Solenoid duty cycle in vending machines: why 100% is a trap

Why this happens

Duty cycle is the parameter most often read as a property of the component and most often wrong by a factor of two in the field. The reason is that the number which actually limits a coil is a temperature, and temperature depends on where the coil is standing.

Start from the physics. The winding fails at a temperature, not at a power level. The temperature it reaches is the ambient plus whatever rise its own losses produce. If the ambient goes up by 30 K, the rise allowed goes down by 30 K, one for one. For a class B winding with roughly 95 K of rise available in a 25 °C workshop, moving into a cabinet at 55 °C leaves about 65 K. Since the temperature rise a coil develops is roughly proportional to its dissipated power, the continuous power it can carry falls to around two thirds:

Ambient	Allowable rise (class B)	Continuous power available
25 °C workshop	~95 K	reference (100%)
40 °C	~80 K	~84%
55 °C cabinet	~65 K	~68%
70 °C near a heater block	~50 K	~53%

So a coil sold as continuous duty is not lying; it is answering a question about a 25 °C bench. Inside a vending machine, with a compressor, a display and a power supply sharing the cabinet, the same coil is a roughly 65% duty device. That is the trap, and it is why a specification that says only “100% duty” is not a specification.

There is a second trap inside the definition. Percentage duty is on-time divided by cycle time, and it is only meaningful when the cycle is short compared with the thermal time constant of the coil. A coil has a slow thermal response — minutes, for anything potted. Thirty seconds on and thirty seconds off lets the winding cool between pulses; five minutes on and five minutes off does not reach the same steady state at all, even though both are 50%. Ratings are conventionally quoted for cycles of a bounded duration for exactly this reason, and a customer reading “50%” as a licence to run five-minute cycles is a predictable field failure.

Check these in order

1. Measure the ambient where the coil is mounted, under worst-case operating conditions. After a full sale cycle, with the machine at temperature, not at first switch-on. Log it rather than estimating it.

2. Convert the rating to the real ambient. Take the allowable rise from the insulation class, subtract the real ambient, and compare that with the rise the coil actually develops. This takes two minutes and settles most duty arguments.

3. Establish the cycle time, not just the percentage. Get the worst-case on-time and off-time in seconds, and compare the on-time with the coil’s thermal time constant. Long on-times need a steady-state calculation; short ones do not.

4. Check the heat path. Is the coil bolted to a metal frame, clamped against a casting, or floating inside a plastic housing? Conduction is the dominant path in a sealed box, because still air inside an enclosure removes very little heat.

5. Separate pull-in from hold. If the drive holds the coil at full current after pull-in, most of the thermal load is being spent doing nothing. Measure the current waveform, not just the average.

6. Check whether the duty rating came with a mounting condition. A continuous rating for a coil in free air is not the same as a continuous rating for the same coil potted into a housing. If the datasheet does not say which, assume the more favourable one was intended.

What actually to change

Finding	What to change	Why not the other thing
Cabinet at 55 °C, coil continuous	Reduce duty cycle in the control software start	Raising the voltage raises the losses too
Long holds, full current	Add pull-in then hold-current drive	A bigger coil costs more and still generates heat
Needs no current when open	Use a latching mechanism	A holding solenoid can never be efficient
Still too hot after derating	Add copper window, or conduct heat to the frame	Providing vents changes the machine’s hygiene rating
Cycle time long, percentage nominal	Re-evaluate as steady state	The percentage alone does not describe the thermal load
Two similar coils, different ratings	Compare ambient, mounting, and cycle time	Different test conditions make datasheet ratings incomparable

When it IS the harder problem

The coil is potted into an insulating housing. Potting protects the winding and removes it from the air at the same time, so heat has to travel by conduction through the compound and the housing to the frame. A potted coil in a closed plastic housing with no metal path can still be climbing after two hours, which means a thirty-minute bench test measures a number that does not exist in service. This is the case where the thermal model matters more than the magnetic one.

The machine has to pass a surface-temperature requirement. Consumer equipment often has a touch-temperature limit, and the coil is inside the machine but the housing it heats may be outside. In that situation the thermal budget is set by the requirement and not by the insulation class, and the achievable duty cycle can be much lower than the winding alone would allow.

The duty cycle and the service life pull against each other. Every degree of winding temperature costs insulation life, so choosing to run at the top of the allowable range trades a shorter life for a cheaper coil. That is a legitimate engineering choice, but it should be made deliberately with the customer, written down as a temperature limit, and verified on the actual mounting — not left as the accidental outcome of a duty figure that was copied from a catalogue.