964 idle gremlins

[iamtheari]

My car recently (about 100 miles post oil change, so over-filling is a possibility although I was careful to let the car warm up enough to move the oil level gauge before I topped it up and I topped it up to about 10:00 on the gauge) developed a noticeable idle gremlin. It's a 1990 C4. The gremlin first appeared while driving home on a cool evening at highway speeds. When I slowed down for the first stop light back into town, it sputtered a bit. It died at the fourth stop or so. I managed to get it home and into the garage and let it sit for a while. The other day, I started it up and it started fine but surged from almost-dead RPM to 1000 as if it was seeking an idle and kept missing it. Barely pressing the gas pedal kills the engine. Opening the throttle a bit more aggressively revs it up just fine but when you let go it drops back down and may or may not actually die.

I blasted the ISV/ICV (different sources call it different names so I put both in here for future people searching) with carb cleaner, let it soak in, blasted it again and scrubbed the visible portions on both in and out sides of it with Q-tips, then blasted it again and let it sit for a couple hours. I put it back on the car, and the symptoms were less but still present. I did not unplug the battery when I did this and I know that doing so will reset the DME which can help. Nor did I take the car for a drive, just let it idle for a minute or two to see how much it surged. It actually didn't much at all for a bit. More between 850 and 900 RPM rather than the broad surges from before.

I swapped the DME relay, no change.

I am halfway through the fuel filter replacement but am giving it a rest for the night because I am at that stage of frustrated where a person starts making costly mistakes on account of the input side of the filter being extremely stubborn. Fresh hands and eyes tomorrow evening should solve that and then the new filter can go on and we'll see if that helps.

But while I'm mucking around, what can I inspect, test, clean, kick, experiment on, swear at, and so forth that might also relate to the idle-seeking? I had planned to do plugs, wires, rotors, and coils over the winter while also plugging a few oil leaks. Maybe I should order those now and get them on the car before I head down the road for a DE weekend later this month.

But before I order anything, should I add anything else to the list? New ICV? (Is there a good way to test the old one short of hooking up test leads to the car battery and seeing it if moves? That method seems potentially hazardous to the ICV itself.)

[porterdog]

I've had a long string of VWs and then Audis and now the P-car; can't recall of an ISV ever having gone truly bad. Not to say it couldn't happen, but it wouldn't be my first guess.

An intake leak of some sort could do that- unmetered air is never your friend.

There was/is a TSB for the AFM (airflow meter) wiring. Might be worth a look.

(not my) TSB reference.

[iamtheari]

Thanks for the response. Do you have any good pointers on helping track down intake leaks? I've never had to hunt those down. Same goes for vacuum leaks.

[iamtheari]

Update: New fuel filter, no effect. Glad I did it, though. Some herculean jerk really put that old one in there. And in my car, for some reason, the power steering hose arches upward from the pump side fitting and forward, which puts it directly in line with the fuel filter input fitting. I almost removed the power steering system writ large out of sheer frustration.

The car did idle in a relatively stable manner for a minute or two but then started to lose stability and seek an idle again like before. No telling if that was a DME reset, fuel filter, or just sitting for a bit.

I drove the car around 4,000 rpm up the road a few miles and back. By the time I left my driveway, the car wouldn't idle at all. Instantly dies when letting off the throttle.

Getting home, I noticed the oil level was at the top of the band when I slowed the engine (won't idle, of course) on my level garage floor. Not past the top, but at it. Dipstick shows oil up to the twist, which is slightly above the high-level notch. Oil overfill is a real possibility, after all.

ICV/ISV is now sitting and soaking up another carb cleaner bath. I tested it by applying 12V and it does pop open without delay and close under spring pressure just as smoothly.

Are there other parts I should blast out with carb cleaner as long as the car is sitting and cooling off before the next attempt?

If anyone needs me, I'll be trying to figure out how to gracefully remove a pint of oil.

[porterdog]

A syringe and long tube will work for the oil removal; do it from the tank right after shutdown. You can get the syringes from McMaster-Carr.

The air leak is tougher. The best way by far is to contrive a way to pressurize the intake- any leaks will hiss. Another method is to spray some type of flammable solvent (or propane from an unlit torch) onto the areas prone to leakage and listen for an idle change.

[iamtheari]

Thanks for the tips. Maybe starting ether? Does anyone have a diagram or at least list of places likely to leak? Hopefully I can even get the car to idle long enough to test that method.

[iamtheari]

One more update: Closed throttle idle switch works properly, registering about 6 ohms at closed throttle and open circuit when throttle is moved. WOT switch also works properly, registering open circuit until pedal nears the floor and then about 6 ohms.

Some old posts say no carb cleaner in the mass air flow sensor. Others say to clean it with carb cleaner. Streather's enthusiast's companion book (p. 184) points out that the throttle housing can be contaminated with oil if overfilled. He specifically states that "This contamination may cause unstable idle ('hunting')." He doesn't say how to clean the contamination out.

My money at this point is on the throttle body or its neighbors having oil in them. That at least seems like a reasonable next step to try, assuming I can figure out the proper, safe way to clean that part of the intake.

I will say that I didn't think I overfilled the oil in part because I never had a cloud of smoke chasing me around after the oil change as most people say is how you know you have oil getting into the intake system. Evidently smoke from the exhaust is not a given.

[Shamus911]

Try disconnecting the O2 sensor wire in the engine bay, if the idle improves you'll know that the O2 is affecting the situation. The sensor is out of the loop until it warms up a bit and that's when this issue seems to manifest.

You can also disconnect the electrical connection on the ISV once the car warms and see if the idle improves, which would then point to a faulty ISV.

Sounds like you definitely have an overfill situation - so I'd start with the air filter and clean it all, right down to the intake ports - there will probably be a fair amount of oil sitting in the throttle body.

