Most Danfoss Pressure Switch "Failures" Were Never Broken — What $8,000 in Mistakes Taught Me

The Switch Wasn't Broken. My Process Was.

I'll say it plainly: most Danfoss pressure switches that get replaced in the field were never broken at all. I know that sounds like a guy defending a brand he specs, but I have the paperwork to back it up — most of it generated by my own mistakes.

I'm a maintenance engineer handling industrial controls and refrigeration orders for seven years. I've personally made (and documented) 12 significant mistakes, totaling roughly $8,000 in wasted budget. Now I maintain our team's checklist to prevent others from repeating my errors. And the pattern across all that paperwork is uncomfortably clear: the switch is rarely the problem. The selection process, the setup, and my own impatience? Those were the problem.

Here's the thing — every one of those $8,000 mistakes was avoidable. Not just in hindsight, either. There were datasheets, numbers, and simple tests that would have caught every single one before I ordered parts. I just didn't take the time. This article is the checklist I wish someone had handed me in 2018.

Mistake #1: I Bought the Right Brand, Wrong Series

In my first year (2018), I made the classic green-engineer mistake: I assumed "Danfoss pressure switch" was one product. It isn't. The Danfoss KP pressure switch family, the RT series, and the MBC series are built for different jobs. Pick the wrong one and it won't just cost money — it'll fail in a way that makes you look like you don't know what you're doing. (Which, honestly, I didn't.)

The case that still stings: a customer with twelve AHUs in a commercial building. They wanted clogged-filter monitoring on their 14x25x1 air filters — a signal that tells the maintenance team when to swap the filter before pressure drop starves the unit. Standard HVAC stuff. I ordered 60 KP15 pressure switches for the job.

Never expected a "perfectly reasonable" pressure switch to be completely useless for the actual task. Turns out the KP15 range is 0 to 6 bar, and its differential — the gap between cut-in and cut-out — is 0.4 to 1.5 bar. And the pressure drop across a dirty 14x25x1 filter, even a totally clogged honeycomb air filter? A few hundred pascals. That's 0.00-something bar. A KP15 can't sense a difference ten times that size.

Those numbers aren't from memory. They're in the Danfoss KP series technical datasheet — a public document that has been available for years. I just never read the differential column.

For differential pressure switch work like air filter monitoring, the right call was the RT260A, the RT-series air pressure switch, with ranges down to a few hundred pascals at the low end. But by the time I figured that out, the $3,200 order had shipped and the customer had already asked for the wiring diagram. I had to explain that the parts I'd spec'd were wrong. Redo cost, expedite fee, a week of delay, and a credibility dent I felt for months. A lesson learned the hard way.

Mistake #2: I Ignored the Differential and Blamed the Switch

The second mistake took three service calls and two replacement switches to figure out. That's the embarrassing one.

September 2022: a refrigeration client's plant keeps losing the low-pressure cut-out — the KP1 is tripping intermittently, shutting the system down. My first instinct? Replace the switch. First call: installed new KP1, rushed away. Next morning: same call. Second call: replaced it again, checked the wiring, tested the contact with a basic continuity check. Two days later: the phone rang again.

The surprise wasn't the switch. It was that the switch was doing exactly what it had been set to do — and I was the one who'd set it wrong. The KP1's adjustable differential (again: the gap between cut-out and cut-in) ranges from 0.1 to 0.3 bar. Whoever installed it had set it near 0.1. In a plant where refrigerant pressures fluctuate, the switch cut out at the setpoint, then cut back in just 0.1 bar above it — so the system tripped, recovered, and tripped again. The contact never "failed" at all.

Why does this matter? Because it's exactly how a reliable switch earns a reputation for being faulty. What I mean is — the KP1 is a solid, predictable piece of hardware; its problem was that I never read the differential column, so I never verified the actual cut-in point. I just checked that the contact opened when I pushed the button. That's like testing a smoke alarm's battery and blaming the alarm when the kitchen fills with smoke.

And this isn't a KP-only issue. I've watched our own technicians throw MBC 5100s at machines, then call them defective because the fixed differential caused the same short-cycling pattern. Not ideal. But the switch was telling the truth the whole time.

Mistake #3: The Multimeter Was Supposed to Be Last — Now It's First

After the third call on that KP1 job, I made a rule for myself: no switch gets replaced until it's been proven bad with a multimeter and a hand pump. That rule has caught 47 potential errors in the past 18 months — and only three were actual switch failures. The rest were settings, wiring, or the wrong series entirely.

The method is boring, and that's the point:

• Isolate the switch from the circuit. (Seriously. Do this.)
• Connect a multimeter in continuity mode across the normally closed (NC) contacts.
• Test the switch at rest: near-zero ohms when closed, or OL when open.
• Connect a hand pump and raise pressure slowly. Watch where the multimeter flips.
• Release the pressure and note where it flips back. Compare both values to the dial setting.

You don't need amp gain math, no special current setting, no signal generator. Just a cheap multimeter, a hand pump, and the datasheet in front of you. The whole check takes about five minutes — before you spend $150 on a replacement you might not need.

The conventional wisdom is: "When the contacts don't hold, swap the switch." My experience with 120+ field checks says otherwise: the switch was almost always fine. What had actually happened was either the differential had drifted from wherever a previous technician "eyeballed" it, or the range was wrong for the application — a clogged filter, a misread nameplate, a spec copied from the wrong job. Put another way: a Danfoss pressure switch is a beautiful, simple machine. It will faithfully trip at the exact pressure you set — great news if you set it correctly, and a terrifying thought if you didn't.

I Know What Some of You Are Thinking

So, the pushback. I've heard most of it from my own team, so let me answer it here.

"Electronic transmitters are more accurate. Why use mechanical switches at all?" Fair. If you need trend data or PID control, put in a transmitter. But for a simple alarm or a compressor cut-out, a mechanical switch has one huge advantage: it doesn't need power to hold its state. Remove power, and the contact stays where it was — it won't reset automatically and defeat your safety circuit. That's a fail-safe behavior electronics don't always give you for free.

"This sounds like extra work every time." Ten minutes with a multimeter versus $150 plus installation plus a possible callback? I'll take the ten minutes. The checklist has saved us real money — we're no longer burning $3,200 orders on the wrong series.

"Maybe your switches really did fail — you just had a bad batch." Could be. But our data over 18 months says otherwise: 47 flagged issues, 3 confirmed switch failures. The other 44 were process issues caused by the person holding the datasheet. Often me.

Bottom Line

So here's my final position, and I don't expect it to be universally popular: the Danfoss pressure switch is rarely your problem. Your selection process, your understanding of differential, and your verification habits are the problem. And that's good news, because all three are fixable.

Start with the datasheet. Match the series to the real-world range — KP for refrigeration and HVAC when you need adjustable differential, RT for heavy-duty and low-pressure air work like filter monitoring, MBC when you need a compact, corrosion-resistant switch with a fixed, predictable differential. Then verify the cut-in and cut-out with a multimeter before you trust it. And above all: don't replace a switch until you've proven it wrong.

$8,000 and 12 mistakes later, that lesson finally stuck. It's cheaper than the alternative — and way less embarrassing to explain at a maintenance meeting.