I got a call about a johnson controls hvac unit in a small data center. Not a critical facility—just a server closet running hot. The complaint: no cooling. My gut said ‘dead compressor.’ I was right. But I was also wrong. The how I got to that answer cost my company nearly a thousand bucks and a week of downtime.
If you're here because you're trying to figure out how to test ac compressor, let me save you the headache I caused myself.
The Surface Problem (What I Thought Was Happening)
The unit wasn't cooling. The condenser fan spun. The contactor pulled in. But the compressor hummed and tripped on internal overload. Classic signs of a seized pump. Or so I thought.
I checked the capacitor (fine), the voltage (within range), and the amperage (high, then tripping). My brain checked out. I announced to the facility manager: 'Compressor's shot. Need a replacement.'
I'd made my decision based on a single data point—the motor wouldn't start. That was my first mistake.
The Deep Reason (Why I Actually Screwed Up)
The real issue wasn't the compressor. It was the Johnson Controls temperature sensor circuit board that had lost its mind. The control board was sending a false 'run' signal, but with a wildly incorrect resistance reading from the sensor. The compressor was trying to start under a load condition the board thought was safe but wasn't.
I assumed 'no start' meant mechanical failure. I forgot to check if the brain was telling the muscle to do something stupid. I assumed 'same specifications' meant identical results across vendors. Didn't verify. Turns out each had slightly different interpretations. (Blatant self-reminder: check the control logic before you condemn the pump).
The compressor wasn't dead. It was confused. And because I didn't trace the signal back to the source, I condemned a perfectly good (albeit expensive) part.
The Real Cost (This is Where TCO Hits)
This is where my total cost of ownership lesson began. The line-item cost of a new compressor for that specific johnson controls hvac unit was $870. Labor was another $400. But the hidden costs were brutal:
- Diagnostic wasted: My time (2 hours) + the service call fee ($150) that got us nowhere.
- Rush shipping: The compressor was 3 days out. The client needed cooling now. That added a $200 expedite fee.
- Dual labor: I had to swap the compressor (a 4-hour job). Then, when we realized the board was bad, another 2 hours to replace the control board and a $250 board.
- Downtime: The server room hit 95°F. Not critical data, but the client spent a day moving equipment. That's not on my invoice, but it's on their trust ledger.
- Don't skip the control logic. Check the johnson controls temperature sensor input to the board. Is the board calling for cooling? Is the sensor reading accurately? A bad sensor can lock out a compressor or send it into a ‘short cycle’ death spiral.
- Test the windings with a megohmmeter (megger). A simple multimeter check for continuity isn't enough. A megger tells you if the insulation is breaking down under real-world conditions. The compressor I condemned had good continuity but bad insulation to ground. (I only learned this after I pulled it out and bench-tested it).
- Look for thermal lockouts. If the compressor is hot, it will trip on internal overload. Let it cool for 30 minutes. If it starts when cool, you have a heat rejection problem (dirty coil, bad fan, or low refrigerant).
The final tab: $1,870 for a fix that should have cost $400 (for a control board swap). My gut vs data moment: The numbers said go with Vendor B—15% cheaper with similar specs. My gut said stick with Vendor A. Went with my gut. Later learned B had reliability issues I hadn't discovered in my research. In this case, my gut said ‘dead compressor.’ The data (voltage, cap, amps) said ‘maybe not.’ I went with my gut. Biggest regret.
The Lesson (and the Cheap Fix)
So, if you're wondering how to test ac compressor properly:
One quick note on water heaters: A similar logic applies. I once spent all afternoon troubleshooting a water heater that wouldn't heat. Replaced the elements. Replaced the thermostats. Nothing. Turns out the control board had a faulty relay (note to self: check the board first next time). The TCO on that job: $200 in parts I didn't need + 2 hours of labor I couldn't bill for.
I still kick myself for not documenting that vendor's verbal promise regarding the compressor warranty. If I'd gotten it in writing, we'd have had grounds to dispute the restocking fee ($150 for a compressor that wasn't defective). If I'd calculated the TCO before making the swap decision, I would have spent 15 more minutes on diagnostics and saved $1,470.
That's the TCO of a bad diagnosis. The unit price of a fixed compressor is high. The total cost of a rushed one is brutal.
