My insulation resistance reads over 100 megohms cold. Why does the coil still fail?

Because insulation resistance is temperature dependent and the failure is usually a hot-state event. Windings expand when hot, marginal lacquer films yield, and a lead-out that is fine at rest can be pulled against a sharp edge. A coil that reads 200 megohms at 20 °C can read a few megohms at 90 °C. Measure hot, or measure in an oven at working temperature.

How do I find a turn-to-turn short? The ground test passes.

Turn-to-turn is a different failure and a megohmmeter cannot see it, because the shorted path is inside the winding rather than to the frame. Compare inductance or impedance against a known-good unit of the same batch, or use a surge comparison test. A coil with a shorted turn typically reads lower inductance and draws more current while producing less force.

What test voltage should I use for the hipot test?

Follow the standard that applies to your product rather than picking a number, because the test is destructive at the margin and an over-voltage test can create the defect it was meant to find. Common practice for low-voltage coils is around twice rated voltage plus 1000 V, but the applied value, the ramp, and the duration all come from the specification you are claiming compliance with.

Where do these failures usually start?

At the lead exit, more often than anywhere else. It is where the winding is least supported, where the wire is bent, and where any movement under thermal cycling concentrates stress. Inspect that region first under magnification, and test it with the lead held in the position it sits in the machine rather than hanging free.

Coil short to ground: how to test it properly

14 September 2026

A cold insulation resistance reading will pass a coil that fails in service. Insulation breakdown is often a hot-state event: the winding expands, a marginal lacquer film yields, a lead-out flexes. Measure at working temperature, test the lead exit specifically, and remember that a turn-to-turn short will not show up in a ground test at all.

Why this happens

Insulation failure is a temperature story with a mechanical component, and testing it cold tests only the easy half. The lacquer film on magnet wire is thin — tens of microns — and it is asked to survive the differential expansion between copper and whatever it is pressed against, across thousands of thermal cycles, while possibly sitting in humid air.

Three things change when the coil gets hot. The winding expands, so any point where the wire is pinched against a form or a lead-out gains pressure. The lacquer softens slightly, so its dielectric strength falls. And absorbed moisture redistributes, which is why a coil that has been sitting in a humid workshop overnight can read differently from the same coil after an hour of running. A cold, dry measurement can miss all three.

The second thing to understand is that “short to ground” and “turn-to-turn short” are different failures with different tests. A ground short is a path from the winding to the housing or the core, and a megohmmeter finds it. A turn-to-turn short is inside the winding, where two adjacent turns touch and bypass part of the coil. The coil still has full insulation to ground, so the ground test passes, but the effective turns count drops, inductance falls, current rises, and force drops.

Check these in order

1. Cold insulation resistance, and treat it as a baseline rather than a verdict. Megohmmeter at the voltage the applicable standard specifies, typically 500 V DC for low-voltage windings. Record the value and the temperature together. A reading without a temperature attached tells you almost nothing.

2. Resistance at working temperature, immediately after the unit is hot. Bring the coil to thermal equilibrium, disconnect, and measure within a minute or two. This is the measurement that catches the failures that matter. If the reading collapses by two orders of magnitude when hot, you have found a marginal insulation system even though it passed cold.

3. Inductance or impedance comparison for turn-to-turn. Measure inductance on the suspect unit and on a known-good unit of the same design, at the same plunger position and the same temperature. A significant drop points at shorted turns. Do not measure at different plunger positions — inductance varies with gap, which will swamp the effect you are looking for.

4. The lead exit, specifically. This is where most of these failures begin. Flex the lead gently in the direction it is routed in the machine while the meter is connected. A reading that flickers under that flex is a real defect that will become an open circuit or a ground short in service. Inspect under magnification for a nicked conductor or a lacquer film scraped through against an edge.

5. Hipot to the applicable specification. Ramp, hold, and duration from the standard you are claiming, not from habit. Note that this is a destructive-at-the-margin test: passing it does not prove long-term survival, and over-testing can create the weak point that fails six months later.

