A compressor that trips its earth leakage, or fails intermittently with no other symptom, gets condemned on the strength of a single insulation resistance reading more often than any other test in the panel \u2014 and that single number, taken without temperature correction or a trend to compare it against, is frequently the wrong basis for the call. By the end of this article you’ll know the correct test voltage for 230 V and 400 V equipment, what 1 M\u03a9 and 100 M\u03a9 actually mean, how to correct a reading for winding temperature, and why the trend across service visits matters more than any single pass or fail.<\/p>\n\n\n\n
A multimeter tells you a circuit is complete. An insulation resistance tester \u2014 a megger \u2014 applies a DC test voltage across winding-to-earth insulation and measures the tiny leakage current that gets through it, finding damage a multimeter cannot see at all: insulation that holds fine at 230 V in normal operation but is already breaking down internally.<\/p>\n\n\n\n
Use 500 V DC on 230 V equipment and 1000 V DC on 400 V equipment.<\/strong> Testing a 400 V motor at 250 V under-stresses the insulation and will pass windings that go on to fail in service \u2014 the test voltage needs to be high enough to find the weakness the operating voltage will eventually exploit.<\/p>\n\n\n\n The absolute minimum for any HVACR motor is 1 M\u03a9.<\/strong> Below that, the winding has an earth fault and the compressor does not get energised \u2014 full stop, regardless of how it otherwise runs. A healthy compressor reads 100 M\u03a9 or better<\/strong>, and a good new hermetic often pegs the tester’s range well above 500 M\u03a9. Treat everything between those two figures as a graded scale, not a binary pass or fail:<\/p>\n\n\n\n A reading of 40 M\u03a9 is not a failure. It is also not the same compressor it was two years ago, and it belongs in the log rather than dismissed because it cleared 1 M\u03a9 comfortably.<\/p>\n\n\n\n Insulation resistance is temperature-dependent, and the relationship is large enough to flip a pass into a fail depending only on when you measured it. Insulation resistance roughly halves for every 10 \u00b0C rise in winding temperature above the 25 \u00b0C reference.<\/strong><\/p>\n\n\n\n Work an example both directions. A compressor tested immediately after shutdown reads 8 M\u03a9 at 60 \u00b0C<\/strong> \u2014 35 \u00b0C above the 25 \u00b0C reference, which is three and a half doublings (35 \u00f7 10 = 3.5). Correcting upward: 8 M\u03a9 \u00d7 2^3.5 \u2248 90 M\u03a9 at 25 \u00b0C<\/strong> \u2014 comfortably healthy. Taken at face value, uncorrected, that 8 M\u03a9 reading sits in the “deteriorated, investigate” band and would send you looking for a fault that doesn’t exist.<\/p>\n\n\n\n The reverse trap is the dangerous one. A compressor tested cold reads 3 M\u03a9 at 25 \u00b0C<\/strong> \u2014 marginal, but technically above the 1 M\u03a9 floor. Run that same winding hot, at 60 \u00b0C, and the true reading is roughly 3 M\u03a9 \u00f7 2^3.5 \u2248 0.3 M\u03a9<\/strong> \u2014 a genuine failure, hidden by the fact that you tested it cold. A cold reading flatters the machine.<\/strong> Always record winding temperature with every insulation resistance reading, and correct the figure \u2014 or at minimum, note the temperature alongside it so the next technician can.<\/p>\n\n\n\n A single insulation resistance reading tells you where a compressor stands today. A series of readings across service visits tells you where it’s going, and that trend is the earlier and more useful warning.<\/p>\n\n\n\n Consider a compressor that read 4,000 M\u03a9 eighteen months ago, 800 M\u03a9 at the last service, and 150 M\u03a9 today.<\/strong> Every one of those readings clears the 1 M\u03a9 floor by a wide margin \u2014 technically, it still passes. But the trajectory is unmistakable: each reading is roughly a fifth of the one before it, and a machine losing insulation resistance at that rate is heading toward failure, not sitting stable at a healthy number that happens to be lower than it once was. A falling trend across successive visits is the real early warning, worth logging every time specifically so the pattern is visible<\/strong> \u2014 not just whether this visit’s reading clears the floor.