surging off idle

[redcoupe86]

First I need to get my cars idle and mixture set to stock before I can begin hunting this problem down but I'd like to get some opions on what it might be:

What happening is if I blip the throttle to say 1200 rpm and let off the gas the revs pick up to 2k-2.5k rpm and just hang ther for like 5 +/- seconds before dropping. Then every now and then once the revs drop they bounce up and down between 1k and 1.2k rpm 3 or 4 times before completely settling down.

I have my stock chip back in (I removed the SW chip to check this), but this issue happens with both chips. I did adjust my throttle linkage before this all started happening and it might be the problem, but I'm not so sure. It seems if the throttle was sticking it would stick at what ever rpm you let off the gas and that's not what happening.

Any suggestions?

[Nitrometano]

I dont know really what can cause that a cheap check that you can do is to buy carburetor cleaner and clean the throttle inside and his plate. This will remove the gunk that acumulates there. This have worked for me in other non porsche cars with idle rpm problmes.

[stlrj]

You might check the idle throttle switch first, then have the base idle speed and mixture set to spec.

[john walker's workshop]

Originally Posted by stlrj

You might check the idle throttle switch first, then have the base idle speed and mixture set to spec.

agreed. too high an idle speed or a throttle micro-switch not closing can cause this. try turning down the idle first. 800-880. that's the center adjustment screw on the throttle housing, lookin right at you. 7mm deep socket, clockwise. break off the yellow plastic cap if it's still there.

[scarceller]

It is very possible that when you adjusted the linkage you may have done something that is not allowing the throttle body to fully close and thus the idle micro switch is no longer closing because the linkage is no longer coming in contact with it.

I would check that the throttle body plate is fully closing, a simple test is to disconnect the linkage from the throttle body and see if this allows the body to close further than it did with the linkage connected.

Do NOT start adjusting the idle screw without first disabling the idle speed control motor! To disable the ISC you need to jumper B to C pins on the plug at the left side of engine bay.

Check other threads for procedure for idle adjustment.

[redcoupe86]

well here is what I did: I pushed the gas pedal several times from several different possitions (with car off) and between each push I went back to the trottle body to see if the lever was in full contact with the mircoswitch and it was - every time. I guess it's possible the microswitch could be out. . .but all this started happening once I messed with the throttle linkage, but everything seems fine with the adjustment.

Also, while I know the procedure for adjusting idle I do have a question: if you adjust idle speed up and down does if affect the mixture ratio at all?

[hcoles]

I also have surging problems and will be working on it soon....
My approach is to first set the micro switch and confirm it is working and maybe reset the butterfly stop. Right now I have the O2 unplugged while I get the energy to work on it.
regarding the adjustment of the micro switch... how is this done? Follow Bentley or a thread here? The same question goes for the butterfly stop.. I was told to never adjust this but there must be a way to do it...again Bentley or a thread...?
Another related question... if the adjustment "mixture" screw under the MAF is incorrectly adjusted will this cause surging? Right now it seems to idle smooth/etc. other than the surging.

[redcoupe86]

ok, did some work on the car and here are some findings:

1. The surge to 2k rpm ONLY happens when the clutch is depressed. In other words I can blip the throttle all i want with the the car in neutral (no pedals pressed) and the car revs up and drops fine. However, if I press the clutch pedal down and blip the throttle then the revs hang around 2k rpm, but will fall as soon as I release the pedal.

2. I jumped B to C to bypass the ICV and set the rpm to 880. It Idled dead smooth, but as soon as I undid the wire the rpm dropped to around 710 and it idled kinda rough. So since my mixture was all messed up anyway I played around with that to see if I could bring the rpm up by making it richer or leaner. Didn't seem to help a whole lot. Maybe got the rpm to go up 20 rpm at most, but the rpm always kept fluctuating about 20+/- rpm. I know your suppose to be able to have the jumper on or off and the rpm remain the same but Im not sure how your suppose to do that. Because if I pull it off it drops, so how to you make it the stay the same? If the mixture is set right will that bring up the revs with the jumber off? And I replaced my ICV several months ago so that should be OK.

