Danfoss Pressure Switch Not Working? It's Probably Not the Switch — Lessons from 9 Years of Field Mistakes
A Wednesday afternoon, 2:30 PM. The phone rings. A maintenance tech on the other end, clearly frustrated: "The Danfoss KP15 pressure switch on the compressor isn't working. Ship us a replacement overnight."
I used to take calls like that at face value. Grab a switch from inventory, hand it to the shipping guy, pay whatever UPS charged for a next-day box, and wait for the inevitable second call. The replacement arrived. The problem didn't go away.
Somewhere around the fifth time this loop repeated, I started asking different questions. The answers reshaped my entire approach to pressure switch troubleshooting.
Quick introduction: I'm a field service technician who has handled pressure switch callouts for HVAC and industrial compressor systems for nine years. I've personally made—and documented—at least fourteen significant misdiagnoses. Combined, those mistakes wasted roughly $4,800 in parts, shipping, and labor that bought nothing. I keep a replacement checklist now. This article is the reason behind every line of it.
The surface problem: "the switch is dead"
The classic call goes like this: a plant operator says the compressor isn't cycling. The refrigeration side is misbehaving. Someone checks the pressure switch, doesn't see the expected behavior, and declares it defective.
The fix looks simple. Order a new Danfoss pressure switch. Swap it out. Done.
Except it isn't done. The symptom comes back because the switch was never the fault.
The deeper problem: how the working principle gets misread
"Pressure switch not working" usually means one of two things: it isn't switching when it should, or it's switching when it shouldn't. Both symptoms regularly trace back to a perfectly healthy switch being misunderstood.
Here's the Danfoss pressure switch working principle in plain terms. A Danfoss pressure switch is mechanical. Inside, a pressure element—bellows or diaphragm depending on the range—reacts to system pressure. When pressure crosses a set threshold, it trips a snap-action contact block. The contacts open or close, driving a compressor, fan, or alarm circuit. It's a control circuit device, the same family of electromechanical switching elements covered by IEC 60947-5-1.
The key word is set. A pressure switch doesn't know what pressure is correct for your system. It has an adjustable setpoint and a defined differential. If cut-in and cut-out points don't match the application, the switch behaves exactly wrong while being mechanically perfect.
That sounds elementary. I've still stood in front of a panel holding a perfectly good switch, one hand on the replacement order form, fully ready to fix a problem that switch was never causing.
One trap deserves extra attention. A fixed-hysteresis switch like the KP15 is not the same animal as an adjustable-differential switch like the RT series. The KP15 has a fixed gap between cut-in and cut-out. Adjust one point, and the other follows at that fixed spacing. If your process needs a narrow, tunable differential, no amount of turning the set screw will turn a KP into an RT. That's a selection error, not a switch failure.
The filter problem that keeps fooling everyone
A big slice of the "failed" switches I've pulled traced back to airflow conditions—not electrical faults.
Take the standard 14x18x1 air filter common in commercial rooftop units. It loads up with dirt. Pressure drop across the filter climbs. Static pressure in the duct shifts. The pressure switch, doing exactly its assigned job, responds to the changed condition. The system acts up. The technician blames the switch.
The switch was telling the truth. The filter was the liar.
I have a specific memory from August 2022: a rooftop unit on a small commercial building. The tech on site was certain the high-pressure switch had failed. The compressor was short-cycling, the cutout kept tripping. I pulled the switch, tested it on the bench, and it was fully functional. The real culprit was a dirty filter and a supply-air flow condition the switch was reporting correctly.
Since then, filter condition is the first thing I rule out. Even before I reach for a multimeter.
The wiring problem: the KP15 diagram is printed on the housing for a reason
The other "failure" that isn't a failure: wrong wiring.
The Danfoss KP15 pressure switch wiring diagram is printed right on the housing. Right on it. Still, enough installers skip it that I've dedicated an entire checklist section to reading it. The KP15 terminal block supports both "making" and "breaking" configurations, depending on which terminals you use. Pick the wrong pair, and the switch operates exactly backwards.
I made this exact mistake in my first year, 2016. I was wiring a KP15 into a pump control panel and assumed I knew the terminal functions. The pump ran on high pressure and shut off on low pressure—the opposite of what the system needed. I spent most of a day convinced the switch was defective.
The switch was fine. My confidence was the defect.
That day I learned to read the diagram and test the switch before forming conclusions. Equipment isn't ashamed of its instructions.
How a non-contact voltage tester works (and why you need one)
While we're on the electrical side: I've stopped trusting panel labels alone. When a tech is in a hurry—and a tech is always in a hurry—the temptation is to assume a circuit is dead because the label says so. Labels are sometimes wrong.
This is where a non-contact voltage tester earns its keep. It detects the electric field around a live conductor without needing to touch bare wire. You hold the probe near a terminal block, a wire, a switch contact. If it beeps or lights up, there's voltage present. No direct contact, no wire stripping, no reference ground needed. It's not a substitute for lockout/tagout, but it's a ten-second verification step that prevents a very bad day.
These days I won't touch a panel with pressure switch wiring without one.
The price of misdiagnosis
Now the costs. These are personal estimates from memory—I'd need to pull the service logs for exact numbers.
Over nine years I've been involved in roughly 200 pressure switch replacement jobs. Maybe 180. More like 210 if you count jobs where I was supervising. Let's settle on around 200.
My best estimate: 40 to 60 of those were unnecessary. The switch wasn't the problem. At $70 to $120 per switch, that's $4,500 to $6,000 in healthy hardware removed for no reason. Just the parts cost.
I once bench-tested ten "failed" switches pulled from different sites in the same month. Nine were fully functional. That comparison forced me to face an uncomfortable conclusion: the problem was often in our process, not the product.
Then there's the shipping and downtime. I once paid $47 for a UPS box and next-day delivery to move a $68 switch that turned out to be fine. The actual fix was a new filter and a setpoint adjustment. That $47, plus the $120 service call, could've been avoided by ten minutes of diagnostics.
Downtime is the real monster. A stopped production line isn't priced in switches; it's priced in lost output per hour. A misdiagnosis that stretches a one-hour fix into a day and a half turns a minor repair into a serious expense. I've done that math the hard way. Twice.
What I check now
If you take nothing else from this, here's the sequence I run before recommending a replacement:
- Check the system first. Filter condition—and yes, that 14x18x1 filter is common in small commercial units—pressure setpoints, and whether the switch is actually sensing what you think it is.
- Confirm the switch type. KP fixed hysteresis or RT adjustable differential? Does the application actually match the hardware?
- Read the wiring diagram on the housing. Confirm the right terminals. Verify no live voltage with a non-contact tester before touching anything.
- Test the switch. A multimeter and a hand pump will tell you if the contacts switch as pressure changes. If they do, the switch isn't broken.
- If it's still unclear, call someone who does this daily.
On that last point, I'll say something about professional boundaries. I'm good at pressure switches and the compressor/HVAC systems I work with. I'm not the right resource for every control system on the market, and I'm comfortable saying so. The specialist who names their limits is more credible than the generalist who promises everything. I've been the generalist. It costs more.
The short version
A Danfoss pressure switch does one thing: it responds to pressure. When it looks wrong, the answer is usually not a new switch. The answer is understanding what the switch is trying to tell you.
Check the filter. Check the wiring. Check the differential. Test the switch before you condemn it.
It's probably not the switch.
I've got the scars and the shipping receipts to prove it.