Four-Wire Cable Resistance Measurement: Kelvin Sensing, Contacts, and Temperature

Predict First, Then Define the Measurement Boundary

The Wire Resistance and Voltage Drop Guide covers conductor geometry, resistivity, gauge choice, and first-pass drop. This guide addresses the next step: what the installed path actually measures. Write down whether the span is a cut conductor, a crimped assembly, an outbound-and-return loop, or the complete source-to-load network. Mark the intended sense points on the wiring drawing before taking readings. A low conductor resistance does not guarantee a low installed drop when connectors or fuses add series resistance.

Place the Kelvin Sense Leads Deliberately

Drive test current through two force leads and measure voltage through separate high-impedance sense leads. The sense pair should touch the exact boundaries whose resistance is required. Because the voltmeter draws very little current, sense-lead drop is small; the often substantial force-lead drop is outside the voltage measurement. Kelvin wiring does not magically remove a contact inside the sense span. Put both sense points on the conductor to isolate bulk copper, or across the assembled joint to include it. Keep polarity and location repeatable, and prevent the sense clips from moving while current is applied.

Remove Stable Voltage Offsets with Current Reversal

Thermoelectric voltage at dissimilar-metal junctions and instrument offset can be significant when the desired drop is only millivolts. Apply +I and record signed voltage Vplus, then apply -I and record Vminus after settling. If the parasitic offset remains approximately constant over the pair, Vplus = I R + Voffset and Vminus = -I R + Voffset. Subtraction gives R = (Vplus - Vminus)/(2 I). Do not take absolute values before subtracting; that destroys the sign information. Reverse quickly enough that temperature and contact conditions remain comparable, but allow the source and meter to settle. For a changing offset, use a suitable zero-current or delta method rather than assuming perfect cancellation.

Worked example: with +10 A the sense meter reads +20.20 mV, and with -10 A it reads -19.80 mV. Their half-difference is 20.00 mV, giving 2.000 mΩ. Their half-sum is +0.20 mV, an estimate of the stable offset. At a 10 A operating current, that measured path would drop about 20 mV and dissipate 0.20 W. If the sense span covers only one conductor, the full loop may be larger. Check Wire Resistance against the predicted geometry, then use the Wire Voltage Drop Calculator with the complete path resistance and operating current.

Control Heating and Compare Like Temperatures

Test current itself heats a low-resistance path by I²R. At 100 A, a 2 mΩ span dissipates 20 W; a long test may change its resistance, so use a safe current and pulse duration appropriate to the instrument and cable. Record conductor or terminal temperature as close as practical to the tested span. For moderate changes over a documented range, use the Wire Resistance vs Temperature Calculator to compare a measured hot value with a reference-temperature prediction. Do not use an ambient room reading as a substitute for the heated conductor temperature. Copper coefficient models are approximations and do not explain a degraded crimp or oxide film; contacts may have nonlinear or time-dependent behavior.

For a screening check, a copper span of 2.000 mΩ at 20°C and alpha = 0.00393/°C predicts approximately 2.3144 mΩ at 60°C, assuming the linear model applies. At 10 A this changes the estimated drop from 20.0 to 23.144 mV. A much larger measured shift should prompt investigation of temperature distribution, joints, probe placement, or material assumptions before it is attributed to copper alone.

Validate the Installed Load Path

A current-source bench test may exclude the source impedance, fuse, switch, return, or connector present in service. Under representative load, measure voltage at the supply terminals and directly at the load terminals, simultaneously if the current varies. Relate the difference to current and identify which segment is responsible with additional sense points. For pulsed loads, scope bandwidth and probe grounding matter; do not treat a DC milliohm measurement as the complete AC impedance of a cable. Check connector heating, strain relief, and current-sharing branches at the worst credible operating state. Record a baseline so later contact degradation can be detected.

• Use a current range that resolves millivolts but remains below conductor, terminal, and instrument limits.
• Verify zero offset and repeatability with a short or known resistance standard appropriate to the instrument.
• Keep force and sense contact positions fixed; document the span on the drawing.
• Repeat at cold and hot operating states, and include the outbound and return paths.
• Separate conductor, termination, and whole-loop measurements instead of comparing unlike spans.
• For safety-critical wiring, follow the applicable product and installation requirements beyond this electrical screening test.

Common Mistakes

• Using two-wire DMM ohms mode on a milliohm harness without a lead correction.
• Putting Kelvin sense points outside a bad connector, then calling the result wire-only resistance.
• Taking absolute values before current-reversal subtraction.
• Using insufficient test current for meter resolution or too much current for a hot cable.
• Ignoring thermoelectric voltage and source settling.
• Comparing one-way cable resistance with a two-conductor loop measurement.
• Applying a copper temperature coefficient to a complete path containing nonlinear contacts.
• Treating ambient temperature as conductor temperature under load.
• Using a DC value as though it predicted high-frequency cable impedance.
• Failing to log contact placement and test uncertainty for future comparisons.

Measurement Sources

Keysight low-resistance measurement guidance covers Kelvin connections, self-heating, and current-reversal treatment of thermal EMF. NI four-wire resistance documentation describes the separation of source and sense lead drops. Instrument compliance voltage and safe lead resistance must be checked against the selected equipment manual.