My 1990 S4 auto has developed a strange idle speed problem and I'm hoping that someone here might have experience of the fault and can point me in the right direction as I've run out of ideas.

On a cold start everything seems fine but when the engine gets to a certain temperature (about 65-70c on the water bridge) the idle raises as if I just pressed the throttle. With the air box off I hear a sudden increase of air intake and the idle jumps from circa 670 to circa 900rpm and stays there. If I restart the engine it is normal for about 3-4 seconds and jumps up again, what is this telling me?
If I block the air hose going to the side of the intake via the venturi the idle doesn't jump, idle is a bit lower than it should be and is lumpy probably from the lack of air. Unblock it, same issue which is very repeatable.

I've been through all the tests I can in the WSM measuring temp sensors, throttle switch and so on - both cold and hot. The MAF is good, I've tried a different LH, I've even disconnected stuff like previously measured temp sensors. Because it sounds idle valve related I've replaced it twice, first with a Lowe unit and then Bosch (they both measure correct resistance compared to a known good one) I've tried two different ventruri pipes. The O2 sensor outputs the correct voltage and makes no difference if it's connected or not. I do not have a vacuum leak (I measured it but I'd also know because the shifts would be harsh), nor any air leaks and in any case I would expect an air leak to cause an immediate problem, not something that changes the idle speed after a few seconds on a warm engine.

So why does blocking the hose to the intake stop the idle jump? It's all measured air even though it bypasses the throttle plate after all. The only other air flow controller is the idle valve, could this be receiving an adjusted signal from the ECU in relation to some other sensor that's telling it more air is required? Then of course more air flows through the MAF and fuel is added. The other problem the higher than normal idle creates is I don't get any shutdown when coasting that I used to, when you can hear the characteristic fuel cut sound from the engine. That sounds throttle switch related but it tests fine on the ECU pins hot and cold.

I've owned various 928s since 2006 and done most of my own work, I'm a seasoned mechanic having replaced torque tube and transmission bell housing bearings, intake refreshes and timing belts etc. and learnt a lot about them but this one has me stumped.

Any valid input from experts welcome as I feel this has now got into expert territory.

Thanks

 

What you describe suggests the "warm up logic" is being reversed- the only way I know of that could do such would be to reprogramme the warm up idle speed map values using my ST2 kit . During a cold start the initial idle should be around 1000 rpm, quickly dropping over several steps until the regular idle of 675 stabilises when the motor temp reaches about 82C and that typically takes about 3 minutes.

The only thing I can think of at the moment is whether the temp2 sensor is functioning correctly- I suppose it could be off or misbehaving somehow. This item has two channels & three pins- a common earth and the channels that feed the LH and the EZ units- both channels should show the same resistance to earth for a given temperature. I can dig out the values if you do not have them - they are commonly published in other threads as well.

It will interesting to see what the "brains trust" makes of your issue.

 

Duplicate in error [somehow]

 

I have had your exact symptom after changing the TPS. It turned out to be a defective one but I don't know for sure. I had done an intake refresh about a year ago but had not changed the TPS and all had been fine until I changed it. I had removed the intake 4 times to try and find the problem since the TPS appeared to work, passing all tests even with Theo's tool. I had also put in a rebuilt MAF at the same time as the TPS so I had that sent to Louis Ott for testing and came back at less than 1% tolerance so I had my LH refurbished just incase my idle circuit was getting flaky but hat didn't work either. I put the old TPS back in on the 4th removal of the intake and it works flawlessly. Matt968 contacted me with the same problem but it turned out to be a leaky venturi tube for him. I would look at the TPS and venturi tube first and if that doesn't work out look for intake leaks elsewhere. Oh I also changed the ISV twice first with Bosch and with Lowe which is still in there. Good Luck because it was very irritating.

 

I'm with you on the irritation front, believe me. I think I've had my intake off at least 6 times looking for air leaks and so on. I'm not convinced TPS but I'll order one in and replace it as I have no other clue to go on.

FredR, I thought 'reversed' too but it's not a gradual climb as resistance on the temp sensor changes as one would expect. I've tested the temp sensor on both sides and also left it disconnected completely and it made no difference to the idle jump I'm seeing.

Thanks

 

Perhaps one thing you can try for purposes of elimination is to disconnect the tank vent line/breather connection and blank that off. If that system is failing open for whatever reason then you may be getting false air that way assuming the hose is not split or something daft like that.

 

You can test your TPS at your LH plug-in at the computer, you'll have to lookup the pin number.

 

Yep, checked that, all good.
I know, been there done that as I said in my opening post but has been suggested that even testing correctly can still not work properly.

