When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
SOLVED! Experts Required - High Idle has me stumped
I've been troubleshooting this on my own as well as with @PorKen for a few days now and nothing we have tried has worked. I am hoping to lean on some experts out there.
The patient is 1992 928 GTS 5spd. The high idle was 850rpm before undertaking a full intake refresh. Now I am wrapping up the intake refresh and the high idle is worse at 1100rpm despite several rounds of troubleshooting and removing the intake three times so far. Before I go and remove the intake a fourth time I want to lean on any experts out there.
and Pressed in one side at a time to prevent swashing the throttle opening)
new seals, vacuum, and breather hoses
new LoweISV (the OEM Bosch unit was functioning properly and no change was observed with swapping)
new injectors (OEM injectors were functioning properly and no change was observed with swapping)
new Bosch O2 sensor (0258003048)
CPS, Hall, and Knock Sensors are all new within the last year.
I have been using the WSM Troubleshooting Guide and been troubleshooting "Irregular Idle" and all the tests seem seemed to have passed
Regarding Vacuum Leaks: I have a smoke tester I have verified the system is leak-free (FYI having a smoke tester is a game changer, I strongly recommend it). This is the unit I have,
Regarding throttle linkage: I have tested to make sure the quadrant linkage is adjusted correctly and tested with the cable to the throttle plate disconnected to make sure the throttle plate is resting on the throttle stop.
Regarding the Throttle plate stop
The throttle stop screw no longer has any factory paint on it as showing in the example photo below. So it was assumed it was messed with at some point.
throttle plate adjustment?
Deformed throttle Body? Take a look it's pretty tight.
Regarding ISV Tests: I can hear the ISV actuating when I conduct the WSM tests and I can hear it actuating as idle settles after blipping the throttle. Also the high idle will do some variation as the car warms up and goes into closed loop mode.
General Stuff
The car is running good, It starts quick, and burns clean. @PorKen connected his Theo diagnostic tool up and we didnt see any O2 lean/rich faults. We ran through the rest of the system tests (eg Tank Vent, ISV, Temp II, MAF, etc.)and they all passed.
What to do Next?
So that is where I am at, I am not sure what else to check or adjust. I sure could use your expert guidance if you have it. One thing I am considering is that the TPS could be acting as an artificial stop, meaning it is preventing the throttle plate from further closing. However, after looking at the photos above it would seem the throttle is fully closed and the ISV should be able to more than compensate if that small gap is in fact an opening.
I am desperate to get this resolved so thanks for helping me dig in.
Last edited by Michael Benno; 04-20-2023 at 03:39 AM.
I think I have a tentative victory. I am able to get the idle down below 700rpm and all other signals look good! Here are some plots from the Sharktuner, these were taken after the car was warm. O2 adaptation was disabled.
RPM is in the black line and you can see it's mostly below 700rpm The O2 signal looks totally normal. The ISV % value is averaging about 37% which is a little lower than I would want. Jim said a target of 45% is what I want.
I'm pretty excited about these results. Thinking about what I changed to "fix" the high idle issue. I did two things. I replaced the original used Bosch ISV with another used Bosch ISV from a local 928 owner. I decided against the Löwe ISVs because the diameter of the openings was not the same and the preset gaps were not the same and I was being superstitious. The other main thing I changed was I sealed the intake side covers with DC111 per the recommendation from @Mrmerlin .
The evidence points to the ISV because when I unplugged the old one the idle dropped significantly to 600ish. Meaning there was not extra air being sucked in. And, the intake passed numerous pressure tests. However the bench testing of the ISVs did not show a significant difference in performance for any of them. Maybe the bench test was not an effective simulation of the Pulse Width Modulation that the LH does.
However, there is a distinct possibility that there was a vacuum-only leak from the intake side covers that did not show up in pressure testing. I had new seals in there and the insides of the covers were also blasted.
In any event, the idle seems much better and all the systems are reading totally normal. So thank you all for this very helpful and supportive troubleshooting. I am very appreciative for your patience.
Next up, I'll be replacing the old OEM injectors for some modern 30lb injectors and some Sharktuning. But that will be another thread.
