Minor boost leak and shrill tone at peak boost: hose clamp, gasket or blow off valve?

Question

Vehicle in question: 2004 Subaru WRX wagon.

This weekend, I replaced the intercooler hose (y-shaped upset the intercooler), the throttle body hose and the blowoff valve hose (routes the over boost back to the intake when the throttle closes suddenly). New hoses are silicone, thicker than stock and have significantly smoother curves. Plus, they're red so that should add 0hp per hose, right?

Install went pretty well: old hoses are tired, have plenty of grease and oil accumulated after 12 years and are ugly. The new hoses are thicker and come with super robust clamps. One of the clamps was a bit too small (blow off valve end of the route-back hose) so I had to reuse one of the old clamps. (Pro tip from future me: think harder about the clamp rotation next time!)

The only problem that I'm left with is a high pitched tone that triggers at nearly peak boost. It sounds much like a balloon with a hole in it and only manifests under load and under a lot of boost (e.g., going up a bridge in high gear at full throttle). It isn't subtle which implies that something is venting to atmosphere.

So, what does this sound like? If it were a bad gasket, I'd expect to hear a rising tone over time. Same if it were a hose clamp or a loose hose. It could be a blow off valve noise but I'm super surprised to hear this much racket: the stock blow off valve has always been very subtle (no big whoosh on drop throttles for instance).

Answer 1

It is quite likely the work performed has introduced a subtle air leak that only manifests itself under high boost.

Pressurizing the intake tract with the car off will help you identify the source of the leak. A hose fitting that isn't honked down all the way will exhibit this kind of behavior under boost.

I like how Jafro figured out the source of his car's leaks in this video.

Answer 2

I Believe It's a Hose Connection Point

Unmetered air whether it's a negative pressure or positive pressure is bad.

That being said, I find it interesting that losing some of the boost isn't detected and triggering a CEL. So that maps to your idea of a smaller leak along with the not subtle high pitched sound.

You said

If it were a bad gasket, I'd expect to hear a rising tone over time.

I completely agree. It wouldn't suddenly release the pressure. It would always be leaking and would therefore rise in tone over time with RPM increase.

Same if it were a hose clamp or a loose hose

I'm not in agreement with this assumption. I can see a clamped hose holding back pressure to a point and then releasing the pressure when the pressure overcomes a weakness in the hose mounting.

Theory

Hose clamps have a way of grabbing a surface in an inconsistent fashion across the entire length of their band. They will tend to pull an area of the hose and bunch it up a bit in other areas. Worm screw type hose clamps are notorious for this.

The hose could be bunched up a bit or pulled together to create a small wave in the rubber to nozzle mating creating a tiny area that allows pressure out under the extreme circumstances as you describe.

Remediation

You could try some of the following if your willing to take the hoses off and put them back on. It was the only real change to the system so logic would dictate that this has something to do with the change.

• Lubricate the hose Lubricate the hose to nozzle mating surface a bit with a light compatible oil that will not degrade the hose integrity. I use spit to get handlebar grips onto handlebars, hoses onto sensors or air boxes, anything that is pushing rubber onto a fitting. It doesn't harm the component and it's a good temporary lubricant. I also never have to worry about if a particular lubricant will degrade my hose or fitting.

• Clean and Double Check Clean all of your surfaces well when you have the hoses off. I'm sure you already did but having a little 300 grit sand paper in your hand or a brillo pad will be handy if you have to knock off any type of buildup or just smooth up a surface a bit. Visual inspection solves most issues IMO.

• Hose Clamps Procure a higher quality hose clamp. I usually use Oetiker type hose clamps. They apply pressure evenly across the entire surface better than the worm screw hose clamp. They tend not to grab the hose in one area and not another area. They accept lubrication, spit, to allow the clamp to more evenly grab across the surface of the hose. They're reliable, and they look quality. The spring action metal hose clamps are also very good at grabbing a hose across its entire surface well but the diameters you are dealing with may preclude them.

Here is an image of oetiker style hose clamps

Answer 3

The culprit was right here:

If you look closely, I made two errors: 1. I put the clamp on at an odd angle. That was dumb because that made it hard to torque down properly. I should have had the main shaft of the clamp pointing up. 2. I didn't torque it down well enough. If you look closely, you can see a hint of a gap between the metal and the hose. Guess what! Air under pressure really wants out!

For those not in the know, that is the hose coming out of the compressor side of the turbo. That's the place where you have your tippy top boost, so proper clamping is important.