I seem to have a really unique issue, and would love to hear your $.02. I think I have read all most of the threads on EVOm on the subject (most all don't apply to this), consulted several friends that have owned and been around the 4G63s since the early 90s DSMs (and that are master mechanics). We're all scratching our heads. Maybe a fresh set of eyes will churn a new idea? Keep in mind - I hit/hold boost at 25+ on the stock turbo, with what I consider normal settings on the boost controller for that. A change in AFRs is what is driving this test, but its in testing, that I am having the issue.

So here is what started happening: I attach boost leak tester, attach fitting from the air compressor (regulator set to 40psi). Over 10 seconds it builds 3-5psi and levels out, never building any more pressure. Engine is manually cranked to precisely TDC (no effect). PCV replaced (no effect). PCV and crank case vent capped (no effect). BOV not leaking. During this time while under pressure, you just hear a lot of air whizzing around in the crankcase, and you feel some air coming out of the exhaust. No sounds of something particular leaking in any way (no bubbles around piping/hoses when sprayed, etc). Totally unlike months before when I was boost leak testing, and the entire system pressurized as normal. Another thing to note, we popped the oil fill cap while testing (also not leaking), and a TON of pressurized air exited -- much more than the 5psi reading on the gauge. Our 'guess' was that it felt like 20-30psi of pressure was on the cap when we relieved it. It seems that there is a ton of pressure in the system that is almost not at all registering on the gauge. It drains a 40gal compressor quickly enough.

So what is causing this??? Any ideas? I am almost back to the drawing board here. Everything I can think of would not have such a huge effect like this (and keep in mind, this is a meticulously maintained 30K mi car that is not beat up), like throttle body seals, turbo seals, etc. Anyway, if you can think of anything, I'd be glad to try it!! I just cannot explain how I can easily boost all that I want while driving, but cannot hold but 3psi on a test, with it all hissing away in the crankcase.

Thanks!!
Erik

 

Your gauge might not be reading right.

You shouldn't have so much air in the valve cover though, that sounds like PCV right there but you said you replaced it.

Where are you attaching your leak tester?

 

I suspected the gauge already, but I have ruled that out. I bought a brand new 30psi gauge from Grainger, and that did not change things.

I have tried the boost leak test 2 ways (no difference either way) > I use a transition that I got from Hose Techniques to attach the tester directly to the turbo inlet, and I have also tried just putting the tester at the end of the intake pipe where the MAF would be attached.

I know something is off because previously (when the car was boost leak testing as it should), the (stock) intake pipe would start to accordion outwards as pressure built up. Now with the gauge reading 3-5psi when boost leak testing, the intake pipe has barely stated to accordion out at all, so I know there is not a lot of pressure in there. But as mentioned, crack the oil fill cap, and you suddenly realize that there is a ton of pressure in the crankcase.

Very weird, love to hear more ideas!

 

Did you do a compression test? Is the Car running fine?

I don´t think that even a bad valve could leak that much throught the exhaust.

 

Yeah, car is running *great* (if it wasn't for the AFR anomaly, I would not even suspect a boost leak). But I suppose you are right, a compression check is worth it just to keep ruling things out. Keep the ideas coming. Thanks!

 

my freinds car did they same thing.. all 4 pistons were cracked.

 

Just my 2 cents. Do a leakdown test. You might see as stated above a ring land/ piston problem. Compression through the intake though should not come through the crankcase unless you have a massive sealing issue.

Also, turbo seals okay? Maybe an oil seal on the turbo is not conforming and now with positive pressure is goes through the oil lines, But that would be a short lived turbo.

Like I said, I would do a leakdown first.

Good luck.

 

Another method I would use is to bypass the turbo itself and try this; take off an intercooler pipe and put the boost leak tester on there. If that doesn't work i would do it straight onto the throttle body.
Look to see if you boost pressure goes up.
Also do you have a boost gauge in the car?

I would have 1 person pressurize the system and have one person read the boost gauge from the ****-pit. This HAS TO WORK..you will definitely build pressure this way!

keep us updated

 

Thanks, I'll give that a shot (though it may be this weekend before I get to it). I am presuming that this version of the test is just to ensure that the engine itself will hold the boost, when you take the turbo and intercooler piping out of the mix. And if such a test is successful (can hold 25-30psi), then it points me in the direction of something else - vacuum/boost hoses, IC piping, or turbo? Would you also recommend capping the crankcase breather and PCV when doing this? And having the engine at TDC?

Though I will still try a leak down test as indicated, I really think the engine is healthy. The car has had an easy life relative to a lot of EVOs, fluids always changed early, no mods until 20K, with only 30K on it (and that was just a boost controller and a conservative tune). The car does not smoke 1 bit, even at WOT, and barely burns any oil (in line with what a healthy turbo engine does burn)

Thanks for the help!

Erik

 

I've noticed a couple times doing boost leaks myself that there's a spot right around tdc that runs all the air out of the exaust and I can't build any pressure. I rotate it a degree or two and it fixes it. May or may not be related to your issue though.

 

Previously, I had just been 'blindly' turning the crank, and boost leak testing when I hit resistance (believing that the increased pressure was the end of a stroke). After that, I just did TDC off the marks on the timing gears and valvecover. I will try a few degrees past next time. Thanks!

 

If you put the engine at TDC you are guaranteeing that a cylinder will be right between the exhaust and intake stroke where there is valve overlap, which is not what you want to do.

Like others have said, the air can enter the crankcase through a bad ring, cracked piston, pcv line on either the back or side of the valve cover, and I guess maybe through the turbo too.

I think I like the idea of testing almost right at the throttle body and seeing what happens.

 

A few things...

Remove the two vacuum lines that go to your valve cover. This will eliminate the pressure on the top end of the engine. Normally, that pressure wouldnt be there anyways considering on a boost leak test, you're supplying the valve cover with positive pressure from the turbo inlet pipe (which wont ever see positive pressure while driving) and you're also cranking air past the check valve on the PCV valve on the backside.

I always remove those two vacuum lines and cap them when doing a test.

Secondly, setting the regulator at 40psi is weaksauce.......I have a 30gal compressor with a twin cyl setup on it, and i set the regulator to its maximum, and just watch the boost gauge as I pump air into it. With my regulator set to max, I can usually pump just shy of 30ppsi of static pressure into the system.

Setting the engine to TDC isnt going to help either, I have yet to check the positioning of the cams/crank while doing a boost leak test

 

I had never set the compressor regulator over 40psi before, and had good results, but that was probably because the system was pretty much leak free at the time.

Is there a consensus here that 100+ psi on the air compressors regulator is a good thing while boost leak testing (provided that you only pressurize the system to the intended boost you will be pushing, and don't just completely overwhelm the car with pressure). Or will 100+ psi really only ever charge the tester to about 30s-40s psi?

How fast should I see the boost drop, once I disconnect the air compressor (provided I no longer hear any leaks, and everything else is healthy?)

Thanks again!
Erik

 

I have a small air compressor in my garage and it works perfectly fine. With your situation you want to make sure that you can get to 25psi and once you get there if there is a huge boost leak you will hear it. If there is a small boost leak than the pressure will drop fast, now if that is the case then get a bottle of water/soap mixed and spray it.


I have never had to put it at TDC and i do a boost leak test twice a month.
Try the method and let us know what happens..