Should I just warm the engine up and let a (smallish) vacuum hose suck in some water or engine oil?
How much is too much?

 

I understand what you're going for but with the car running you're not going to achieve what you're trying to do. Without a smoke machine or even a fog machine like what you would find at a Halloween party, you're going to have to hunt for EVIDENCE of an exhaust leak. So sut, hot spots, the feeling of air coming from gasketed surfaces, or even areas of non uniform discoloration.

Even with the twins, there's few places you can actually have an exhaust leak.

To answer your question, you can put small amounts of 2 stroke into the intake manifold while it's running and you'll get the smoke you're looking for but it's not going to come out of where the exhaust leak is like you're thinking. The engine needs to be off for that to work.

I would suggest finding a creative way to make some smoke or fog. You're going to drop the midpipe and feed the smoke in through the downpipe under the car. Put some rags to seal it so it can only fill the downpipe and ultimately the turbo(s) and manifold.

 

So, you don't think the smoke would emanate from any crack in an obvious way like I think it would?

The car doesn't make the boost it's supposed once it's up to full operational temperature, which is why I suspect an exhaust leak. Probably a small one.

There's intermittently a tiny bit of smoke in the engine bay, but it may be oil contamination. The engine has a sort of percussive timbre when you listen from the engine bay when the car is cold and the silencer is in, but it seems to go away when the engine is warm and the silencer is out. So that's kind of a wash.

I actually already found a small leak in the down pipe at the flange, which I had welded. It didn't make much of a difference, though.

I also pulled the turbo *for reasons* and replaced the metal turbo gaskets while I was at it, but I didn't take the next step and pull the manifold in order to inspect it for cracks. I suppose it might also be an issue with a flange not being perfectly flat, or maybe the downpipe isn't perfectly aligned with the outlet...

 

For a compromise in any of the sealing surfaces not large enough to be obvious, it won't smoke itself out with the exhaust pressure behind it. It will simply continue on and go out the the rest of the exhaust. Any trace amount that finds its way out will be inconclusive. For single turbo cars you only have 3 places: manifold to block, manifold to turbo, downpipe.

Unless the chra is not fully seated or the exhaust housing has a crack or something in it, finding the leak will be fairly straightforward. Flanges do warp and cracks do change in size with heat. Whether it be to get louder or quieter, the phenomenon is real. It makes for finding them a little more difficult as you've explained. Scenarios like this are exactly what smoke tests are for.

Checking your flanges for flatness is easy. At a minimum you would be able to rule out the manifold. Then you would check the exhaust housing flange for straightness. Depending on your v band connection type, you can do the same there as well. Wouldn't hurt to replace the clamp too. Don't rule out the wastegate as being a possibility either.

if you can take the manifold and turbo into a very dark space and use a very bright light, pin holes, cracks or otherwise will show themselves. You may be over thinking this. In a single turbo set up, exhaust leaks are very easy to find using the proper methods.

 

 

he lives in japan

 

Yeah, if I take everything off again, finding cracks should be pretty easy. The setup has seen a lot of track time, so it's probably time to inspect the manifold, anyway... But god damn, that job is a pain in the ***... Every single nut and bolt is a challenge. Wonder if I can't just pull the turbo, manifold, and wastegate as a singe unit... Probably not.

I already checked the wastegate, which seemed fine. Unless the fire ring isn't sealing once the whole thing gets hot (not entirely unlikely) or the spring is getting weak once it's hot. Or it's placed in a bad spot and it's getting hotter than it should. It basically works perfectly for a few pulls, and then drops 2-4 PSI, where it stays. It's pretty inconsistent.

I would consider replacing it with a more modern piece, but it has an old-school four-bolt flange and not a modern v-band flange, which would make the process more complicated and expensive.

 

Admittedly, I did not notice that. But those smoke tools are sold all over the world.

 

I could buy one, but they are a bit too expensive here.
maybe if I can find a used one.

 

made one similar to trace a vacuum leak. this will work

 

What happened to briefly blocking the end of the exhaust pipe with a sneaker (or similar) and see where the exhaust comes out.

 

I tried this, but I think the exhaust (especially on rotaries) is just too hot to feel around for small leaks. Listening just isn't enough.

