3.2 low idle and popping unless I disconnect AFM?

[Javers]

Hi all,

Just started having a problem with a low idle (600rpm ish) and popping on my 3.2 after warm up. Always has been 800rpm rock steady up until now. So far I have checked:

- Cleaned contacts on DME
- Re-flowed solder joints on DME relay
- Cleaned cap and rotor (both are less than a year old)
- Installed new HT leads
- Verified ICV valve operation by applying voltage to pins (it moves from fully open -> fully closed)
- Verified correct operation of both throttle position switches
- Verified AFM resistance varies (with no dropouts) as the vane moves (was rebuilt less than a year ago)
- Verified resistance from the AFM temp sensor is within range recommended by workshop manual
- Verified resistance from the CHT sensor is within range recommended by workshop manual
- Verified alt sensor is open circuit (I'm about 20ft above sea level)
- Checked all vacuum lines are attached, sprayed carb cleaner everywhere I can think of, can't find any leaks
- My car is Euro spec so no 02 sensor feedback system


So far it feels like an air leak somewhere but I can't find any, and these symptoms have got me baffled:

- With the car idling low, I can manually open the AFM flap about 20 percent and idle raises
- The idle adjust screw seems to have no effect whatsoever on idle speed with AFM connected
- If I disconnect AFM the idle shoots up to around 1200rpm and idle adjust screw will now adjust rpm
- If I open oil cap to introduce more air idle does not increase, it drops lower
- Sadly I have no way of verifying idle mixture atm, but if anything smells too rich

I would say AFM vane is open about 5% ish when idling, surely disconnecting it would cause less fuel to be injected? and therefore my mixture is overly rich? But some of the above would indicate too lean? Maybe I have a different problem altogether?

All thoughts welcomed.

Jon

[greglepore]

You didn't mention that you checked the idle position switch, ensure proper operation of that.

I'd also check for leaks around manifold gaskets, common problem.

If needed I have a spare afm you can try for the cost of shipping.

[scarceller]

I've done much work on these AFMs and I know that your AFM signal voltage should be about 1.0 to 1.3volt at idle. Now if your idle is way down in the 600RPM that voltage would be lower if lean and much higher if rich. I suggest a quick poke at the voltage line on the AFM just to see what you have. If you really are rich the voltage would be higher than 1.0 volt. Testing the voltage value from the AFM while at idle will give you a general idea of what the AFM is signaling, above 1v is richer below is leaner, should at least give us some more clues and it's a simple test. I know you said you don't run an O2 sensor but can you stick one in? do you have a spot for one? If so put in a simple narrow band O2 and measure the voltage on it's signal line, if it above .7v your rich if it's way down below .4v your lean, another cheap trick to find lean or rich indication.

Here is a picture to help, the signal line is the green/red stripe:



I also have a detailed AFM doc I started sometime ago that may be worth taking a look at, you can open the AFM but I DO not recommend changing any of the adjustments, not yet anyway. You need to look everything else over first. Here is the doc:

http://members.cox.net/widebody/ then find the link at the bottom, it's a WORD doc.

[scarceller]

You mentioned that at idle your AFM is 5% opened, this indicates LARGE air leak somewhere after the AFM. You then said that if you open it to 20% by hand the idle comes up, the 20% open door is correct for idle. See below.

At Idle your AFM should be about 20% opened, see this picture, it should be at pos #2 (more or less):



The 20% open door is not a guess on my part, I actually looked at my properly running car with the cover removed from the AFM. The voltage was about 1.0v at idle in my car.

[3.2 CAB]

Hi Jon, you said that the idle lowered when you removed the oil fill cap from the tank. On every 3.2 that I have seen, this is normal for the idle speed reduction, when the oil fill cap is removed. I have also seen where a bad sealing gasket in the cap, also caused the idle RPM drop. I can only speak of what I have actually seen, and there might be others that have a different view on this. So far, it looks like you are getting some VERY useful information from Sal, that most everyone can benefit from. This one is going in my file. Good luck!! Tony.

[scarceller]

Tony,

Something I learned about removing the oil cap and why it can vary on some cars:

If you remove the oil cap it allows false un-metered air into the system downstream of the AFM if the amount of air is not much then the idle goes up. But if the amount is very high the idle actualy goes down.

Now, here's the tricky part. If you take the air tube of the top of the oil filler neck. This is the tube/pipe that goes back to the Throttle Body area and look inside the tube right at the end that connects to the oil filler neck (you are looking inside the rubber pipe) you should find a Brass cone shaped device. This device is in the pipe to actually restrict the amount of air that goes down the pipe. So usualy if the brass device is in place and correctly installed (it sometimes gets crossed up) and you open the oil cap the car should idle slightly higher, mine has this device and idles higher when I open the oil cap.

