Compressor teardowns are cool.

I’ve known that it’s important for any HVACR equipment training course to include a hands-on portion; you learn it the way you learn catechism. (“… And do not let them learn about the compressor, ’lest they lay hands upon the compressor and so discover its very depths.”)

Teardowns that are part of Copeland’s COSS (Compressor Operation and Service Seminars) training almost have the feeling of mortuary science. Contractors and technicians have already been through a few days covering refrigeration fundamentals, applications, system components, specialized compressors, single- and three-phase characteristics, and then the classroom portion on compressor failure causes, troubleshooting, correcting system problems, and failure analysis.

The “failure analysis” brings contractors and technicians to the warranty teardown lab at Copeland’s headquarters in Sidney, OH, where they are given randomly selected compressors that were returned under warranty; or, if the COSS is being held at a different location, they are sometimes asked to bring in their own failed compressors (not under warranty) to tear down.


The core of the training is to get contractors and technicians questioning whether the problem was:

  • Actually a failed compressor (say it was manufactured without an essential pin, which the trainers thought they discovered to be the problem with a Scroll in the lab); or

  • Due to an operational problem that led to a compressor problem (such as flooded starts, which can result from migration if the compressor is off for too long).

    The other option is that nothing is actually wrong with the compressor. If this were the case, if you put it back together, it would work. (Incidentally, that Scroll failed due to system problems. I was standing there while the lab employee drained out the oil and sifted the bits inside the compressor. He found the pin he was looking for.)

    It is very hard for some contractors and technicians to admit that they misdiagnosed a system or two. It’s probably much easier to blame the manufacturer, but after a few compressors fail in the same system, blaming the manufacturer gets to be expensive — and a waste of time.

    Likewise, if the manufacturer keeps saying that the compressors being turned in have nothing wrong — or, nothing was wrong with the compressor that caused it to fail — of course contractors are going to be skeptical. That’s why teardowns are so important.


    Good troubleshooting is often compared to detective work. You are given a set of physical clues and with your knowledge you trace them back to the most probable cause. It’s cool, and it’s nonjudgmental.

    Next year, The News and Copeland’s Training Department are going to publish a series of articles on compressor detective work. Like another detective series from TV, we want to present “just the facts,” as well as possible solutions.

    The articles will be based on actual teardowns. Let me tell you, the inside of some of those compressors is ugly. One of the semi-hermetic units at the November COSS had a valve plate that was discolored on top; the suction side had cooked-on oil.

    Parts were broken inside, and there were signs of scoring. The overload had tripped, and the suction screen was clogged and wedged up in the corner of the head.

    The culprit was heat. System heat had done that compressor in.

    A look at another semi-hermetic showed that there had been some tripping, but the rotors, clips, bearing, oil — all were in good shape. The valve plate was good, and the compressor was only about two years old. The overload protector was open.

    Was this a frustrated or inexperienced tech who was in a hurry, and just had to get the system working again? It’s possible. It’s also possible that the system will need another compressor soon, because the cause of the tripping wasn’t tracked down.

    Some people call the compressor the heart of the refrigeration system. Like the human heart, compressors are affected by many other problems in the system; it’s just the last thing to stop working.

    Barb Checket-Hanks is the service/maintenance and troubleshooting editor. She can be reached at 313-368-5856; 313-368-5857 (fax); (e-mail).

    Publication date: 12/02/2002