3.2 stalls when cold

[C3larke]

I have a 1985 Carrera 3.2. It's all stock except for a Steve Wong chip, SSI headers and a K&N type air filter setup. I recently did a full rebuild of the engine and transmission. Before the rebuild, I could not get the car to run properly because the exhaust valves were burned and the engine was losing compression. Now the car runs fantastic with the exception of the engine stalling when cold.

When I start up the car in the morning, it fires right up and idles around 1000-1200 rpm as it should on a cold start. I typically pull out of the driveway, coast down a hill for about 1/4 mile, and come to an intersection. If I let it coast down to ~1000 rpm so the fuel injection kicks back on and then hit the clutch, the engine will promptly stall. If I pop the clutch before the fuel injection kicks back on, roughly 1300 rpm or higher, the engine speed will drop to a normal idle and it will continue to run perfectly. If I coast down to ~500 rpm and pop the clutch, the engine speed will pop up to normal idle speed and it will run perfectly. It's almost like the idle control valve gets stuck shut as the car coasts and can't open up again in time.

Once the car is fully warmed up, the engine generally does not stall when I put the clutch in but the engine speed still tends to dip to a near stall if I don't time my use of the clutch carefully. Until the engine is fully warmed up, it really likes to stall when I pull up to stop signs and signals. I've gotten so accustomed to it, I routinely pop the clutch a second time to bump start the engine as I pull up to a stop before changing gear.

Here is what I have tried to correct the problem so far:

Confirmed the O2 sensor voltage fluctuates from 0.1 to 0.8.
Confirmed the idle switch on the throttle body is working properly.
Set the CO level to ~0.9%. [It seemed like I had to turn the adjustment screw too far out to achieve this. I think it required about 6 turns.]
Set the idle such that engine speed is unchanged when a jumper is connected to put the ICV in the neutral position. [confirmed CO setting with idle speed set.]
Replaced the cylinder head temp sensor when I rebuilt the engine.
Replaced the idle control valve.
Replaced crankshaft speed and position sensors.
Replaced vacuum lines to fuel pressure regulators.
Replaced cracked breather lines.

I am suspicious of the air flow meter. I opened it up to check for wear on the wiper/resistor strip. It had worn through the resistor strip in a couple of areas, so I repositioned the wiper to run on an unused portion of the resistor strip. This made a noticeable improvement in the way the car runs, but did not solve the stalling issue. I also noticed someone had carved a notch into one of the teeth on the adjustment wheel which makes me concerned that a previous owner may have messed with the spring setting. Could an improperly set air flow meter spring cause my problem? If so, how can I restore the factory setting? How many notches do people generally adjust the spring? How far could it be adjusted before the engine would no longer run properly?

I'm also unsure how to rule out vacuum leaks. I replaced most of the vacuum lines and breathers, and the engine speed changes when I take off the oil filter cap. When I rebuilt the engine, one of the intake runner flanges cracked when I torqued it down. I took it to a machine shop, had it welded and then had the runners surfaced. I installed the manifolds with sealant on the gaskets around the spacers. I am pretty confident that it does not leak, but it would be nice to be able to confirm that. What else should I do to confirm there are no vacuum leak issues? Get a propane hose and check for idle speed fluctuation when I move it around the engine?

There is one other potential concern. The engine makes an intermittent tapping noise at idle. It fades in and out at idle, and its speed seems to keep pace with the engine. Once the engine speeds picks up past 2000-3000, I can no longer notice it. I have always assumed it was a typical 911 engine sound and a fellow 911 owner listened to it and also considered it to be normal. Should I be concerned about this?

I would really appreciate any input on this problem. I have been reading through all the threads I can find on stalling engines, but have so far been unable to correct my problem. The car is running perfectly except for this stalling/dropping RPM issue and I would really like to get to the bottom of it.

Thanks for your help.

Chris

[steely]

Welcome.
This is a lot of information.

What is your idle rpm?
Any / all rubber lines in the vacuum system should be replaced.
What did you have done during the rebuild?
Why did you replace the ICV? Was it bad? How did you determine this? Exactly? Is it being driven properly by the DME or was this done out of a blanket update?

[C3larke]

With the Steve Wong chip, the idle is supposed to be 880 rpm. The tach indicates it is idling around this speed when warmed up, but I don't have any other instrument to confirm this. The idle speed does not change when I connect and disconnect a jumper to set the ICV to neutral. I believe this indicates the DME thinks it is set to 880 rpm. Should I get an instrument to measure the RPM off the ignition wiring and confirm the idle speed?

I have replaced all the rubber/plastic vacuum lines on the rubber intake boot as well as the boot itself. I also replaced the two ~1" diameter crankcase breather lines that run to the oil tank. I have not replaced any of the other lines.

I did the entire engine rebuild myself. I cut fresh valve seats, replaced all the valves, guides, and springs, replaced all the bearings, installed new piston rings, new timing chain and ramps, all new timing gears, new intermediate shaft gear, etc. I went into it planning to do a top end rebuild only, then thought I might as well do the rings and hone the bores, and before I knew it I split the case and did everything. I also initially thought I would just open up the transmission and replace the synchro rings to make it shift better. Then I discovered metal fragments...... Ended up replacing every single bearing in the transmission, first through third gears, all the synchros and sliders, sent out the case to get the spun mainshaft bearing bore sleeved and the preload/backlash set. Shifts great now!

