High Idle trouble (91 5mt turbo sedan)

[ciper]

I'm having trouble with a hyperactive idle. So far I have -
Cleaned IAC valve and installed with new gasket
Replaced or sealed hard/broken vacuum lines and pcv piping
Replaced leaking coolant line to IAC valve
New PCV valve, spark plugs,
Removed throttle body to verify adjustment of throttle stop screw
Adjusted throttle bypass screw on top of throttle body
Reconnected neutral switch on top of transmission (causes a similar problem on newer cars)
Cleaned MAF sensor
Reset the computer

When starting cold the car the rpm will often shoot up to 2k then eventually settle to 1.5k rpm. Once warmed up the idle will settle at 1.2k rpm. Pumping the brakes many times or quickly revving the engine to 3k will make it drop to 1k and then over ten seconds raise back to 1.2k rpm.
I tried closing the screw on top of the throttle body reducing the idle down to 800 but it starts to shake like a mofo and seems like its missing. No detectable missing or loss of power over 1.2k rpm. There is currently no check engine light.

I'm not sure how to proceed. My only thought is that the intake manifold gaskets are leaking?
Details on the ongoing project http://bbs.legacycentral.org/viewtopic. ... 5&p=378462

[jefferson]

Make sure the diaphragm in your bypass valve is good. It can be leaking and causing issues.

Jeff

[Inlinesicks]

I had a similar issue, the TPS wasn't adjusted properly.

[BSOD2600]

I had a similar issue, the TPS wasn't adjusted properly.

Me too.... well mine just started going bad. Eventually the damn thing would "idle" at 3K. lol everyone at lights thought I wanted to race. Anyways, re-adjusted it per the FSM and it was fine again.

[Inlinesicks]

I had a similar issue, the TPS wasn't adjusted properly.

Me too.... well mine just started going bad. Eventually the damn thing would "idle" at 3K. lol everyone at lights thought I wanted to race. Anyways, re-adjusted it per the FSM and it was fine again.

Yeah when I pulled the engine to change the clutch everything was fine but when I put it back in it would only idle at 3k. Checked for intake and vacuum leaks for days and finally realized the the chain I used to lift the engine knocked the tps way out, was very relieved. It was a bitch to drive in traffic with a stage 2 wrx clutch and idling at 3k lol.

[91Beater]

Mine also idles high, at 1,500RPM. (NA Ej22)
Adjusted TPS, cleaned MAF, chased down vacuum leaks by spraying starting fluid along lines, replaced PVC and no improvement.

[ciper]

Adjusting the TPS is next on my list. Found these scans but is one for turbo and one for NA? I wonder since one of them focuses on connector b56 while the other focuses on b58 with seemingly the same steps.

http://www.main.experiencetherave.com/s ... sting1.jpg
http://www.main.experiencetherave.com/s ... sting2.jpg
http://www.main.experiencetherave.com/s ... sting3.jpg

How can I verify the BPV is ok? Should I just block it off and ignore the compressor surge while testing?

[91Beater]

Adjusting the TPS is next on my list. Found these scans but is one for turbo and one for NA? I wonder since one of them focuses on connector b56 while the other focuses on b58 with seemingly the same steps.

Should be the same. You can only adjust the idle switch position. Use a 0.7mm and 0.9mm mech pencil lead for gauging the bottom TB stopper if you don't have a feeler gauge.


To check intake manifold leak, spray the suspect area with engine starting fluid and monitor RPM fluctuation.

[jefferson]

Just unplug the vacuum line from the bpv and plug it. You won't have to drive the car to see if that is the source of your vacuum leak.

Jeff

[Legacy777]

TPS instructions are the same for turbo & non-turbo.

There is supposedly a typo in the FSM. The instructions on my site are correct.

http://www.surrealmirage.com/subaru/engine.html#tps

[ciper]

I was able to improve the idle. First I adjusted the TPS to be connected at .7 and open at .8 (I didn't have a .9mm feeler).

Secondly I had a throttle body from my other car which had not been molested and I measured the depth of the adjustment screw on top. I used an electronic caliper with a depth gauge on the bottom. The good throttle body was 3.9-4mm so I adjusted the one on the car to match.

I then loosened the IAC screws and reconnected all the PCV hoses. I started the car and waited for the fans to cycle once (indicating normal temperature). I nudged the IAC magnet over until the rpm was a little over 800rpm. I turned everything off and tightened it.

