Pagnobito

Thanks Greg, I'm going to attempt to leak test the entire intake today while it's off the car although when it was running I sprayed brake cleaner all around it and it was good. I blocked the vapour line, it made no difference to the stepped idle increase.
When I put it back on the car I'll set up the test you suggest.
I took these screen shots yesterday.
I should have written something down as I'm not absolutely certain but I think the difference between the two photos from my phone is software disabling the idle control to bring the idle down and this then shows a +3.9% O2 adj. The time between photos didn't allow me to change anything on the engine like blocking hoses.

FredR

Your ST2 photos clearly suggest you have a vacuum leak as does your previous info around the venturi. What is leaking and where remains to be seen.

If Greg says the minimum value for the ISV is 20% and the max value is 80% then common sense should tell us that the normal position for the ISV would be somewhere around the 50% mark and once the ISV has hit its lowest value clearly it cannot nip up any further and the rpm's increase. As I recall the ISV value is usually in the range 40 to 60%. You say that you have blocked off the line from the tank vent but remember that line disappears under the manifold after joining with the vent from the side of the filler neck. If there is a split in that section of hose you will get false air.

I take it you have no wideband installed and thus why it shows the 10.7 AFR [the value shown before it gets input] .

Pagnobito

Ignore the AFR.

I had the entire intake on the bench. sealed every hose and opening and it doesn't leak, I was pretty sure it didn't as I couldn't find anywhere that it could. Everything is plumbed correctly. As are all the other hoses or entry points on the engine, even the o-ring on the dip stick is a new one.
So I looked at the throttle plate and although I can't flow test it, the plate seems to be as closed as possible and the paint on the stop still in place. Even so I backed off the stop and the plate doesn't close any further. I put it back to where it was.
I'm confident there is no unmeasured air entering the intake.
I've fitted the intake back on the car double checking everything as I went and it's just the same as before.

I had to get fuel because the range light was on so I took it for a longer run as I've only run it in N a few times which isn't too good for it. From cold idle seems about right and in D it also seemed right even after warming up, knock it into N and the idle goes up and sounds too high, back in D seems OK so I'm not sure it's actually a problem in normal driving. One difference I noticed on this drive, previously it wasn't cutting fuel when coasting, didn't get the engine note I was used to but now it does cut fuel and instant drops to 0.0. Performance is very good, pulls well and smoothly, no issue there.

I'll get to monitoring the O2 voltage another day so I can check it from cold, didn't today as I was concerned about the low fuel and continuing running the engine with no load. When I do that I'll also take it for a drive with logging enabled.

In the meantime a video from the monitoring screens, hot engine after the drive, in N and showing high idle, disabling idle control (adjusting ISV percentage doesn't change actual reading, nor does turning off idle control) putting in D, back to N and D again. Should be able to see what's going on. Even when revs increase or decrease the ISV position doesn't change. As I say I'll get data from the O2 tomorrow if I get a chance.

GregBBRD

Wow. Strange problem.

If that 11% idle stabilizer value is with the idle stabilizer running in completely "normal" mode...and not "fussed around with" by Sharktuner, you are getting "extra" air from somewhere.....

With the butterfly closed as far as you say it is and the idle stabilizer at 11%, you should have about a 400 rpm idle....tops.

No "extra hole" drilled into the butterfly, right?

There's a nipple with a hose coming off the right side of the oil filler neck. That hose goes directly to the intake system. The filler neck should have a restrictor made into that nipple (pull the hose off and look inside.) No one drilled that one out, right?

jcorenman

The ISV% on Sharktuner should be around 50-55% with engine warmed up, at normal idle, and higher when cold. The active range is about 35-75%. Yours is stuck at 11%, which means LH has reduced the ISV as far as it can, and the engine is still idling too high. So exactly what Greg said: it is getting extra air, from somewhere. The O2-adjust (right-hand side of the fuel-monitor screen) is also bouncing around zero (+/- a few percent), which means that they LH is happy with the mixture (AFR) being monitored by the factory O2 sensor.

Which this means is that the extra airflow is likely metered air, i.e. flowing through the MAF and not from some other air leak e.g. a hose under the intake, tank vent canister, brake venturi, etc etc).

So either the ISV is stuck open (unlikely since you indicated that it has been replaced, solving the cold-idle problem), or someone drilled an extra hole in the throttle-plate, or someone with small hands and no drill has messed with the mechanical idle-stop.

