[solved] Very high idle

[BSOD2600]

edit: solved. re-adjusted idle switch that was out of FSM spec.


This past week, after getting done playing around with spark plugs/gapping, now my idle is way out of spec. Just to be sure that the ECU hadn't learned something wrong with all the changes I reset it (and verified it as reset by the fuel trims being back to 0%). Zero CEL's are showing either.

Post-ECU reset, idle is still high. In the driveway, it never gets lower than 2000 rpm. After driving around town for a bit, if I come to a stop light, the RPMs will sit between 3000-4000 rpm -- not good! I sprayed electronics cleaner into the IAC tube, let it sit for a bit and then tried things again -- no change. I then marked and took the top of the IAC off (damn that back screw is a pain to get to!), and felt that the shaft had zero friction while turning.

I then connected the laptop to the ECU to see that would shed any more light on what is going on. While idling in the drivweay at 2K rpm, the IAC duty cycle is around 30%. Comparing this to past months data logs, I see this should be around 15%. After going for a drive and then the RPMs are stuck in the 3-4K range, the IAC is around 55%, which seems way too high to me.

What's next to check/do? Buying a new IAC (22650AA033) for $250 is something I'm dreading, as the TD05 in that area is WAY too damn crowded... will have to take everything off just to get at it.

[Legacy777]

The duty percent on the IAC valve should be between 30% - 45%.

If you look at my post here, mine was around 40% and was idling normally.
http://bbs.legacycentral.org/viewtopic.php?f=19&t=24376

Does your throttle body have a screw on the top of it?
http://www.main.experiencetherave.com/subaru/images/TB/

If it does, screw it down. If not, I'd suspect a vacuum leak

[Apex3]

Well typically the first thing you should look at when something messes up after doing something to the engine, is the thing that you changed, it's likely related to the plugs, or possibly something you bumped while you were in there, maybe knocked off a vacuum line.

Personally I wouldn't mess with the idle control screw unless you absolutely have to.

[Legacy777]

I was referring to the needle valve screw on the top of the throttle body, not the stop screws.

[BSOD2600]

I think I fixed the problem... will report back in a few days.

I started to look through the FSM in section 2-7c for the various diagnostic procedures for the components. I found the IAC (aka air control valve) portion, printed it off and dig into the engine bay doing all the stuff from my OP. I then too realized that the FSM states the IAC should be in the 30-45% range while idling... but as mine never has, even when the car was working 'properly' I think it's ok.

After getting everything put back together, I was looking at various parameters with RevScan. I noticed that the throttle voltage was umm 4.5V and position was 0.4%. That made me think about looking at the Neutral, idle and throttle closed switches -- I found that the idle switch wasn't set! I started to re-adjust the IAC position to see if I could manually bring down the idle and happened to bump/touch the throttle sensor wires/area and all of a sudden the rpms went down from 2K to ~1K AND the idle switch was now set. ah-ha, so something is wrong with the throttle sensor...

Printed off more diagnostic instructions this time for the throttle sensor and then idle switch. Throttle sensor voltages and resitances on the various pins were a bit lower than the FSM ranges (forget numbers offhand). Next, to test the idle switch, I made sure that:
1) No 1 and No 2 pins were conducting when throttle closed -- check
2) No 1 and No 2 pins were conducting when 0.7mm thickness gauge inserted between stopper and stopper screw -- no!
2) No 1 and No 2 pins were NOT conducting when 0.9mm thickness -- check.

I then proceeded to take apart all of the PCV / intake stuff so I could get access to the throttle sensor screws. Spent the next 5 minutes playing around with its position with the continuity tester and the different sized thickness gauges. Re-assembled everything and fired it up. Idle was initially around 2K due the car being off for the past few hours. But fairly quickly (as the engine temp was still ~60C), I could see the rpms slowly drop down. The IAC duty cycle was also slowly going down from the ~30% it started at. Took it for a spin around the block to fully warm everything up. Several times I checked how things responded when getting off the throttle and the rpms stayed well below 2K. Pulled into the driveway and it continues to idle at 1200 rpm. I figure this a combo of me:
A) re-setting the ECU so things are still re-learning
B) my adjustment of the IAC.

