jerome951

Well, I think I fixed my cutting-out issue. I've driven the car about 10 times with nary a stumble.

I replaced the main engine wiring harness with another used one and this seems to have done the trick. Now, let's hope it goes at least 1 track season without causing me problems.

So, moral of the story. After you've exhausted all logical options to resolve a problem, look at the illogical ones.

Special thanks to Chris Prack for advice and the wiring harness, and thanks to all those who offered ideas on possible causes.

Later,
Jerome
***************************
Hey Guys,
Hoping someone can shed some light on my engine cutting out problem. I've been battling this for ~2 years and this past weekend was the worse it's every been.

Here's some video from this past weekend where I've focused on the tach. Hopefully this sheds some light.
http://members.rennlist.com/jwelte/engine_cutout.wmv

The car:
'89 951. Stock engine controls (DME, AFM, etc.), setup for club racing

Symptoms:
*Random engine cutout. Sometimes is temporary (only a hesitation or brief cut, sometimes engine dies). When engine dies, warning exclamation (!) light comes on. Sometimes will restart itself if I'm still rolling, sometimes key has to be recycled and it will restart.
*Frequently tach swings when it cuts out but not always, especially when it flat out dies
*Mostly happens under full throttle, but has happened at idle and while cruising
*Twice, the RPMs jumped ~500-600 very quickly (as in instantaneously, much faster than a slipping clutch) while accelerating then dropped back down instantaneously after a second.
*While moving around the main engine wire bundle at the back of the firewall (where it branches off to the engine grounds and speed/ref sensor), I did get the car to die while idling. However, I was unable to replicate so don't know if it was a random occurrence.

What I've tried:
moving around wiring and plugs for DME and KLR
bypassed temp sensor w/ resistor
bypassed kill switch to coil
bypassed alarm
swapped DMEs and KLRs
Tried 2 different AFMs and coils
New cap & rotor
New TPS
new plugs
New speed/ref sensors, verified gap to ring gear, checked continuity from connectors to DME plug
replaced DME relay
Lifted relay/fuse panel, tapped, moved, twisted everything
Laid DME/KLR on floor rather than mounted vertically
replaced fuel filter

What I'm thinking:
* The tach bouncing suggests a problem with either speed/ref signal going to the DME or a proper signal that is not getting processed properly. Since I've tried 2 DMEs, that rules out the latter.
* DME ground could be bad. I have cleaned the ground points at the back of the engine recently but ground wire could be frayed, broken, or corroded.
* FPR could be intermittantly dieing, but doesn't support the jumping tach

Any ideas on where to look next would be appreciated.

Is there a way to tap directly into the DME wiring harness where it plugs into the DME without cutting wires? The connector fit to the DME box is too tight to stick a wire in there. I was thinking of adding some indicator lights to determine where power may be getting cut.

Van

Yes. It's outlined in the service manual, I believe in the section for testing the AFM with a multi-meter. You can slide the cover of the big DME plug back and reach the tops of the pins.

Do you have a fuel pressure gauge you can monitor from the cockpit? That may tell you if a fuel delivery problem is plaguing you.

dillon410021

thats is very weird on how it will just cut out yet when It regains power, it is at a higher rpm!

blown 944

Pull back the covers on the reference sensor connectors (the side that goes to DME) they may be bad inside there. If they are bad Lindsey sells these for ~175 IIRC

Airflite40

it looks like an electrical problem to me, I would start by replacing the speed & reference sensor part of the harness, then go from there. Lindsey sells that kit for $100. good luck!

Tom M'Guinn

+1 on the harness -- inspect/jiggle the speed/ref sensor connectors at the rear of the intake, as well as the fuel injector connectors and harness. By chance do you have any kind of data logging on that car? If so, post up a log of when that happens. Also, check its wiring. I had a very similar problem when the wire used to tap into the RPMs started pulling out of my zeitronix box. It shorted the ignition when it would make contract with ground, causing nearly identical symptoms.

jerome951

Thanks for the ideas, guys. I have tried moving the speed/ref wiring around but only got it to hiccup once. Couldn't reproduce it but am leaning in that direction. I'll visually check the wiring and try testing for continuity again.

Unfortunately no fuel press readout or data logging (too cheap for that). The way the tach jumps around I'm leaning away from a fuel delivery issue.

jerome951

Tom, are you saying the signal exiting the DME to the tach was shorting and causing this problem? I haven't tried wiggling wires under the dash except for those in the passenger footwell but could try those behind the display panel.

Ski

Wires 25 and 26 are from the reference sensor, it almost sounds like ignition cutout. If you can, ohm out those wires, then get wire 27 for the speed sensor. While you're there, check 14, 15, and 16; the first two are FI signals and the last one is the common ground for all four.

Just by chance, you didn't mention anything about plug wires - how old? It's a simple thing, just checking.

Papamurphdog

See my "must read" post from today... while the car is running, pinch the injector boots on the female side, wiggle them, etc., and see if the car dies, burps or whatever. If so, cut open the connector and follow my post... might be the problem, might not be. Worth a try...

shaheed

faulty injector harness can also cause cutouts, those wires take a lot of heat and typically are not in good shape.

Tom M'Guinn

In my case it was the wires exiting the zeitronix unit. Your problem looked just about the same, however, making me wonder if your ignition signal is cutting out. Have you tried wiggling the trigger wires to the coil and/or have you tried a new coil? The way the tach is responding, it seems much more likely to me that you are having a problem on the ignition-triggering side rather than a fuel issue. The most likely source of that type of problem is in the 20 year old harness -- ref sensor connections, coil connections, dme, etc.

jerome951

Rick, thanks for the heads up. I read your post and know how frustrating that was to chase.
I had a few minutes to let the car idle last night and wiggle/pinch the injector connectors and harness. No change.

Tom, I replaced the coil (only 1 year old) with my old coil over the weekend and no change. I moved the wiring near the coil but not further upstream.

I won't have a chance to conduct more diagnosis for a couple weeks (going on vacation in a couple days). I did pick up a code reader (thanks Chris) so will use that next time it cuts out.

Bret: wires are Magnecores. They're about 8 years old but still seem to be in good shape. While idling, I moved them and the coil wire around with no change. Also tried misting water over them and didn't see any leakage. When the engine cuts, it's all cylinders, so unless they all fail at the same time or the coil wire fails, I don't think they're causing the problem.
I did check continuity on all the wires from the speed/ref sensors to the harness connector. I haven't had a chance to probe the FI and ground wires but will when I'm back from vacation.

Thanks everyone for the ideas. I'll test them in a few weeks.

fast951

You did try a different DME, have you tried a different KLR? The Ignition signal goes DME ==> KLR ==> DME ==> Coil.

jerome951

Hey John,

Yes, have tried 2 different KLRs, DMEs, and coils with no change. Have also tried moving/wiggling the KLR connector with no luck.