Frustrating 959 Problem -> Need suggestions...
 

Okay, let me start by saying that I am aware of the proper diagnosis procedure for the engine, and it's quite involved. I'm hoping to get some ideas on what could be the problem with the system before I start from step zero and check *everything*.

Here's the problem. When we came back from SEMA, I took Tom Woodford (the fellow who helped me write the Engine Rebuild Book) out for a drive in the 959. I drove first, and then I let him drive. Before I let him drive, I noticed that the gas tank level was low, so we filled up with some 91 octane at the local Chevron (this station closed for "maintenance" a few days later). After about 2-3 miles of driving, we got back to the shop, and the car stalled as I was putting it away.

Next time we took the car out, it wouldn't start without feathering the gas, and giving it a little bit of throttle when starting it up. It's been the same way ever since. Here are some more details:

- I ran the tank down to 1/2 full, then filled up again with 91, and then some 100 octane from Willow. I also added some of the Swepco fuel system cleaner that we sell. So far this has had no effect.

- The car ran pretty fast at Willow - the car has plenty of power, particularly at the top end. If there is a power loss at the top end, it's news to me.

- When you let off of the throttle and let the clutch out (pedal in), the engine rpm drops really, really low, and then bounces up to about 900 or so. Occassionally, the car stalls if the idle gets too low, and particularly if the car is cold. This dropping down below normal idle happens 100% of the time since this problem started.

- I have swapped out and replaced the idle control valve (ICV) with 2 other ones, and this has no effect on fixing the problem.

- The fuel injection system is not showing any fault codes. This in particular is odd because I did run the car briefly without the ICV hooked up (it ran worse). I would have thought that this would have tripped a fault code. Maybe it has to run for a certain length of time before the code would have been tripped.

- No obvious vacuum leaks, although I have not yet put my vacuum gauge on there to check (that will probably come tomorrow). Still, I doubt that there are any leaks because the upper operating range seems good, and the boost levels are good as well.

- The car runs very rough at cold idle, and if you get on the throttle, you can hear it missing when you rev it to about 2000 rpm. When warm, the car seems to have lots of power. When the car is warm and idling, the idle seems to miss slightly (much less apparent than when cold).

- Oil cap on tight.

It may be a clogged injector from grime in the bottom of the tank, but that doesn't seem likely to me either. I will replace the fuel filter regardless (ordered one today - it's the same as the 928). I don't really think it's a temperature sensor, because the engine runs rough at warm too.

Any thoughts from outside the box? I have a gut feeling, it's something real simple, instead of something real complex...

-Wayne

I would have thought ISV (ICV) too, had a problem exactly like that with a 964 that was solved with a replacement. But you have already checked that, so I am out!

I have to admire a guy who asks the rest of us for advice on a truely unique machine!

I wish I had an answer.

Cheers

Did you steal my Zeniths? This sounds really familiar :D

(P.S. if you did steal them, you can keep them!)

Not exactly outside the Box since you also may have suspected it…

Wayne, from your description my take is that the engine is running on the lean (will not be noticed at high RPM) side. I would think false air (not metered) may be the culprit…

Spray some ether or carburetor cleaner near the intake manifold ports and see what happens... (?)

Jascha

1989 911 (at last, fully sorted!)
1996 993 (for sale, I think?)
http://mysite.verizon.net/connexin/993_Sale.html

green wire from hell???

reminds me of my 911SC acting up. I took the shotgun approach though. :( Injectors, injector seals, fuel distributor, and even went to msd. Also did the fuel filter, accumulator. It ended being the green wire.

:dunno:

Pure guess, but would the fuel pressure regulator be diverting too much fuel away at high pressure i.e. low rpm?

If it was a flow limitation such as a blocked filter I would expect you to have a misfire or lack of power at high rpm, not at idle. If the pressure regulator was opening too far when the pressure got too high and you weren't developing sufficient fuel pressure in the fuel rails then I can see that it might give you starting and idle problems.

The problem I see with this is that I can't see how you would get normal fuel pressure back again unless the regulator sticks open once open or returns too slowly...

Like I said this is a pure guess as I don't know how the pressure regulator works.

Oxygen sensor?

