Compression test results - diagnostic help please
 

In brief: I installed my MS last weekend, and when trying to tune it, I could not get a smooth idle. Eventually I dropped the oem ECU back in and had the same idle. A little exploration (checked plugs, clipped a timing light to the spark plug leads) and I found that the rough idle was due to misfires in cylinder 3. I chased down a number of potential causes, but eventually picked up a compression tester and indeed:

Cylinder | 1 | 2 | 3 | 4 |
Dry PSI | 140 | 136 | 70 | 144 |
Wet PSI | 160 | 166 | 85 | 170 |

Please note, I forgot to block the throttle open for this. Hopefully the results are still valid. 'Wet' means I poured a half-cap of oil in and then re-tested.

I'm still holding out hope this might be a valve so that I don't have to pull my engine. Any experienced opinions out there? In the meantime, I'm off to start pulling the head.


I've posted details on my setup below (all work done by the previous two owners). I've been driving the car for maybe a thousand miles. It's been running very rich (and failed emissions) due to the FPR+1.8 injectors. Also, I was running a Bipes to pull timing with the oem ECU.

There are two other details I think worth mentioning:
1. At an early point trying to get the MS COM port functioning properly, I turned the key to ON and heard a loud pop or bang. Sort of sounded like a misfire (but the car was off, obviously). It ocassionally since then has idled smoothly, but mostly pretty steady misfires.
2. I drilled and tapped my intake manifold for the AIT with the IM on the car. I did take a vacuum to the hole in an effort to clean out any chips, but I'm concerned a chip might have dropped in and fragged piston ring or cylinder wall.


My setup:

Stock Bottom end at 163,000 mi
Greddy Turbo kit at 6 psi
Vortech fuel pressure regulator
Manifold has relief cuts
Bowser Heat Shield
Stripes Intercooler kit
Turbo Smart BOV Type 1
2.5 Inch Down Pipe Mild Steel (turbo tony)
2.5 Inch Race Cat Stainless Steel
2.5 Inch Enthuza Turbo Cat back Stainless Steel
1.8L Tan Top Miata Injectors
K&N Cone Air Filter
HKS Headgasket 2.0mm
ARP Headstuds
Dual Core Radiator

Maintenance done in the last 6 months or so (by the PO, quite possibly questionable work):

New Water pump/gasket
New Timing Belt
New Heater Hoses
New Thermostat/ gasket
New Oil Return Line
New Throttle cable
New Intake/Exhaust Manifold gaskets
New Valve cover gasket
NEW head (cleaned & inspected as well)
Multi-angle valve job

I'd re do it with the throttle open. Number 3 will probably still be lower than the others. Judging by your wet test you have a leak in the top end and a head refresh would probably be worthwhile, but there is a bit of ring wear.

JMHO.

if it were rings the compression would most likely have shot up more than 15psi, and it went up evenly across the board.

At this point I'd probably want to do a leakdown to test to pinpoint the loss of compression exactly.

leakdown test > compression check

second the leakdown test. compression tests are kind of quick and dirty in that they'll tell you if there is a big problem, but not really why/where.

Harbor Freight sells leakdown testers for cheap, and you'll also need access to an air compressor (also pretty cheap at HF).

F. I've already spent $300 more on this Megasquirt install than I originally planned. But, I did know you were going to tell me that. I started shopping for air compressors yesterday.

So I could probably get some leakdown results this weekend. You definitely think it's premature to skip that and pull the head?

Loading up the parts shotgun before you know whats wrong is a bad idea...Been there done that. Sometimes you get lucky and your right, other times you waste more time and money than you should have. I'd say diag it right and be sure.

Alright, at long last that harbor freight cylinder leakage tester came in tuesday, and my neighbor's finally home today and let me borrow his compressor. Freaking summer is flying by.

I just ran the test at 80psi, warm engine. I'm going to go back and run it at 100psi after reading some of the other posts (my leakage numbers were ~2-3% on cylinders 1, 2, and 4, and ~5% on cylinder 3. This seems surprisingly low).

So preliminary results are that cylinder 3 has about twice the leakage of the others, and I can distinctly hear air flow out the intake. I can't hear any leak out the oil dipstick on any cylinders.

Results are the same with the compressor at 100psi. The leakdown tester is I guess calibrated to about 15psi, so the cylinder pressures were unchanged from the previous try.

On cylinder three, the needle bounces around quite a bit.

Anything I might be doing wrong to produce such low numbers? I had the throttle blocked open, the radiator cap off, and the oil dipstick out.

From what I have read, anything less than 10% leak down is within spec.
Some compression testers I have used are synsitive to junk in the spark plug bore and do not seal completely. Clean out #3 to the best of your ability and redo the compression test.
You are giving the same numbers of cranks/turns of the motor on each cylinder, correct?
However your later post at 3:43, where you hear air escaping out the intake on #3, points to a bad valve. It is possible there is something on the #3 intake valves that will not allow them to seal, carbon build up or aluminum frag. I would just think there would be more than 5% leak down if that was the case.
Anyone else?

