Interpreting the results of leakdown test

[kevinrse]

Thanks as always for the help. Love these forums. 

 

On my 1999 Legacy Outback (2.5L DOHC), I recently replaced my timing belt & components after a slipped tooth resulted in poor compression and misfires on cylinders 1 and 3.

 

Before the timing belt replacement I got 90 and 80 psi compression on cylinders 1 and 3, respectively. Now I get 210 and 60 psi respectively. (On cylinders 2 and 4 I got 190 and 180). I detailed this info in this thread.

 

A friend advised a leakdown test next. I put Cylinder 3 in TDC, held it in place with a socket and breaker bar, dialed input pressure up to 75 psi and got 72-73 on the other gauge, i.e. approximately 3-4% leakage. I could hear a quiet hissing through the oil fill cap. 

 

1) Do these numbers seem right---i.e. are the results of my leakdown test consistent with the compression test? Want to make sure I'm doing it right.

 

2) Assuming it is right, this means I have worn piston rings or cylinder walls, right?

 

 

 

 

[idosubaru]

you're absolutely positive you didn't install the timing belts with one cam one tooth off?

 

ring and cylinder wall wear is not a typical subaru block issue, but of course anything can happen in a cylinder containing explosions.

[kevinrse]

Well, I can't say I'm 100% positive of course. But I was very careful. I checked all 5 marks on belt and corresponding sprockets, counted teeth to make extra sure, re-checked marks and re-counted teeth after...etc...

 

Plus---if I installed it off by one tooth, that would affect the action of both cylinder 1 and 3, right? (since they share camshafts).

 

By the way, I forgot to mention it's relatively high mileage, 228,000 miles---I don't know if incidence of ring/cylinder wall wear increases for high-mileage vehicles?

[Fairtax4me]

Normally a leak-down test is done at 100psi, but only 4% leakage is pretty damn good. Not enough to make a difference.

 

I've leak checked the 2.2 in my 96 and had leakage of 10-15% in all 4 cylinders. Compression test numbers were all above 170.

 

I would have to question the results of the compression test.

Have you started the engine yet? Does it run ok?

[kevinrse]

Yes, it runs the same as it did before the timing belt slipped---but the compression test made me worry that I might have had an underperforming piston for some time. The engine does run fairly smoothly, but there have been occasions since I bought the car when I wasn't sure if I still had full power. So the results of the compression test made me think I discovered the reason.

 

I can re-do compression test. I also spent quite some time worrying about TDC position. E.g. if someone put the mark on the crank pulley in the wrong place or something, that would affect the test.

 

I used compression gauge when turning pulley and looked for compression to begin occurring, then continued turning until mark was lined up with "0" on timing belt cover. So I'm pretty sure I did that right, though not positive.

 

I did notice one thing when turning crank: When I attached compression tester to cylinder 1 (good---210 psi) and approached that cylinder's TDC, pressure bumped up a bit each time I turned the crank, and it remained up. In contrast, when I attached compression tester to cylinder 3 (bad---60psi) and approached TDC, the pressure on the gauge bumped up a bit each time I turned the crank, but also went down again immediately---i.e. the pressure dissipated. This was behavior would seem to be consistent with the bad compression test performance.

[1 Lucky Texan]

this is a way-out-there idea, but what if the injector is leaking fuel into that cylinder and washing out the oil from the rings? Could that so severe as to reduce compression to this level?

 

any sign of the plug being wet with fuel when pulled from that cylinder?

 

swap injector with cylinder one and see what happens?

 

a bore scope might also be able to see cyl. wall scoring if a ring or piston land is broken.

[kevinrse]

I'll be damned.... just now repeated compression test on cylinders 1 and 3. 210 and 200 PSI!!!

 

I guess I did screw it up somehow the first time around. Only thing I can think of is maybe I forgot to push the throttle in while cranking? Hard to know, it was a couple of days ago.... but I'm not gonna complain.

 

Thanks again everyone for the help. Someday I will be a true Subaru Master. 

[czny]

Sometimes a bit of dirt or carbon gets into the Schraeder valve of compression tester resulting in inconsistent numbers.

 

There is a way to check for a bad valve seal, like if the seat has come loose in the cylinder head, or a bad valve. If your tester is the screw-in type, remove the Schraeder valve, test the cylinder in question again cranking with starter. Pressure ought to be consistent with each compression stroke(squeeze).

 

If not, squirt a little engine oil down the spark plug hole, crank & repeat the test. Rings - pressures will rise. Valves - same inconsistent pressures.

 

Just my $.02

Edited May 24, 2014 by czny

[Fairtax4me]

I did notice one thing when turning crank: When I attached compression tester to cylinder 1 (good---210 psi) and approached that cylinder's TDC, pressure bumped up a bit each time I turned the crank, and it remained up. In contrast, when I attached compression tester to cylinder 3 (bad---60psi) and approached TDC, the pressure on the gauge bumped up a bit each time I turned the crank, but also went down again immediately---i.e. the pressure dissipated. This was behavior would seem to be consistent with the bad compression test performance.

This is the result of a sticky, dirty, or loose Schrader valve in the end of the hose.