I am having trouble sourcing a TPS in the UK, seems it's now a Porsche only supplied part so yet to find out stock and price situation but also emailed Roger as another option.

 

IMHO I doubt the TPS could cause the issue you describe. If the closed contact is not functioning when cold the engine will not support an idle and one has to work the throttle to keep the engine alive. Once the motor is fully warmed, the ISV will not function if the contact in the TPS is not functioning but it will support a stable idle as the power off position of the ISV allows an air flow to pass roughly consistent with that needed to support a stable idle with the a/c switched off. I have run my motor with the TPS disconnected and used this to test whether the idle was more or less where it should be or whether the engine raced a bit to determine whether I had false air- that and measuring the AFR with my WBO2. The idle was stable and the AFR was correct thus no false air.

Perhaps you can try this test if you have not done so already.

 

Mike I'm seeing the Lucas SEB183 listed on eBay UK as equivalent to Bosch 0280120322, Porsche 92860615700.

https://www.ebay.co.uk/itm/New-Genui...cAAOSwzC9bSg2W

 

Interesting. I can't check AFR as I have a standard O2. Although the warm up cycle seems normal right now it was only a short while ago that it was the other way round and it wouldn't stay running without slight throttle input from cold.

I think I saw that one and even though it has the right numbers doesn't say it fits my car (not that those lists can always be trusted). I'm getting mine from Roger now.

 

As "Taguid" mentioned I have the exact same issue. Like Taquid mine started after replacing the TPS and idle speed valve. Mine is an 89 S4 automatic with 57K miles and has had a full intake refresh within the past two years.

I thought I resolved it with the venturi but it did not fix the problem. Same symptoms as you. Starts fine cold or hot. Warm up idle is correct...starts around 800 and will slowly drop in a few minutes to normal. Drives fine at all rpm's. But after it warms up sometimes it idles at 675 and sometimes it's around 800 or so (this is with the car in Neutral or Park).

I smoke tested it and no leaks. I replaced the vacuum valve that vents the carbon canister. Replaced the venturi pipe. Replaced the O2 sensor. I also replaced the cable that goes to the throttle plate and set all the cables to the correct tension based on the Greg Brown procedure thinking it might be a mechanical issue and not an electronic one. I even pulled the A/C control unit and disconnected both multipin connections thinking the A/C system was calling for an increase. Nothing has made a difference.

I took it to the my local shop and they connected their scan tool (don't know which brand) and it did not show any codes and the TPS was working fine.

It's almost as if it is sometimes it is coming out of closed-loop and going to an open-loop default idle setting. Know idea if it is or isn't but that's what it is like. It's as if the LH is telling it to hold one idle or the other for some reason.

Currently my LH is at Louis Ott's waiting for return and my MAF was calibrated within the past 3 years.
If the LH does not fix it I will probably put another TPS on just to see if I get lucky.

Should have my LH back Saturday or Monday and can follow up if it made a difference.

 

I read your other post about the venturi and that's when I tried blocking mine and finding it stopped the idle increase but I don't know why it would do that.
When you start it warm does it idle at 675 and jump up after a few seconds like mine or sometimes start normal and others go straight to 800? I don't understand why the sudden jump but stable jump up.
My LH is a JDS 'upgraded version' but I tried a another good one which made no difference.
Good luck with it and update this thread if you find something please.

 

When the problem first developed, I would start it and when it got to operating temperature (meaning I would start it and just let it sit and warm up) the idle would just change from 675 to 800....something was switching or causing it to make the nice transition. However, it no longer does that. But I have also removed the intake manifold several times by putting my old TPS in and then removing it again to put the new TPS back in plus other tests - so there is no telling why it does not do that anymore. I'm assuming it's because after I do any work on it now I try to drive it for 20-30 miles to reset the LH (I always disconnect the negative terminal before working on it).

One other item I can add, which makes no sense, I drove the car about 30 mins to my local shop. It was idling at 800 when I arrived at the shop. I left it running and they hooked up the scan tool. After about 2mins with the car still idling and the tool connected the idle would drop back down to normal. Then after a minute or so it would go back up to 800 or so. It would cycle back and forth like this with the tool attached. After that I drove it home, parked it and pulled the LH and sent it off to Ott's.

 

maybe the TPS you have is adjusted just close enough to click open once the throttle body gets hot.
Next time try adjusting it so the throttle plate has to move a bit farther from fully closed to cause the TPS to click open.

Also make sure before you remove the intake that you have enough slack in the throttle body cable( the one that goes around the wheel)