It just occurred to me that I could just use something like a shop vac in blower mode to pressurize the exhaust. That way I could feel around and use soapy water.

 

Thought you were suffering from this a couple of years ago, still the same problem?

You might be able to. Do it here from time to time (minus wastegate). Depends entirely on the manifold design, w/g and turbo size, but there's additional parts that need removing before you use a floor jack to gingerly lift the motor asymmetrically (not on the bottom of the sump ) to avoid stressing those bits..

I'd wager it's the turbo flange on a track car, if you haven't done that recently....can be a near every event task to true the flange on fabricated manifolds!

Not sure about a vac, if you have a small compressor regulated way, way down, a few hose fitting into a plastic cap that fits the end of the pipe and some soapy water/dishwashing liquid, that might serve as a poor man's smoke generator.

 

I only get to drive the car a few times a year, and only work on it every once in a while, so I never really found the exact cause.

The car still drives fine, it just bugs me that the car doesn’t do what I expect it to.

I will be moving it closer to my house soon, so I’ll be able to work on it more.

I use a portable compressed air canister to do boost leak tests, but for something like they exhaust system, constant flow would probably be ideal.

It’s a Greddy / Trust turbo kit, but it’s well over ten years old at this point.

 

Have you try shop vacuum with the blower/exhaust port? Hook that to the exhaust tip and seal it up, with the car rise up … all you have to do is spraying soapy water to all the exhaust components(any leak area will bubbling up). I haven’t try it personally, but I saw a video on YouTube from chris fix?

 

I haven't even seen my car since last year, lol.

 

amazon sells "smoke cones" you basically stuff in the tailpipe and and pressurize it by running it briefly and use soapy water

 

I already tried plugging the exhaust and using soapy water. You're probably underestimating how fast a rotary exhaust manifold gets hot.
The backpressure does significantly change how the engine sounds, though.

I am pretty sure if an exhaust leak is causing my problem, it only happens (to a significant extent) when the exhaust is fully hot, so soapy water isn't going to be an option.
I will probably need to feel around with my hands.

It might be a better use of my time to just pull the manifold and inspect it for cracks.

Of course, it ultimately might be a completely unrelated issue.

 

I pulled the turbo blanket off, and found there was a bit of soot on part of it, so there’s definitely a leak somewhere.

 

The blower trick worked. Turns out I am an idiot and didn’t properly torque the downpipe flange because I didn’t have a tool that would fit. I’m gonna need to go buy a c-shaped box wrench.

 

I might check the housing and surface the flange face a bit, if this has been ongoing for a period of time, rather than just torque it up. We've even had v-bands fail smoke test when the alignment was slightly askew - far from the first to find problems in that area.

 

As soon as I am able to torque the bolts properly, I will check it again.
I previously had an undetected exhaust leak from a crack in the flange, which I got welded shut, but then I didn't reinstall the downpipe properly because I didn't bother to buy the right tool. Sigh.

The nozzle of the shop vac fit perfectly into the hole on my muffler's silencer.
I'm actually surprised how well it worked...

If I fix the exhaust leak and still can't maintain consistent boost, then I will just have the throw up my hands....

 

Come to think of it, a post-turbo exhaust leak shouldn't be effecting boost, so I guess I'll need to check for boost leaks one more time.
I might have to replace one of the elbows, which is at a weird angle because it was repurposed from an FMIC kit.
That, or chop it in half and install a flexible elbow.

 

Exhaust leak is fixed. The lower bolt, the one I originally couldn't properly torque without a half moon wrench, was completely loose. Can't see any other exhaust leaks.
Wonder if it was affecting my AFRs...

I just noticed that the vacuum cap on the nipple that I have been using for boost leak testing was torn. Hmm. That may or may not have been part of the problem. No idea when it tore.

Once again, at least one of the couplers hooked to the badly angled 120-degree elbow wasn't holding pressure perfectly.
Not sure if that counts as a boost leak, unless the leak gets worse when the engine gets hot (or when it moves under full throttle?).

But I think clocking the turbo up actually made the problem worse, so I've ordered some beefy snap ring pliers so I can clock it back down, and then some.

Will probably have to relocate some things so I can get clearance from the frame.

Wonder how much a new section of custom charge pipe would cost...

 

In Japan? They will $zing$ you for sure.
.