UPDATE: Although I originally thought this could be IACV or ECU reset issues, it seems more consistent with a sticking PCV valve.

Original post below:
I thought I'd post my recent dramas in case it helps anyone. My car is stock and if you have issues with wild rpms while stationary, some of this may help. Also I'd be interested in any expert views of ECU resets because I went into this thinking IACV and came out the other side thinking its ECU gremlins or the PCV (and maybe a bit of all 3).

I'd loaned out my car and it was returned with reports of a CEL which self extinguished after a little while. It was just code 24, based on CEL flashes. That's IACV so I swapped out the valve for a known 'OK' one but the one coming off looked fine. Reset the ECU via the 'proper' under dash plugs method and it cleared the code. The car then drove fine. Put the original IACV back on as it idles at 100 rpm less and I wanted to test it again. Problem solved!

A couple of days and 200 km later (while I wasn't driving again!), it threw another code - but this time the idle was all over the shop. Near stall, 2500 rpm, down to 1000 and back up again, then almost stalling. But fine when driving, this is only on idle. CEL is on. ECU reset does not eliminate the problem this time, but car drives fine apart from being demonically possessed when stationary.

Checked codes with Evoscan and only had stored code 24, stuck IACV indicated. So I swap for a known 'OK' one. No change in behaviour which really freaked me out. Thinking I must have a MAJOR problem I set out to check what I could. I'll spare you the methodical 3 hours of work, suffice to say everything was good and between spraying brake cleaner through the PCV valve, powering down the ECU completely (and removing and replugging the ECU connectors - no evidence of water ingress/condensation as feared), the problem was gone. 850 rpm idle again, smooth as.

I had a suspicion the problem may be the PCV because it seemed to make sense that the PCV could be getting stuck, the IACV works to correct rpm, the PCV unsticks, revs change, IACV tries to alter them, etc, etc. Counting against this, the IACV wasn't that cruddy, I had cleaned it pretty well only a couple of months ago. Last possibility I can think of is that a complete ECU power down is somehow different to the under dash plug ecu reset 'the proper way'. I have seen others suggest that in other forums, and the P1 manual supplement does speak of an IACV reset (separate from an ECU reset) so I guess it's possible that the complete power down I did might have cured some electrical gremlins. Anyway, it seems these symptoms are pretty common, so I add this in the hope it saves someone chasing an IACV. I'll update if the problem returns, but remember the moral of this tale: DON'T LET ANYONE ELSE DRIVE YOUR CAR!

Hi,
Did your problem ever return? I have exactly the same symptoms of hunting idle when warm in my completely stock wrx sti version 5. I have since changed the air filter, cleaned the iacv and throttle body, changed to a new AFM via eBay and new Ryco fuel filter. Bought a second hand good iacv but the problem still exists. I have done an ecu reset via plugging the black and green connectors under the dash. Any ideas as this problem is doing my head in.Should I look at changing the pcv?
Appreciate any advice.
Cheers

Try checking your position switch mate, and your idle control valve can start going over time so maybe you might be up for a new one. And get a new idle control valve gasket as well as over time they tend to expand.
Hope it helps buddy


using the Perth-WRX mobile app

You may have missed it but he did say he had bought a second hand IACV which was working, and the problem still exists.

__________________
A man in nature;

"All in relation to nothing. Nothing in relation to infinity. A central point to both, And infinitely far from understanding either."

He needs to reset his throttle position switch so that with the throttle closed it is sending <.5V to the ecu, over time they tend to wander due to wear.

It's hunting because the ecu can't see that the throttle is in the idle position (the TPS is sending over .5Volts to the ecu) & usually, above idle the IAC is closed to prevent over-run backfiring. The ensuing situation confuses the hell out of the ECU - hence the IAC CEL. The ECU thinks the IAC is sticking, when it actually isn't.

__________________
[CENTER][SIGPIC][/SIGPIC][/CENTER]
[CENTER][COLOR=#ff0000]Built Ej207: Semi-closed RA Cases - Wiseco Forgies - Manley rods - BC 272 cams - GDA Heads - billet f/whl - Vi-pec - WbO2 - 800cc Yellow tops - TD05-20G - 409Hp ath @ 1.75bar on E85[/COLOR][/CENTER]

You are just pure GC8 knowledge Rossco!!

