Compression tests do not tell all.

[76GMC1500]

So, my 1987 CR250 made 185 psi on a compression test as of last night. Pretty good, huh? Well, it made good compression but I could grab on to the flywheel and turn the motor over easily by hand. Last night, I put new rings in the bike and after a short ride this afternoon, it made 200 psi. Not that much higher, but the blow-by was so reduced I cannot turn the engine past TDC by hand at all. Now, it took something like 1.5 hours to get the bike started. It started within 5 kicks last night, I idled it for a few seconds, but had to shut it down as it was midnight. I had to pour a little gas in the spark plug hole to get it going it was so badly loaded up, I think I put too much oil on the piston before I installed it. The bike has a ton more midrange, I haven't run it much on the top end yet. It like to point the front wheel for the sky over the slightest of bumps.

The rings were so badly worn, the gaps were about 4-5 times what the new rings were. The moly band around the outside of the ring was so thin you could barely see it. All this and the motor still made 185 psi.

 

[Ol'89r]

That's why it's better to do a leak-down test.

 

[76GMC1500]

Yep.

 

[whenfoxforks-ruled]

Other than the leak down, now you have a base line to monitor your readings, and if there is a drop, time to open it up and have a look. Qualified auto mechanics can diagnose problems in cars so whats the difference, less cylinders to mess with.

 

[darringer]

Compression testing alone can be very misleading. The only other way than a leakdown test is to pull the top end down and measure it. Compression testing is great for ring seal issues, but won't tell the condition of the piston. Even totally clapped out top ends can give decent psi readings, as you found out.

 

[RM_guy]

A leak down test won't tell you how good the rings are. It's only good for checking seal and gasket integrity. The rings can’t hold back the pressure for an extended period of time so unless you came up with some kind of timed leakage past the rings you’re wasting your time. Use compression reading as a gauge and measurement as the way to know for sure if things are out of spec.

 

[adam728]

A leak down test won't tell you how good the rings are. It's only good for checking seal and gasket integrity. The rings can’t hold back the pressure for an extended period of time so unless you came up with some kind of timed leakage past the rings you’re wasting your time. Use compression reading as a gauge and measurement as the way to know for sure if things are out of spec.

There's another kind of leak down where a constant psi is inputed to the leak down tester, then the pressure the cylinder holds is used to read the leak down.

Example - if you regulate 100 psi to the tester and when opened to the cylinder it stays at a constant 90 psi you have 10% leak down.

Downsides are that the measurement is much like a compression measurement, irrelavent without a previous reference. One engine might be healthy with 18% leak down, another might be half dead at 12%. The other downside is that it only tests with the cylinder in one position. So if there is damage on the cylinder wall half way through the stroke it will never show on a leak down unless you lock the engine in place mid-stroke. And on a 2 stroke you need to pull the reeds when doing this test (which isn't a common way to test 2 strokes). On a 4 stroke the crankcase is vented, so leakage past the rings can escape. On a 2 stroke reed-valve engine the crankcase should be fairly well sealed, so you'd be possibly applying a lot of pressure to the lower end.

I agree that compression and leak downs don't tell the whole story. About a year ago my bike made 160-170 psi on the compression tester, yet had a 1" x 1/2" section of plating missing above the exhaust port. I ran the top end way too long and carbon was building in the ring lands and didn't allow them to move properly, so they started scraping the cylinder. You just can't beat a tear-down-and-measure session when you need to know how your engine is doing. :cool:

 

[darringer]

Use compression reading as a gauge and measurement as the way to know for sure if things are out of spec.

I have had bad luck using compression readings alone to tell how much a top end is worn. For example, my '99cr250 would pull consisitent 210-215psi readings over the course of a year and a half. At teardown, the rings and piston were out of spec. This has also been the case on my '02cr as well. Although the compression did drop from 180 to 175psi over about a 9 month span. That is still only about a 3% drop. Hardly noticeable. I believe the only sure way to tell how things are is to open it up and measure. For the short amount of time it takes, and the inexpensive cost of gaskets, it's cheap insurance against catastrophic engine failure.