Question: Checking for Vacuum Leaks in Intake Tract

[NolimitZ32]

Ok, so I have seen and heard of people pressurizing their intake system to check for vacuum leaks by removing the MAF housing and capping it, then filling the system with compressed air at fairly low pressure and listening. I attempted it this weekend and could not get it hold pressure in the intake on both cars. I would sit there with the compressor going, pumping air into the intake and got absolutely zero increase in pressure, it just stayed at atmospheric pressure. I'm fairly certain that at least my Subaru doesn't have any vacuum leaks so tell me is there something I'm missing or doing wrong or is this method just that inconclusive?

Thanks
Anton

[AZhitman]

Maybe I'm missing something, but that would require all the intake valves be closed at all times. Not gonna happen.

I use starting fluid. Run engine at idle and spritz starting fluid at all the joints. When the idle fluctuates, that's where your leak is.

[NolimitZ32]

Ive done that and I cant find anything, I tried to pressurize the intake to make sure. As far as all the intake valves being closed thats not entirely true if I understand correctly, each cylinder is either going to have the intake or exhaust valves closed so the pressure should be held by either the intake or exhaust valves on any cylinder and should pressurize. So yes the cylinder itself may not be able to hold pressure due to open exhaust valves but if on a given cylinder the exhaust valves are open than the intake valves are closed and the pressure should stay on the intake manifold side of the valve, you see what I'm saying? I'm not disputing just trying to see if my logic is correct.

[AZhitman]

That may well be the case, but I've never counted on it when pressure testing a turbocharged car.

The other alternative is disassembling / re-sealing, but that's a nightmare on a Z32.

[NolimitZ32]

Yeah pretty close to a nightmare on a Swapped WRX as well, I actually have more vacuum lines in there then I do in the Z haha.

Im going of the same basis as a leak-down test only there you obviously have to make sure that the cylinder is fully sealed but the logic of valves holding pressure would be almost the same except the pressure is on the opposite side of the intake valves in this case.

[GerryO]

http://home.swipnet.se/e-solutions/Vacuumleak.html

[NolimitZ32]

Thanks Gerry and yes I've seen this page plenty of times. That site however doesn't say anything about how long you should pump. I figure if my Compressor puts out 2.6 CFM at 40 psi it should be putting out at least a quarter of that at 5psi and the entire intake system on my WRX is very small since it has a TMIC so it should only fit about 3 or 4 cfm at atmospheric pressure which should only take about 5 minutes to fill. but it wont fill so either I have a huge leak and the air is escaping immediately or I'm not getting something.

[GerryO]

Imagine the best way to leak test a head would be to bolt/gasket it to a flat surface with the camshafts removed so that all of the valves are fully closed, and then to pressurize each cylinder separately.

Can't imagine pressurizing an intake system other than by continuously and quietly supplying a volume of air that is larger than the volume that is escaping through normal pathways (at least one open set of exhaust valves) and listening at/feeling/soaping specific locations to detect air that is possibly escaping in the form of a leak.

[robomatic12]

It sounds like you have a fairly large leak somewhere. I built a tester out of a hockey puck, tapped it for my air compressor quick-disconect and a gauge in small increments (<15psi). I set my regulator on the compressor to 5psi and hook it up. The gauge will hold 5psi and slowly bleed off if I remove the air supply. The intake system should stay pressurized and not bleed down right away.

The best way I found is if you fill up the air compressor tank, shut it off so you don't listen to the noise of it trying to turn on and listening to leaks at the same time.

[NolimitZ32]

Robomatic thank you so much, this is exactly what I needed. I guess I gotta remake my plug to use a quick connect rather than a shrader valve fitting. thanks again man.

And Gerry: what I was saying earlier is that even if the exhaust valves on one of the cylinders is open, the intake valves would be closed and since you're pumping air in from the manifold side it should not be escaping into the cylinder therefore it should not matter if the exhaust valves are open, do you follow what I'm saying?

[GerryO]

And Gerry: what I was saying earlier is that even if the exhaust valves on one of the cylinders is open, the intake valves would be closed and since you're pumping air in from the manifold side it should not be escaping into the cylinder therefore it should not matter if the exhaust valves are open, do you follow what I'm saying?

There's a bit of valve timing overlap, bigger with performance cams, where both intake and exhaust valves are open at the same time, at the top of the exhaust stroke. It's animated pretty nicely here (scroll down and play):

http://www.howstuffworks.com/camshaft.htm

[NolimitZ32]

Hmmmm. Hmmmm. ok so if I'm not building pressure that means one of the cylinders might be at this stage in the timing, then technically if I turned the engine little by little and tried again i should be able to find a sweet spot. Thanks Gerry.

[jcv]

Alternate ways to check for vacuum leaks: DRY ICE, vacuum cleaner, compressed air and insense
Here is a Link:
http://www.bimmerfest.com/forums/showth ... ?p=5165240
http://www.m5board.com/vbulletin/e39-m5 ... d-s62.html

[NolimitZ32]

Wow, zombiefied thread lol. I ended up using a party fog machine and some homemade sleeves to transfer the fog from the machine to the intake.

[t.mcginley.jr]

Did you end up finding the leak??? lol the 1+ year suspense is killing me

[NolimitZ32]

I found the leak, then blew the turbos, blew the HGs, tore tha car apart, installed Carillo/Eagle, bored, blueprinted, HKS2530s, 3" full exhaust, so on and so forth.