how to find intake leak

[ivymike]

I seem to have an intake leak upstream of the throttle, and I'm looking for suggestions at to how I might find it. The vehicle is a 2005 scion xb.

I first suspected that I might have a leak when I started hearing a "squealing" noise that was throttle dependent - no noise at idle, and no noise at ~10% load, but at very slight openings there is a whistle/squeal which has only recently become noticeable (but is annoying). I did a pretty quick visual inspection of the intake system and didn't find anything obvious... I decided that if there was a leak it must be a small one, so no big deal (except that damn squealing).

I've just gotten back the results of an oil analysis, and (surprise, suprise) it shows high silica and various wear-related metals. Seems like my engine is getting dusted. Now I need to find that leak.

When I'm under the hood with the engine running, the leak is not loud enough for me to get a bead on it even if I monkey around with the throttle. It doesn't help any that I've got some halfway-bad alternator bearings joining the chorus (I'll replace that when it stops charging).

 

[gt6racer2]

First, if it's upstream it really doesn't matter unless it's after the flow sensor.
You might even get better gas mileage ! ( OK - it's a joke - just thought I'd start that whole discussion again... )
The classic check is to have the engine running and spray a little carb cleaner around suspect areas. The rpm will go up if you have a leak.
Do it at your own risk - carb cleaner is very flammable.

 

[ivymike]

I believe it's somewhere between the filter and the throttle, hence the silica, the wear metals, and the need to fix it. If it was downstream of the throttle, I would have guessed that I'd have noticed a higher idle, or that some emissions check or other would have caught it.

I thought about the ol' spray cleaner trick... while that approach seems to work fine with relatively small areas (gasket interfaces, for example), in this case I'm talking about a few feet of convoluted plastic crap, and I'm a bit worried about both fire and inhalation.

Any other suggestions?

 

[patprimmer]

Pressurise the tube with the throttle closed and feel for jets of air escaping.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[ivymike]

I was thinking about something along those lines... any suggestion for how I might seal between my air line and the filter box (or duct connection if disconnecte from filter box)? maybe a towel with a latex glove over it and my air nozzle wedged in alongside?

 

[patprimmer]

A towel would be my first shot. You do not need a perfect seal, just enough to build up some pressure inside. It might take 2 people, one to hold the towel and airline and one to check

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[patdaly]

Use a hand held propane torch, valve open, not lit obviously. With the engine running, slowly pass the torch over any suspected areas. Usually, the additional fuel will change the idle RPMs enough to find the source area.

 

[ivymike]

can try that too.. I assume "MAP" gas will work just as well (I don't have a pure propane torch).

 

[franzh]

Most full service shops use something called a Smoke Machine. It pressurizes the area to a couple inches Water Column and fills it with smoke. Most of them use hot naptha or kerosene for smoke, but I used to have a man working with me who would smoke all day, every day, and he would just blow into a vacuum hose connected to the suspect area and look for smoke trails.

So, find a smoker and use their talent, or find a smoke machine.

Franz

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[carnage1]

how about skip the test and go right to sealing? seven or eight rolls of duct tape ought to do it

 

[swall]

Instead of an airline, you could use the exhaust end of your shop vac. Stick the shop vac hose over the exhaust and stick the other end into your intake tube. Seal with duct tape.

 

[evelrod]

Isaac, if you have a stethoscope handy, or just a small diameter plastic tubing such as a quarter inch fuel of vacuum line, stick it in yer ear and listen for your whistle. If you still cannot find it, do all the hard and often messy stuff the other guys suggested, any and all of which may work...Then again, they may not.

Rod

 

[izzmus]

The propane torch method works well but unless it's a major leak, the computer will react fast enough that you won't be able to detect a change in idle, so you'd need to monitor fuel trims on a scantool.

 

[ivymike]

I didn't find anything w/ the torch method. I haven't tried the other methods, because the ducting is shorter than I implied above (my recollection was a bit pessimistic). A thorough visual inspection didn't turn up anything, but I can still hear the obnoxious noise when I'm driving. Can't seem to find it (or hear it at all) when the car is in the garage. I'm wondering whether there is some reason that the serpentine belt load would increase between say 5% and 15% throttle and drop off after that - the noise could certainly be incipient belt slippage. The sound is loud enough relative to other underhood sounds (at least at part load) that I'm sure it's not simply being drowned out when load goes up. It's not a normal "squealing" like a constantly-slipping belt, it's more of a continuous twitter-chirp.

No good alterate explanation for the dusted oil... will have to have another analysis done in a few k-miles to see if the problem persists.

 

[izzmus]

Scion... does it have the 1ZZ engine, with the tubular header intake manifold?

There is a resonance in the manifold at idle or just off-idle that sounds remarkably like a vacuum leak. They all do it.

A "false air" situation (upstream of throttle) never makes a whistling noise.

 

[ivymike]

it's the 4-cyl engine in the Scion Xb. I'm not enough of a Toyota fan (or car fan at all, really) to have any familiarity with brand-specific engine designations. The thing I'm unhappy with in the resonance scenario is that I only relatively recently started noticing it, and I've had the car since '04. My hearing has certainly gotten worse in that time.

 

[patprimmer]

I don't know the design of your engine, but some designed use an engine oil pressure activated hydraulic ram to adjust belt tension.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[catserveng]

From the industrial engine world, we will take an old air filter and seal it with duct tape, install it, pressurize the intake with air and use Snoop or soapy water to find the leaks. Never tried it on a small car engine but I would think it would work as well as it does out in the patch on the jack engines.

Hope that helps.

 

[mcgyvr]

If its just a hole in the intake tube then just take it off the car, seal the ends with tape/saran wrap or whatever, drop it into a bathtub filled with water and squeeze the tubing slightly and look for the bubbles..

How many engineers does it take to find a pinhole in 2 feet of tubing.....

 

[evelrod]

Belt slippage is out. Not the right sound. Beside that, you could certainly see evidence of belt slippage by now. Why worry yourself about a leaking piece of tubing, just replace it and be done with it!

Oh yeah, that may not fix the problem??? Love engines, just love 'em. ;-)

Rod