A lot of people will tell you to do a leakdown test, but few really know how to do it. In my opinion if you are confident with your compression #'s you don't need the leakdown tester.

The percentages a leakdown tester will give you are great if you are trying to determine weather or not to replace/overhaul the motor. In your case, you know something is wrong with the #3, you just need to know where.

Most compression testers have a hose with a quick connect air fitting attached. Set the #3 at TDC. Make sure the #3 is on the compression stroke, and intake & exhaust valves are closed. Take your compression tester hose and screw it into the #3. Connect the hose to an air source and bring the pressure up to 100 psi. Listen carefully.

Air leakage out of the tailpipe means an exhaust valve is open/bad. Air leakage out of the intake manifold means an intake valve is open/bad. Air leakage through the oil filler means the rings are bad.

FWIW, I'm on my second HF leakdown tester. I put 100psi to the first one and ruined it, so I exchanged it for another. This time I carefully raised the pressure a little at a time and mine also calibrated at about 15psi. In my opinion this would increase the sensitivity greatly over testing at 100psi so with compression as low as 70, I would expect the the tester to read >50% leakdown.

As a comparison here were my #'s using the HF tester:

CYL.--DRY/WET---LD

1-----176/200----28%
2-----174/195----35%
3-----154/177----50%
4-----180/195----30%

I was getting air leakage through the exhaust so it was obvious I had a problem with the exhaust valve. As it turned out, some debris got into the oil and ruined some of the lifter bores as well as the camshafts. My machinist thinks one of the lifters was getting jammed and not letting the exhaust valve close all the way. It was a PITA. I replaced the head and rebuilt the bottom end since I had to pull it to check for damage anyway. I guess the car is better off, but if I had to do it all over again, I'd have just driven it until it died.

I don't know where you live but I have an old 1.6 motor in the garage.

Ok, had trouble posting on Saturday. Hopefully this works now.

Despite the broken leakdown gauge, I'm pretty confident at this point that it's an intake valve broken or stuck:
1 - Misfiring in cylinder 3
2 - Easily audible hissing from cylinder 3 with leakdown tester
3 - No air from dipstick tube in any cylinder
4 - Sketchy headwork done by previous owner

Any other opinions out there?

A bit off-topic: this little summer track-toy idea has degenerated into a dollar-sucking beast (what's going to happen when I actually get it on the track!?)

I've been working so far entirely with my roomate's (fussball.biturbo) little bag of emergency tools and a few essential tools i've picked up. It's a huge pain in the ass, and most projects involve one or more trips to the hardware store to pick up missing tools. Sunday was thunderstorming, so I spent the day picking up a basic Craftsman toolkit and setting up a little workshop in our (formerly) extra storage room.

If you're looking, it's not too late to catch the tail-end of some sweet father's day sales at Sears.

Check it out:
http://tonic.smugmug.com/photos/317999235_bzvmR-L.jpg

http://tonic.smugmug.com/photos/317998026_zuNbz-L.jpg

http://tonic.smugmug.com/photos/317998843_Ecg8V-L.jpg

It sucks but welcome to the hobby of modifying cars. :)

Sounds to me like you're on the right track. I'm learning about diagnosing compression issues by reading your post...Good luck.

Well ... wow. Life goes on and it's taken almost 4 weeks for me to get to pulling the head. I finally got it off today and found the following:
- the cylinder walls look extremely smooth and clean
- I cannot find any evidence of a broken valve or valve spring
- with the intake and exhaust cams both at TDC, I can see a bright light shining through both intake valves of cylinder #3 (all other valves appear tightly closed and show no light)

This is a different result than I expected. I talked with the previous owner, who replaced the head only a few thousand miles ago. Apparently the head has ~100,000 miles on it and received a 3-angle valve job by a shop in Connecticut.

So, the questions:

A. What might have caused both valves in the same cylinder to stop seating correctly?

B. What's my best course of action?
1. Strip the head down and re-build
2. Bring it in to a shop to have them get those valves seated correctly
3. Buy a new head


Hopefully, I'll get some photos up tomorrow for visual aids

I think the cheapest course of action is to source a used head.
It's also the riskiest. At a minimum, you'd need to check that the replacement head isn't warped (and I'd have that done by a machine shop).

And you're already in for a new gasket, and it's a good idea to put new studs in. Once you've dumped that much, it's really tempting just to rebuild the head.

Yeah, I'll do head gasket of course. And I plan to take it to a shop to have them check it for straightness. I'm hoping to re-use my head studs (ARPs with less than 5,000 miles on them? Is that a no go?).

You can reuse the ARPs, they do not tend to stretch out.

Don't buy another head yet, take the head to the shop that did the work and tell them that they didn't seat the vaves properly and that they have cost you a lot of money, and if you are lucky and intimidating enough, they'll fix it for free. It would be great if you could go with the previous owner, since he is the one who got the work done.