Leak down test worth doing or not ?

[wildo105e]

Morning all,

Just a general enquiry but here is the background. My duratec install is complete except awaiting my exhaust from Ashley. So whilst bored i decided to do a compression test on the engine cylinders 1,2 and 4 pulled 100 with throttle shut cylinder 3 pulled 50 the numbers i think are irrelevant but the difference in the numbers to me seem very relevant.

Tested again with wide open throttle and 160 -170 and 100 for cylinder 3. So obviously cylinder 3 on my £150 gamble engine is down.

Is it worth doing a leak down test with a cheap tester and will it tell me any more ? it may not be the cylinder but a valve ?? sort of thing.

I have a spare engine now tested that and it pulled 160 on all 4 cylinders when tested on the garage floor so not crucial i getting the installed engine perfect i just hope it will run enough for initial testing etc

Thanks

[Graham]

if the engine has sat around a long while without manifolding on it your low compression might just be corroded valve seats, test no 3 a couple more times it may improve.

leak of testers are handy in that if valves are leaking you can feel the air coming out a port, or hear it rushing out the crankcase/breathers.

[rallyrob]

Years ago I had an Escort 1.8d (98 on an "S") having had it from new it had covered just under 50k in about 2 years when it developed a diesel misfire. I ran it like that for another year, then when changing the cambelt with the help of an old friend at the time he seemed to think it was due to the pump timing being out slightly, so he messed about with it trying to get rid of the misfire.

After that failed he suggested we did a cylinder leak down test. One cylinder was definitely down on cylinder pressure by about 30% over the rest, but it never stopped it running. In the end, we came to the conclusion that maybe it had a broken piston ring on one cylinder. But it never really affected its performance, not that it ever had any really anyway.

It went on to do about another 60k before I finally sold it when it was about 6yrs old.

[wildo105e]

cheers guys sort of confirmed my logic which is what i was after. I was thinking of running it it to see. The enegine came in 2 halves head and short block and all looked good. New valve seals and head gasket etc maybe a quick valve lap might have helped at the time.

Or maybe it was sold because of the issue !! It would be nice to know though as swapping it out costs money in oil gaskets etc just for the sake of a dodgy compression test.

[RS2OOO]

You could still put a drop of oil down bore which may help confirm the valve seat theory.

[apmonty]

Got a cheap tester off ebay worked a treat. I had low compression on no 1 100psi and 180 on other 3 with a compression tester. Suspected a bent valve as I had buzzed the engine. Before I striped it down I performed leak down test on all cylinder. As I thought gauged showed amber meaning too much of a pressure drop and loads of air hissing out of no1 inlet port. But surprisingly hissing to a lesser extent from the other 3 cylinders inlets though green. When I checked the tappet clearances all the inlets had closed up to less than I could get a .0015" feeler gage in and could have been in effect less than 0 holding the valve open.

Exhausts were fine at 12 / 13thou. Inlets are a waisted stem type I wonder if they could have stretched to close the gap? I never built the engine but the builder said they were set to 8thou and I don't doubt it. Very strange and the seats look fine, some sort of seat settlement may have gone on, but looks like i will replace at least all the inlets.

But I think the leak down tester is a useful tool in diagnosing the state of an engine.

[wildo105e]

You could still put a drop of oil down bore which may help confirm the valve seat theory.

ok so got some oil in the bore which is harder than you think in twin cam engine. A tube and an oil can did the job. A few hand cranks, let it soak and then plug than back in. A few spins on the starter and then re-test the compression gauge. It now matches the previous readings for the other pistons on both open and closed throttle

I think it will be good now especially once it runs under it's own power. For curiosity i will still leak test it a cheap tester is about £20 which is less than the engine oil for it cost.

Thanks guys and watch this space for the leak test results.

[wildo105e]

So my ebay leak tester turned up and a quick change of fitting to match my air line and ready to go. I tried it on my now spare engine and a weird result on cylinder 1. Read the instructions and you need to calibrate it first 3 pistons showed around a 20% loss which according to the gauge is plenty acceptable although cylinder 4 seemed a bit lower but still in the green zone on the gauge. Which was weird as on a compression test showed great.

Repeat the test on the in situ problem engine and nr 3 was showing great which was the one in question. Ironically cylinder 4 showed lower than the other 3 but still in spec. Maybe it is something about duratecs ?

So it seems i probably have 2 good engines (patience may have been a better plan before getting a new engine). My plan was always to by a second engine to rebuild with rods pistons etc it now seems a shame to tear down a good engine to throw most of it in the bin !!

[Graham]

20% loss is actually ok, its common to see 30% engines which run perfectly well, 10% and under is very good. its like a lot of measuring systems you have to get used to whats good and what isnt.

one thing you need to be aware of with leakdown testers is results can vary depending on exactly how/where the pistons are. sometimes a cylinder leakage will drop dramatically if you put a spanner on the crank and rock it a bit as rings seat

[wildo105e]

Thanks graham, I did notice some things that you mention. When using the tester if the psi went down it gave funny results. Putting the air nozzle on untill the compressor fired back up and charged again cured this. As you mention with rings when i got the result for the cylinder 4 it was interesting to turn the engine and see the results vary as it rotated.

[Graham]

Thanks graham, I did notice some things that you mention. When using the tester if the psi went down it gave funny results. Putting the air nozzle on untill the compressor fired back up and charged again cured this. As you mention with rings when i got the result for the cylinder 4 it was interesting to turn the engine and see the results vary as it rotated.

its better to leak test around TDC, as thats where the most wear occurs and ring seal is most important. i always leak test with the compressor connected

[katana]

its better to leak test around TDC, as thats where the most wear occurs and ring seal is most important. i always leak test with the compressor connected

Exactly correct - we used to have a 2 foot long spanner on the crankshaft nut and put some effort into holding it at TDC and 100psi - Scary ! Air source needs to be constant supply as gauge is effectively measuring the difference between supply and retained = leakage