Almost dying at idle and deceleration

[broox]

I have an auto 95 talon tsi FWD with 167k miles on it. over the last... long time, it's been acting weird.

1. when i'm sitting at lights or at a stop it almost dies, like it just cuts out and studders. this doesn't happen 100% of the time, but definitely a good portion of the time.
2. driving along normally (not boosting - just driving) and i let off the gas. when i let off the gas the rpms should normally just drop down to like 1200 or something. but they drop down to 700-800 and then jump back up to 1200. again, it like stutters when i let off the gas.

i thought it was the sparkplugs, so i went and got some new plugs. but it's still doing the same thing... my old plugs had about 20k miles on them.

Some other things I've done

• Replaced Idle Air Control Valve - 2,400 miles ago
• Water Pump, belts, hoses - 2,400 miles ago
• Plug wires - 15,000 miles ago

Any ideas on what could be going on would be great.... Fuel filter? I imagine it's something fuel related.
Thanks a lot

 

[diablos991]

Are you venting your BOV?

 

[broox]

You must be logged in to view this image or video.

 

[Maverick3169]

Try doing a boost leak check.

 

[92EclipseGS]

I'm getting a similar problem on my 92 n/a. At idle it will randomly drop in rpms and sometimes even stall out but this doesn't happen all the time. Sometimes I can drive it a whole day without problems then randomly it'll start happening again. Not sure if this is related but sometimes if i push the gas very slightly just to cruise the engine stutters pretty bad as if its choking and can't get enough fuel or something.

 

[talontsi9095]

Sounds like a boost leak somewhere. If your driving down the road at a steady speed does it studder then? Even if it doesnt i would still test it for boost leaks.

 

[broox]

i haven't noticed any misfires or whatever happening during constant speed. i'll do a boost check tonight.

 

[majik2k5]

The O-ring around your BISS screw could have gone bad. When they go bad your idle is completely uncontrollable and you're cursed with constant stalling.

 

[toybreaker]

I don't know 2g's that well.

On the 1g, there's a closed throttle idle switch that tells the ecu the throttles closed.When it doesn't work, the cars have some weird idle issues.

On a 2g, I *believe* it's in the tps itself. It would be worth a quick check. If'n the ecu doesn't know you want it to control the idle speed, it probably won't do a very good job.

Also, check the throttle cable adjustment, and throttle stop. Ensure the throttle plate isn't hanging open a schosche..a little gentle cleaning of the throttle body may help as well.

 

[Smevans]

What about your TB? Maybe it needs to be cleaned. Do the boost leak test first and if that doesn't do anything, check the TB. The BISS should be a last resort

 

[majik2k5]

What about your TB? Maybe it needs to be cleaned. Do the boost leak test first and if that doesn't do anything, check the TB. The BISS should be a last resort

On the contrary. The BISS screw should be one of the first things to be checked. It's a piece of cake to check and can be checked in a matter of seconds. When my O-Ring went bad, I put my finger over the BISS hole and my car corrected itself, simple. Always check the small things first, because with DSM's that's normally what comes back to bite you in the ass later on.

 

[DGajre777]

Are you still on the factory fuel pump? It sounds like fuel overrun cause by a big fuel pump and the lack of an AFPR.

 

[broox]

as far as i know it's the original fuel pump. i haven't replaced it since i've owned the car.

i've gone through 2 water pumps though, heh.

 

[broox]

Alright, I've finally gotten around to doing a boost leak test.

I recently replaced my stock turbo with a 14b. So it's got that elbow that wraps around to go down to the SMIC. There is a small tube coming out of the top of the elbow that just goes to absolutely nothing. It's not plugged or anything, haha. Just air going straight out.

I have no idea wtf it is, or if it was ever supposed to be connected... This is probably the culprit. Should I just plug this or hook it to something? I'll go look around for more hoses.

Here's a pic of what's up.
http://derek.broox.com/photos/problems/4473/

 

[toybreaker]

On a 1g 14b, that's the reference line to the wastegate.

It goes to a solenoid (that the computer controls) and then out to the wastegate actuator diaphram on the side of the turbo.

Right now, it's a vacum leak off boost, and a pressure leak under boost.

I sincerely hope there's a line leading over to the wastegate actuator from somewhere!

Otherwise..unregulated boost...and expensive noises.

I'd get on that ASAP!!!

 

[oldman]

That is suppose to go to the inlet of your boost controller/stock bcs. If you're tapping the BOV line for your mbc, switch it back to the fitting in the picture with new hoses then cap off the Tee on the BOV line. You'll also need to finish your boost leak test, sounds like you're just doing a visual inspection.

 

[oldman]

Right now, it's a vacum leak off boost

The vacuum system doesn't start until after the TB plate, otherwise good post.

 

[toybreaker]

I meant "false air". As in air not measured by the airflow meter...which has the same effect on the mixture. Any air not measured and accounted for will make the mix a little leanish, which will give a somewhat errratic idle and etc,,,

I'm sorry...

I will now banish myself to the garage....

 

[broox]

If you're tapping the BOV line for your mbc, switch it back to the fitting in the picture with new hoses then cap off the Tee on the BOV line. You'll also need to finish your boost leak test, sounds like you're just doing a visual inspection.

That is exactly what I'm doing. I've got a T right by my BOV that has a line to my MBC and to my wastegate.

You think I should make the line to the BOV go straight through again. Then run a line from the fitting in the photo to the MBC to the wastgate?

Thanks for the help!

p.s. i capped off that hose in the pic and drove around for about 20 minutes with no stuttering! Hopefully it is cured. I'll drive a lot more tomorrow to figure it out for sure.

 

[oldman]

I meant "false air". As in air not measured by the airflow meter...which has the same effect on the mixture. Any air not measured and accounted for will make the mix a little leanish, which will give a somewhat errratic idle and etc,,,

I'm sorry...

I will now banish myself to the garage....

No problem and yes, it's a source of unmetered air. Banishing yourself to the garage isn't a bad thing, I work on cars on my driveway.

 

[oldman]

You think I should make the line to the BOV go straight through again. Then run a line from the fitting in the photo to the MBC to the wastgate?

That's a definite yes but I'm so sick of explaining why. If you trust me, just do it. If not, do a quick search as this has often been discussed lately, BOV lines is practically the worst place to tap for you mbc.

p.s. i capped off that hose in the pic and drove around for about 20 minutes with no stuttering! Hopefully it is cured. I'll drive a lot more tomorrow to figure it out for sure.

Sounds like you're trying to skip the real boost leak test, don't be lazy. Considering what you have found so far, who knows what else is out there.

Thanks for the help!

You're welcome.

 

[broox]

Cool, I trust your word and will fix that line... it takes 2 minutes so why not?

But what do you mean I'm being lazy with that boost leak test? You can see in the photo I linked that I clearly have an IC pipe tester hooked up. Here it is again: http://derek.broox.com/photos/problems/4473/

Am i missing something?

 

[oldman]

My bad, I missed it, I thought you were just visually inspecting it because you didn't respond to my previous comment.

You'll also need to finish your boost leak test, sounds like you're just doing a visual inspection.