But because these car are now 20+ years old and have seen much service, folks sometimes change this pipe without knowing that this brass device is in the pipe. If the pipe does not have this device and you open the oil cap then LARGE amount of air enters the intake and the RPMs drop. Hope you get the idea.

Check your pipe!

I'm amazzed at the things I find in these cars! I found another one just like this on a fuel breather line in the front trunk!

[mhinch]

Javers, same symptoms as my 3.2, checked all the obvious things (as you have)
bit the bullet and removed the intake manifolds to find 6 out of the 12 inlet manifold gaskets had been sucked in. Replaced with new gaskets, reset tickover and CO went for a celebratory thrash and 3rd gear went bang.....ho hum. Strange thing about the inlet gaskets is that although it was obvious when I stripped the manifolds off they showed no signs of leakage when liberally doused in carb cleaner .

Mark

[3.2 CAB]

Sal, that could be it. I have had the car for right at 10yrs now, and I still learn about something or other about it from time to time.
Just like I learned something about my BMW 750iL V12, while driving in the rain yesterday. All of the other times it had rained before, I was on the interstate. Yesterday I was on city streets. I found that, you can have the windshield wipers turned on to the lowest continual wiper speed, BUT, when the engine and speed of the car drops below a certain RPM or speed, or at a stop, the wipers automatically go to the intermittent wiper speed!! I have been learning something about that 750iL ever since buying it. Tony.

[scarceller]

One very good way to test for air leaks in the intake is to have the intake 'Smoke Tested' I had this done at a local tuner in my area. He has a simple machine that makes smoke and then they remove the air cleaner and housing from the car and plug the smoke machine directly into the intake side of the AFM with a special adapter. Next, the machine pumps smoke into the intake. You then fully open the throttle plate and wait. In my car I had smoke coming out of tiny air leaks like the shafts on the throttle body. Way cool. Only cost me $40.00 to do. Highly recommended for the money.

[rusnak]

you can quickly check the hoses visually and the clamps for tightness. I found and corrected vacuum leaks around my ICV clamps, which I tightened with a pair of sidecutters. I also found leaks in the cruise control hose ends, and finally cold air valve between the throttle body and the crankcase. This valve is very small, has a small hose, but can easily be missed when looking for vacuum leaks visually.

[Steve W]

If you don't have access to a smoke machine, try engine starting fluid (ether). It is more volatile than carb cleaner, and if you have a leak, the rpms will rise sharply. Before doing this, disconnect the plug to the ICV and O2 sensor if the car has one to eliminate their autocorrection into the idle speed.

It may be running lean causing a misfire, and raw gas going out the exhaust, and give the impression of running rich - a bit difficult to say without an air/fuel ratio monitor.

[Javers]

Hi everyone, I'm amazed at the amount of replies in such a short time! My thanks goes out to you all, here's where I'm at:

Greg - I've checked idle switch and its working correctly. Thanks for the offer of spare afm, i'm in the UK so will pass for now, but its offers like this that make Pelican an amazing place to be a member.

Sal - fab info on the inner workings of the afm, I've downloaded your word doc and have also taken a reading from my afm at idle, its hovering between 1.36 - 1.37v which is a bit higher than your findings of 1v. Makes me think either I was wrong and my vane is opening a lot more than 5% or that my afm is way out of calibration.

Tony - thanks for the tip about the oil cap, I've had a close look at mine and the seal is in good nick, but I'm glad you bought it up because I never checked before.

Mark - my inlet gaskets certinaly haven't been changed since I've owned the car (probably never have) so even if they are not the cause of my problem, probably a good idea that I do so anyhow.

Rusnak - i will do some more checking for leaks, don't put it past me to have missed some.

Steve - sadly no access to a smoke machine, will see if I can pick up some starting fluid and go over the engine again.

Bearning in mind all the advice so far, there's quite a few things i'd like to achieve during sorting this out:

- Get an 02 sensor installed so I can actually meter my idle mixture level
- Crack open my afm and log the vane positions / voltage out against sal's baseline
- Hunt / eliminate any leaks I can find

Hopefully (weather permitting) I'll get cracking with this lot over the weekend. Will report back soon.

Jon

[scarceller]

Jon,

You said:

Sal - fab info on the inner workings of the afm, I've downloaded your word doc and have also taken a reading from my afm at idle, its hovering between 1.36 - 1.37v which is a bit higher than your findings of 1v. Makes me think either I was wrong and my vane is opening a lot more than 5% or that my afm is way out of calibration.

At what RPM are you seeing 1.37V? My numbers are what I see in my car, but if your RPMs are low as you say and you are seeing 1.37V signal from the AFM my bet is it's running rich. But, please don't guess here. Get that O2 installed and see if it reads above .7v that will tell you for sure.

If you are at 1.37V or so then that AFM is about 30% opened.

