Air Compressor Won’t Stop Running Unloader and Pressure Switch Diagnosis

Diagnosing an air compressor that won't stop running starts with one thing: look at the pressure gauge. If the tank pressure has reached or exceeded the cut-out value on the nameplate and the motor is still running, the problem is on the electrical side of the pressure switch. If the tank pressure is nowhere near the cut-out value and climbing slowly or not at all, something on the air side is leaking. That one glance cuts the search area in half, and all disassembly and testing should happen only after confirming which direction to go.

A surprising number of people start tearing things apart without making this determination first.

The Mechanical Linkage During Shutdown

When tank pressure reaches the cut-out pressure, the sensing element inside the pressure switch pushes a linkage, the linkage separates the contacts, and at the same time pushes the unloader pin out to unseat the unloader valve. Power cutoff and pressure relief happen within the same linkage stroke.

This shared-stroke design buries a long-term interaction that doesn't get enough attention. When carbon buildup makes the unloader valve spool stiff, the pin needs more travel and more force to push it open, which eats into the travel available for contact separation. The contact gap shrinks, arcs can't extinguish cleanly, and erosion accelerates. The Condor MDR11 and Lefoo LF10 are two of the most widely installed pressure switch models on the market, and plenty of people have cracked them open. The two have subtle differences in how their linkage mechanisms allocate stroke. The MDR11 has slightly more pin travel margin, so it tolerates increased unloader valve resistance a bit better. The LF10 is more compact in design, which means even a small increase in valve stiffness on that side can start affecting the contact side. This is not a knock on the LF10. It just means the LF10 is more sensitive to unloader valve maintenance condition, and the cleaning interval for the unloader valve should be shorter when paired with one.

Pressure Switch Contacts

Lifespan Is Calculated, Not Guessed

Contact electrical life is measured in switching cycles. A typical pressure switch is rated for around one hundred thousand cycles. That number only means something when you put it next to tank capacity and air usage pattern.

Take the Fusheng V-0.17/8 as an example. 24-liter tank, driving a nail gun with intermittent use, cycling roughly twenty-five to thirty times per hour. Four hours a day, three hundred days a year, that's about thirty thousand cycles per year. Contact life is used up in just over three years.

Now take a Jaguar 7.5kW screw compressor with a 300-liter tank feeding a paint booth. During continuous draw the tank pressure drops slowly, maybe three or four cycles per hour. The contacts last many years.

The same pressure switch, installed on different machines, can have a fivefold or even tenfold difference in service life. The deciding factor is not the quality of the switch itself. It is tank volume and usage rhythm.

So anyone running a small-tank compressor should treat the pressure switch as a routine replacement part, same category as the air filter. Don't wait until it won't stop running to remember this thing exists. High-frequency cycling on a small tank is hard on every component in the machine. The pressure switch is just the first one to go down.

Welding and High Resistance Are Two Different Failure Modes

Contact erosion in its later stages goes one of two ways.

Contact welding is when the contact surfaces partially melt and fuse together. Spring force can't pull them apart. The circuit won't break. This is the classic direct cause of a compressor that won't stop. Silver tin oxide alloy has lower yield strength at elevated temperatures, so welding probability tracks with heat. It's more likely in summer and after prolonged high-load operation. A compressor installed in a poorly ventilated corner with the pressure switch right next to the pump head exhaust is on a fast track to welding.

High-resistance contact is when the carbonized layer on the contact surfaces gets so thick that closed-contact resistance climbs. The motor doesn't get enough voltage, runs sluggish, output drops, pressure builds slowly. This gets misdiagnosed as worn cylinders or failed valve plates all the time. The pump head gets torn down, everything looks fine, it goes back together, and the problem is still there. Measure the voltage drop across the pressure switch contacts while closed. It should be near zero. A high-resistance contact will show several volts or more.

Welding causes won't-stop-running. High resistance causes what-looks-like-efficiency-loss. Both symptoms point to the same part. The conditions and presentation are completely different.

Should You Sand the Contacts

Filing down eroded contacts with fine sandpaper and putting them back in service is a practice that shows up constantly in forums and videos.

The conductive layer on silver alloy contacts is thin. One pass with sandpaper removes the carbon but also removes part of the silver alloy plating. With thinner plating left behind, the next round of erosion arrives faster. And because the surface is now rougher, arcing concentrates on the high points, uneven erosion makes localized overheating worse. Contacts that have been sanded once have significantly shorter remaining life.

If it's an emergency, there's no replacement switch on hand, and the machine needs to run for a few more days, sanding can buy some time.

As a regular maintenance practice it doesn't make sense. A new pressure switch costs the equivalent of a few dollars and solves it cleanly.

Diaphragm and Springs

Diaphragm hardening and spring fatigue look similar from the outside. Both cause the cut-out pressure to drift upward. The way to tell them apart is the adjustment screw.

Spring fatigue: turn the cut-out adjustment screw back to lower the setpoint. If normal shutdown resumes, the spring is still within compensable range. If the screw has to be turned close to its limit before the switch operates normally, the spring is significantly fatigued and the usable adjustment range of this switch is almost used up.