You pretty much cannot have a high idle without an air leak, but they can be very hard to find. It feels like I have spent years of my life searching for them and that its always a crack in a hose in a turn where its hidden from view. I really recommend an inexpensive smoke tester. Oops, re-read your post, and you have smoke tested. Do it again? Try blocking the ISV? Excess air is getting past the throttle valve somehow.
Last edited by karl ruiter; 03-31-2023 at 08:18 PM.
Did you rebuilt the throttle body? If so, could the throttle plate be slightly off center, and so it appears closed but actually catches the tiniest bit?
Could there be a leak with the brake booster hose or the booster, that might not show up with your smoke test? Could you plug the booster hose and see if the idle changes?
The ultra sonic tool may be good but not while the engine is running.
If I cant find a vacuum leak using a smoke machine, then there is no leak or it is so small that it would not have an impact on idle as @Mrmerlin has pointed out previously. The idle at 1100rpm is way outside of the range of 770rpm. I can tell you when I pressurize the intake to 5psi. It will hold this for several minutes as the pressure leaks very slowly.
Originally Posted by fiatrn
Did you rebuild the throttle body? If so, could the throttle plate be slightly off-center, so it appears closed but actually catches the tiniest bit? Could there be a leak with the brake booster hose or the booster, that might not show up with your smoke test? Could you plug the booster hose and see if the idle changes?
Yes, I rebuilt the throttle body. When I re-assembled, I put the throttle plate in the closed position to center it before tightening, as you suggest. However, this may be something good to double-check. The idle is unaffected by disconnecting and plugging the large vacuum line to the brake booster. Separately I tested for vacuum leaks in the booster and HVAC. None found.
Last edited by Michael Benno; 03-31-2023 at 08:49 PM.
So that is where I am at, I am not sure what else to check or adjust. I sure could use your expert guidance if you have it. One thing I am considering is that the TPS could be acting as an artificial stop, meaning it is preventing the throttle plate from further closing. However, after looking at the photos above it would seem the throttle is fully closed and the ISV should be able to more than compensate if that small gap is in fact an opening.
Did the Theo tool indicate that the idle circuit was closed “at idle?”
Did the Theo tool indicate coolant temperature consistent with conditions?
Into what did you inject smoke for the smoke test?
During your ownership has it always had a high idle?
I think so. I have had the car for 2-years and I cant remember a time the idle was not at 850rpm. I thought maybe the needle on the gauge was off, but it sounds high when comparing it to other S4's in the area. So I am sure the needle is about correct.
Originally Posted by worf928
Did the Theo tool indicate that the idle circuit was closed “at idle?” MB: yes, and validated it was open just off idle, and validated WOT
Did the Theo tool indicate coolant temperature consistent with conditions? MB: Yes, we tested it cold (5 degrees C) and hot (80 degrees C) and both values were within specifications.
Into what did you inject smoke for the smoke test?
MB: I put the smoke tester tip into the large vacuum to the brake booster
Last edited by Michael Benno; 03-31-2023 at 10:13 PM.
Micheal, you certainly have seemed to covered all the bases. I don't think this would cause the idle discrepency as large as yours, but have you checked the gaps on your spark plugs? There are some brands that come out of the box way off, causing an idle issue and rich conditions. Worth an easy look IMO. (Along those lies, what is visual condition of plugs telling you?)
You’ve done a great deal of investigation but I do not see any mention of the charcoal canister that is part of the fuel tank vapor management system. If the canister has never been replaced, there is an internal membrane which prevents the charcoal pellets from escaping into the vacuum lines leading to the purge valves which control the routing of tank vapors. If the membrane fails, the pellets escape and travel through the hoses to the purge valves. These pellets can cause the purge valves (one vacuum actuated, one solenoid actuated) to become stuck open causing a vacuum leak you won’t find with a smoke test. See my thread “fluctuating high idle on a 90GT” for details. Hope this helps.
You’ve done a great deal of investigation but I do not see any mention of the charcoal canister that is part of the fuel tank vapor management system. If the canister has never been replaced, there is an internal membrane which prevents the charcoal pellets from escaping into the vacuum lines leading to the purge valves which control the routing of tank vapors. If the membrane fails, the pellets escape and travel through the hoses to the purge valves. These pellets can cause the purge valves (one vacuum actuated, one solenoid actuated) to become stuck open causing a vacuum leak you won’t find with a smoke test. See my thread “fluctuating high idle on a 90GT” for details. Hope this helps.
