Hi. I'm brand new to your forum. Thanks for having me! I have a 2002 Accord SE 2.3L VTEC automatic with about 130k on it. Bought it new in 2002 (leased then bought off lease actually), so I know where it's been and how it's been treated. My issue is that I have zero PSI in cylinder #3 using a compression tester, dry and wet, but when I do a leak-down test, it holds 80 PSI of air with the same amount of "bleed" as the other three good cylinders, all three of which were at 120-135 during compression test. Can the intake valves be NOT opening on just this one cylinder and therfore there is "nothing to compress"? I have the rocker cover off and do not see any obvious damage, loose rockers/followers (whatever) or anything else. Other ideas? Next steps? Thanks for your help!

A little (more) back history. Daughter drove it home, said it's running terribly. I started it, shaking badly, rough idle, rougher under load. OBDII read "lack of compression" (or whatever the actual wording is) on cylinder #3, hence the compression test, leak-down next, yada, yada ....

 

I suppose you've tried the compression test multiple times; making sure it's not some aspect of the way you install the compression gauge? Wide-open throttle? Disabled fuel & spark?

First I'd check the valve clearances (all 4) for that cylinder. Just so you aren't taking anything for granted.

You've got the valvecover off; have you rotated the crankshaft (forward only please) to watch the action of the valves? Pay especially close attention to that cylinder #3.

Another thing you can try, IF you can find a good way to lock the crankshaft in position. Try the leakdown test at different positions through the cycle. Listen for WHERE it leaks.
But BE CAREFUL! When the crank isn't at BDC, the pressure will cause the crank to rotate. Make sure it's locked in position, strongly-enough so that it doesn't move. Especially during the compression stroke, where the pressure will try to make it rotate backwards & mess up the timing belt.

 

Ditto JimBlake's questioning of holding TDC, especially on #3.

Compression testing is dynamic: the engine is spinning and if there's zero compression on one cylinder then there's a massive leak somewhere, most likely valvetrain related or - unlikely - a big hole in the piston; bad rings couldn't cause zero compression. If all the valves were staying closed on #3 then that cylinder would have compression. Most likely one of the valves in #3 is stuck open. Did the car ever sit not running for months at a time?

Leakdown testing is static and everything has to be lined up perfectly. The "UP" mark on the camshaft should point to the front of the car when #3 is at TDC.

 

Thanks JimBlake and Roader. My confusion is that the events I'm experiencing contradict each other. If a valve is stuck open (zero compression) then it should bleed down fast, right, through the intake or exhaust manifold? To JB's point "wide open throttle?", no, but I'll do it again with WOT. "Disabled fuel?" no, "spark?" yes (at distributor harness). Wasn't warmed up either, should I warm it up to operating temp before any of this? To Roader, I can hear a very minor hissing at the oil drain back hole in the head on the driver's side/end, and this hissing is the same during bleed-down on all cylinders. I agree there's (probably) no big hole in the piston. The car has never sat more than ten minutes (exaggerating, but no it's never been out of service). It takes me forever to get time to fool with it, so unless I get lucky and find some time soon, it may be awhile for me to update. Thanks for your help and suggestions, and I'll keep y'all posted!

 

I've heard one person (out of all the others) mention low compression after the fuel-spray has washed down the cylinder walls. Don't remember the discussion very clearly, but I bet it wasn't ZERO compression. So it's still kinda confusing especially since you repeated the compression test wet.

Yes, compression test should be done with the engine warmed up. So if you repeat it, get the engine warmed up then unplug the wires from the injectors when you do this. Hold the throttle wide-open and crank the starter long enough so the compression gauge quits rising any higher. (Any chance you're touching the pressure-release valve on the compession gauge to let out the pressure?)

Minor hissing into the crankcase is normal. Pistons rings are never completely airtight.

 

Sure. The cylinder wouldn't even pressurize. The compressed air would just blow through the valves.

Your test results don't indicate closed valves either. Variable displacement engines close all the valves to deactivate cylinders resulting in no combustion, but there's still compression in those deactivated cylinders; the last normal cycle's exhaust gases stay in the cylinder acting like a spring.What's the actual code? The only cylinder-specific codes that I'm aware of are P030N: "No. N Cylinder Misfire". For number 3 the code would be P0303.

 

If the valves for cyl#3 are staying closed (all 4 of em?), then the compression test would behave pretty flaky, but I don't think it would stay at zero.

But you should be able to watch those valves open & close as you rotate the engine through a complete cycle.

 

JB - No, not accidentally pressing pressure release on compression tester (but I appreciate the "don't ignore the obvious" vein; that's happened a zillion times before!). Also I understand rings can't be 100%, so we're good there I think. Roader - I'll have to OBD it again and get that exact code. I forgot and can't find where I wrote it/them down. That only take a second so I can zap it soon (whatever "soon" means!!). One of these days I'll get some time to fool with it again. Thanks for all the help, follow-up, and encouragement!

 

Yeah, I hate to sound insulting, but from a distance I really don't like to take ANYTHING for granted...

 

Did you obtain a good compression test on a cylinder other than #3 to verify your procedure/tool is correct/functioning? I saw no mention of another cylinder results.

Also, leak down test is usually stated as 80/100, a regulated delivery pressure of 100 leakdown to 80 psi through the orifice. When a leakdown test is done, the air must go somewhere and you should be able to hear the leakage path; intake manifold (intake valves), exhaust pipe (exhaust valves), or crankcase (rings). Where was leakage path?

good luck