conditions; a lower reported number does not settle the issue.
- A result close to the acceptance boundary: apply the agreed decision rule and uncertainty treatment. Do not replace them with extra decimal places or a rule invented after seeing the result.
The review should record the actual disposition and any retest authorization. A Kelvin result alone does not approve ampacity, a bolted installation or lifetime durability. Request a cable quotation with measurement evidence, attaching the sense-point drawing and the report fields your reviewer needs.
Buyer FAQ
Can a continuity check replace a Kelvin resistance report?
A continuity result does not provide the defined low-resistance measurement needed for this comparison. Request actual resistance values, units, measurement boundaries and conditions when resistance is an acceptance characteristic.
Does four-wire measurement remove every contact resistance?
No. It reduces unwanted lead contributions when connections are correctly arranged. A product joint or fixture segment between the voltage-sensing points remains part of the measured path. The drawing must identify which interfaces are intentionally included.
What maximum resistance and test current should I specify?
Request values approved for the exact assembly, application and measurement method. Instrument resolution, self-heating and the intended test purpose influence current selection. This guide supplies no universal milliohm limit or test current.
What should suppliers quote for additional testing?
Request separate scope and pricing for fixtures, method verification, sample quantities, recurring inspection and reports. Ask whether work is performed internally or by an identified external laboratory, with availability and timing confirmed for the project.
Is 200 µΩ the same as 0.20 mΩ, and does either mean pass?
They are the same resistance: 1 mΩ equals 1,000 µΩ. Neither value is a universal pass limit. Compare a result only with the requirement and decision rule for the specified assembly, measurement span and conditions.
Send the measurement scope with your RFQ
Send the cable drawing, marked measurement span, proposed operating conditions and required report fields. Request written confirmation of the offered construction and testing scope before ordering.