These cars are sometimes real buggers when it comes to oil overfilling!

[iamtheari]

How much disassembly am I looking at to get the intake cleaned? I'm not particularly excited about removing the entire intake manifold, but if I have to do so then I guess I'd better get started. Not sure if there is a DIY article out there for this. Any hints?

Disconnecting the O2 sensor looks like it requires the heater blower removed first.

I have the airflow meter off the car right now and, assuming I can find the right stuff to seal it back up in my tiny little town, I will remove the cover and clean the electrical portion of it with contact cleaner (fortunately there is some crossover between my amp-building hobby and my Porsche). I'm still not sure on using carb cleaner for the mechanical side of the AFM as I have seen some people say it's fine and others caution against it.

There was a minimal film of oil in the next part of the intake after the airflow meter. I'm not sure if it's enough to cause issues. The air filter (my car has a cone filter of unknown brand from a previous owner--replacing that with a known brand is another goal at some point in the near future) didn't appear to have any oil in it.

Today's plan:

1. Get longer hose for oil removal, silicone gasket, and throttle body cleaner from the hardware store and/or FLAPS
2. Clean AFM and what I can reach of the throttle body without removing more of the intake
3. Try to remove oil while cold since warming up the car to the point of opening the thermostat is a major chore when it refuses to idle
4. Try running the car to confirm no CEL and continued idle troubles (establish control for experiment)
5. Get O2 sensor disconnected and run the car to see if idle gremlins go away
6. Possibly order an O2 sensor from Pelican and hope it arrives soon; possibly buy a case of beer and start disassembling the entire darn intake on my car

Suggestions are still welcome. So far they've been very helpful and appreciated.

[iamtheari]

Update: Found silicone gasket maker, forgot where I put contact cleaner so I did not do that step. But I did clean the throttle body the best I could without removing it from the car, cleaned the mechanical side of the AFM, and reassembled the car. Then I spent a lot of time with a turkey baster and length of tubing to remove 1 pint of oil through the fill neck. Started the car and let it idle up, which it did rather well. It idled long enough to warm up and move the oil level gauge, which came to just about exactly 9:00. If I feel ambitious, I'll remove another cup of oil next weekend for good measure. The car reacted properly to throttle inputs both small and large, when cold and when warm. I will call this a success for the time being.

Next time I order from Pelican, I'm going to get a new O2 sensor as preventive maintenance. And I will never overfill the oil again. What a pain! Thanks for all the help. I probably would have eventually reached "overfilled oil can do this all on its own" but probably not within the weekend or without ordering a few hundred dollars worth of panic-driven parts purchases.

[gordon52]

Had similar issues, change out the fuel pump relay.

[iamtheari]

I think the official list of most likely causes of rough, surging, and dying idle in the 964 engine goes something like this, in order of the cost (in combined dollars and hassle) to eliminate each as the potential cause:

1. Failed DME relay
2. Failed fuel pump relay
3. Dirty or failed ISV/ICV
4. Dirty or failed AFM
5. Fouled throttle body
6. Clogged fuel filter
7. Spark plugs
8. Distributor and coil issues
9. Cylinder head temperature sensor
10. O2 sensor
11. Spark plug wires
12. Failed fuel pump

Fortunately, I was able to stop at #5 (I had a fuel filter on hand and was due to change it anyhow so I did test that earlier). And of course items #3 through #5 for me were all likely contributing causes and it all stemmed from an oil overfill.

[nickyv]

I have a 1990 964. Most definitely not a car guy, but thought I would throw this out to you. I had the exact same issue with idle surge. At times I would be at a stop and it would just die. I would wait a while (it would vary, 5 minutes to 45 minutes) and it would start up again. I went to the Porsche forum and the guy recommended changing out the battery cable. I inspected it and it looked fine so I didn't do it. Finally, I caved in and ordered one from Pelican. Installed it and problem solved. Haven't had a stall or an idle surge since. He had warned me not to be fooled by how the cable , but I didn't listen. Wish I had done it sooner. Hope this works for you as well.

[nickyv]

I should have reviewed the post.

I meant to say "He warned me not to be fooled by how the battery cable looked".

[iamtheari]

Fortunately, my problem was resolved without spending a penny as it was simple fouling of the idle control valve, airflow meter, and throttle body which I was able to clean and, by pulling out about a pint of oil, avoid recurring problems so far. But the battery cable is a good one to note that I did not see other suggestions of, so it's good to note it here for future people lucky enough to have the same symptoms you and I did and search the forum for ideas.

[bazar01]

Originally Posted by iamtheari

Fortunately, my problem was resolved without spending a penny as it was simple fouling of the idle control valve, airflow meter, and throttle body which I was able to clean and, by pulling out about a pint of oil, avoid recurring problems so far. But the battery cable is a good one to note that I did not see other suggestions of, so it's good to note it here for future people lucky enough to have the same symptoms you and I did and search the forum for ideas.

I keep my oil level between the lowest line and 8 o'clock.

When I got my 1993, the lowest section of the intake plenum was full of oil dripping on top of the shroud. It used to have idle surges from 800 to 1000 rpm at stop lights. When I dropped the engine for clutch change, I pulled the intake out and cleaned it really well including the TB and AFM. It was full of oil.
Now my idle is rock steady.

[Liam964]

Thanks for these great instructions - after a recent oil change, and some spirited driving, I was able to resolve my non-starter. Fouled AFM sensor. Bought a DME relay initially, which did not help... So now I have a spare.

In the process sheared off the rubber bushings connecting the airbox to the intake manifold & also the vacuum connector for the resonance valve... All the signs of an amateur mechanic, but learned a lot about the car in the mean time!