84-89 Carrera: Motronic AFM

[WydRyd]

STILL in the midst of troubleshooting my low idle/stalling problem, I was told to do the following test on my AFM to verify whether it's working or not.

1. Remove all plumbing up to the inlet of the AFM.
2. Start engine and push flapper slightly open with my finger.
3. If this improves engine idle and stops it from stalling, then the AFM is faulty.

Can anyone confirm this hypothesis for me?

If I leave the flapper alone, engine idle drops to ~300rpm's, pulsates for 3-5 seconds and dies/stalls.

If I push the flapper open slightly (say approx 10-20% open), the engine idles perfectly at around 850-900 RPM's.

If I push the flapper TOO WIDE open, mixture goes way too rich, engine drowns and nearly stalls. Black smoke out the exhaust.

Applied a 9v battery to the AFM's pins and found the following (excuse the formatting):

~% Flap Deflection Uv Up Up/Uv
0 10.41 0.52 0.04
5 10.41 0.52 0.04
10 10.41 1.03 0.09
20 10.41 1.93 0.18
30 10.41 2.83 0.27
40 10.41 3.73 0.35
50 10.41 4.73 0.45
60 10.41 5.73 0.55
70 10.41 6.63 0.63
80 10.41 7.53 0.72
90 10.41 8.33 0.8
100 10.41 9.43 0.9

With the flapper under ~20% deflection at idle, it made no difference to the idle quality. Once it passed the 20% threshold, the idle impromed dramatically and stayed solid at 850-900.

Thoughts anyone? Is my AFM faulty here?

ICV checks out OK. CHT is OK. Compression looks OK. Spark and Fuel Pressure looks OK too ONly thing I can point at is the AFM, at this stage

[WydRyd]

This might be a better table.



I also read this URL, http://frwilk.com/944dme/ and made special note of the refurbishment mod one can do to the AFM to get it performing "like new" again.

He suggests that by moving the wiper assemply down/up the shaft, it will move the wiper contacts to a new location on the resistive track, resulting in a more reliable connection.

I might attempt this mod and see whether it makes any difference, since mine has pretty deep wear/grooves in the resistive track

[Wil Ferch]

Others can be a better help ( is Chris Bennet here?) , but there is a micro switch that is activated "on-idle / off idle" as the throttle moves from rest. Maybe this microswitch lost it's position or is otherwise not functioning properly.

Wil

[Rot 911]

Merv, I have a spare AFM at home that I was going to do a refurbishing on per the FRwilk tech article. It is a working AFM. If you want I can test it with a 9v battery and see what readings I get.

[bfegan]

If the micro switch fails, the idle will go to 1900-2000 RPMs. At least that was my experience while using a Steve W chip.

If you are running a stock chip, the idle will still be acceptable... though a little fast. As I understand it, the Steve W and other chip upgrades are more sensitive to sensor inputs (like the throttle micro switch) than is the stock chip... hence the difference.

[WydRyd]

Hmmmm... babying the flapper to richen the idle mixture can either be:

1. The AFM is fcuked, OR
2. The engine is running lean at idle, which STILL could point to a manifold leak somewhere, which is introducing un-metered air in.

I tapped the ICV with a heavy wrench to make sure it wasn't stuck or anything. That made no difference. I guess it's still worth removing it and cleaning it out?!?

[ChrisBennet]

Originally posted by Wil Ferch
Others can be a better help ( is Chris Bennet here?) , but there is a micro switch that is activated "on-idle / off idle" as the throttle moves from rest. Maybe this microswitch lost it's position or is otherwise not functioning properly.

Wil

Chris "The Clueless" Bennet is here but didn't want to embarrass himself by asking what "Uv" and "Up" means.

Like Wil said, I would confirm that the idle switch is being activated. The "look-ma! no hands!" method is to pull up on the gas pedal with your toe. Sometimes the throttle cable hangs up on something (cruise control cable, new floor mats?) or is misadjusted. Flooring the gas pedal without floor boards to stop the pedal may bend the throttle linkage. If it's close, pulling up on the gas pedal up will be enough to let the throttle close all the way.

If that doesn't work, (with the motor off) go into the engine compartment and make sure that throttle lever is all the way closed. Pull up on it a little and you should here a faint click (like a ballpoint pen) indicating that the off-idle switch is being activated.

