A milliohm meter can screen switchgear connections when its usable resolution, test current, and repeatability match the resistance being measured. It cannot qualify a joint merely because the display reaches its lowest count. The commissioning decision must come from a defined measurement boundary, controlled four-wire connections, comparable test conditions, and an acceptance limit established by the equipment manufacturer or maintenance program.
Measurement objective and instrument capability
The term repeatability means obtaining closely grouped results when the same measurement is repeated under unchanged conditions. Accuracy describes closeness to the true resistance. Phase-to-phase comparison depends heavily on repeatability, but an absolute acceptance limit also requires suitable accuracy and calibration.
A claimed lower reading of 0.01 mΩ converts to 10 µΩ, not 100 µΩ. Display resolution is not the same as accuracy. If a breaker or bus joint is expected to measure around 10–25 µΩ, a 10 µΩ increment represents only 1 to 2.5 display counts. That is inadequate for resolving small differences even if repeated readings appear stable.
| Test objective | Required capability | Decision |
|---|---|---|
| Compare three equivalent phases | Repeatable readings under identical conditions | Use one instrument, range, current, connection method, and temperature basis |
| Apply an absolute resistance limit | Calibrated accuracy and resolution comfortably below the limit | Use the specified test method and instrument class |
| Measure a joint near the meter's lowest count | More usable digits and adequate test current | Select a micro-ohmmeter rather than treating an unresolved result as acceptable |
- Check 1: expect the instrument specification to state a range, resolution, accuracy, and test current suitable for the anticipated resistance—not merely a milliohm label.
Test boundary and equipment state
Define exactly which interface the result represents: one closed contact pair, one bolted joint, or a longer path containing several connections. A reading across multiple interfaces cannot identify which interface produced the resistance. Parallel conductors can bypass the intended joint and force the result artificially low.
De-energize and isolate the circuit under the site's electrical safety procedure before connecting a low-resistance meter. Stored energy and induced voltage can corrupt the reading or damage an instrument that is not designed for energized testing. Place breakers or disconnects in the operating state required by the test plan; contact-resistance testing normally requires the tested contact path to be closed.
Record the equipment position, endpoints, conductor arrangement, ambient or conductor temperature, and whether parallel paths remain. These conditions become part of the baseline used for future maintenance.
- Check 2: expect the approved absence-of-voltage test to indicate a de-energized circuit and the defined path to contain only the interfaces listed on the test sheet.
Four-wire Kelvin connection
A Kelvin measurement injects current through two force leads and measures voltage through two separate sense leads. The meter calculates resistance from R = V/I. Because almost no load current flows through the sense circuit, voltage drop in the force leads and force-clip contacts is excluded from the calculated resistance.
Four-wire construction does not automatically remove every connection error. Place each sense contact inside its corresponding force contact so the measured voltage boundary includes the joint but excludes the force-contact drop. Attach clips to clean conductive surfaces rather than paint, plating contamination, oxidation, or moving hardware. Keep the sense points fixed while repeating measurements.
The described compact meter has Kelvin leads about 24 in. long. Extensions may improve access, but preserve four separate conductors through the extension and keep the sense connections independent of the force path. Ordinary two-wire extensions defeat the measurement principle. Added lead resistance may also reduce available current if the meter reaches its compliance-voltage limit.
- Check 3: short the Kelvin contacts together using the instrument's prescribed zero or relative procedure; expect a stable result at the meter's stated zero capability before moving to the joint.
Test current and signal level
Test current determines the voltage available for measurement. With the reported current of approximately 100 mA, a 10 µΩ joint produces only 1 µV, while 25 µΩ produces 2.5 µV. Such small signals make offset, electrical noise, thermoelectric voltage, surface contact, and lead movement significant.
Higher-current instruments produce a larger voltage across the same resistance. Instruments discussed for this class of work span 1 A, 10 A, 100 A, 200 A, and 600 A; those values describe different test capabilities, not interchangeable acceptance methods. Compare results only when test current and connection geometry are controlled. Current can also heat a small contact area, so follow the instrument and equipment test procedure rather than extending the injection arbitrarily.
Motor-winding measurements add inductance. Use an instrument designed for inductive loads, allow its current and displayed resistance to settle, and follow its discharge indication before removing leads. A meter that performs acceptably on resistors or switchgear joints is not automatically suitable for a large motor winding.
- Check 4: expect the displayed test current or instrument status to reach its normal measurement state and the resistance to settle without persistent drift or overload indication.
Comparative measurement procedure
- Identify equivalent paths, such as the three phase poles of the same disconnect, and label identical measurement endpoints.
- Condition the equipment consistently. Use the same contact position, mechanical state, and connection preparation for every path.
- Connect force and sense clips in the same physical order and orientation at each pole.
- Select one range and test-current setting. Do not change either setting during the comparison.
- Take repeated readings without disturbing the clips. Then remove and reconnect the clips to expose connection-placement variation.
- Record every reading rather than only the most favorable value. Calculate each path's average only after rejecting a result for a documented setup error.
- Calculate comparative spread as
(Rmax − Rmin) / Rmin × 100%. Compare that result with the percentage limit defined by the maintenance plan; no universal percentage is established here.
A double or triple resistance difference is readily visible, but a large difference still requires localization. Move the sense points inward across individual interfaces to determine whether the excess resistance belongs to a contact pair, bolted joint, conductor termination, or another element in the original path.
- Check 5: expect repeated, undisturbed readings on each path to cluster within the instrument's specified repeatability and equivalent phases to meet the documented comparison limit.
End-to-end commissioning verification
Repeat the complete sequence after disturbing and restoring each Kelvin connection. This separates true path resistance from a favorable clip position. Compare the new readings with the first set, the approved absolute limit when one exists, and historical baseline data taken with the same method.
If the display remains at its minimum count, report the result as below the usable measurement threshold rather than zero resistance. Substitute an instrument with finer usable resolution or a test current appropriate to the target range when the acceptance decision falls below that threshold. A stable reading on a known low-value reference can check repeatability; traceable calibration is required when the recorded result supports an absolute acceptance decision.
- Check 6: expect the reconnection results to reproduce the initial ranking and remain within the specified limit, with no unresolved minimum-count value used as proof of a good joint.
Frequently asked questions
Can I use a low-cost milliohm meter on switchgear?
Yes, for comparative screening when its test current, usable resolution, and repeatability cover the expected range. Use a calibrated higher-capability instrument when an absolute micro-ohm limit determines acceptance.
Does a higher test current always give a better result?
No. Higher current increases measured voltage, but the equipment procedure must permit that current and the result must be compared with data taken under the same conditions.
Can I approve a joint from one resistance reading?
No. Reconnect the four Kelvin contacts, repeat every equivalent phase with the same range and current, and verify that the reproduced readings meet the documented absolute or comparative limit.