Why would bolting a solenoid down change how it works?

Because bolting it down deforms the housing and the plastic bobbin between the coil and the mounting face. If the mating face is not flat, tightening the fasteners bends the assembly and changes the air gap. In the region where the solenoid works, force varies roughly as the inverse square of gap, so a change you cannot see can cost a large fraction of the force.

How much temperature difference matters?

Copper rises about 0.393 %/K. A coil at 90 °C has roughly 27 % more resistance than at 20 °C and, at a fixed supply voltage, carries about 22 % less current — and force tracks current. If the inside of the machine runs 40 °C above your bench, that alone can be enough to tip a marginal design over. Measure the ambient inside the machine, not the room.

Is a bench test even worth doing?

Yes, but it answers a different question. A bench test tells you the unit is functional in isolation. It does not tell you whether the installed air gap, the installed temperature, or the installed harness match the assumptions the design was made against. That is why a unit that passes on the bench can still fail in place.

What if it fails immediately on installation but works again when removed?

Then it is mechanical almost certainly: mounting-face flatness, fastener torque deforming the housing, or a linkage imparting side load. The magnetic design is not the variable here. Start with the flatness of the face and the torque you are applying.

Works on the bench, fails in the machine: the three things to measure

14 September 2026

If it actuates freely on the bench and fails once installed, measure three things before touching the solenoid: mounting-face flatness, ambient temperature inside the machine, and voltage at the coil. Mounting-face flatness changes the air gap when the unit is bolted down, and it is the single most frequently missed item.

Why this happens

A unit that works on the bench and fails in the machine is the most misread failure pattern in this trade, because the bench result is used as evidence that the solenoid is good. It is not evidence of that. A bench test shows the unit works in isolation. The machine is a different set of boundary conditions, and three of them move the result enough to matter on their own.

The aggravating factor is that these three rarely act alone. Each one individually sits inside specification. Together they consume the margin, and the customer feels the failure, while every inspection you run comes back clean.

Measure these three, in this order

1. Mounting-face flatness. Measure the flatness of the face on the machine, not just the mating face of the solenoid. Tightening fasteners against a face that is not flat bends the housing and the bobbin, which changes the air gap. In this operating region, force varies roughly as the inverse square of the gap, so a change you cannot see by eye can remove a large fraction of your force. This is the single most frequently missed item in field failures, and it is also the hardest one for a customer to self-check, which is why it survives so long.

2. Ambient temperature inside the machine. Not the room temperature — the temperature where the solenoid actually sits. Copper resistance rises about 0.393 %/K, so a coil at 90 °C has roughly 27 % more resistance than at 20 °C and, at fixed voltage, carries about 22 % less current. Force tracks current. If you tested at 25 °C on a bench and the machine runs at 65 °C, you have already given away a meaningful share of your margin. Measure it hot and running, not after the machine has cooled.

3. Voltage at the coil, under load, after the harness is installed. Not at the power supply. Long thin runs and installed connectors routinely drop a volt or more, and because force tracks current, that loss goes straight through to the result. Measure at the coil terminals while energised.

Two more worth checking while you are there

Side load from the linkage. If the mechanism pushes the plunger off-axis, you get friction plus a reduced effective pole area. It only appears once assembled, and it looks exactly like a weak solenoid.

Fastener torque. If the housing distorts under torque, more torque makes the failure worse. Try the unit in place with the fasteners at a lower torque and see whether the symptom changes — it is a fast, free diagnostic.

What actually to change

Finding	What to change	Why not the solenoid
Mounting face not flat	Machine or shim the mating face; specify a flatness tolerance	A different solenoid bolts to the same crooked face
Machine ambient 30–40 °C above bench	Reduce duty cycle, or increase copper window, or derate the force target	Raising supply voltage shortens life
Voltage at coil more than 5 % low	Shorten the run, upsize the wire, replace the connector	Cheapest fix, largest effect
Side load from linkage	Add a guide, or re-align the linkage	Re-specifying the solenoid does not remove the side load

When it IS the harder problem

Every inspection passes and it still fails. This is the signature of tolerances stacking, not of a single out-of-specification item. Face flatness at one limit, ambient at another, harness drop at a third — each individually acceptable, together past the margin. The fix is to design against the worst plausible combination rather than against each limit separately. I have been caught by this twice, and both times the mean looked comfortable while the lower tail did not.

It fails, then recovers when removed, then fails again on reinstallation. Usually mounting-face distortion or fastener torque, and the customer’s reinstallation re-creates it. Ask for the torque value they use, and for a photo of the mating face — the answer is often visible in the photo.

It fails only after running for a while. Thermal drift, and it will never reproduce on a short bench test. Run the unit for the full duty cycle in the machine before judging it, and measure the coil resistance at temperature so you can work backwards to the actual current.

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 works on bench fails installedsolenoid fails in machinesolenoid mounting flatnesssolenoid ambient temperaturesolenoid voltage drop

Frequently asked

Why would bolting a solenoid down change how it works?: Because bolting it down deforms the housing and the plastic bobbin between the coil and the mounting face. If the mating face is not flat, tightening the fasteners bends the assembly and changes the air gap. In the region where the solenoid works, force varies roughly as the inverse square of gap, so a change you cannot see can cost a large fraction of the force.
How much temperature difference matters?: Copper rises about 0.393 %/K. A coil at 90 °C has roughly 27 % more resistance than at 20 °C and, at a fixed supply voltage, carries about 22 % less current — and force tracks current. If the inside of the machine runs 40 °C above your bench, that alone can be enough to tip a marginal design over. Measure the ambient inside the machine, not the room.
Is a bench test even worth doing?: Yes, but it answers a different question. A bench test tells you the unit is functional in isolation. It does not tell you whether the installed air gap, the installed temperature, or the installed harness match the assumptions the design was made against. That is why a unit that passes on the bench can still fail in place.
What if it fails immediately on installation but works again when removed?: Then it is mechanical almost certainly: mounting-face flatness, fastener torque deforming the housing, or a linkage imparting side load. The magnetic design is not the variable here. Start with the flatness of the face and the torque you are applying.