I replaced the ICV because a number of threads suggested it can cause the sort of problems I was experiencing if it is not operating correctly. How can I confirm the DME is sending it the correct signal?

[DRACO A5OG]

Yes, 880 is correct but did you set the Base Idle via jumpering B & C in the engine compartment. car needs to be at operating temps when you do this.

Also, just a vacuum test, when warmed, leave her on and remove the oil cap, if idle drops you are good, if not, you have a vacuum leak.

Oops, edit, you set the base idle correctly. Do the vacuum leak test.

[steely]

apparently pretty rare - but refer to this thread if the ICV drive is suspected.
3.2 cold start issues

[C3larke]

Thanks for the replies.

I verified that removing the oil cap makes a significant difference in the idle speed and character. It eventually compensates for the change to some extent. This seems to suggest I do not have any major vacuum leaks.

I am suspicious of the idle control valve drive. It seems respond too slowly in some circumstances. I just can't understand why the car stalls when cold, or dips to low rpm when warm, if I put the clutch in from ~900-1300 rpm. It smoothly transitions to idle speed if I put the clutch in anywhere outside of that rpm range.

It feels like the idle control valve thinks it needs to slow the engine down in this rpm range and closes all the way. When I disengage the clutch, the engine speed falls off and stalls before the idle control can open up again to restore a normal ldle speed.

How should a properly functioning idle control valve respond to this situation? It seems reasonable the valve would close when the car is coasting between ~900-1300 rpm to try to slow the engine speed back down to idle. Is this correct? Is the valve supposed to instantly open up again when the engine speed begins to fall? If so, could a partial idle control drive cause my problem but still allow normal operation under most circumstances? What else might cause this problem?

I think I may have discovered another clue. The other day I tried connecting a jumper to set the idle control valve to a neutral position before I started the car to see if this would prevent the stalling problem. The car would not start cold with the idle control jumper in place. Is this normal, or should the car be able to start cold with the jumper in place?

Thanks

Chris

[john walker's workshop]

the micro switch on the throttle housing needs to be depressed on closed throttle to tell the computer to activate the ISV. you should hear it click upon slightly opening and closing the throttle in the engine compartment. i've had issues with some chips that caused the ISV to be slow to function or not function at all. i've reinstalled the stock chip and it worked fine. manufacturer to remain anonymous.

[C3larke]

Thanks for the quick reply. I have checked the idle switch to confirm it clicks, and I have checked it on a multimeter to verify it is opening and closing the circuit as it should. I can also hear the engine cut in and out as I open and close the switch by tapping the gas pedal while coasting when the engine is turning at a speed above ~1300 rpm (fuel cut off speed).

It sounds like my chip could be the culprit. I will have to try re-installing the factory chip to see if that corrects the problem. It would be unfortunate if my Steve Wong chip is causing the stalling issue by somehow slowing the idle control response. I will swap the chips over the weekend and will post the results.

Any other simple tests I can try before switching chips?

Thanks,

Chris

[C3larke]

The factory chip did the trick! I reinstalled the factory chip after work yesterday and have driven the car several times since then with the engine hot, cold and in between. I can now operate the clutch at any rpm I choose while coasting to a stop and the engine will not stall, or even dip to a near stall speed.

This problem has been driving me crazy ever since I rebuilt the engine and installed the Steve Wong chip. I put in the chip before I had all the post-rebuild bugs worked out which was probably a mistake in retrospect.

Now that I have become accustomed to the Steve Wong chip, I have noticed a number of differences in the way the car runs on the factory chip. I had to adjust the idle speed back down to 800 of course, which makes the car sound very different at idle. The fuel injection cut off speed is a lot lower too; the factory chip shuts off the injectors at around 900-1000 rpm while the Wong chip does so at 1200-1300. I also noticed the engine pops and crackles a lot more than it did with the performance chip when going through gear changes or other abrupt rpm changes. The idle speed also bobs around a lot more, which I understand is normal for a factory chip. I experienced idle surging followed by a bit of rough running and even jerkiness when I started driving on a cold engine. With the Wong chip the engine would smoothly ramp up to a high idle when cold. The car seems to run perfectly once it warms up on the factory chip though. Lastly, I felt like the power may be down a little with the factory chip in the lower half of the rev range, but it's really hard to make a judgement on this. I think it might be be a good idea to check the CO setting again now that I have altered the idle setting.

I will drive the car for a little while with the factory chip and get a better feel for how it runs. Other than the stalling/dipping RPM issue, I really liked how smoothly and nicely the car was running on the Wong chip.

Thanks again for the help. I never would have suspected switching back to the factory chip could eliminate my stalling problem.

[steely]

Sounds great, but even with the SW chip, you shouldn't have issues.
I would continue to check the ICV, idle sw, mixture, and vac lines - while you have the factory chip in her in order to baseline it.
I have a SW chip an overcame similar problems - effer runs soo good.