What are acceptable compression test readings for a 4 cylinder engine?

[zoobyshoe]

I've never done this before, and undertook it for the first time today, guided by books, google, and YouTube.

My truck's a 4 cylinder. I did the first test after driving it for a while to warm it up. I got 150-150-146-140 psi for the four respective cylinders 1-2-3-4. The instruction I was following was to crank the engine for about 5 seconds on each cylinder.

After reading some more, I found a site that said you should only crank 6 revolutions per cylinder. Otherwise you can falsely build up acceptable pressure in a cylinder that might actually be worn or have worn valves. So, I redid the test, trying to count 6 revolutions by ear.
This time I got 120 plus or minus 2 psi on all four cylinders. However, the 6 revolutions seemed to me to take about the same time as my previous 5 second count. The difference may well have been due to an hour of cooling off time: I didn't warm the engine up before the second test.

Anyway, my real question is about acceptable readings. I have two separate manuals for the car, one Chilton's, one Haynes, and neither says what acceptable readings are supposed to be. (I checked them both over for a long time.) Both books explain how to do a compression test, but neither tells you what the readings you get may mean.

A website I found says anything above 100 psi is acceptable. 135 or more would be considered excellent. The implication was that this is true of all cars. Is that the case?

 

[AlephZero]

A website I found says anything above 100 psi is acceptable. 135 or more would be considered excellent. The implication was that this is true of all cars. Is that the case?

The bottom line is that for all fairly modern cars that fill their tanks from the same pump at the gas station, the engine compression ratios don't vary much. High performance engines may be an exception. Ultra high performance engines running on exotic fuels WILL be an exception.

I think the main criterion is that all the cylinders are the same. If one is way out of line with the rest, there is a reason for it that probably needs fixing. If the are all the same but low, you just have a worn-out engine.

The actual readings can depend on a lot of things, including the speed the engine is cranked, which may explain why you got higher readings from the hot engine than from cold. In fact the "old school" way to do a quick test was just crank the cold engine, and check (by listening) if the cranking speed was even.

 

[AlephZero]

After reading some more, I found a site that said you should only crank 6 revolutions per cylinder. Otherwise you can falsely build up acceptable pressure in a cylinder that might actually be worn or have worn valves.

I don't really follow that logic. At cranking speed, the cylinder pressure should drop back to atmospheric on every exhaust stroke, unless your exhaust system is totally blocked.

But pumping unburnt fuel into a catalytic converter is a very bad idea, and that's a good reason to minimize the amount of cranking time.

 

[Averagesupernova]

Having done plenty of compression tests I can tell you that it can take multiple strokes to get the top reading. An engine in excellent shape should not take more than 2. Compression ratio makes less difference than you would think. Just did a Deutz diesel a couple of days ago on a weak battery and got about 150 PSI on all 4. Have done gas engines and got the same spinning a bit faster. There are always a fair number of leaks in the combustion chamber. The valves are not perfect, the rings don't seal perfect when the motor is not spinning fast enough to throw oil on the cylinder walls. Want to check for excessively worn rings? Let the engine set for a couple of hours, do the test. Immediately after do the test again after squirting a tablespoon of motor oil in the combustion chamber. If the reading is considerably higher, rings are worn.

 

[zoobyshoe]

I think the main criterion is that all the cylinders are the same. If one is way out of line with the rest, there is a reason for it that probably needs fixing. If the are all the same but low, you just have a worn-out engine.

A lot of sources stress the importance of them all being close in pressure. However, what I'm trying to accomplish is to pass California's stringent smog test. On my first test at the smog station, I was far out of specs on hydrocarbons and CO. (Strangely, the NO was way below the maximum and I passed that third of the test with flying colors while failing the other two badly.)

Worn valves and rings are one possible cause googling provided. It's probably the most expensive to fix, as well, so I want to rule it in or out. So far it seems I can rule it out, but only if that info about anything over 100 psi being acceptable for just about any engine is accurate. If that's the case, it doesn't look like worn valves and rings are what caused the test failure.

Having done plenty of compression tests I can tell you that it can take multiple strokes to get the top reading. An engine in excellent shape should not take more than 2. Compression ratio makes less difference than you would think. Just did a Deutz diesel a couple of days ago on a weak battery and got about 150 PSI on all 4. Have done gas engines and got the same spinning a bit faster. There are always a fair number of leaks in the combustion chamber. The valves are not perfect, the rings don't seal perfect when the motor is not spinning fast enough to throw oil on the cylinder walls. Want to check for excessively worn rings? Let the engine set for a couple of hours, do the test. Immediately after do the test again after squirting a tablespoon of motor oil in the combustion chamber. If the reading is considerably higher, rings are worn.

What jump in psi would you consider "considerably higher"? I have come across this oil squirt test on most sites, but they describe it being done with the engine at operating temp. I suppose I could try it both ways tomorrow.

 

[zoobyshoe]

Here's a guy on this car forum claiming that the proper reading is 20 times the compression ratio:

"ok4450
March 2009 edited March 2009

Yes, the engine is worn, It's due to either rings, valves, or both. A wet compression test could be done to determine how much of a factor the rings are in this.

The 150 is somewhat tolerable; the 100-130 is low and it appears to run smoothly because all of the readings pretty much suck.

The compression readings should be roughly 20 X the compression ratio. (9 to 1 means 9 x 20= 180 for example).
This will vary a bit based on altitude, barometric pressure, engine wear, etc.

