I bought a high compression JDM B20. I put it on an engine stand and adjusted the valve lash. It was out of spec on all valves (loose).

Next, I did a leakdown test on the cold engine with correct valve lash. Crank pulley was secured so it didn’t move when each cylinder was tested at TDC.

These are the results
cylinder 1 = 51 of 75 psi, 32% leakage at intake valves
cylinder 2 = 50 of 75 psi, 33% leakage at intake valves
cylinder 3 = 60 of 75 psi, 20% leakage at exhaust valves
cylinder 4 = 74 of 75 psi, 1% leakage past rings

I think the test should be performed on a hot engine, but I didn’t want to install it only to find out it is no good. I think acceptable leakage is 10% max. Do you think a hot engine would be less than 10%?

Thanks in advance for your consideration.

 

Hot or cold the engine should be sealed against leakage.

Since you called out specific valves for leaking,I assume you are hearing them leak from their respective areas.

Looks like a rebuild is in order......

 

That’s what I was afraid of...thanks. I might install it anyway and try again on a hot motor.

 

Since it's valves, you could possibly get away with hand lapping the valves. That would mean pulling the head and replacing the head gasket but it's still a lot less parts and work than a full rebuild.

Also, typically 10-15% is considered okay on most engines except full time track cars. Track cars, they tend to rebuild constantly to be 5% or less. However, what I've read is you are normally looking for a variance of greater than 15% between cylinders. If they all have X loss (typically past rings) the engine is still considered okay just worn to some degree.

 

There is no acceptable variance in a leak down test. All cylinders should be 10% or less.

Additionally,he called out specific valves, which indicates to me that he is hearing the air escaping from the exhaust or intake manifold.
That is never considered ok.

 

My leak down tester instructions indicated a variance between cylinders of 15% or more as quite bad. Vary much in line with typical compression testing parameters.

As for a 10% only or less, I disagree with you. An older semi tired but still good running motor can be 15% as long as it's past the rings and still be a solid daily runner. 10% or less tends to be a brand new engine or track car stats where 5%-10% is the typical thresh hold depending on the venue.

Also, he's doing it on a cold engine as such the percentages will be higher.

Leak Down Tester

Also, you missed my () where I said "past the rings" meaning not valves or water jacket. The latter two are bad.

The other part never asked or clarified, what order was the testing done. Or are we just assuming the testing was done properly?

I say this as at first I went the wrong order and of course valves were not closed... Once I did #1, crank 180, #3, crank 180, #4, crank 180, #2 I got accurate results.

And lastly, from what I've researched, beyond just the baseline bit, is you do the test multiple times and average your results to get a clearer picture of the test(s).