This last possibility seems like the most likely: With a history of stalling when cold (from a stuck-closed ISV), an "old school" mechanic might try opening the mechanical idle screw. That would help that problem, but the idle is still uncontrolled (stuck ISV) and would be too high when warm. Better than stalling, but...

So do the checks that Greg outlined, then pull the MAF and find the idle screw-- it is on the right side (car's right side) of the throttle assembly, near where the throttle cable attaches. It is a slotted screw-head with a locknut, and it is possible to adjust without pulling the intake (but not with the MAF in place I think). Back it off half a turn and then check what Sharktuner shows for ISV%, you are looking for 50-55% ISV% on Sharktuner. It may take 3-4 iterations to get it right, keep track of how many quarter-turns you have backed off. (You may also be able to use a feeler gauge on the throttle plate (bottom edge to throttle body) but I don't know what that clearance should be).

Edit: When idle-control is disabled (on fuel-param page) then that set value is used instead of the displayed ISV%. In the video, you increased it to 40% but that is still below normal opening-- try increasing it to 70% with idle-control disabled, that should increase the RPM's significantly (1200-1500). If not, then the ISV, wiring or LH is suspect.

Cheers, Jim

Pagnobito

I need to check but it's a new filler as the original one was cracked, didn't get time today but can spend all of Friday on it. Definitely no holes in the butterfly.

Thanks for your input Jim. I understand what you are saying. It has to be metered air as the MAF signal increases with RPM and duty cycle on the injectors and LH is happy with O2 so I don't think I need to test the latter. Several LH computers do the same thing.

I'll address the butterfly first. The ISV was running at 20% before I lifted the intake to check for leaks and butterfly position. I know there's no figures on adjusting the butterfly, I think it was flow tested and set at factory? I backed off the stop screw, the butterfly didn't move much and doesn't completely reveal the hole at the top however it won't close any further, I set the screw just touching so now the butterfly in theory is going to let less or no air through. I put it all back on and we see 11%. How can that be? I could have changed something during removal and fit other than the butterfly but I'm not sure what. If I remove the intake again I'll get some pictures of the butterfly, maybe it's just a bad fit or something.
I think I increased the ISV percentage as you say in idle control disabled mode and it did indeed increase RPM, I don't think the ST2 ISV reading changed in correlation but I'll run the test again to double check.

If I block the venturi air route or keep my foot on the brake during hot start it stops the idle increase I would otherwise get a few seconds after starting, I've tested the booster it doesn't leak, it held vacuum all night. The hoses around it are good and everything is plumbed correctly. As Greg says it could be that on start idle is normal until the booster reaches vacuum and then idle jumps. Although holding brake on start the idle will then increase after a short drive. Of course I have to hold the brake when putting in D and idle drops but is that because of pressing the brake? I'll see what the ISV reading is with the venturi hose blocked as that figure should go up to compensate for less air?

I'll carry out some more tests, take notes, get some video or photos and report back, if you see the butterfly you might be able to say if it's good or bad.

Edit; you say increase the percentage on the ISV with idle control disabled. In my video I increased it from 34% to 40% and idle slowed which seems to be opposite to what it's supposed to do? Also, if in normal idle control the ISV is at 11% and 900rpm how come at 34% the revs drop to 700rpm.

Pagnobito

Some things are better explained by video, I’m no videographer though so excuse the shakiness and bad focus at times, all done on my phone. Some of it is quite loud so be careful if you have headphones on.
The post is quite long and will take a while to go through it all. I very much appreciate all of your help. I'm going on a driving trip to Germany soon and I really need to get to the bottom of this problem as taking the 928 is the entire point of the trip.
Not recorded was taking the oil filler cap off with the engine running at normal idle made no difference so the hose running from the nipple to the intake doesn’t draw much if any air, or it’s full of oil, no not really as I can blow through the hose.

The object is to monitor what happens to the idle during the warm up phase and when the idle jumps and manually controlling the ISV percentage. I connected my DVM to the signal line on the O2 and a vacuum gauge on the front damper via a T piece.

There are a few videos and a data log at the point idle increases for some unknown reason.