6. Humidity conditioning before the hot test if the field is humid. A coil that passes dry and fails wet needs moisture conditioning in the test plan, because that is the condition it lives in. Dry-state testing of a product installed in a condensing environment is a test of the wrong thing.

What actually to change

Finding	What to change	Why not the other thing
Fails hot, passes cold	Change insulation grade, improve impregnation, reduce coil temperature	A thicker wire does not raise the film’s thermal limit
Turn-to-turn short	Review winding tension and wire handling at the nozzle	Ground insulation is not the problem here
Lead exit damage	Add strain relief, change the exit radius, specify a sleeving	Rewinding will fail the same way at the same place
Condensation in service	Pot or varnish the winding, add a drain path, seal the housing	Class upgrade does not stop water ingress
Winding tension too high	Reduce tension, or specify a wire with a tougher film	Tension damage is created before the coil is finished
Field failures after thermal cycling	Add thermal cycling to the qualification test	A single hot test does not reproduce cycling damage

When it IS the harder problem

A failure that only appears after hours of running. A unit that measures clean cold and clean at equilibrium, but fails in the field after several hours, is telling you that the failure condition is a combination rather than a temperature. Thermal cycling moves the winding slightly each time; cumulative movement eventually abrades a film or a lead. A single hot measurement does not reproduce this, which is why a qualification test based on one steady-state point can pass a design that will not survive a year. Add cycling to the test plan, with the unit mounted the way it is mounted in the machine.

The lead exit as the real defect and the winding as the innocent party. I have seen several coils replaced over a complaint that turned out to be a lead-out routed across a sharp edge with no sleeving, where the failure was mechanical abrasion rather than anything to do with the winding or the insulation class. The tell is that the resistance or insulation reading changes when the harness moves. Once you have seen one, you check the lead exit first on every subsequent case, and it costs two minutes.

A margin that only exists on paper. Insulation class figures are qualified on specimens under defined conditions. In a real coil, moisture, abrasion, thermal cycling, and the mechanical stress at the ends of the winding all combine to make the practical limit lower than the class figure suggests. Designing to the class number with no allowance for the environment is the same mistake as designing to a nominal duty cycle with no allowance for the longest hold: it works until conditions are worse than the assumption.

A note on what this page is

This is a personal notebook, not a product page. I write down the checks that actually decide the outcome and the order to run them in, including the ones I got wrong first.

solenoid coil insulation testsolenoid short to groundsolenoid coil megohmmetersolenoid hipot testsolenoid turn to turn short

Frequently asked

My insulation resistance reads over 100 megohms cold. Why does the coil still fail?: Because insulation resistance is temperature dependent and the failure is usually a hot-state event. Windings expand when hot, marginal lacquer films yield, and a lead-out that is fine at rest can be pulled against a sharp edge. A coil that reads 200 megohms at 20 °C can read a few megohms at 90 °C. Measure hot, or measure in an oven at working temperature.
How do I find a turn-to-turn short? The ground test passes.: Turn-to-turn is a different failure and a megohmmeter cannot see it, because the shorted path is inside the winding rather than to the frame. Compare inductance or impedance against a known-good unit of the same batch, or use a surge comparison test. A coil with a shorted turn typically reads lower inductance and draws more current while producing less force.
What test voltage should I use for the hipot test?: Follow the standard that applies to your product rather than picking a number, because the test is destructive at the margin and an over-voltage test can create the defect it was meant to find. Common practice for low-voltage coils is around twice rated voltage plus 1000 V, but the applied value, the ramp, and the duration all come from the specification you are claiming compliance with.
Where do these failures usually start?: At the lead exit, more often than anywhere else. It is where the winding is least supported, where the wire is bent, and where any movement under thermal cycling concentrates stress. Inspect that region first under magnification, and test it with the lead held in the position it sits in the machine rather than hanging free.