I would think you should start with the ICV valve and you've removed that as the issue. Next I would try the fuel filter. Since you know it started after addition of new fuel and you've got that station being closed I think you can't rule that out. I agree, it does sound like you've got a lean condition or a fuel issue. Once you've checked with your vac tester you'll be in better shape in regards to determining if that is the problem.

But since it was running fine prior the fuel change, you've either got dirt at the bottom of the tank that has gotten in your fuel system OR bad fuel. You're probably right, it is something simple.

i had a motronic-engined BMW that did something almost identical including temperature sensitivity, surging idle with occasional dying, and i did the shotgun approach over 2 years (including cylinder head, ICV, vacuum lines, wires, plugs, distributor, coil, cold start injector, fuel pressure regulator, O2 sensor, ECU etc) and the problem finally went away when i plugged in a new main relay.

Backdate it to CIS. :D

Seriously - just a guess: spark. Have you run the engine in the dark to see if its throwing blue light from the plugs. Don't know much about the 959 engine.

Well, at the risk of redundancy, that sounds a lot like a vacuum leak.
regards,
Phil

Sounds a bit like what I'm going through with my 86,, my first question is what is the A/F indicate?

sounds a LOT like the vacuum leak that I just fixed on my 911. Actually it was more like lots of little leaks that added up to one huge leak. The biggest leaker was the vacuum line between the heated air vent from the oil tank to the throttle body. The vacuum line at the thermo valve at the back of the throttle body was split at the end, which caused at varying leak, depending on ambient air temperature. When the weather got colder, the leak got worse. Mine ran ok at high rpm too. It was low rpm that was the pits.

Wayne, Some of the items I would look at:
- TPS: If it's not signaling ilde to the DME, the idle algorithm will not kick in.
- Vacuum leak: The engine is getting unmetered air (AFM/MAF only). The 959 may be using a speed density/MAP system, so a vacuum leak may not affect it. However if a MAP sensor is in use, a altitude sensor might be used as well. Either MAP or altitude sensor can cause this issue.
- When cold, the DME enrich the mixture, a faulty eng temp sensor (reading rich) can cause the lean condition when cold.

- A faulty O2 sensor (sending the incorrect data to the DME - Reading a incorrect rich condition here) will cause the DME to lean the mixture. If the car has a test plug (round 5-pin plug), usually there is a procedure to set idle, where 2 pins are shorted. This disables closed loop and the idle control algorithm. If possible, run it in this mode and let us know what you find.

- A clooged injector(s) will cause a lean condition as well. However, it becomes a serious issue under boost (lean AFR).

Hope this helps, feel free to drop me a note if I can be of any help.

John

Typical 959 problem they all do that when driven hard...Just Kidding, good luck finding that gremlin !:rolleyes:

+1 on a vacuum leak or lean when cold condition.

IMO: Wayne . . . as far as knowing what kind of power your 959 is supposed to develop, I think you should try to find someone with a sorted example and drive it as well, just for comparison. I'd be happy to let you borrow mine, but Im in Colorado, and I've got it in storage for the winter season right now.
http://forums.pelicanparts.com/uploa...1196135369.jpg

Nobody has mentioned the distributor yet. It could have had an issue while spooling up the rpm's and either freed up or threw a spring,if it has one. Does the distributor have vacuum advance/retard? If it has vacuum retard, try unplugging the line from the distributor and plugging that and fire it up. Also check timing to spec.

My guess.. water in your fuel. I've had it a few times where a bad tank of fuel will plug up a filter. It was so bad for a while that i was running a racor filter on myt 4 runner and 1 bad load of fuel did plug it up on a 600 milr road trip (the filter was new when I left and I had to replace it duriing the trip).

Try putting in some methyl hydrate (methanol) to dry up any water in the tank.

cylinder head temp sensor??? at least that prevents my 87 911 from starting ;-)

Turbo boost places all vacuum lines and intercooler tubing under pressure, and can pop them off or split the lines, especially if the rubber gets old. I'd make sure all the rubber vacuum lines were fresh, and zip tie them to the nipples for insurance.

Tempting as it is to suspect that the gas is related, coincidences do happen... [edit] Just saw the suggestion about water in the gas - do what pilots do - draw off some gas from the bottom of the tank into a glass bottle and let it sit. Any water contamination should be readily visible.[/edit]

Still, checking the fuel pressure/flow rate at the rail, looking for a clogged injector or two and eyeballing the plugs is cheap insurance and good diagnostic procedure.