Diaphragm hardening: after lowering the setpoint, shutdown behavior is inconsistent. Sometimes it stops, sometimes it doesn't. Diaphragm hardening is not a fixed offset. It's affected by temperature, oil saturation level, and deformation history, so the response varies from cycle to cycle. This intermittent behavior is a fairly distinctive signature of a diaphragm problem.

Diaphragm perforation is yet another matter. The sensing chamber opens directly to atmosphere, and the switch essentially loses its ability to sense pressure. Test method: remove the switch, blow gently into the sensing port while blocking the relief port. If it won't hold any pressure at all, the diaphragm is perforated.

The probability of these three failures (spring fatigue, diaphragm hardening, diaphragm perforation) correlates with age. Within the first two or three years, the vast majority of failures are contact-related. Diaphragm and spring issues generally start showing up on machines that are four or five years old and beyond. This timeline distribution helps narrow the search.

Sensing Port Blockage

Compressed air coming out of the pump head contains oil mist, moisture, and particulates. That two-to-three-millimeter sensing port is going to get clogged with oil sludge eventually.

The signature symptom of a blocked sensing port: both the cut-in and cut-out actions become sluggish at the same time. The cut-out value drifts high while the cut-in value also drifts low, with the tank dropping to very low pressure before startup kicks in. If only one direction is abnormal, look at springs and diaphragm. If both directions are sluggish simultaneously, the sensing channel itself is obstructed.

After clearing the blockage, don't blow compressed air back through the port. Oil sludge fragments will get pushed into the switch body and coat the diaphragm face. Use a thin steel wire to clear it mechanically, flush from outside in with solvent, then dry with low-pressure air.

Unloader Valve

Unloader valve diagnosis is relatively straightforward. Not nearly as many twists as the pressure switch side.

Stuck Open

The spool isn't seating. Compressed air bleeds out the unloader port continuously during operation. Tank pressure can't build up. Hold a hand near the unloader port and you'll feel a continuous stream of warm air. Warm is the key detail here. Air that just went through compression is significantly hotter than ambient. Leaks at pipe fittings downstream of the tank are cool because the air has already shed its heat through the tank walls and piping. Temperature tells you whether the leak is on the pump head side or downstream of the tank.

During operation, no airflow at the unloader port. At the moment of shutdown, one short burst of exhaust. That's normal. Hissing during operation is not normal. Complete silence at shutdown means the unloader didn't open.

Stuck Closed

The spool is jammed and won't push open. After shutdown the line pressure between the pump head and check valve can't bleed off. Next startup, the motor has to turn against that trapped pressure. Single-phase motors have limited starting torque. Stall current runs five to seven times rated current. The start capacitor takes the hit first.

This has come up on Jaguar small piston units. Two start capacitors burned in short succession. Before installing the third one, checked the unloader valve. The spring had snapped partway through. The spool could push out but couldn't return. Replaced the spring, capacitor problem disappeared. One spring segment worth a few cents, and before that, two capacitors had already been burned through.

When a machine is burning through start capacitors in quick succession, before installing the next one, reach over to the unloader port and feel for exhaust at shutdown. If there's no exhaust pulse, the capacitor is not the root cause. The unloader valve is.

Pin Length

An easy trap to fall into when replacing a pressure switch. Thread matches, electrical ratings match, it goes on, everything seems fine at first. A one-millimeter difference in pin length might not show its effects for months. Too long and the spool can't fully seat, causing run-time leakage, while also stealing travel from the contact side and accelerating erosion. Too short and the spool doesn't open, unloading fails.

The Condor MDR series and Lefoo LF10 have pin length differences between them. Before swapping, measure the pin protrusion on the old switch and compare it to the new one. If calipers aren't available, set the old and new switches side by side on a flat surface and eyeball it. If the difference is obvious, don't install it.

Combined Faults

The check valve leaks back slightly. The pressure switch cut-out has drifted up a few PSI. A quick-connect fitting somewhere has a micro-leak. None of the three constitutes a fault on its own. Stacked together, either the pressure never quite reaches cut-out, or it reaches it and the switch action isn't clean.

There is no shortcut for diagnosing combined faults. The only approach is isolation under zero-leak conditions. Close the tank outlet ball valve. Disconnect all downstream piping. Let the compressor pump into a sealed tank. If it shuts off normally under these conditions, the entire problem is downstream. If it still won't shut off, at least downstream leakage is eliminated as a variable, and focus can shift to the pump head side and the switch side. If pressure on the gauge starts dropping after shutdown, the check valve is leaking back.

This test should be done before taking anything apart. Takes two minutes.

A Word on Safety

Disconnect power and drain the tank to zero before putting hands on anything. The range spring inside the pressure switch is under preload. Go slow when removing the cover. Clean unloader valve spools get a thin coat of silicone-based grease before reassembly. Don't use motor oil. The rubber seals can't handle it.

Air Compressor Won't Stop Running Unloader and Pressure Switch Diagnosis