Leak Down Test Results

[No Crust Racing]

Car is an S13 with a CA18 in it, original 145ks on the odometer, unopened.

At the last track day I filled my catch can twice and wasn't making the power I thought it should given the recent mods (I realise the power component of this equation could be tuning alone). A compression test (cold/dry) cam up with 155-160 across the board, new is 170 or so (warm and wet) according to the FSM.

I've since bought a leak down tester and had some odd results. When dialing in some pressure from the regulator, 100psi,The leak down % gauge (it's a twin gauge model) reads off the chart whether it's hooked up to a cylinder or not, and by off the chart I mean 0 is a fully blown cylinder and the other end of the gauge is perfect seal, it's going well beyond the perfect seal portion. As it's a cheap gauge I'm not surprised.

That part aside, if I ignore the % gauge and just use te tool to fill a cylinder and listen for leaks, at 100psi I have noticeable air flow from the valve cap hole (or the dip stick if I put that back in), or the pod filter if I seal that back up and let the air flow from the rockers to the catch can and back into the intake piping.

I realise there will always be some air flow through these channels but how much air should really be flowing past the rings and into the sump, pressuring everything else. Should this be minimal? I'm confident I'm at TDC on the compression stroke also.

To add to the mystery, the car has also been using coolant but there was no air flow in the radiator during my testing of cyl 1, I will test the rest, but I thought I'd stop and clarify what sort of air flow I should hear coming from the valve covers. It seems like a bit to me given it was only 100psi of pressure.

Is it as simple as needing to do the test on a warn engine and am I using the thing wrong if the % leak down gauge climbs up to the set position with only 10-20psi of air in it and then climbs way beyond if you add more? Seems like there should be some way to add the amount of air you want then set that gauge so it knows this is your starting point.

Assistance always appreciated.

[The Mafia]

It sounds like your "original" 145,000kms is actually 345,000kms.

It doesn't matter how much you convince yourself it has done what's on the odometer, 99.9% of the time it has been wound back.

Have you done a proper compression and leak down test?

[No Crust Racing]

I guess my point was missed, I don't think 145ks is a small amount on a 23yr old turbo car, I also couldn't care less if it was 445ks, I put that detail in there more for the "unopened" part. The car at the very least has a full service history since coming into Australia which was 70ks ago, all signs show it to have been well looked atfer and a fairly original car, so my "guess" is it still has the original, and no doubt tired, rings in it.

As mention in my first post, compression test came back as 155-160 across the board. cold/dry, new is 170-172ish.

My leak down tester is playing funny buggers on the % leak gauge, as mentioned above, and I can clearly hear air coming from the valve covers, while I know there will always be "some" amount of blow by, should it be that obvious at 100psi on the regulator.

Edited May 20, 2012 by ActionDan

[51NNA]

I cant see anyway around a rebuild, you could try and cheat and take the rocker covers off and close up the gap on the baffles (not sure if you can on CA engines, I used to do this on Toyota's when turboing a NA engine) this would limit the blow of oil blow but not the gas. The lower power figures would suggest worn rings but without seeing the car or knowing the mods its just a guess.

Edited May 20, 2012 by 51NNA

[KiwiRS4T]

You have heaps of compression.. sounds like you should try another leak down tester.

[No Crust Racing]

That was my next question, if it's shot, why would static compresssion read nicely?

I'm certainly going to return the tester, but in the meantime, my question boils down to how much air flow i should hear/feel from the oil cap and dipstick on a healthy engine. I realise that's a very hard question to answer, but given it spat out 2L of oil at the last track day (I don't think it was over filled at the time) it made me think there's just too much boost and the rings can't hold it all in, or maybe the valve seats are shot and it's escaping into the rockers directly?

[No Crust Racing]

I've just had a bit more of a fiddle, disassembled the leak down % gauge, adjusted the needle so now it sits at the starting position properly when there's no air supplied. I'll do some more testing when I can get away with running the compressor for extended periods.

[r33_racer]

You definitely have more then 100psi coming in from your compressor too when checking and are you certain you are holding it on TDC on whichever pot you are checking? I know when we do it, it can be difficult to hold the piston at TDC accurately when waiting for the leak down gauge to stabilise as the pressure builds. Once you have the gauge reading properly let us know what you get.

But typical issues would be there for sure given the age of the engine. Rings would need replacing, valves and valve seats would be carboned up and leaking. Good chance valve guides and valve guide seals are shot and they too can be a cause of leakage. If you spat 2L of oil out then its more then likely a blow by issue from the lack of ring seal.

[No Crust Racing]

"You definitely have more then 100psi coming in from your compressor too when checking" <-- is this a question or a statement?

"are you certain you are holding it on TDC on whichever pot you are checking?" Yep, if I rotate the engine backwards the intake port opens implying it's on the comp stroke, if it was TDC on the exhaust stroke I'd have to rotate forward to crack the intake valve. I also removed the upper time cover and checked the marks,

I should be able to do some more testing tonight or tomorrow.

[r33_racer]

Sorry mate, meant as a question.

Ok cool.

Will be interesting to see what kind of leak down an old CA has.

[No Crust Racing]

There are two other things that have come to mind that bother me.

