Menu Close

Compressor Diagnosis

By Mike Gray of Refrigeration Specialists

Low pumper compressors can be devilish to diagnose properly. None of us wants to replace a compressor that is not bad. With more than 40 years of experience, I want to offer some thoughts.

First, calculate the actual compression ratio (CR). Convert high- and low-side gauge pressures to absolute by adding atmospheric pressure to each reading. I find 15 psi to be close enough for my purposes (e.g., 5-psi suction + 15 psi = 20-psi absolute; 85-psi discharge + 15 psi = 100-psi absolute.) Divide discharge absolute by suction absolute to obtain compression ratio to 1.

In this example, 100/20 = 5:1, on the low end of the range. In time, a technician can create a mental history of CR he finds on various healthy systems and use that for reference. This is another good reason to log measured pressures on each visit, even of healthy units.

Discharge valve leaks are very easy to identify. Just close the suction service valve tight and observe the compressor low-side pressure (with the power off). A slow, steady rise, which also can be pretty rapid sometimes, to discharge pressure is conclusive evidence of a leaking or broken discharge valve. A leaky suction valve will not exhibit this at all. Generally, a small valve leak will not reduce capacity of a properly sized unit enough to be noticed until the problem has grown very serious.

Failure to Start

A leaky discharge valve can also cause a failure to start in some circumstances. If the piston with leaky discharge valve stops at top dead center (TDC), the compressor will restart just fine. If it stops anywhere else, the leak will drive the piston to bottom dead center (BDC) and hold it there. And, just as a reminder, TDC and BDC refer to the points where the piston is at the top or bottom extreme of its travel and the crank pin is centered with the connecting rod centerline exactly parallel to the bore. If the piston travel is measured with a dial indicator, it will show the piston does not move any further in the direction referenced, bottom or top.

A low-torque motor (nearly all single phase, some three phase) can’t overcome this until the high-side pressure slowly drops off and equalizes with the low side. (With properly wired oil safety controls, this can show up as nuisance trips of the oil safety.) I have seen compressors run for weeks with this problem before they were replaced.

It is important to identify a discharge valve as the bad valve. If the compressor has run many hours with a bad discharge valve, it is possible that a valve replacement will not solve all of your problems. Just a few days running with the bad discharge valve allowing high pressure atop the piston at all times denies the wrist pin lubrication. Normally the wrist pin moves slightly in its bore, allowing oil into the clearance. This can’t happen with a leaky discharge valve so the pin runs dry, elongates the hole, and causes loud knocking noise after the bad valve is replaced. Remaining compressor life will be short. And you still must replace the compressor after wasting the effort replacing the valve plate, not to mention expense. When I find a leaky discharge valve, I usually just change the compressor.

Cylinder or ring wear often shows up as excess oil flow rate. A system that has had a stable oil level for years will begin showing low oil level in the crankcase. Eventually, the oil safety will trip. Adding oil will extend the time before compressor replacement, but it should be noted if the complaint later becomes low capacity.

There is a valve between the motor compartment and the crankcase of some compressors. Excess blow-by as caused by cylinder/ring wear can cause the valve to be forced closed and not allow oil to return from the motor compartment to the crankcase until compressor shutdown. A considerable rise in oil level will be seen just after shutdown. If run cycles are not too long for the oil capacity of the crankcase, compressors can run like this for a long time but the warning of serious wear issues is still there.



Causes of valve failure are usually slugging or high discharge temperatures. Occasionally, very long service merely wears the reeds paper-thin, resulting in failure. A few high-temperature causes are:

• Condenser fan failure(s);

• Dirty condenser;

• Wrong fan motor; (How many 1,075-rpm motors do you suppose I have found installed where a 1,550 or 1,625 rpm was correct?); and

• Wrong fan blades. (Too small of a diameter, too flat of a pitch).

Noncondensables such as air will cause very high discharge temperatures as will shortage of charge or other causes of too high of a superheat. (Suction line heat exchangers, poor suction line insulation, and TXV maladjustment are common culprits.) You can be sure that too high of a suction superheat will show up as too high of discharge temperature. The wrong compressor for the application where the suction pressure is well below the minimum acceptable for the compressor will cause high compression ratio and high temperature.

Failed head fans on compressors in low-temperature applications is another cause of valve failure or cylinder/ring wear. Low-temperature compressors need continuous air over the head. If a fan cycling control is used, a head fan becomes mandatory. Anything that elevates discharge temperature to the break-down point of the oil will result in rapid cylinder wear.


The refrigerant in use contributes as well. CFC-12 was the most forgiving, often allowing 40 years or more of service — even in low-temperature applications. HCFC-22 was the very worst, prior to Copeland’s Demand Cooling version of liquid injection. R-502 was somewhere in between. Low-temperature applications are more troublesome than mid- or high-evaporating temperature systems.

Liquid is essentially not compressible. Flooded starts on refrigerant-cooled compressors or flood back of running air-cooled compressors breaks things. The valves do not just leak, they are utterly destroyed. A connecting rod can be broken and show up as a low pumper having lost one of two, three, four, or six cylinders. Depending on just how the rod breaks, the thing can stay running with little or no noticeable increase in vibration. There can be many possible causes, of course, but failed crankcase heaters and systems that do not use a liquid line solenoid valve and operate in continuous pump-out mode are high on the list. Other villains are long off times, such as a power failure, customers that turn off power for a long time to clean a case, frosted evaporators, and failed TXVs, to name a few. Suction accumulators really help — if they are sized properly and installed correctly.

We all learn from accumulated experience. Every semi-hermetic compressor I replace is torn down prior to it being returned for credit. In most cases, all I need do is remove the head and look at the valves and piston tops. Broken valves are obvious. In most cases, so is a broken rod. (The piston will be rotated a few degrees from normal or it will not move when you push down on one of the others.) It will also sound different if tapped down lightly. High temperatures are indicated by carbon formation in the discharge area.

Broken discharge reeds, particularly broken valve backers, and broken or stripped hold-down bolts are indications of trying to do the impossible — compress liquid. Once in a great while you will find a discharge valve and backer bolt unscrewed, not broken or with stripped threads. That is a manufacturing defect. It may take years, but it can happen if the lock plate is not secured properly. If a suction reed locating pin wears through or breaks, field repair is extremely difficult if not impossible. This problem has been the subject of a bulletin by at least one manufacturer — proving how important it is to stay current on the information they pass along to us! If corrected before it fails, it can save a compressor replacement. A broken reed that has floated around in various cylinders can do a lot of damage to pistons and cylinders, making a valve replacement a bad idea. After a number of teardowns, you will see what I mean.