How hot is too hot: solenoid coil temperature limits
The number that matters is not an absolute temperature, it is the allowable temperature rise for the insulation class. Subtract the real ambient first — a coil in a 25 °C workshop and the same coil inside a machine enclosure at 55 °C have completely different allowable rises. Class B allows 80 K by the resistance method, class F 105 K, class H 125 K.
Why this happens
Almost every solenoid overheating dispute comes from comparing the wrong two numbers. One side quotes the insulation class, which is a maximum. The other side quotes a measured temperature, which was taken somewhere convenient. Neither number answers the question.
The insulation class describes the temperature the winding insulation system can survive. Class B is 130 °C, class F is 155 °C, class H is 180 °C. Those are absolute winding temperatures, not rises. A rise is what your current produces; temperature is where the coil ends up once you add the environment.
The standard expects a specific ambient. In practice, the allowable rise by the resistance method is quoted as 80 K for class B, 105 K for class F, and 125 K for class H. Those figures already carry a margin for the hot spot being hotter than the mean. So a class B coil in a 40 °C ambient is allowed to rise 80 K to reach 120 °C mean, with the hot spot landing somewhere near the 130 °C limit.
The consequence is simple and widely ignored: the same coil has less allowable rise in a hotter place. In a ventilated plastic housing in a 25 °C workshop, a class B coil has room. The same coil bolted inside an enclosure that sits at 55 °C in summer has 25 K less room. Nothing about the coil changed. Complaints of “it burns out in summer but not in winter” are usually this, not a coil defect.
There is a second reason the question goes wrong. Copper resistance rises with temperature, about 0.393% per kelvin. Once the winding is hot, the same voltage produces less current, so the coil partly limits its own heating. The temperature it settles at is a balance point, not a runaway, unless something else is feeding it — a higher supply, a stuck plunger, a duty cycle the designer never saw.
Check these in order
1. Know the insulation class and the allowable rise it implies. Class B: 130 °C / 80 K rise. Class F: 155 °C / 105 K. Class H: 180 °C / 125 K. Write both numbers down before you measure anything, because you will need the rise budget, not just the ceiling.
2. Measure the ambient where the coil actually lives. Not the workshop, not the test bench. Put a thermocouple or a logger inside the machine enclosure, near the coil, with the machine running under normal load. I have seen a 25 °C workshop and a 52 °C interior of the same machine on the same afternoon. That 27 K difference is most of a class B budget.
3. Measure cold resistance accurately, at a known temperature. Four-wire if you have it. Record both the resistance and the winding temperature at that moment. Twenty degrees of error in your cold reference becomes twenty degrees of error in every rise you calculate afterwards.
4. Run to thermal equilibrium, then measure hot resistance. Every 15 minutes until three consecutive readings agree. Then compute the mean rise: rise = (R_hot / R_cold − 1) / 0.00393. For a coil whose resistance grows 27% from cold to hot, that is a mean rise of about 69 K.
5. Add the hot spot allowance. The resistance method gives you a mean winding temperature. Standards assume the hot spot runs above the mean, and the allowable rises already include that margin — so compare your measured rise against the class figure directly, do not add a second allowance on top. But do compare against the ambient, not against 20 °C.
6. Compare against the rise budget, in the real ambient, on the worst day. Rise measured plus ambient inside the enclosure on a hot afternoon. If that sum plus a sensible margin crosses the class limit, the design has no margin, even though today’s reading looks acceptable.
What actually to change
| Finding | What to change | Why not the other thing |
|---|---|---|
| Rise within class but ambient is high | Ventilate the enclosure, move the coil away from other heat sources | Raising the insulation class treats the symptom and costs more |
| Rise itself too high, continuous duty | Reduce duty cycle, or enlarge the copper window | Raising supply voltage shortens life instead of extending it |
| Rise too high only during long holds | Add PWM hold or a mechanical hold | Cheaper than re-winding, and it removes the heat source |
| Coil potted in plastic, no conduction path | Provide a metal path from the coil to the frame | Insulation class upgrade does not fix a thermal bottleneck |
| Measured with IR, number disagrees with resistance method | Trust the resistance method | IR reads the surface, and surfaces are not windings |
| Hot spot concern at the end of the winding | Improve winding distribution rather than raising class | End turns run hot because of geometry, not material |
When it IS the harder problem
The duty cycle your customer reports is not the duty cycle running. This is the single most common cause of a coil that heats beyond any calculation. The specification says 50% duty. The PLC program has one branch that holds the coil for 30 seconds at startup, or a fault condition that latches the output, or a manual override an operator leaves engaged. Current-clamp the coil for a full cycle and time the on-period yourself. Percentage duty is on-time divided by cycle time, and a “50%” that actually runs 70% changes the thermal picture completely.
A plastic housing that traps the heat it should be shedding. The copper loss is the only heat source, but copper is rarely the bottleneck — the bottleneck is whatever sits between the winding and the outside. A potted coil in a closed plastic housing with no metal path to the frame accumulates heat in the winding while the outside of the housing still feels warm rather than hot. That gap between surface temperature and winding temperature is exactly why IR readings mislead people into thinking there is no problem.
Equilibrium you never reached. Coils have thermal mass, and a coil with poor conduction paths can take three hours to settle. Test benches run for 30 minutes precisely because that is how long the operator is willing to stand there. A reading taken at 30 minutes on a slow-settling coil can be 20 K below the true equilibrium, which is more than enough to flip a pass into a fail on a hot day.
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 temperaturesolenoid coil temperature risesolenoid insulation classsolenoid overheatingsolenoid coil hot to touch
Frequently asked
- How hot is too hot for a solenoid coil? Give me a number.
- There is no single number, and that is why this question causes so much trouble. A class B winding is rated for 130 °C, but that 130 °C includes the ambient. In a 40 °C ambient the allowable rise is 80 K. In a 55 °C enclosure it is effectively 65 K. Same coil, same insulation, different answer.
- Can I just use an infrared thermometer on the outside?
- Only as a rough check, and only to confirm something is wrong. An IR reading gives you the surface of the housing or the exposed winding end, not the mean winding temperature, which is what the insulation class is defined against. The resistance method is the one that maps to the standard: measure cold resistance, measure hot resistance after thermal equilibrium, and the ratio gives you the mean rise.
- How long until the coil reaches thermal equilibrium?
- Longer than most test benches allow. Small tubular coils on a metal frame can settle in 40 to 60 minutes. A potted coil in a plastic housing with poor conduction to anything can still be climbing after three hours. If you measure every 15 minutes and the resistance is still rising at the hour mark, your test is not finished.
- The coil is rated class H, so I have 180 °C to work with. Right?
- That is the trap. Class H means the insulation system survives 180 °C, not that running there is good design. At 180 °C you have no margin for the ambient inside the enclosure, no margin for a hot day, and no margin for the resistance rise that your own current is causing. Design against the allowable rise, then subtract the hottest ambient the machine will ever see.