Video 1: Warm up part 1; This is from a cold start, or almost as I had a technical issue so the engine had been on for about 10 seconds before but I don’t think it makes a difference. You can see how it's running, O2 not warmed enough to be included in idle.
Video 2: Warm up part 2; After restarting the DVM which was the reason to cut video 1 short the engine continues to warm, all looking good with ISV about 45% and O2 starts to work. Looks normal at this point, right?
Video 3: Warm up part 3 - Idle jump and hot starts; This is where it gets interesting the engine continues to warm and pretty much gets to operating temperature and the ISV drops to 20% and idle increases to 900rpm. O2 still appears to be in range on the DVM. Just before the idle increases vacuum drops a couple of Inches. I was logging this session so you can see where the increase happens.
Check the attached warming.txt file to go along this this video as I data logged the point where the idle increases, not that it tells us much as ISV opening is not logged - it wouldn't need to be for tuning.

I had to leave it here for a little while and when I returned the engine had cooled.

Video 4: warming up again and clamping venturi hose under normal idle. By clamping the hose (you can’t see me do that but you can see where when I take the clamp off) the ISV changed from 45 to 50% (backwards?) so it took account of less air to keep idle correct.
Video 5: Fast idle and disabling idle control; The engine is now hot and the idle is fast, the ISV dropped to 20%, notice the O2 adjust. If I clamp the venturi hose idle drops, the ISV doesn’t change on account of this like it did before, it sticks at 20%. I remove the clamp and disable idle control in fuel parameters setting the ISV at 45% so expect it to idle at normal speed, note immediate choke noise and indeed idle drops back to where it should be. I switch it back and forth a few times but didn’t change the percentage from 45, forgot to do that.
Video 6: hot start; A short video so you can see idle and ISV is good for a few seconds then the ISV changes to 20% and idle increases.

Why does the ISV change to 20%, is the LH telling it to do that? I think it must be as I can override it and return the idle to normal if setting it at 45% If so what are the computers sensing that causes it to do that? I don’t think there is too much air entering the engine as the idle is perfect up to the point the ISV moves then more air enters, MAF signal increases and adds more fuel, mixture is still controlled via seemingly normal O2 sensor.
The ISV also seems to be backward as more air would mean larger percentage. When it was idling normally and I clamped the venturi hose the ISV changed from 45 to 50% allowing more air in to compensate, is that the right way around?

Hopefully you can make sense of this and have some idea what’s going on and a direction to go in finding a solution!

jcorenman

I haven't had a chance to go through the videos, so just a couple of quick thoughts:

Have you checked the throttle cable (from bellcrank assembly to throttle body)? It should have a bit of slack. If it were tight then that would pull the throttle open, which is what this sounds like. If the stop screw is backed off, then I think the throttle plate would be closed to the point of almost (or literally) sticking.

With the idle control disabled in Sharktuner, then it is possible that increasing the manual ISV setting would decrease (not increase) the idle speed, but only if the manual ISV setting is less than 35% or so. The ISV valve is actually max-closed at 35%, and then starts opening again below that. This is so that a broken connection (or dead ISV) will still pass enough air to idle badly.

In any event the ISV% should never be anywhere near 11, 20 or even 35%-- that is outside the range where the LH should ever need to go, and the motor shouldn't run. So this still all points to air bypassing the throttle plate, somehow.

Pagnobito

If it were the case that the throttle plate was being held open wouldn't I have too much air all the time? I'm not seeing that, something is making the ISV change position.. you see during monitoring today the ISV stayed around 45% and normal rpm up until the point it suddenly changed to 20% and idle went up. When it was at 45% and around 700rpm I clamped the venturi feed hose the ISV went up to 50% and idle remained. When it was idling high around 900 and ISV was stuck at 20% I clamped the same hose, ISV didn't change but idle dropped back. I can manually overrride in the ST software setting ISV to 45% and idle returns to normal. Watch video 5 to see what I mean. Video 2 shows the engine running normally if you want to check that, only need to see a few seconds of that one.
I'm not disputing too much measured air getting into the engine but that only seems to happen when the ISV suddenly goes to 20% as all other times when idle is normal ISV is at 45%. Something is causing the ISV to do that, I don't think it's a bad connection as I wouldn't be able to override it and set it back to 45%.

I didn't have this problem before I changed the ISV and I changed it again because I blamed the first new one. The only thing I haven't done is fit the old one back on so that will be my next step, just to see what happens.

Cheers.

GregBBRD

I have no way to count the number of times you've had the intake on and off....but I know it is several times...and for that reason, alone, I want to help you solve this.

I'm incredibly tenacious and patient when things,are going wrong....but I'd be "postal" if I had one intake off that many times!