The combination of those three things made a HUGE difference. I drove it for a few hours with many on/off cycles and 80% of the time the idle was where it should be. It still rested at 1100rpm at random times. It misses alot at idle so once I solve that I will probably do some fine tuning to the previous steps but so far I'm happy.

Tomorrow I will replace the coolant sensor, disable the BPV and probably replace the intake manifold gaskets. If the BPV is found to be the problem Ill try to retrofit an 02+ WRX unit in the same location.

[Legacy777]

You can remove the BPV and blow into it. It should not leak any air. If it does, it's bad.

[ciper]

I did not get around to testing the BOV because I FIXED THE PROBLEM!
Not only is my idle in spec but its rock solid smooth. Smoother than any other Subaru I've owned. I have not checked the bypass valve yet but it wasn't the primary leak.


I used no force to remove this gasket. It slid right out like this. The metal looking portion has lost all of its gasket material and "coincidentally" is right next to the power steering pump which often leaks harsh fluids...

BTW it is possible and somewhat easy to replace the gaskets without removing the manifold. I literally only removed the radiator, the ac to manifold bracket (behind alternator, two bolts) and the lower AC mounting triangle (three bolts). Once all eight manifold bolts were removed I was able to gently pry the manifold up enough to slide the gasket right out.

I didn't disconnect a single part of the turbo/intake except one PCV line right up front near the cam gear because it was sort of blocking the front bolt. EASY PEASY

- If anyone else has idle problems and know that the screw has been fucked with use a depth caliper (most have it on the bottom) to set the screw at 3.9mm, adjust the idle position switch, fix any intake leaks then if necessary after the fans turn off once bump the IAC over slightly until it sits at 800 rpm.

[91Beater]

I think there's a proper way to adjust it that we haven't figured out yet. If it is supposed to be 3.9mm every time, they wouldn't make it an adjustable screw me thinks.

[ciper]

I agree. But what if the proper method mounts the throttle body to a flow bench to measure pressure difference? Since these were made in Japan I image the tolerances are pretty damn close so starting at 3.9mm should get most close to where they should be? I get a more precise measurement and post it. I think it was actually 3.94mm

[91Beater]

I agree. But what if the proper method mounts the throttle body to a flow bench to measure pressure difference? Since these were made in Japan I image the tolerances are pretty damn close so starting at 3.9mm should get most close to where they should be? I get a more precise measurement and post it. I think it was actually 3.94mm

I could be totally wrong but its a wild guess.

---
\
---

Unlike engine cylinder, its a dry fitting and the rim of the plate is not supposed to have a friction along the wall. So, if it's too loose, you'll have excess air. If you get it too tight, I'm thinking it may drag along the wall excessively and cause sticking or scoring of the rim or bore.

The purpose of calibration adjustment is to compensate for manufacturing variance. What's a perfect adjustment for one unit isn't the right unit for another batch or even another unit of same batch.

[Legacy777]

I would agree. They are bench flow tested/calibrated, and while they'll likely be very close, they won't all be the same.

[ciper]

Today I took a 400 mile drive. After about 60 miles I noticed my idle was at around 1100. It settled down eventually so I thought nothing of it. 100 miles later it was at 1500 so I looked around under the hood and thought "the idle screw seems higher, I must be crazy." 100 more miles and now it was 2000rpm wtf? Looked under the hood and the screw was MUCH higher?

Come to find out there is an O-Ring around this screw and mine was hard and not holding the screw. I didn't have a problem before because I wasn't in boost for very long but in I-5 it is easy to go over 100mph for extended periods. The boost and vibration was forcefully unscrewing it! It had almost unscrewed to the point that it would have shot out.

Since I was in the middle of the desert I cut a small strip of my Gatorade bottle label and used it like teflon tape. Not only did it work perfect and my idle is back to stable.

As a side benefit I realize this was also a boost/vacuum leak. Previously I hit 63 centimeters of mercury during engine braking and now I hit 65-66

[Legacy777]

You're talking about the screw on top of the TB correct?

http://www.main.experiencetherave.com/subaru/images/tb

[ciper]

Yep, as shown in DCP_1512.JPG of your link. It was automagically loosening itself.

[Legacy777]

Damn automagically loosening!