3. As a side note I did replace my O2 sensor with a generic one (as suggested by this message board) that I got from Pep Boys many months ago.

[redcoupe86]

bump for posterity

[scarceller]

Very interesting that when you hold the clutch down is when you get the problem. I suspect that the pedal cluster may be lose or flexing and actually pulling on the throttle slightly but just enough to pull it off idle.

This should be easy to check just have a helper depress the clutch while you look at the idle switch in the rear.

Maybe your pedal cluster is comprimised (cracked and weak) or maybe just loose, meaning not properly mounted to the floor. I would check this out before hand.

As for your idle adjustment you should never adjust the screw without b&c jumpered. Set the idle to 800-825 and that's it.

[scarceller]

I just checked the chip data for a 86 Carrera USA and the idle in that chip is set to 800RPM. I do a lot of my own chip burning and tuning so I'm rather familiar with these chips.

So this means the Computer will target to keep the idle at 800RPM so you should set the base idle (with B&C jumpered) to 800RPM then remove the jumper.

One other note, don't trust the dash tach for setting idle RPM use a tech tach for this if possible.

I really think your problem has somthing to do with a faulty idle micro switch or bad adjustment. But I'd really look at the pedal cluster for starters.

[redcoupe86]

I'd don't believe its the pedal cluster but I will certainly check into it to be sure!
Let's for the moment concider this a separate issue to be fixed at a latter dat and loock at the idle adjustment. Here is how I did it:

1. warmed up car
2. Jumpered (b+c)
3. set Idle to 880 - with multimeter
4. pulled jumper and car fell to 710 +/-

If I pull the jumper and the idle doesn't pick back up what should I do next?

[DonMo]

Did you take the ICV off and clean it? Make sure it rattles when you shake it. You may have a bad ICV. Also, the DME does some funny stuff if you have cold solder cracks on the board. Easy to spot and if you have a friend with a 3.2 DME that works, a quick swap will tell you if it is yours or not.

[scarceller]

How to bench test the ICV? I covered this topic a while back, it's very easy to bench test you ICV. All you need is a 9Volt battery. See this thread for the details.

Bench testing Carrera Idle Control Valve

However, I suspect your ICV is working just fine. Usually they just completely stop working but since you see idle change when you pull the jumper from B&C this indicates the valve is functioning.

[scarceller]

I'm giving you the idle adjustment procedure right from the Motronic manual titled "911 Carrera DME Testing Plan" this manual covers the Motronic system and howto fix it.

Here is the page on setting the base idle:



Please be aware that you CAN NOT alter the idle by setting the idle screw! The DME controlls the idle in the Carrera and the only way to change the idle is to alter the Chip. I know this for sure since I burn these chips often and my 84 Carrera idle is set to 920RPM because of the 20/21 Web Cams I had to increase the idle. I did this by re-programming the chip.

The 86 USA chip has it's idle pre-set in the chip to 800RPMs. The idea for seting the base idle with B&C jumpered is to get the base idle in the ballpark. I ussually set the base 25RPMs higher than the chip idle. So for your car dissconnect the O2 sensor (O2 MUST be disconnected) jumper the B&C pins and set the idle to 825RPM or so then pull the jumper and the DME will command the ICV to pull the idle down to 800RPM.

When you jumper the B&C pins it tells the DME to center or park the ICV at it's center point (50% opened). This insures that the ICV is half opened and has headroom to pull the idle down or up.

Many folks think you can somehow alter your idle by just adjusting the idle set screw but this is not possible! The idle set screw is only used to get the idle in the ballpark when the ICV is parked or centered.

Also, notice that the Procedure for base idle set should be done along with the CO setting but setting CO is a little more detailed and I'll assume you are OK here so just try to set the idle as described.

If you really want the details on a simple CO set procedure that can be done with just a Digital Volt meter and the existing O2 sensor I can post it later. You can get the CO in the ballpark with just a digital volt meter. But the best method is with a wideband O2 sensor, I have one of these installed in my car but they are expensive.

Hope this helps you better understand your idle.