As mentioned, the readings should all be close to each other."

http://community.cartalk.com/discussion/2119316/significance-of-compression-test-results

 

[Ranger Mike]

Zoo, your readings are excellent. On doing a comp test - If you had the throttle wide open ( WOT) on every reading, and use the same revolutions on EVERY cylinder you will have adequate readings. The readings you have, I would not even pull off the valve cover on this engine. Drive it and change oil and you are good to go for years. The 100 psi reading is also good if it is on a Ford tractor with 6.2 to 1 compression and a very mild camshaft ( factory spec is 118 psi at 1000 ft ASL). Engine with those readings will last for a few years. Mine has! I use a 20% variation to determine if it is worth a re-ring job on the engine and 15 is a little over 15%.
The odd thing about your readings is the low cylinder is number 4 which should be a high reading cylinder (maximum cylinder oiling as its closest to the oil pump) because it is closest to the bell housing ( minimal deck and cylinder flexing) and one side is not next to another hot cylinder. Usually the inside cylinders are low readers because of all the heat build up. Heat is a piston rings biggest enemy. Still a 10 psi drop is well within limits and I would not tear it down for rebuild.
A useful rule of thumb states that cylinder pressure in an engine in good condition with a mild or
"street-type" cam should range (at sea level or a little above) from 17 to 20 times the specified
compression ratio. So if we apply the rule to the 134-inch OHV engine introduced in the NAA with a CR of 6.6:1,using an average of 18.5, we arrive at a nominal cylinder pressure of 122 PSI for an engine in like-new condition.
But...here it comes...reality has a way of bringing down theory. Our 12:1 four cylinder should have a comp ration of 222 psi and we never have been close to this! The camshaft lift, over lap and where the cam is dialed in ( advanced or retard) all bleed off static cyl. pressure big time.

These are only guide lines , let your pocket book dictate your actions.

 

[Highspeed]

Zooby, It sounds like your truck is running rich or has a misfire. You could also have dirty oil or a plugged air filter. If it's FI then even a slow or dead O2 sensor, or some other cause of a lean o2 reading. There are some other possibities too, but hard to guess without more info. A good tech with a 4-gas analyzer, a scope, and some tuning knowledge should be able to fix you up. I'd agree with Ranger Mike that the engine is probably quite sound. Good luck!

 

[Nugatory]

On my first test at the smog station, I was far out of specs on hydrocarbons and CO. (Strangely, the NO was way below the maximum and I passed that third of the test with flying colors while failing the other two badly.)

That makes sense. Hot and complete combustion minimizes CO and HC production, but that's also what produces NOx, so it's very common to find low NOx levels in a high CO/HC motor. The challenge for the engine designer is to get combustion efficiency high enough to keep the the CO and HC levels down without spiking the NOx levels. This requires keeping the fuel-air ratio very close to stoichiometric; then a modern catalytic converter can clean up what comes through. +1 to what Highspeed said about you're seeing and how to fix it.

 

[Nugatory]

Here's a guy on this car forum claiming that the proper reading is 20 times the compression ratio: ...

I don't buy that. I've seen different compression testers give repeatable readings more than 20 psi apart on the same motor, so wouldn't believe the absolute calibration. This is also why everyone tells you (correctly) to focus on differences between cylinders, not the absolute value.

Some other thoughts on compression testing:
- If you really want to know about the health of a motor (from the rest of this thread you don't - you just have a tuning problem, which is good news) you need a leakdown tester and a vacuum gauge. A compression tester is a bit of a blunt instrument.
- The engine's static compression ratio is a red herring. The static compression ratio is the ratio of the cyclinder volume when the piston is at the bottom of its stroke to when it is at the top of its stroke. But because the intake valve remains open well after the piston has passed bottom dead center (BDC) and is moving up on the compression stroke the actual compression(the "dynamic" compression) of the fuel air mixture will be different (and critically dependent on the dynamics of the air flow, which are completely different under operating conditions and at cranking speeds). The higher-revving the engine design, the more pronounced this effect is.
- When compression testing, block the throttle open and run the test with all the plugs out. You want the engine spinning and breathing as freely as possible.
- The compression test results vary with the cranking speed, and it's the battery that's doing the work so you're counting on the battery output to remain constant as you test all the cylinders. When I'm compression-testing a four-cylinder, I get a total of seven readings: 1, 2, 3, 4, 3, 2, 1 in that order. That way I can tell whether my results are trending down just because the battery is weakening.

 

[AlephZero]

Having done plenty of compression tests I can tell you that it can take multiple strokes to get the top reading. An engine in excellent shape should not take more than 2.

The basic reason for that is because the engine has to pump some air into the gauge and its connecting pipe (through a non return valve) so on the first one or two strokes you inevitably see a lower pressure. But as everybody has said, the absolute values are not so important as the differences. I would have said differences up to 25% are not conclusive proof that anything needs fixing, but unlike RangerMike I'm not too concerned about getting race-car engine performance.

As well as leaking rings, if you have two adjacent cylinders both reading low, most likely the head gasket is blown between them.

@Zooby, what did the spark plugs look like? If some or all of them were black, oily, or damaged, that might be all you needed to know about the cause of the problem.

 

[kazzy]

Like others have said, its more important for the cylinder compression readings to be within a close range than the reading itself. I read somewhere that up to 10% difference in cylinder compression is normal but any higher and there may be some leaks.

Its important to do a 'wet' compression test to identify where the leaks are occurring (if any). If compression is higher after the wet test (it will be to some degree, testing compression on a warm engine will produce more accurate results), it indicates worn piston rings. If a cylinder shows low compression and pressure doesn't increase during the wet test, that indicates leaking valves.

 

[zoobyshoe]

Thanks very much for the replies, everyone!

Something has come up and I can't spend time on this thread just now, but I'll get back to it. I've read and appreciate everyone's input!