It does sound like it might be a vacuum leak possibly improving as it heats up though (they didn't use phenolic injector blocks on the 959 did they?).

Boost would tend to easily overcome a minor leak in the intake, you might never notice the wastegates opening slightly later than usual - so it could seem to run fine as soon as things are spooling (which starts to reduce manifold vacuum on WOT/under load long, long before the factory gauge moves off the stop representing atmospheric pressure).

If this is what is happening, the loss of metered air in what should be a closed system may make it run richer when spooling (like running a BOV instead of a recirc valve does), while the leak would make it lean lower down when cold/not spooled.

A wideband O2 should show precisely what's going on with overall AFR's in a snap (although this won't show up a single cylinder running lean).

Much safer, more convenient (and very likely legal) than a full-throttle run under load in a high gear, followed by an ignition kill/clutch in and a plug chop on a very hot motor.

My Motronic 1983 BMW E21 320i had similar symptoms. It was a vacuum leak (and stuff that needed a head job- valve adjustment, valve guides, valve seats). My way to proceed would be to check for vacuum leaks that would cause lean running at low engine speed- after the fuel metering plate.

I am sure you have heard all this before, though, and I am not an MIT graduate engineer. Hope you find a simple fix.SmileWavy

New Info...
 

Update - New Info!

Okay, I wasn't convinced that the ICV was not the problem, so I swapped it out again, and it still didn't work. So, I sat staring at the engine (trying to find that loose vacuum line), and then I decided to unplug the electrical connection to the ICV. The engine started and idled perfectly fine. Well, almost, the idle was a bit erratic but it started right up.

So, I think that the computer thinks that the car is not at idle. When the car is not at idle, it should completely close the idle valve and only let air through the main throttle body. At startup and at idle, this will cause problems. The only way it would think that it's not at idle, is if if the TPS (throttle position switch) was not indicating the proper values, or the "throttle home" position switch was inoperative. I had to unplug the TPS when I was replacing the oil pressure switch the other day, it's a likely culprit since it's been disturbed.

Any other thoughts besides this?

-Wayne

Wayne,
I would think your first guess that it would be one or more partially clogged injectors, causing poor atomization of the fuel at idle and less noticable at higher speeds. If it was running well before the new tank of gas the problem most likely would be fuel quality related, and any crud that got in there will stay in there. Try a shot of carb spray while it's idling and see if it smooths out.

Don

Wayne,
I would think your first guess that it would be one or more partially clogged injectors, causing poor atomization of the fuel at idle and less noticable at higher speeds. If it was running well before the new tank of gas the problem most likely would be fuel quality related, and any crud that got in there will stay in there. Try a shot of carb cleaner spray while it's idling and see if it smooths out.

Don

I watched Scott help Chris diagnose a similar problem with Chris's 3.2. They tried unplugging the ICV and it helped but the solution turned out to be unplugging the o2 sensor. The engine ran fine with the ICV plugged in as long as the faulty o2 sensor was unplugged. I'm not sure what was wrong with the o2 sensor but it would be easy to try out unplugging it. The ECU doesn't look at the o2 sensor at full throttle so that may explain why it pulls nice and hard but can't idle or run well at low rpm, ie during closed loop operation.

I wonder if bad fuel could have ruined your o2 sensor.

Interesting. This car is not supposed to have an O2 sensor, but the coding in the computer has been modified and adapted to include it as per California smog rules. I don't even know what O2 sensor is in there - I think I'll have to call the guys who did the conversion to find out. It's probably some generic one (they never made a "959 O2 sensor"), but it's not as easy as looking it up on PET.

Good thought though - worth taking a closer look at the O2 sensor...

-Wayne

GIAC did the programming for the new code in the 959 CPU in or around 2002. I just read this on their website:

This would seem to indicate that it doesn't read the O2 sensor when the car is cold. So, erratic and non-operative engine problems when the car is cold might not be caused by a faulty O2 sensor?

-Wayne

All cars with non-heated o2 sensors ignore that part of the engine management system until the temp sensor (oil/air/ or otherwise) indicates that everything is up to operating temperature. GIAC activated that part of the ECU for the US conversion but just because it's ignored doesn't mean it can't cause problems outside of it's spec'd operating capabilities.