If the idle continues to behave 'normally' the next few days, I'll re-set the IAC to its original position which I marked.

[Legacy777]

Sounds like you got it. Yeah....if that idle switch isn't set just right....it can give you problems.

[keelhaul]

hello. sorry for capping this thread and not making a new one

the last days now, when my engine is at operating temp, my idle wont go down to 6-800 like it used to.
its now on 11-1300, even after much driving.

any ideas on why this can be?

[BSOD2600]

Well declaring this problem of mine fixed, thanks to re-adjusting the idle switch. This leads me to a few other things though...

1) Idle is now a solid 1200 rpm when fully warmed up. Not really sure I want to adjust that lower stop screw since the FSM says not to.... doesn't really seem like the ECU is doing much else to adjust the idle lower. No, don't have a screw in the top of the throttle body. No, I'm not talking about the visible 'idle screw' on side either. There is another set screw below which the umm throttle plates levers touch.

2) IAC duty cycle is pretty much always 15.6% while idling. I tried moving it around again, but it only increased the idle to 1500 rpm and the duty cycle didn't change. During startup or driving, it ranges from 30-50% so at least it appears to be working -- idle doesn't really fluctuate once warmed up. Looking at my data logs from last year and with the 94 TW (same engine though), the IAC was ~20% while idle. Guess that's the norm for this engine.

3) Since the idle switch was the issue, why didn't the ECU throw the code?

4) Why did the idle switch position move? The screw for the senor itself was damn tight, so something internal/electrical must've changed. Sign the throttle sensor is on its way out?

5) Octane factor (OF) is still at 75%!! While I haven't done any highway onramp romps... I still drive spiritedly in town occasionally. At least from past experience, the OF would've reduced itself by now. I'll be keeping my fingers crossed. I wonder if the ECU knock sensor learning does some sort of baseline after a reset.

[ericem]

I am pretty sure the tps plays a role in the amount of fuel dumped. If I recall correctly, when you hit the gas initially the MAF doesn't react right away and the engine runs lean, but when the ECU sees the TPS signal it reacts to that initially when you hit the gas so it does not run lean. Maybe it is just the TPS causing the car to run lean when you gun it therefore it knocks initially. I think my car had a similar issue, because with this other engine I don't get any knock. Both engines seemed fine though!

Guess the TPS is more important than you would think. Still the IAC is running at 15% which is odd :S What is your CTS showing in revscan warmed up?

[BSOD2600]

CTS? Coolant Temperature? The typical 82-94C range. Reflects what the dash coolant gauge shows, thus highly doubt the coolant temp sensor is bad.

[ericem]

Ok then. I guess replace the top part of the IAC. So you don't need to replace the whole unit.

[Legacy777]

I highly doubt it's the IAC. It's reacting to the high idle and trying to bring it down, but the 1,500 rpm is as low as it can go. It's at the max of it's control. You've got some other issue there. There is either a vacuum leak somewhere, or the throttle body is off. To be quite honest, I'd suggest trying to find another used throttle body that has not been played with. Check the TPS for the idle switch adjustment and that the TPS resistance changes smoothly. If that's all good, swap the throttle body and see if that changes things.

One thing you may want to check before you swap the throttle body is to verify that the bypass valve is indeed closed at idle. If it were open a little, it could be allowing air to bypass the throttle body and IAC valve.

[ericem]

Technically at idle the BPV is always open or whenever it is not in boost. I think it is is closed it might do this actually. You sure you didn't accidentally unplug the vacuum line on the bpv?

[Legacy777]

You're right eric....had a brain fart when I posted that.

[BSOD2600]

Update:

Ever since I re-adjusted the idle switch position, the RPMs never went below ~1300. I finally got tired of it and got a throttle body from the junk yard. It's somewhat out of spec too for things and this is the 2nd one I've tried. bah! I was noticing that its stopper screw (the thing that Portion G rests on) was considerably more screwed in that mine. Since I haven't anything to loose with the idle getting messed up more, I went ahead and screwed mine in more... and noticed how loose it is already. I wonder if it just vibrated itself loose over a while and then suddenly it was too much.

Anyways, after screwing it in more, now when warmed up it idles in the 850-900 rpm range. finally! IAC duty cycle still sings down to the min of 15.6% though, so something else is up.