Next, You can open the AFM just to have a look but DO NOT touch any adjustment! If you muck with it you'll find it very hard to get it back to original settings. Look closely at the AFM to see if it has ever been opened by PO. If you think it has it's a red flag! Many folks go into these AFMs and losen the spring thinking somehow it will help performance, it won't! The only way to tune these cars is with a Wide-Band O2 monitor and the services of someone like Steve Wong to burn you a new chip.

Sometime ago I tried to get folks to measure the AFM voltage at idle on well running 3.2L Carrera to get a baseline from several cars. I had no luck, folks just did not respond. I think this is very easy non invasive test that would really help others. If you wish to measure the voltage please send me the reading to sal.carceller@cox.net I'd love to have more numbers.

I'm suspecting you are rich at idle based on that AFM signal.

Please send the RPM value that went with the 1.37V signal.

[john walker's workshop]

pull one plug at a time and see if one cylinder is dead at idle. you can also pull the injector plugs one at a time for the same result. 3.2s suck in intake gaskets which kills the affected cylinder at idle and low rpm. hard to find by spraying stuff around the intake because the potential leak is way down below the edge of the fan shroud at the head's intake surface.

[scarceller]

John,

I was thinking about why the starter fluid does not get to the intake gaskets and it MAY be because of the huge positive pressure in the shroud from the engine fan. Meaning that the large pressure inside the shroud just does not allow the fluid into the shroud. So what about taking the belt off the fan just for a few moments at idle? That would allow the fluid easily in.

Just a thought.

Originally Posted by john walker's workshop

pull one plug at a time and see if one cylinder is dead at idle. you can also pull the injector plugs one at a time for the same result. 3.2s suck in intake gaskets which kills the affected cylinder at idle and low rpm. hard to find by spraying stuff around the intake because the potential leak is way down below the edge of the fan shroud at the head's intake surface.

[Steve W]

Sucked gaskets are very common. Preventing this by tightening the manifold bolts only makes it worse, because it warps the manifold flanges, increasing the possibility of the gaskets getting sucked in. I had to redo mine a second time because I didn't realize mine were already warped from previous tightenings, and the new gaskets with Curil-T got sucked in within 2000 miles. Starter fluid easily detected them because the high intake vacuum sucks it through. That's probably why the 964s and later came with the three bolt intakes. So my recommendation if you do redo the intake system is to:

- check and replane the intake manifold flanges
- use new phenolic spacers as they have concentric ridges to grip the gaskets, and the old ones warp, and there's really no way to remove the old gasket material out of them without scraping off the ridges
- use a good gasket compound such the sticky dark brown Permatex aviation compound on all surfaces
- replace the injector o-rings
- replace throttle body o-rings
- replace intake manifold rubber boot
- check and r&r the rubber hoses/vacuum lines above the intake manifold - in particular behind the throttle body, and the ones going to the fuel pressure regulator on the left, and especially the pressure damper on the right. Sloppy oil changes dripping onto the rubber vacuum nipple of the fuel pressure damper swells and cracks rubbet and will cause havoc with the fuel pressure, and all kinds of idle and off idle drivability issues (too rich at idle, too lean off idle)

[Javers]

Right quick update, its pouring with rain here atm so not got as much done as i'd hoped however here's where I'm at:

Sal - In answer to your question, yes my afm has been opened before (by me a year or so back when moving the fingers onto fresh track and also adding the extra wire that was present in the later year afms) but scouts honour I didn't fiddle with any of the adjustments! I've just cracked her open again and baselined my voltage using the guide in your word doc. Here are my results compared to yours:



I couldn't comment on the tension of the spring as I've never changed it, but my adjustment "A" looks in quite a similar position to the photo of yours in your word doc:



So since my results are so similar to yours i'd say my afm is working properly, but since I'm getting around 1.37v with a low idle my guess is either I need to adjust my idle mixture to allow a bit more air to bypass the vane, or my spring tension is too weak. Will know more once I get an 02 sensor installed and can start taking some readings.

John - Will try pulling plugs / injector connections one by one as see what affect it has.

Steve - Think you've inspired me to swap out my intake gaskets, will order up a fresh set tomorrow. If I'm going to take intakes off may aswell swap out that leaking oil pressure switch while I'm at it, I feel a case of "while your in there" coming on.


Jon

[Javers]

Guys its all over. At approx 18:30 tonight this porsche had its last journey, in true 911 style I went off the road backwards and into a tree I was only doing about 40-45mph. It had been raining and while taking a fairly graceful corner the car never straghtened out and (from what I can remember) whipped round, and went hurtling backwards off the road into the nearest tree. I don't have a scratch on me but the crash had enough force to buckle the steering wheel, break my seat and pop both the front and rear windscreens out.

Guess I'm still in a bit of shock and totally gutted.

Jon

[mhinch]

Jon, Gutted for you mate, the best crashes to have are the ones you walk away from.......

Mark

[3.2 CAB]

Damn Jon, are you sure that you are OK? I really hate to hear this about your car. Tony.