[16v] Cold hesitation

[archive@porknet.fdns.net]

I've been having this annoying cold running as long as I've owned the car.
What happens is after cranking once or twice and having the motor finally
start, it revs from 1000 to 1300 slowly in about 1 second then dips down
slowly in about 2 seconds to 1000 rpm. After idling somewhat roughly at
1000 rpm for about 4 or 5 seconds, it drops to 850 - 900 rpm and then
begins to stumble and shake. If I step on the gas, the rpms drop. As soon
as I let off the gas, they climb again to about 1200 then drop to 900
again within 1 second.

This continues for at least 3 or 4 minutes if i just let it sit without
driving. If I try and drive it, I have to rev the engine to 3000 rpm and
feather the clutch just to get it moving without killing the engine...
then as soon as I let the clutch all the way out it will dive back down to
1000 rpm and shudder, slowly moving the car. After a few seconds
(sometimes as many as 10 seconds) of holding down the gas with no change
in rpms, it will catch slowly and creep up to about 1500 rpm where I will
finally get some power and be able to go. Very annoying and embarassing.

As soon as I get to the first stop sign and take my foot off the gas,
though, the engine will drop to 500 rpm and die within a second. After I
restart the engine, I can usually pull away from the stop slowly within 2
or 3 seconds and after the engine temp needle has moved "at all", the car
is warm enough that the problem is pretty much gone.

The coolant temperature sensor checks out okay (correct
resistance/temperature values), the idle stabilizer buzzes and the
voltages at the harness for it are correct. I recently cleaned it out with
carb cleaner. The air flow sensor potentiometer appears to be reading
correct values as well. I did almost all the CIS-E electrical tests at the
25 pin brain connector, they all passed.

I'm stumped, though I have a new hypothesis. Not long ago I was trying to
test injector spray patterns. I had the fuel pump running and tried to
lift the air flow sensor plate with a small magnetic tool. It budged about
1 - 1.5 mm then the magnetic tool pulled off as if it wasn't strong enough
to lift the plate. But this thing is pretty strong - I thought for a while
that maybe the plate doesn't go any higher, but it seemed crazy to have a
big "V" shaped opening if the plate didnt go up that high.

Yesterday I had things apart and I decided to try a bigger magnet. Sure
enough, the plate did go all the way up to the top. But there was some
resistance after the first 1 - 1.5 mm. After that initial binding, though,
it moved smoothly all the way up to the top. There was no restriction on
the way down. But there remained this "bump" after the first little bit of
movement. I put it all back together and didn't think much of it. But
today I began to wonder if this isn't causing some of my problems.

If I step on the gas and the air flow plate doesn't go up because its
binding, vacuum will still pull more air around the plate and the mixture
will be lean and rpms will drop. As soon as I let off the gas the mixture
will return to normal and the rpm will go back up. And if I get enough
vacuum to get the plate up past the place where its binding, the engine
revs fine (which is the case - it revs fine if i get it up above about
2000 rpm).

Maybe after warming up the engine bay and working the plate up and down a
few times to loosen up old lube, the plate starts moving freely again and
my problems go away undil the engine cools and the grease solidifies for
the next cold start. Even if that wasn't the case, perhaps either the O2
would correct for it (but it might be too slow for that) after warming up
(there does still seem to be a very slight hesitation and drop in rpms
after warmed up), or the plate frees itself sooner and mixture returns to
normal.

I know this sounds pretty unlikely, but the more I think about it and my
other running problems, the more I think it might be possible. Hopefully
someone can tell me if their plate also has this resistance or if I have a
problem here.

Input, please!

Thanks,

-Toby

--
'87 16v

--
Email problems to: scirocco-l-probs@scirocco.org  To unsubscibe send
"unsubscribe scirocco-l" in the message to majordomo@scirocco.org