You definitely deserve the "Patience of the Year" award!

BTW....I will look at your videos, when I have something bigger than my cell phone.
​​​​

​​​​​

Pagnobito

It's probably about 8 times now, I've lost count. Any input welcome.

Mrmerlin

Is their any way you could take a picture of the throttle plate as its sits in the closed position.
to note also the position of the port hole at the top edge of the plate. This should be visible.
It might be possible that the throttle plate isnt centered in the bore.

to adjust it loosen the screws that hold the plate a half turn and open and close the throttle arm about 3 MMs this should center the plate .
note if you make the screws too loose or open the arm too far the plate may lodge and thus not fully close.
I will get a picture of the intake plate to show you the hole

Pagnobito

I've seen the picture on another thread, The hole you mention is almost entirely visible, the throttle plate is on the very edge of it. If I take the intake off again I will get a photo but I don't think there's anything wrong with it. The reason for this is if the throttle plate was on some way comprised and letting in too much air I would see evidence of that right from the start wouldn't I? I also checked the tension on the cable to the plate again (done it already) and if I pop it off the ball on the quadrant it's about 1mm past the ball, ie not tight at all.
On running the engine today and watching the monitoring tab within ST I could see from a cold start that idle was between 750-800rpm, ISV between 45-50% and when the O2 loop started with correct voltage the idle dropped a little and ISV gradually dropped to 40-43% That's normal.
The engine reached operating temperature and the fans came on, it didn't do the idle increase, odd because I haven't changed anything,
ISV still reading 43% and idle speed good. I switched off idle control manually setting the ISV to 43% and idle didn't change but in the monitoring screen ISV read 70% when yesterday it was 20%
Switching back to automatic idle control and immediately more air entered the engine as the ISV changed to 20% and idle jumped to 900, switching to manual and less air, ISV at 43% idle back to 700.

The problem definitely seems to be either the ISV or the signal controlling it. Simply for some reason the ISV changes from 43-45% to 18-21% (still fluctuating and maintaining higher idle, so not failed to the stop as I'd expect idle to go to 1200rpm in that case). The ISV is the item letting more air in (I can hear it rushing through the restriction in the Y piece on the MAF elbow. You can see that in video 5 on a previous post. It only seems to happen when hot so perhaps the automatic control is unable to drive it properly maybe if resistance changes from a bad wire when hot?

I think I need to make a test lead and get an oscilloscope to see what's happening with the signal and voltage. At least that will tell me if the ISV is working correctly or it's the signal.

John Speake

Maybe the new (alternative) ISV is letting more air in when in it's nominal closed position ? Is there an adjustment on the ISV for this ?

Pagnobito

Photos for Stan of my throttle plate.

Three months, three. All that time I've been trying to solve the problem of why my 1990 S4 goes to high idle when it warms up. I first of all fitted a Lowe ISV to cure a cold start problem (wouldn't run without throttle but was OK when warm). I had the high idle problem after that. I looked for leaks, tested everything I could via the WSM, nothing seemed out of spec. I thought maybe it's the new ISV so sourced a Bosch one, unfortunately they don't make one but supplied what they described as a compatible one. On the first run out it seemed to be good but after stopping for lunch the car wouldn't start - tracked down to a failing crank sensor. Replaced that and Bosch ISV reverted to same problem of high idle when the engine was hot.
I checked everything again, looking for vacuum leaks or any leaks at all and testing everything again, I had removed the intake several times during this process and then I saw a post about the venturi so I borrowed another one which didn't make a difference but blocking the hose to it did but why?
Then I borrowed a Shark Tuner in an attempt to see what was happening and I could see the ISV was for whatever reason suddenly changing position and letting a crap load of extra air in.
Lots of posts here and today.
I refitted the now cleaned up original ISV and took the car for a 25 miles drive. It was absolutely perfect, even better than it was before as the new crank sensor has eradicated a slight miss every now and then and more importantly a perfect idle. In N or D it sits around 675 rpm.

All three ISVs measure 8 Ohms so there's apparently no difference, I've had the intake off maybe 8 times so can't see how I could have changed anything else that might affect idle. I'm happy it appears to be fixed and I'll have to assume it was the ISV (both) unless one of them is fitted to a good working car which exhibits the same problem afterwards but who would do that!

I need to take the car out a few more times just to be damn sure but it looks good so far.

Thanks for all the suggestions and help.

Mike