Gotta get back to work now,
-Chris

[WydRyd]

Sorry Chris, "Uv" is the input voltage into the AFM (I used a 9v battery). "Up" is the output voltage at the loom connector on the AFM.

Yeh I tested that the microswitch is engaging. I can hear it click when I let the throttle mechanism go (engine off).

Engine is definately running too lean at idle. Playing with the flapper proves this. So it's either the AFM or a huge vacuum leak. Finding where is going to be a b!tch because all the vacuum pupes look OK. Must be lower down in the manifold. Maybe where the manifold runners meets the heads?

[ChrisBennet]

Merv,
Have you disabled the ISV by jumping the B & C on the test socket under the engine fuse box cover?
Finding vacumn leaks - how ?
Disabling the ISV will let the engine idle where it "wants to" without correction from the ISV. This should help in diagnosing vacuum leaks.
-Chris

[WydRyd]

Thanks Chris. I'll try jumping those two ports and see what happens.

If the engine idles properly with the ICV/ISV disabled, does that mean the valve is faulty?

[ChrisBennet]

Originally posted by WydRyd
Thanks Chris. I'll try jumping those two ports and see what happens.

If the engine idles properly with the ICV/ISV disabled, does that mean the valve is faulty?

I don't know if jumping sets the ISV to some neutral value, say 50% open or if it closes it all the way.
The ISV motor doesn't spin it "hovers". If it hovers someplace where the brushes are making good contact it might seem fine.
The thing is, when you aiming the propane as suspected leaks, the idle will change without the ISV trying to correct it.
-Chris

[WydRyd]

OK, well, I've checked all the vacuum hoses & ports behind the throttle body etc, so they all seem secure Only other place is the mating point between intake manifold and heads. Grrr!!!

[toolman]

My intake gaskets were bad when I got my 911. It made a strange whistle sound when I shut it off. I knew something wasn't right, but it did not have any low idle problems.

[greglepore]

I have a spare good afm-I'll send it to you for testing purposes if you pick up the shipping both ways.

[Rot 911]

Originally posted by WydRyd
OK, well, I've checked all the vacuum hoses & ports behind the throttle body etc, so they all seem secure Only other place is the mating point between intake manifold and heads. Grrr!!!

This is the most common vacuum leak on the 3.2 engine. The intake gaskets dry out and shrink.

[WydRyd]

Yeh, but thinking about it, IF there was a vacuum leak, the engine would be idling HIGHER, not pulsating at 300-400 for 5 seconds and stalling. Bizarre & frustrating!

Might re-visit the ICV/ISV and perform the test Chris suggested above.

Greg, thanks for the offer, but it'd probably be cheaper for me to just buy a second hand unit than ship yours down and back up

I sent FR Wilk my measurements and he said my AFM numbers look good, so it's unlikely it's my AFM He also suspects my ICV, as does Souk.

Time to rip the bastard out and pull it apart!!!

[Ned nyna 11]

Merv,
Pull the plug on the ICV and block off the bypass hose. Then you can dink with the idle speed with that 7mm screw on the throttle body.
This will get the ICV out of the diagnosis. You do know that it adjusts the bypass flow to maintain a controlled speed.
The idle switch on the throttle body switches the ICV when the throttle lever passes the closed butterfly position.

[WydRyd]

Thanks Ned. Will give it a try and see what happens.

[ianc]

Yeh, but thinking about it, IF there was a vacuum leak, the engine would be idling HIGHER, not pulsating at 300-400 for 5 seconds and stalling.

No. If you have a vacuum leak, you are pulling extra air that no fuel is being metered for, and the car will run very lean and will probably idle lower. You might also hear popping from the exhaust if the car is warm enough. The pulsing or surging you describe is very common with a lean condition.

The idle switch has two functions; when on, it activates the ISV, and it will stop the flow of fuel to the injectors until engine RPM hits 1100 (on engine overrun).

How does the change from cold-running to warm-running affect the problem (if you can keep it running that long)?

ianc

[WydRyd]

Originally posted by ianc
How does the change from cold-running to warm-running affect the problem (if you can keep it running that long)?

Does the same thing whether cold or hot. Temperature independant, so it's not the CHT. I tested that too and it's within specs.