[hcoles]

Sal, I think this is sinking in for me... setting the ICV at center and the "idle screw" turned to get the rpm's at the target idle rpm (I think for me it is 880, I have an 89) when this is done the "system" now has proper head room to let the DME control the idle. Right?

Yes, if you have the time/etc. post a how to get a close adjustment for the CO.

I have a variable power supply at work so I'll try the variable method... hopefully I can see if the valve movement has "glitches" indicating bad spots. My valve is almost 20 years old now... so likely has some worn spots on the reostadt... if that's what's in there.

thanks.

[scarceller]

You are correct, you now understand the theroy just fine.

Also, I just looked at the 89 chip data (Both USA and EURO chips) and you are 100% correct, the idle is set to 880RPM. So for your car I'd set the base to 900RPM (but setting to 880 is just fine also, the spec says +/- 20RPM from the DME setting) and then pull the jumper out. The ICV should then pull the idle down to 880RPM.

Are you running the stock chip? If not you really need to find out what the idle value is set to in the custom chip and the only way to do that is to talk to the supplier or have someone read the data off the custom chip.

Now if your idle drops to values much lower than the 880 you have an issue else where. Air leaks in the intake are common but these ussually tend to pull the idle up not down unless the leak is large. You can test for air leaks easily with simple starter fluid, run the car at idle and spray the fluid near the intake tubes where they connect to the heads. If you have leak at one of those rubber gaskets (they often get sucked in) it will draw the fuild in and introduce false fuel and the idle with surge up. Start by looking for the air leaks. One other source is a faulty oil cap, if you pull the oil cap off with the car at idle the idle should go up because of extra (un-metered air) introduced into the system. If you pull the cap off and nothing happens this is a red-flag that something is wrong. I suggest you start looking for air leaks first if the idle procedure does not work.

After this we can discuss CO settings but it makes no sence to even look at CO untill you are certain you have no air leaks.

And of course a faulty ICV could cause your issue as well so please test it as I mentioned in my thread.

Originally Posted by hcoles

Sal, I think this is sinking in for me... setting the ICV at center and the "idle screw" turned to get the rpm's at the target idle rpm (I think for me it is 880, I have an 89) when this is done the "system" now has proper head room to let the DME control the idle. Right?

Yes, if you have the time/etc. post a how to get a close adjustment for the CO.

I have a variable power supply at work so I'll try the variable method... hopefully I can see if the valve movement has "glitches" indicating bad spots. My valve is almost 20 years old now... so likely has some worn spots on the reostadt... if that's what's in there.

thanks.

[john walker's workshop]

so when i turn the speed screw and hear the idle raise or lower and the tach says the same thing, then it's just my imagination?

[scarceller]

John,

First before I explain let me say, I do not ever mean any disrespect to anyone on this board and I know I'm not always 100% correct but in the area of this topic (Carrera Idle and how it works) I have done many many long hours of research and learning.

OK, it is not your imagination, you can alter the idle even if the B&C is not jumpered. But you should never do this here's why:

If you start to crank the idle screw 'IN' you richen the mixture and then the ICV is commanded by the DME to 'OPEN' to lean it back out. Now if you continue to crank the screw in the ICV will bottom out and fully open, at this point it is out of headroom and if you still continue to crank the screw in the idle will change.

Same holds true for cranking the screw 'OUT' leaning the mixture. This will cause the ICV to fully close in an attempt to richen the mixture and eventually it will bottom out on the other side (Fully Closed).

So, it is not your imagination but if you do this you just completly messed up the ICV and drove it to bottom out. Not good.

Always use the procedure I showed above that came right from the Motronic manual to set the idle.

Hope this helps.

Originally Posted by john walker's workshop

so when i turn the speed screw and hear the idle raise or lower and the tach says the same thing, then it's just my imagination?

[redcoupe86]

I have replaced my DME and ICV within the last year so I don't think they are bad but I will take another look to be sure.

Originally Posted by DonMo

Did you take the ICV off and clean it? Make sure it rattles when you shake it. You may have a bad ICV. Also, the DME does some funny stuff if you have cold solder cracks on the board. Easy to spot and if you have a friend with a 3.2 DME that works, a quick swap will tell you if it is yours or not.