1 - To get the % leak gauge to the set position only requires 20-30psi of air, when I connect that to the cylinder of course it's going to show a huge difference straight up because it's now connected to a large volume of space it has to pressureise, so I'm guessing I should be waiting for the tool top pressurise that newly available space first?

2 - In the end of the hose that connects to the spark plug adaptor, there's a one way valve, like on a bike tyre. I can see now way for that valve to ever be acted upon by the compressed air and there's nothing in the spark plug adapter to act on the needle either. The only way I could get air into the cylinder was to remove that valve.

[No Crust Racing]

I spent some more time on it tonight and with the leak down % gauge on set it's only reading 10-20psi on the regulator side (ideally I would like to swap out the second gauge for just another 0-150psi gauge that way i can just compare the two readings and do my own % calculation as the leak % gauge is dodgy as hell).

Never the less, I tested each cylinder cold/dry on both TDC positions (just before TDC actually) so I could be sure I was on the comp stroke TDC as the exhaust stroke TDC obviously allows plenty of air to escape the valve prior to true TDC.

In each test the gauge gave me approx 30% leakage, and that was at only 10-20psi input pressure. In each example I could clearly feel/hear air escaping from the dipstick.

What I can't work out is how the comp test could read so well if it's that shagged. (When i did my comp test it was a dead cold motor that had been sitting for a week or more so it was dry/cold and it came back around 155-160 each cylinder.)

Keeping in mind the reason I started down this path is because the car spat out a heap of oil at the last track day, which it has never done before, and only after I swapped to a JDM S15 turbo and put in some more boost. So my guess is the rings just can't hold the extra pressure, but as I said if it was shagged why did the comp test come back good - and where is my coolant going! If it was a head gasket, surely I'd have seen bubbles in the radiator? Or maybe not when it's cold/dry and not enough pressure to show up a headgasket leak.

My other thought is it's so little input pressure that it's just cruising around the rings and not supply enough pressure to seal them up properly.

Perplexed!

[KiwiRS4T]

I can't help feeling that you are not doing the leak down test right - if leakage really was 30% would your compressions be so high? Perhaps you should repeat the compression test with a hot motor and then do the oil down the spark plug hole trick to see if there is any change to give you a clue as to ring condition.

[r33_racer]

I dont think your gauge is upto scratch. You may need to get another one or exchange it to compare results. It is not making any sense.

[The Mafia]

I haven't had time to post more but I needed two people to do this.

Someone needs to be under the car with a socket on the harmonic balancer bolt, rotating the engine back and forth ever so slightly while you are watching the gauge peak and drop off. The gap between TDC and a valve opening is 2 tents of f**k all and is hard to get.

The pressure you put into the cylinder will also push the piston down enough and cause a valve to open f**king up your readings.

My mate was having trouble holding against 170psi so in assuming 100psi is even enough to push down a piston.

[No Crust Racing]

I can't help feeling that you are not doing the leak down test right - if leakage really was 30% would your compressions be so high? Perhaps you should repeat the compression test with a hot motor and then do the oil down the spark plug hole trick to see if there is any change to give you a clue as to ring condition.

Chances are you're right and I've been asking myself that same question about leak vs compression test results.

I've got a buddy with a freshly built CA that I'm going to test for comparison.

I dont think your gauge is upto scratch. You may need to get another one or exchange it to compare results. It is not making any sense.

No doubt the leak down % gauge is suspect, I was thinking of replacing it with a standard 0-150psi gauge and I can just compare the two readings.

I haven't had time to post more but I needed two people to do this.

Someone needs to be under the car with a socket on the harmonic balancer bolt, rotating the engine back and forth ever so slightly while you are watching the gauge peak and drop off. The gap between TDC and a valve opening is 2 tents of f**k all and is hard to get.

The pressure you put into the cylinder will also push the piston down enough and cause a valve to open f**king up your readings.

My mate was having trouble holding against 170psi so in assuming 100psi is even enough to push down a piston.

I've mulled this over as well and spent considerable time testing in and around both instances of TDC in the cycle. I'm using fairly low pressures, see above, so I can work it myself, when I rock the piston back and forth the results get worse (implying a valve opening) but never better than about 30% and on rare occassions 20% .

If you watch this clip, this is the method I'm using and almost identical gauge, mine doesn't have the sticker down the side but given everything else is identical my guess is that's just another branded version of what is a mass produced item.

I've tried turning the reg pressure up also (ignoring the set gauge) to see if the extra pressure helps seal the rings, nope, just more hissing from the dipstick.

[No Crust Racing]

I'm replacing the leak down % gauge with a standard 0-150psi gauge, will be much easier to use.

So, thoughts on the air in the sump?

Can a car comp test well and leak down test poorly?

[51NNA]

Rebuildddddddddd itttttttttttttttttt, the time spent playing with the leakdown you could of had the engine out and stripped lol.

[No Crust Racing]

Yeah no worries, you fund it and I'll get going

I need to know if it's a legit problem first, then I can decide if I just plumb back the catch into the sump, run thicker oil and turn the boost down to get a few more track days in.

[51NNA]

You cant plumb it back to the sump or you'll get oil cavitation, what oil are you running?