I like this hypothesis and it should be simple enough to test - just remove the hose leading to the tank vent solenoid and blow through it. The rest position should be closed. The WSM test calls for audible clicking of the solenoid but does not test the opening and closing of the airflow.
Originally Posted by Jason89s4
Micheal, you certainly have seemed to covered all the bases. I don't think this would cause the idle discrepency as large as yours, but have you checked the gaps on your spark plugs? There are some brands that come out of the box way off, causing an idle issue and rich conditions. Worth an easy look IMO. (Along those lies, what is visual condition of plugs telling you?)
I replaced the plugs 11months ago (4k miles) with new Bosch WR 7 DC which come pre-gapped at 0.7mm which is the factory specification but I still confirmed the gap before installing. I'll pull one of the plugs and inspect.
Last edited by Michael Benno; 04-01-2023 at 12:45 PM.
On the fuel tank vent system you could disconnect the line at the solenoid valve and close off the solenoid valve to see if it makes a difference.
You could also disconnect the ISV and see what happens with a fully warmed up motor and then if the idle is stable [as it should be] engage the ac and see if the motor dies.
Michael, I think your leak test (and other checks) are valid, as far as they go. By blocking the intake at the MAF and pressurizing the plenum, you are checking for leaks from anywhere in the intake system to atmosphere-- all the hoses under the intake, etc. But you are not testing any of the normal idle-air paths, those are all inside your "test chamber". The idle speed is high because the engine is getting too much air through or around the throttle plate, with the throttle closed.
There are three paths for idle air:
(1) The throttle plate obviously, which is held slightly open by the mechanical idle-stop. You backed that off without change, so that's not it.
(2) The ISV circuit, a valve that bypasses the throttle plate and is used by the LH to control idle speed. If idle speed is low, the LH opens the ISV and lets more air bypass the throttle. Those hoses are under the intake, and often soften and split-- but that creates a leak to atmosphere which shows on a smoke test.
3) The brake venturi, the "Y" connector and the hose loop shown in the photo where you connected the smoke. The end you can't see (to the right in the photo) connects to a Y-fitting in the rubber boot securing the MAF (along with the ISV), the other end goes to the side of the intake plenum. This loop does two things: provides another air bypass around the throttle plate for idle air, and increases the vacuum to the brake booster via a venturi.
So one of those three paths is supplying too much air. You've pretty much eliminated the throttle plate and the ISV loop, so disconnect the venturi and give it a good look. Intake vacuum pulls air through a small venturi from the MAF side of the throttle body (i.e. atmos pressure), increasing the vacuum on the booster connection. The hole through that venturi should be 4mm, check it with a 5/32" drill bit--- should barely fit.
One way to solve a too-low idle problem (caused by someone messing with the idle-stop screw) is to drill out the brake venturi to compensate. Then when the original problem gets fixed, you are left with too large a flow through the venturi, and a high idle. If in doubt, swap it for another.
Blocking the booster connection doesn't do anything, that's a dead-end anyway. If you block or pinch-off the hose connection to the plenum then the engine should die, it needs that additional idle air. And opening that plenum connection to atmosphere should kill it also, due to a too-lean mixture from the large unmetered air leak.
I spent this morning testing the tank vent system. In summary, the tank vent system is functioning properly and does not seem to be contributing to the issues.
Here is a review of the tests I performed:
Continuity tests between the LH and the solenoid passed
Activation of the solenoid by bridging terminals in the LH connector (Test Point 20 - WSM 1A page D24-26)
Power supply (12v) to solenoid valve through the LH passed Test Point 20 - WSM 1A page D24-26)
Checked airflow open/closed with activation by blowing air into the solenoid valve using a tube.
free airflow without voltage (ignition off)
closed with voltage at idle (blew air into the valve)
pulsed-open with the throttle open and engine running
checked the running of the engine with the solenoid hose plugged (no impact)
Checked operation of the diaphragm valve
verified to the diaphragm valve (no vacuum at idle / vacuum just off-idle)
verified opening and closing of the diaphragm valve with airflow to/from the carbon canister
Verified free airflow from the carbon canister
Last edited by Michael Benno; 04-01-2023 at 06:09 PM.
03-31-2023, 08:03 PM
SOLVED! Experts Required - High Idle has me stumped