Faulty o2 sensors can cause a lot of weird problems, kind of like the head temp sensors on 3.2s. A 964 won't even start with a shorted o2 sensor. If you use the general application Bosch sensor instead of the Porsche part it doesn't have a long enough boot at the sensor end. So it works great as long as you don't drive through water or in the rain, as soon as the thing gets wet the car won't start.

Hopefully it's something as simple as unplugging the o2 sensor for testing! I assume it's a two wire sensor like a 3.2?

No clue. It's a non-standard part, so I have no idea. I know roughly where it should be located, but that's about it.

Tomorrow, I will test the TPS and also try unplugging the O2 sensor to see if that solves anything. It's a bit like shooting in the dark, when the car has been modified from original form. Even in it's original form, there's not too much information on it. Perhaps I will give those guys a call at GIAC to see what sensor they used.

-Wayne

Yeah they are right down the 405 from you. I would think they just used a Bosch two wire narrow band sensor but I'd be curious to see if they used a Porsche harness. Hopefully with the limited nature of the cars modified they went ahead and used OEM connectors as well as sensors.

does the closed throttle micro-switch make and break properly? rough run cold can also be too rich as well as too lean.

I'll second the moisture theory. I chased it with fuel additive and then new gas. I finally drained the gas, replaced the fuel filter and added yet more fuel additive(to remove any remaining moisture) Something about the specific gravity of water vs fuel it winds up in odd places and won't "move on".

Have you thought of asking the Dept of Agriculture (which governs gas pumps in my state at least) for some info on the station in question? ie: Was there something in the gas that caused the closure?

Good question, I can hear it clicking, but I have not tested the actual electrical connection. I did not have my Fluke with me today...

-Wayne

That makes sense. I had a similar problem on my 964 and found that a new throttle cable was not adjusted properly and thus not allowing the TPS to be activated.

Sounds like you are on it.

i am going to sound like a noob for suggesting this but my 964 have been doing about the same things for a while and i try to look at everything and nothing help (clean isv,etc) . it's stall sometime, low idel is rough but top end power is fine.
It was a clog up air filter. i have k&n that wasn't clean for a while, lot's of dust around here. Clean it and every thing was A ok again. car ran great at low idle and very steady idle.
could it be something this simple?
cheers

I would make sure that the fuel filter is changed and completely drain the fuel system. It might be a good idea to back flush the old filter just to see if anything is in there. I would also inspect the plugs to see what they look like, while your at it you might just want to replace them. Make sure you have good fresh fuel in there. I think this will give you a clean slate for diagnosis. Also if the car has a tps make sure it is reading the correct resistance in the closed position. Just my two cents, good luck. Also check to see if you have power at the ICV and the correct voltage.

It almost sounds a little bit like the problem some early 944s used to have off throttle - RPM was known to drop WAY down when throttle was released and then "spike" back up which made the car seem a bit erratic. Turns out it was related to the fuel-shut-off subroutine within the DME (It didn't react quickly enough to how quickly RPM would fall off with the throttle closed and in neutral, so RPM would pass the target RPM of 900 before the system caught up with it).

The "fix-it" procedure to correct simply is to set the TPS position slightly above the shut-off position so that a closed signal is never sent to the DME. I realize this is a bit of a "fake" solution but short or re-mapping chips it's the only way I'm aware of to remedy the problem. This affected a number of 944s - I thought that it was remedied in the later models (about the same era as the 959) so I don't know why it would be a problem. Maybe your crank has been knife-edged or flywheel lightened resulting in quicker-than-expected RPM drop that the DME can't keep up with (?)

Idle problems on turbo cars are an absolute pain to track down - I was still fighting with a cold idle problem on my 951 before it blew up - and this was after two different ISVs. It's worth noting that the ISVs themselves can develop leaks at the electrical harness connection. There's a good procedure for rebuilding them on the 951 forum - I'll see if I can find it for you. I suspect the 959 uses the same ISV (?) or similar.

+1

Since you suspect the fuel may have been "bad," the filter is where I would go next.

Ask the gas station what happened. They might say, bad gas, water in the the system, it was scheduled maintenance for a long time.....It might shed some light on it. Personally I would drain all the gas, change the fuel filter, add a fuel line conditioner to the first new tank. Just to be safe.