It was a Tuesday afternoon in late September 2022. I was staring at the display screen of an Atlas Copco GA 37 VSD compressor, and it was flashing fault code 73: Low net pressure. Simple enough, right? I’d seen it before, a few times. The textbook fix is to check the intake valve, look for a leak, or adjust the setpoint. I followed the manual, swapped a sensor for good measure, cleared the code, and walked away. Felt good about it. (Note to self: that was a mistake.)
The compressor ran fine for about four hours. Then, around 7 PM, production called—the line was down. Air pressure had dropped to 60 psi, and the machine was locked out on the same fault. That night cost us a full shift of lost production. The root cause? A partially clogged separator element that the fault code didn’t flag, because the differential pressure switch was still technically within range. I’d treated the symptom, not the cause.
The Problem With Relying on Atlas Copco Compressor Fault Codes Alone
I want to be clear: Atlas Copco makes excellent machines. Their fault code system is comprehensive. If I remember correctly, the GA 37 VSD has over 150 distinct fault codes. But here’s the thing—a fault code is just a symptom, not a diagnosis. It's tempting to think that clearing the fault code fixes the machine. But the machine doesn't care if the code is clear; it cares about real-world conditions. The oversimplified version is: if the code says 'A', check 'A'. It ignores all the nuance of how components degrade over time.
In my first year (2017), I made an even more classic error. I misread a fault code 50 (motor overload) as a genuine electrical issue. I spent an entire afternoon checking motor windings and contactor contacts. Nope. The real issue was that the ambient temperature in the compressor room had hit 95°F, and the machine was derating itself. The code was accurate—I just lacked the context. A code is a pointer, not a verdict.
Building a Practical Spare Parts and Diagnostics Checklist
That bad experience in September 2022 was my turning point. After the third similar rejection of a 'simple fix' in Q1 2023, I created what I call my 'M3' checklist: Measure, Manually inspect, then Match the code. The match is the last step, not the first. I also carry a specific set of Atlas Copco compressor spare parts that I’ve learned are the common 'false alarm' triggers.
Here’s the list of the top five parts I now keep on hand—not because I expect them to fail, but because swapping them is the fastest way to eliminate a variable when a code is misleading:
- Pressure transducers (P1 and P2): These drift over time, especially in dirty environments. A drifting P2 sensor can mimic a real pressure drop fault. I keep two in my truck. (I really should order a third).
- Intake valve diaphragm kit: A leaking diaphragm can cause fault code 73 (low net pressure) long before any other sensor catches it.
- Minimum pressure valve (MPV) rebuild kit: This is a classic one. The MPV starts weeping, the system pressure drops slightly, but the machine compensates until it hits its limit. The code it throws might be for high temperature, not low pressure.
- Start/check relay and fuse pack: Intermittent electrical faults are the hardest to find. A corroded relay can trigger a cascade of unrelated codes.
- Separator element (the one I failed to check): The differential pressure gauge on some models is mechanical and can stick. A new element is a fast swap that eliminates the possibility.
I should add that not all spare parts are created equal. I buy genuine Atlas Copco parts for anything internal to the airend or control system. But for filters and separator elements, I’ve had good experience with some high-quality aftermarket options. The cost difference for a separator element can be 30-40%, and as of late 2024, I haven't seen a difference in lifespan on the three I've tested. (Take this with a grain of salt—your mileage may vary depending on your air quality).
Understanding the Thermostat Analogy: How a Compressor Regulates Itself
I’ve found it helpful to think of a compressor's behavior using the same logic as an HVAC system. People ask me all the time, 'how does a thermostat work?' It’s simple: it measures temperature and turns a system on or off. But the system it controls is complex. A compressor's control system is like a much smarter thermostat. It measures multiple variables (pressure, temperature, current, vibration) and makes decisions.
If a thermostat says the house is cold, you don't immediately assume the furnace is broken. You check if the pilot light is on, if the filter is dirty, or if the thermostat itself is calibrated. The same logic applies to an Atlas Copco controller. It doesn't always know why the pressure is low—it just knows it is. The code is the signal; the checklist is the method to find the source. Don't trust the code; trust the process.
Lessons Learned from a $3,200 Mistake
The mismanaged separator element incident cost about $3,200 in expedited freight, a new separator cartridge, a filter technician's overtime, and roughly half a day of lost production. That’s not counting the embarrassment of having to explain to my production manager why I didn't catch it the first time. Since implementing my M3 checklist in Q1 2023, I’ve documented 47 potential 'false diagnosis' situations that I caught proactively. I’m not 100% sure of the exact savings, but it’s easily in the five-figure range in avoided downtime. A supplier once told me, 'The vendor who says, "This is my specialty, but for that issue, you should call a different guy" earns trust for everything else.'
That’s been my experience, too. Knowing when not to trust a simple fault code—and building a method to look deeper—is the most useful skill I’ve developed. It’s not about memorizing every Atlas Copco code; it’s about understanding the machine behind it. And keeping a good stock of spare parts helps, too.