First, clean the ICV really well then bench test it quickly, see this post:
http://forums.pelicanparts.com/porsche-911-technical-forum/360766-bench-testing-carrera-idle-control-valve.html

Once you know the ICV fully opens and closes then you know the valve is not burned out.

Then test that the ICV can bring idle up and down properly, here's how.

1 - get the car warm and idling
2 - unplug the harness from the ICV
3 - apply load to engine, AC, heat blowers, lights whatever it takes to cause the idle to drop noticeably
4 - with load applied plug harness back into ICV the idle should instantly return to your baseline (84 cars are 800RPMs)

Next continue testing in other direction,
1 - ICV plugged in and load applied (from where you left off above in step 4)
2 - Unplug the ICV
3 - remove all load
4 - Idle should go above 800 when all loads removed
5 - plug ICV back in, idle should return to 800RPMs

If those steps above work then the ICV and the DME portion that controls the ICV are functioning properly.

A hunting idle as you describe is most likely not the ICV, it could be the idle stop micro switch test it with an ohm meter (or test light) at the connector for this switch. Switch should close when throttle plate closes. It then opens if throttle open very slightly.

The other cause is a poorly set base mixture, most likely running lean. Also if you have a O2 sensor, disconnect it and see if hunting stops with it disconnected, could be bad O2 sensor sending bad readings in closed loop.

Lastly, an air leak in the intake. This will cause lean mixture

I think I have found the socket - rubber 1" dia, about 2" from the coil....

but not the reference in the Bentley manual. which page??? which pins are b and c?

My idle sits at 1000 rpm when cold and bounces between 1000 and 1200 when hot.

Thanks for your help

I'm inclined to believe John Walker's microswitch theory. If it hasn't been checked yet, it should.

I just breezed through a lot of the suggestions, but I don't believe I saw anyone suggest the O2 sensor. Also, Draco is spot-on about base idle.

Personally, I would concentrate on the air/fuel mixture. Everyone knows that a rich mixture in the 3.2 is a major contributor to idle surge. The other is the O2 sensor...I replaced mine recently and eliminated my surge problem.

.02 cents from someone who's still learning the game.

A bad idle microswitch or one that does not close (make contact) when the throttle plate returns to closed (idle position) will produce the behavior you explain and cause for high idle.

Also, shorting the WOT switch by unplugging the harness from the WOT switch and jumpering the 2 pins in the harness is exactly the same as jumpering B&C pins on the sidewall connector. I do this all the time to set base idle on these 84-89 cars. Simply the procedure for setting base idle is that the idle switch must be closed and then the WOT switch as well, this is a condition that could never occur under normal operation. But be very careful, if the idle switch is not closed and the WOT switch is then this will put the DME into the WOT maps and the AFR will drop to the very lo 12.0 range.

This is all really good information, but the problem I'm having I can't seem to get answered, is what direction to mount the valve?

I've downlaoded the facotry manual and on page 20-7, it's in the diagram, but it doesn't show the orientation. The arrow on the ISV body implies it is uni-directional - but which direction?

In some photots of the cleaning and testing process, I seen the ISV mounted in one direction, but when i pulled my ISV, it was rotated in the mount 180deg, pointing the other direction. Perhaps this is part of my rough cold idle problem, but I would like to know for sure which direction is proper.

Can anyone definitively say what is right and what your reference source is?

Jim
1985.5 944 NA

Jim,

My ICV has only one arrow and it points in only one direction. The air flows from the top of the Throttle body through the valve to the bottom of the TB. Simply this device by-passes air around the TB. In my car the Valve is positioned so I can see the Arrow pointing down with the electrical connector to the left side. I'll try to find a good picture.

Here's the picture:

http://forums.pelicanparts.com/uploa...1247060260.jpg

Source for this picture is from "Carrera DME Testing Plan" book.

Sal,

Help me verify this thought:

Way to test the idle switch:

Start car, go to throttle on engine, lift throttle just enough that the idle stop micro switch opens. car should be around 2000 rpm. With the other hand, try to use a small tool to close the idle stop switch, while keeping the throttle up.

This should cause the DME to get an idle signal, see the high rpm and cut the signal to drop rpm. This should induce the fluctuating idle posted earlier in this thread. If this happens, shouldn't it show the idle stop micro switch is working?
(As long as DME and other things like the wiring are ok)

Neal

Neal,

I'd simply test the idle switch by disconnecting the harness from the switch with the car at idle, the minute you disconnect the harness the DME will transfer into the PT maps and the IdleControlValve gets parked. You will also notice your RPMs go up somewhere between 1000-2000RPM most likely around 2000RPM. If this test works it means the idle switch is closed when at idle and unpluging the harness from the switch causes the same thing as opening the switch.

If you have an ohm meter or continuity test light you could also test the switch by disconnecting the harness from the switch and testing for continuity across the switch. You can test with engine 'Not Running', simply test for continuity with throttle plate closed in idle position and then slightly lift the plate off idle (about 2mm or so) and the switch must open.

Sal,

Thanks, did not know unplugging the switch caused the PT maps to be selected.
Regarding my ramblings, just proposing a way to do a "field test" for those away from home or a shop. I agree that a meter and actual manipulation of the switch is the best way to diagnose any issues, 'specially with the car not running, much safer.
Appreciate all your tech expertise.

Neal

Neal, the quick field test is simply with car running at idle simply unplug the harness from the idle switch (this is like opening the switch) it then moves you out of idle maps into PT maps and the RPMs rise instantly to close to 2000RPMs. That's the simplest field test I know.

Here's another trick many may not know: Jumpering pins B&C at the test connector near the fuses is exactly the same thing as closing the WOT switch. So simply unpluging the WOT switch harness and jumpering at the WOT switch harness does the same thing as jumpering B&C. So I have a old WOT switch that I soldered always closed and I use this to set base idle. The test condition for parking ICV circuit to set base idle is simple, You have to have both the idle and WOT switches closed, this condition is never possible under normal operation but this is exactly what jumpering B&C pins does.

So the next common problem is what happens if you jumper B&C and the idle switch is not closed because it's not adjusted properly or simply bad? Here we have the WOT switch closed but the idle switch open, this one will drive you nutts because what you have is the WOT maps engaged at idle and the AFR plummets to the low AFR=12.0 range I have had this happen and drive me nutts! Those 2 switches MUST be functioning correctly in our cars!