Basic Leakdown Test Question - Plugs In?

[Smoove1010]

My understanding is that for a compression test, all plugs stay in except on the cylinder being tested. Is this the same for a leakdown test, i.e., plugs out one at a time?

Thanks in advance,
GK

[jess p]

For a compression test I take out all plugs. The idea is to rotate the engine as
fast as the starter will allow to attain a maxiimum reading, keep the throttle
fully open. This technique is cited in most of the literature.

For a leak down test it dosen't matter, but I would take the other plugs out
on some engines. An example would be to see if the head is warped. You are
just watching for air flow and looking to see where it escapes. As well as looking
for total air loss. Anything up to 10 percent is acceptable.

To evaluate an engine fully I first do a compression test. Look at the total on
each cylinder and if any cylinder is 20 lbs less than any other cylinder pour a little
oil into that cylinder. If the compression increases that indicates worn rings. If
compression does not improve it could mean worn,leaking valves. This gives you
more information of what to look for when you move on to the leak down test.

[Peter Zimmermann]

During a leak down test never remove a spark plug until the engine is on TDC of the cylinder to be tested. Then remove only that plug. The reason is the way the 911 2-valve head is constructed, and the exhaust valve is directly below the plug. If the plug is removed, when the exhaust valve is open, a good chance exists that a piece of carbon can fall from the tip of the plug and land on the exhaust valve seat. Then, during testing, when the engine is put on TDC for that cylinder, the exhaust valve will close against that very hard chunk of carbon, and not seal against the valve seat. Result? Faulty number.

What is the chance of this happening? Probably pretty slim, but my shop had cars come in "with 40% leakage in cylinder #x" asking for a second opinion. Of course, the car had been driven since the "bad" test at the other shop, and the carbon that was there was gone, so our test came up clean. The benefit? We gained a new customer - every time.

[mreid]

I agree with the above, but will add that you run a risk with all plugs out of dirt getting sucked into the other cylinders. On a 911 engine with the intake plugs angled up, this is a bigger risk than you think. On a compression test, as long as you test all cylinders in the same way your results will be consistent. Also, the gauge captures the compression reading, so engine cranking speed is not as important.

[Smoove1010]

Thank you Mr. Zimmerman for passing that along. Would you care to weigh in on the compression test procedure, i.e., plugs in or out?

For the reason you cite, it seems like it would be best to perform the leakdown first, followed by the compression test.

As I recall, the engine has to be at operating temp for both tests to be accurate. I'll have to work fast and wear long sleeves!

[Peter Zimmermann]

Originally Posted by Smoove1010

Thank you Mr. Zimmerman for passing that along. Would you care to weigh in on the compression test procedure, i.e., plugs in or out?

I've always done compression with all plugs removed, but I really don't do compression checks on 911s.

For the reason you cite, it seems like it would be best to perform the leakdown first, followed by the compression test.

Yes, definitely do the leakdown test first. If you know the rate of oil consumption you don't need a compression check, because the leakdown tests everything but valve guides. When you have compressed air entering a combustion chamber, and the second gauge on your tool indicates more than 10% loss, listen for air noise at the exhaust (ex. valve problem), the intake (intake valve problem), or the crankcase (piston/cylinder/ring problem).

If your car uses a quart of oil, or more, in 300 miles, you need a valve job. If your car uses a quart per 1,000 miles, all is OK. If your car uses a quart every 3,000 miles you have a rare 911! Because the valves are closed during leakdown, the valve guides are not exposed to compressed air, therefore aren't tested.

If you must do a compression check, use a mirror to check how much crud is at the base of each spark plug. If it's not too dirty, pull the plugs. Use full throttle to test (ignition DISABLED!), have a helper crank the engine while you watch the gauge. You want to use about 8 engine "cycles" per cylinder; you can hear and count the cycles. You want to watch how quickly the gauge builds pressure, which will indicate overall health. If each cylinder builds to about 160 pounds, but one cyl takes four more cycles to get there, a problem might exist.

One more thing, if you suspect broken head studs, when the leakdown tester is pressurized, lay under the car and listen at the joining point of the cylinder and head. You will usually hear air escaping if the problem is in an advanced stage.

As I recall, the engine has to be at operating temp for both tests to be accurate. I'll have to work fast and wear long sleeves!

NO! Not operating temp. Cool enough to work comfortably in the engine compartment. The difference in leakdown readings between hot and cold engines is so minimal that it won't matter, mainly because your combustion chambers are being evaluated against each other. It doesn't matter if you get 4lb and 5lb readings on a hot engine, and 5lb and 6lb readings on the same engine when it's cool. The engine, in either case, is sound.

I do leakdown tests in the firing order (1,6,2,4,3,5), and always put the engine on the next cylinder to be tested before removing that cylinder's spark plug.

[Quinlan]

So, could a reliable leakdown test be done on an engine on an engine stand?

[Don LaVance]

So, during the compression test it is mentioned to "disable" the ignition and the fuel system.
What is the best/safest way to accomplish this on a 1988 3.2.
Also, do I need an extended reach coupler for the deeply recessed spark plug hole .
Enjoyed all the above.
Thanks.
Don