944S: Cuts out when very cold (a tough one) xpost

gtroth · 12-06-2005, 01:35 PM

Searching for new ideas. This is a three year old, seasonal, intermittent problem for me that has returned today -

Car runs just about perfectly. But a few times when very cold (usually when <10°F, but a couple of times when only ~20°F like this morning), it will sometimes cut out for a second when under load, refusing to push past, say, 3500rpm, and dipping to, say, 2800rpm. When it comes back it runs great. But it will cut out every 10 seconds or so. Doesn't stop doing this when car warms up. Pull it out of gear with the clutch disengaged, and it will rev beautifully as high as I want. Once (today) got off the highway and at the ramp stop (in gear, clutch pedal in) the idle bounced pretty chaotically (dumping fuel to prevent stalling? Maybe pointing to ignition, not fuel?). This afternoon (~32°F) no sign of trouble.

Known good – FPR, TPS, ISV, DME relay, vacuum lines under the manifold, plugs, cap, rotor, wires, fuel filter, injectors WitchHuntered, clean grounds that I was able to locate (including the two at the back of the engine).

Ref sensor waveform looks great on the oscilloscope, AFM voltage vs opening closely matches what FRWilk has on his site, NTC I and II read consistent with what’s in FSM.

So, not a lot left, is there? Ideas?

AaronM · 12-06-2005, 01:57 PM

O2 sensor? That can cause a very rough cold idle and some hesitation issues.

KY944S · 12-06-2005, 02:48 PM

Could be temp sensor in AFM?

gtroth · 12-06-2005, 03:02 PM

Yes, that temp sensor could be intermittent (that's NTC I, which had about the right resistance when I measured it). Can its failure cause such a dramatic effect, I wonder (same for O2, for that matter)? I am thinking of forcing NTC I failures (open and short, with a switch, see if it's even possible).

KY944S · 12-10-2005, 06:08 AM

any results from your test?

AaronM · 12-10-2005, 02:19 PM

Quote:

Originally posted by gtroth
Yes, that temp sensor could be intermittent (that's NTC I, which had about the right resistance when I measured it). Can its failure cause such a dramatic effect, I wonder (same for O2, for that matter)? I am thinking of forcing NTC I failures (open and short, with a switch, see if it's even possible).

O2 sensor can make a hell of a difference. At least, it sure did on my 951. It went from stuttering/stumbling/no power when cold to running just like any other time when cold. With the failed O2 sensor the engine was running lean and it wasn't enjoying it at all.

gtroth · 12-14-2005, 04:50 PM

KY944S - that test is going to take more attention than I have for it right now. Plus, the car's been running perfectly in the 17°F~23°F range the last few days. I did find the time to replace the coil (there's a guy over on rennlist whose S2 behaves somewhat like mine, and a bunch of guys sugggested the coil...). That didn't solve it for me, though. Today (7.9°F) I did feel a little bit of stuttering when under load. It went away, which it doesn't usually do. The car ran pretty well right after start-up and only began its thing after driving it for a few minnutes, which is typical (engine warm-ish, air very cold).

AaronM - O2 sensor over-leaning might me a good explanation, actually (it's the original sensor). And it has a warm-up time. Car smells a bit rich, however, when it's not too cold out.