Low compression-- Most likely cause?

Status
Not open for further replies.
Joined
Feb 9, 2005
Messages
2,290
Location
Arizona
I have a '91 Subaru XT-6 with the ER27 flat 6 engine and 103,000 miles. I was compression testing this morning and found 1 cylinder low. A second cylinder is at the very lower limit of factory spec. The rest are well within spec.

Is low compression more likely a function of ring seal or valve seal? IOW, am I more likely to need a valve job to fix this or ring/cylinder work? The car runs well aside from an intermittent minor electrical irritant, and the spark plugs in all 6 look very nice and even, though the low cylinder has a few extra white deposits on it.

Any help or ideas are appreciated.
 
No. Opposite banks and not directly opposing cylinders. How does a leak-down test distinguish between poor ring seal and poor valve seal?
 
A cylinder leakage test puts compressed air into the cylinder (through the spark plug hole) when it's positioned at TDC on the compression stroke. Then you can listen for where the air escapes. You'll also see the percentage of leakage.
If one of your valves is leaking I would think you might notice the engine backfiring either out the tailpipe, or back out the intake.
 
SBC350 was seeing if you had a bad head gasket between the two cylinders. It could read as a bad valve on both if the alternate valves were open while you were checking the agacent cylinder. This happened to me in auto school. We read one cylinder as having a bad exhaust valve ..the one next to it a bad intake. Removing the head showed the head gasket between the cylinders had failed. It was a great lesson that day.

Since you only have ONE dead cylinder I would think that it's valve related. It would be odd that one cylinder was very low and all the other were in spec (even your next lowest cylinder is within limits) due to ring wear.
 
Wow, thanks. I never thought of listening for the source of the leak (in the leakdown test).

I had another thought last night. The car has hydraulic lash adjusters. If one of them is leaking down its prime after shutoff, could that make an intake valve not open enough during the compression test and therefore show low? Or is that not a big factor? Normally, the lifters do not rattle at cold startup. I did run the engine a minute or two before I pulled the plugs and ran the test.
 
quote:

Originally posted by bulwnkl:
Wow, thanks. I never thought of listening for the source of the leak (in the leakdown test).

I had another thought last night. The car has hydraulic lash adjusters. If one of them is leaking down its prime after shutoff, could that make an intake valve not open enough during the compression test and therefore show low? Or is that not a big factor? Normally, the lifters do not rattle at cold startup. I did run the engine a minute or two before I pulled the plugs and ran the test.


As a general rule, a cylinder with a leaking lifter/lash adjuster will actually show higher compression than a normally operating one. A leaking lifter cuts down lift and duration and that causes increased compression at cranking speeds.
 
Did you try a wet compression test putting oil in the cylinder to see if the compression went up? If it goes up it might be the rings.
 
quote:

Originally posted by punisher:
As a general rule, a cylinder with a leaking lifter/lash adjuster will actually show higher compression than a normally operating one. A leaking lifter cuts down lift and duration and that causes increased compression at cranking speeds.

I'm not sure I can agree with that. If the intake valve isn't letting enough air into the engine you'll have decreased volumetric efficiency. The less air in the cylinder the less heat and pressure is produced, you'll get lower compression readings. This is why cranking compression is higher than running compression. When cranking the valve is open for a longer period of time, allowing the cylinder to fill up. When the engine is running the valve is only open long enough to partially fill the cylinder.
 
quote:

as a general rule, a cylinder with a leaking lifter/lash adjuster will actually show higher compression than a normally operating one. A leaking lifter cuts down lift and duration and that causes increased compression at cranking speeds.

quote:

I had another thought last night. The car has hydraulic lash adjusters. If one of them is leaking down its prime after shutoff, could that make an intake valve not open enough during the compression test and therefore show low? Or is that not a big factor? Normally, the lifters do not rattle at cold startup. I did run the engine a minute or two before I pulled the plugs and ran the test.

The context is when the engine is not running, and only cranking before starting, as in when doing a compression test. At 100-200 rpm cranking speed, the cylinder will fill up with air just fine if the lifter is not opening all the way. If anything, the bad lifter (bled down, collapsed, having too much lash) will help because that will ensure the valve is closed fully. The biggest thing is to make sure the throttle body is open all the way (wide open throttle). This will hinder air flow the most when doing a compression test. The purpose of a hydraulic lifter is to compensate for temperature and valve-pushrod clearance when the engine is running. Every lifter will bleed down when the engine is off and this should not affect valve opening one bit as long as the lifter is adjusted right to start with, and therefore "not a big factor" and should not affect compression #'s at all when cranking it such a low rpm.

the opposite though, if you're lifter is pumped up from too much preload on the lifter, then the valve won't close all the way and you will see poor compression #'s.

When the engine is running, I don't know if a collapsed lifter would actually cause more compression. It would cause the valve to not open all the way, and/or stay open for as long as its supposed to. In that case, with rpm a big factor, it's not going to pull in as much air resulting in less compression and richer mixture. I've read & heard about adjusting hydraulic lifters, where there is no preload I think, and they operate similar to a solid lifter to help with high rpm performance. You also get into the realm of volumetric efficiency > 1 with inertial ramming, acoustic ramming I think it's called, and a whole other mess of cylinder head & intake design. Then you will certainly see higher compression #'s than at cranking speeds, for the most part the compression # will be variable and dependent on rpm.

I agree with the oil squirt when doing the comression test. That'll definitely pinpoint piston rings, and much easier to do than setting up a leak-down test so do this first before a leakdown.

[ July 11, 2005, 01:36 PM: Message edited by: 1 FMF ]
 
Keep in mind he has a flat 6 (or was it a 4). squirting oil into the spark plug hole may mask a bad exhaust valve ..or even an intake if the valve is open when he squirts the oil it.

If he can take the valve covers off easily and loosen the valves up on the bad cylinder ..all he needs to do it put air to that cylinder. If he hears it out the intake ..intake vavle...exhaust ..exhaust valve (or both intake and exhaust) ..if he hears it out he top of the head (he should be there at the uncovered head) ...rings.
 
I like that idea, Gary. And you're right, this is a flat-6.

Trouble with loosening the valves is that Subaru uses a valve cover that doesn't have much function. Below that is the "cam tower." It is a separate housing that just holds the cam. Below that are the lifters and the heads. So, I have to take a bunch of crap off (including the timing belt on that side) and then get a special sealant because there's no gasket between the cam tower and head. I just don't have the room or tools for that kind of thing anymore.

I'm thinking I don't have a simply stuck lifter (as in stuck in the fully extended position therefore not letting a valve close completely), because I don't have any backfiring of any kind at all, and the car seems to run fine and the plugs look fine.

I'm such an idiot. I forgot to block the throttle plate open for the test.
pat.gif
 
Steveh: I did not put anything in the cylinders before the test. I did run the car a minute or so before disassembly to try to make sure everything was well oiled and "pumped up."

I have a bottle of ARX that I intended to use in maintenance dosages only in this car and perhaps a flush in the tranny in a while. I really don't want to put it all into the crankcase unless folks with more/better knowledge or experience than I think it is likely a ring issue. I ran a Schaeffer Neutra flush as directed just a couple hundred miles ago. I guess you could say this is the "rinse" oil... ?

Thank you for all the replies and assistance. I will hem and haw about it a bit and see what I'm going to do.
 
Status
Not open for further replies.
Back
Top