NOT my Idle Control Valve - car revving when hot.
 

I have some idle problems with my '84 911 Targa. It revs from about 800 to 1000 every second or so when the car's temperature is at about 3/4. It generally happens in the summer when waiting in line to autox OR when I am stopped at a light after some spirited driving. It never happens when the car is running lower than 3/4 temperature (so it does not happen often.)

I replaced my ICV this winter and thought that would solve the problem. I let the car heat up this weekend (parked and running after some driving) and the problem came back.

The ICV I installed was brand new - so I can't imagine that is the problem.

Any other suggestions?

jcsjcs

IMO, if what you're saying is that the temp gage needle is 3/4 up, your car is probably overheating and starving for cooling...

be sure the closed throttle micro-switch is depressed, at closed throttle. you should be able to hear it click as you open and close the throttle. by hand, back at the engine of course. also, if the idle is too high, it will do that. 800.

Do a search on "hunting idle" or "surging idle", the possibilities are boundless. From the CHT, to the WUR, to the o2 sensor, to air leaks around the intake manifold.
Happy reading

Well...

My knowledge of the ICV is hands on, played with my past '88 quite a bit chasing down a problem, so I will try to help.

If the ICV is good, (and the old one was doing the exact thing) we can take the ICV out of the culprit status, for now.

But not the role of the ICV. From your description it seems like:
The idle switch is closed, telling the DME your foot is not on the accelerator. So, if the engine speed gets over a set speed, it cuts the engine for a moment. (Does it do this by closing the fuel injectors, or the spark? Don't know, any others know?)

This should not happen, the ICV is sent a signal from the DME to open or close, adjusting air to regulate to the proper idle speed.

For your condition, my guess is one of the three problems.
1.) There is an air leak, the ICV closes all the way to adjust the speed down, but there is still too much coming in, so when the idle speed gets to high, the DME cuts the signal.

2.) The DME has a short in the idle circuit preventing proper control of the ICV.

3.) Both ICV's are bad

Not a trained mechanic, all suggestions above are from memory, any others with more clarity, please correct me!

Techman

Looks like i have some things to check out this weekend.

A couple of notes:

1) See pic of a gauge with arrow of where I am seeing the problem. Want to dispel the concern that my car is running too hot. My temp only reaches the arrow if the car has been running without moving for a long time OR when it is REALLY hot out and I have been driving aggressively and I am stopped for some reason. The car NEVER gets any closer to the "RED" - so please confirm that this is an okay place to be on the guage.

http://forums.pelicanparts.com/uploa...1239206599.jpg

2) BOTH ICV's are bad???? Where is the 2nd one?


Will report back as I check out the suggestions above.

Jonathan

Both- Referring to your old one and the replacement. Did they both behave similar?

I am 99% sure the car is acting the same with either ICV installed. If some of the other ideas don't pan out, i will try switching them (easy to do...)

HMM, thoughts on how to check for intake leak if the DME cuts out when revs too high. This would prevent using the normal way of spraying something around intake to see if the revs change.

Take the idle switch out of idle, maybe by wedgef it open just enough to open the microswitch. This should also allow it to over-rev, the DME would think the increased revs is caused by the throttle, and not cut the signal. Should not get above 1200-1400, unless you have a huge leak, but be ready to kill it.

your ICV should not cause it to hunt. if the ICV is making it hunt, it is because something else is telling the computer to tell it to change RPM. when ICV's go bad, they stop working and the idle never changes, such as when the AC is truned on.
this is from my experience with other cars, i have not doen much with the 84 and newer P cars.

Yep, a malfunctioning ICV will not change the idle... and they will not prevent it from changing.

ie. Car is set to a static 800 rpm when cold, ICV failed half open. Car warms up to operating temperature, idle is now 900-950. The malfuncting ICV will not close to adjust the idle.

Next example
Car is set to a 950 rpm idle when cold, ICV failed. Car warms up, idle now at 1100 plus. Throttle position switch is closed, telling the DME "I am at idle", DME sees the revs hit 1100 or so, cuts signal to drop rpm. Turns back on, idle climbs again, DME cuts again etc.

A bad ICV could cause it to hunt, in this situation, as mentioned in the first post.

a dead ISV would not cause the hunt, because it's dead. the hunt is a working ISV doing what the computer tells it to do. like lower the too high idle. it's also not in the system until the throttle closed micro switch is tripped.

Is it possible for someone to do a quick test for the original poster?

Unplug the ICV to simulate it's death. Hold in the idle microswitch, lift the throttle just a little, and have a partner start the car.

For my clarification and the OP, tell us what happens.

Thanks all!

edit

+1 on Base Idle.

When I was messing with my idle, I adjusted it too high and got "hunting".

If you have a stock chip, your Base Idle should be 780-800. If you have a Steve Wong chip the base idle will be 880RPMs

Jumper with a piece of wire at B & C (see Bentely Guide) on that open round Relay socket on the driver side wall of the engine compartment then verify your base idle.

I would also check the AFR on the AFM.

Ghetto Check by connecting a DVM to the O2S and known good Ground. You should be reading between 0.20-0.60 volts, fluctuating. If not, adjust to proper levels, up to 0.80 is rich but tolerable depending on the condition of your O2S.

Of Course a LM1 is the best tool for this. you want to be between 14.2-14.7

If you do adjust your AFR, you need to verify Base Idle again.

my 2 cents.

Draco, I agree the baselines all need to be properly set.

What can you tell me about "Unplug the ICV to simulate it's death. Hold in the idle microswitch, lift the throttle just a little, and have a partner start the car"

Will that not get the result the OP presented? Won't the idle of the car rise to about 1100, drop and rise again to about 1100?

Hey Techman1,

That sounds correct but I have not done that procedure.

When I replaced my ICV, she was in fact faulty.

When she hunted, I took a screw driver and tapped on the old ICV and she cleared right up. Replaced and no more hunting.

Make damned sure that idle switch is doing it's thing (measurably,..not necc. "audibly)


Best,

Doyle

OK - so a friend helped me do some testing over the weekend.

1) Ran car without ICV - idled perfectly

2) completed all Bently tests around idle components - and all checked out. Checked various gaps, etc.

3) Used a jumper on on Relay socket to check base idle. Coming up at about 750 rpm - looking at the car's tach. Is this reliable enough?

4) When we look inside the ICV while the car is running, we can see it closing / opening halfway - but then it vibrates a bit - not really closing or opening all the way. Is this normal?

Be very careful when monitoring the ICV, I believe a short there will dmaage the DME (ask me how I know!)

While watching it, have your friend engage the AC and turn the headlights on. This load on the engine should drop the idle, and the ICV should try to compensate.