__________________
Now handles like a cat on carpet

Sorry to hear about your dramas bteoh - yep, whatever I did the problems have not been back. It's been running fine (by my inexperienced standards) since, but I wouldn't say that it was definitely because of anything I did.
Rossco is God so far as I'm concerned and so if he says its the IACV being driven by the ECU getting false readings from the TPS, then I reckon it was. I will get my Digital Volt Meter out tomorrow and check the idle voltage from the Throttle Position Sensor and report back. I didn't do anything to the TPS other than disconnect the wiring harness and spray the plug with electrical cleaner. That would also fit with the ECU not getting a good read of the throttle position...
Any views on the 'wear' - is it worth putting a new TPS (because the wear is something that results in variation in 'closed throttle' signal every time the throttle plate is moved), or is it just a gradual drift in the offset and resetting the idle value to under 0.5V all that's needed? My idle does change between 850 and 950 rpm, but it's constant during the stationary stint at whichever value it picks. At the next period at idle it might be slightly different but it doesn't fluctuate DURING a stationary period any more. I put the slight variation down to changes in thermal loads in the electrics and mechanicals and worn everything, but again, could well be TPS related. There's the odd lump and shudder every now and again, but it's very occasional and I put that down to any/all of the boxer, the heads, and an engine that could have done 270k. If I knew more of the engine's history I would take it to a workshop to get it hooked up to an SSM but I figure putting in a new short block would be a better investment than spending money asking a workshop to troubleshoot and tell me it's an old engine/dodgy heads!

Hah yeh my bad I didn't read it properly


using the Perth-WRX mobile app

Hi Rossco and everyone that chimed in,
Thanks for the heads up. Sorry for being inexperienced but how should I reset the TPS and how do I measure the volts to determine if indeed 0.5 volts is being sent to the ecu at idle?
Any advice is much appreciated.
Cheers

So I’ve checked the TPS voltage. Have a look at Alternate’s great post here to see how to do it: http://www.perth-wrx.com/vb/mechanic...on-sensor.html. It’s not clear from the P1 Supplement instructions I was looking at, but I did this without the engine running (ignition on obviously), but he indicates the manual says it should be off.

With engine running, the voltage is higher and quite erratic too, which I find weird but must be linked to some duty cycle I guess?
Engine off, I had 0.52 V throttle closed, 4.12 throttle wide open. P1 Specs say 0.5V is standard for closed, with 0.2 V to 0.8 V acceptable. Wide open should be 4.0 V to 4.6 V. The return to closed gave consistently 0.52 V although I could get 0.49 to 0.54 on occasions, blipping the throttle plate open then closed by hand. Being an inquisitive person and figuring I could always put it back, I thought I’d try sliding the TPS with the engine running. It clearly affects the idle speed, but also seems to bring in the IACV valve into play and at one point I managed to bring on the CEL light and the RPM sat at 1,000. Presumably this is the ECU going into limp mode and playing it safe. I’m afraid I didn’t bother checking the code as it cleared when I moved the TPS back to mid range. I ‘judged’ what was the smoothest idle at the lowest rpm (couldn’t get under 850 rpm), switched off the engine and checked the voltage: bang on 0.5V! (It had drifted back to 0.52 after a test drive).

My take-home from this is my TPS is fine, though seeing how sensitive the whole idle system is to it, definitely something to consider as a failure point. While I was fiddling with the engine running, the idle was only changing when I moved the TPS, or very slightly just afterwards. The symptoms I had at the start of this thread seemed like at least two systems working at cross purposes with a bit of a lag. Moving the TPS here was just like opening and closing an old fashioned idle screw. I guess it would depend on exactly how the resistor fails.
Anyway, after fiddling, it drives just like before. Maybe very slightly smoother at idle, but probably just in my head.

What sort of rpm variation are you seeing bteoh, and over what sort of timescale?

PS
Don’t forget to tighten the TPS locking screws again if you do loosen them. And you’ll need a good quality long shaft Phillips screwdriver and some care or those screw heads are going to strip. The lower screw is hard to get to – I guided in the screwdriver head using an arm under the intercooler, then once seated in the screw head, took my time to get a good grip on the handle to break the screw 'seat'. “Do, or do not. There is no try!” Because once you mash the screw head it’s going to be a pig to get to it properly because space is tight. It seems the TPS pivots on the lower screw, both screws need to be just loose, not lost! Of course, if you enjoy taking the intercooler off, go for your life (and please pm me so I can get you to do mine next time)!

It never occurred to me before that you might want to get the TPS off to clean that area of the throttle body. I reckon it must seat directly onto the throttle plate spindle so I can't see it holding any crud there that could affect anything - unlike the IACV and associated air passages. But does anyone know about that?

As I said, Alternate has a great post on here about how to actually check the voltage. A dress-makers pin through the harness wire (white and red) worked a treat with alligator clips on the DVM wires. I disconnected the TPS plug and harness to get a better angle to get the pin in, then reconnected the plug. I left the pin in for test drives but it had an insulated head and the shaft was nowhere near any metal, be careful with your wiring folks! I earthed the DVM to the dimple on the top of the throttle housing – I’ve checked before that is a good earth with negligible resistance between it and the negative battery post. Once you're done, remove the pin!

Attached Thumbnails