<\/p>\n\n\n\n Log winding temperature alongside every reading for exactly this reason: without it, you can’t tell whether a lower number this visit is genuine deterioration or simply a hotter compressor than last time.<\/p>\n\n\n\n This is an absolute rule, not a guideline. An inverter drive’s output stage contains IGBT switching devices and their anti-parallel diodes, along with MOVs and Y-capacitors bridging to earth for EMC suppression. Applying 500 V or 1000 V DC through that path destroys the drive<\/strong> \u2014 semiconductor devices rated for the drive’s normal switching voltages have no tolerance for a megger’s test potential, and you’ll put a new drive on the invoice for a test that should never have touched it.<\/p>\n\n\n\n The correct procedure:<\/strong><\/p>\n\n\n\n The same rule applies to any panel with electronic controllers, earth leakage units, soft starters, or surge protection devices in the current path \u2014 test the motor alone, never the motor through the electronics.<\/p>\n\n\n\n Safety note:<\/strong> a drive’s DC bus can hold a lethal charge for many minutes after isolation regardless of how quiet or dead the panel looks. Confirm bus voltage below 50 V with your meter before touching any bus terminal, capacitor, or output conductor \u2014 never rely on a stated wait time without measuring it, since a failed bleed resistor leaves the bus charged well past the expected window.<\/p>\n\n\n\n Insulation resistance to earth is only half of a proper winding test \u2014 the other half is confirming you’re actually measuring the compressor’s windings and not a miswired terminal. On an unmarked or relabelled three-phase or single-phase compressor, get the terminal identification right first; see how to identify C, S and R terminals on an unmarked compressor<\/a> for the resistance method that does it without guesswork. On three-phase machines, also run the winding-to-winding balance check \u2014 all three pairs within 5% of each other \u2014 before you conclude an insulation failure explains a trip: a winding imbalance and a genuine insulation fault can produce overlapping symptoms, and a compressor tripping intermittently sometimes turns out to be voltage imbalance on the three-phase supply<\/a> rather than anything wrong with the winding insulation at all.<\/p>\n\n\n\n A three-phase compressor was suspected open-circuit after a technician found unusual voltage readings at the contactor. Before condemning the windings, insulation resistance was tested at 500 V DC, winding to earth, with the compressor isolated and at 24 \u00b0C: over 500 M\u03a9 on all three windings<\/strong> \u2014 no correction needed at that temperature, and well clear of the healthy threshold.<\/p>\n\n\n\n Winding resistance confirmed it: run to common measured 3.8 \u03a9, start to common 11.2 \u03a9, and run to start 15.0 \u03a9. Since 3.8 + 11.2 = 15.0, both windings were continuous, in series as expected, with no shorted turns anywhere. The compressor was sound. The actual fault, traced separately, was a burnt contactor pole feeding a low reading that looked like a wiring problem \u2014 insulation resistance and winding resistance together are what cleared the compressor before anyone spent money replacing it.<\/p>\n\n\n\nThe Numbers: 1 M\u03a9 Is a Floor, Not a Target<\/h2>\n\n\n\n
Reading<\/th> Verdict<\/th> Action<\/th><\/tr><\/thead> Over 100 M\u03a9<\/td> Healthy<\/td> Return to service<\/td><\/tr> 20\u2013100 M\u03a9<\/td> Ageing or slight moisture<\/td> Record, retest at next service<\/td><\/tr> 5\u201320 M\u03a9<\/td> Deteriorated<\/td> Investigate: oil contamination, moisture, acid<\/td><\/tr> 1\u20135 M\u03a9<\/td> Marginal<\/td> Do not return to service without investigation<\/td><\/tr> Under 1 M\u03a9<\/td> Failed<\/td> Do not energise \u2014 winding-to-earth fault<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n Temperature Correction Is Not Optional<\/h2>\n\n\n\n
The Trend Matters More Than Today’s Number<\/h2>\n\n\n\n
Never Megger Through an Inverter Drive<\/h2>\n\n\n\n
Reading the Windings Correctly Before You Test Insulation<\/h2>\n\n\n\n
Worked Example: A Butchery Compressor Cleared on the Numbers<\/h2>\n\n\n\n
Final Thoughts<\/h2>\n\n\n\n