Compression issue / smoke on decel....

[meatpile]

I finally ran a compression test. Temp about 150-160. The results were:

1 - 172
2 - 179
3 - 80
4 - 171

I'm thinking - but unsure - that the #3 has a bad valve seal (or is it a guide?) because she smokes on decel. I did a good bit of visual testing - heavy acceleration up hills, no smoke. However, if I accelerate to about 5000 in second, let offf accel but leave in gear, there IS trace smoke with the throttle off, and then a poof of it when I accel. This was what led me to do the test.

I'm speculating that the leaky seal/guide is allowing oil to collect on decel - burning some if there is alot with no throttle, and then blowing it all out at first throttle. Because there is no smoke ever on acceleratiojn without deceleration, I'm thinking it's not rings.

Is this a 'common' ailment on these 2 liters? Also, is it possible that I could adjust the valves and remedy the problem?

How can I perform a leakdown test? I'd like to confirm it's not rings. Also, anybody use those new heads from Centerline?

[Roadtrip]

JC - It looks like the engine may have a bad cylinder, but first confirm how you did the compression test before we start pointing any fingers.

In general, you're looking for numbers between the cylinders to be within 10% of each other. The compression on 1, 2, & 4 look very good. #3 is very low. With one cylinder that bad, I would suspect that the engine does not idle or run as smooth as you expect an Alfa engine to run. IMHO, when compression gets below 120 or so, it's time to buy parts. For comparison, the compression in my engine (90k miles, stock pistons) runs about 145-155 right now . . . . showing wear, but still has much life left in the engine.

Confirm:

1. Engine sounds like it was warmed up enough, although I rather see it done at full operating temp.

2. All spark plugs removed?

3. Throttles wide open? (Take the fuse out of the fuel supply pump to prevent injecting fuel into the engine and disconnect the long rod). If you don't open the throttle butterflies you'll get false readings on the build-up of compression.

4. Fully charged battery?

5. Turn engine for 5-7 compression cycles?

6. Note how the compression builds. Do you get a nice big boost in compression on the first couple strokes, or does it start out weak and slowly build? A weak build-up would point towards weak rings. Low and not building would point towards a leaking valve.

7. Do the first test "dry", then "wet" on questionable cylinders. On cylinders where you get low compression, squirt about a teaspoon of oil through the spark plug hole and retest. If the compression markidly improves, it's likely that a compression ring is weak or broken.

I think you're analysis of the smoke is correct, but for review:

1. Blue smoke is oil
2. Black smoke is fuel
3. White smoke is coolant
4. Bluish smoke on heavy accel is probably rings
5. Blusih smoke on decel is probably leaking intake valve guides. (During throttle off decel, the cylinders are trying to suck in air, but the throttle butterfiles are closed, so it creates a big vacuum in the intake and sucks whatever it can past the valve stem and seal.)

However, with compression that low on one cylinder (if true), it's more than just a valve guides and seals . . . . . Most likely a bad valve. Really no way of telling until you get the cylinder head off the engine.

Do your compression test again as specified above, and try and narrow down the culprit.

Try not to get discouraged. These are high performance engines and they're getting old. In addition, POs may have treated them badly. It's up to us to get them back into shape they deserve to be in.

Maybe Papajam, John M, or Alex can chime in and make any further suggestions on troubleshooting. Papajam's the pro and I'm sure he's seen hundreds of Alfa engines with the same symptoms.

[meatpile]

Hmmmm.

My throttle was not wide open. How do I remove the fuse from the fuel pump?

I thought it was a strange 'build up'. If I just turned the key and help it, it stopped building. If I turned the key, then turned it 'off', then turned again, then turned 'off', the comp built up nicely. Except #3.

The engine doesn't idle very nicely, and the valves, to me, seem loud.

So - it seems like a good way to check for ring wear is to drip some oil into the low cylinder?

Other than the above, I did everything by your specs.

[Roadtrip]

The throttles have to be open or you'll get screwy readings.

Adjusting the valve probably won't do anything. If a valve (especially an exhaust valve) was not being allowed to close completely by too tight adjustment, it'd like get burned in very short order. Noisy valve trains usually are too loose of adjustment, although, don't expect Alfa valve trains to be silent. . . they're inherently a little noisy.

The fuse for the fuel pump is located under the very left side of the instrument panel. You'll see the big fuse box, then a small one with only 2 positions, only one of which will have a fuse in it. That's the fuel pump fuse. Alternatively you can go into the trunk lift up the mat and you'll see a wire running across the trunk with a connector in it. Just disconnect the wires and tape up the ends to make sure they don't short out. After that you shouldn't hear the fuel pump running.

[papajam]

Roadtrip lists a great procedure for doing a compression test which, if not done properly, could result in a wrong diagnosis. Ensure that the engine goes thru 5 - 7 compression strokes (or until the gauge stops rising) before releasing the key. Turning the key on and off will not yield accurate results.
There are two types of compression tests; dry and wet with the dry test always being done first. A dry test simply means that the cylinders are dry (no oil). If the dry test results show good readings that are within 10% of each other, the wet test doesn't even have to be done. If however, the dry test shows one or more cylinders more than 10% different, then a wet test should be done. The wet test will help determine if the lower compression is valve related or piston related. To do a wet test, pour a capful or so of engine oil thru the sparkplug hole and into the cylinder. Crank the engine for a second or two to distribute the oil. Now retest the compression on that cylinder. If the compression reading goes up by only 5 - 10 psi, then there's probably a valve related issue. If the reading goes up a lot, then there's probably a piston compression ring problem.
Although there are a lot more subtlies in interpretating compression test results, a least we'll know where to look next.