loss of power at 3-4K rpms; stalled out; no codes

[ericinboca]

Will see what else it does Sunday.

It shut off on Sunday as a session was coming to an end, on entry to turn 17 at Sebring. Fortunately, the entrance to pit road is at the apex of 17, so I made it into the pits, but it would not refire. I left it sitting for a good 1.5 hours, and it still would not start. It would turn over, but not catch.

We loaded it on the trailer, and after 2.5 hour drive home, it started up like nothing was wrong. No codes, not check engine light.

I leave tomorrow morning for two weeks, but I am thinking about removing the DME and having my wife take it to a place that can bench test it...I don't know it is the DME, but it is a chance, and I will only have a week to get it fixed after coming back - it is at least something I can get fixed or rule out while I am gone.

Any thoughts?

[Topless]

Yes. Failed DME or O2 sensor to DME harness and connectors. This is where your error codes are leading. Good luck!

[ericinboca]

Yes. Failed DME or O2 sensor to DME harness and connectors. This is where your error codes are leading. Good luck!

update

Last night, when I removed the DME, I dropped a nut into one of those places no human hands can fit. I had to pull back the trunk liners, etc. to find it, and found significant heat damage under the carpet near the DME. The plastic linings and the foam pieces that line the floor of the trunk are: missing, melted, or brittle. There is even a large hole, which I suspect is supposed to be filled or part of the mount point for the thick foam insulation, that is wide open, allowing heat in.

I now remember at Sebring, in December, a bracket broke off on the top of the race muffler. In the heat (pun intended) of the battle, like an idiot, I kept going. I remember finding the heavy plastic on the left side that lines the heat shield by the rear bumber melted - amazing so. I cleaned all that up, but did not have the presence of mind to look and see if anything else had melted. The problem started after that event.

I just talked to the DME guy. He is going to try and bench test it using both vibration and heat as intentional variables to mimic track use. He says excessive heat can screw it up, but puts his money on a software problem over hardware. He says he has seen several 986 DMEs "lose" their software. Hoping I will know something before end of week. If not the DME, then I will crawl all over the motor tracing wires.

[Topless]

Bingo!

[ericinboca]

Update

Considering the amount of heat damage I found in the rear trunk, I felt pretty good about having found the cause of the problem, and when the DME doctor reported the DME passed, but was suspect, I told him to go ahead and burn me a new one.

I know it sounds like a dumb move, but everything added up and pointed to a faulty DME. (I'll post a photo later of the heat damage.)

I installed the new DME today. It fired up and idled. As it was idling, I had Durametric connected, and, while everything was not completely hot yet, I got valve lift control errors in both banks (P0026 & 28), plus my old nemesis P0050, 02 sensor.

I would clear them all, but they kept coming back. Shortly after the car idled down in revs a notch, it shut off, and would not/will not refire. Just like it did when it expired a couple of weekends ago at the track.

I have crawled all over the motor - tracing the wiring harnesses. I've been all up under it (fortunately, I have a lift in my garage). I took the air distributor off to have a better view. I checked all the harness bundles in the rear trunk.

I cannot find a thing out of the ordinary on the harnesses. No chew marks, no fraying, no melting, no loose connections, no pinching, nothing. I can't even find the factory wrap coming undone on any of it.

I checked the voltage on the valve solenoid - it is within spec.

I checked the resistence across the pins on the solenoid - it is within spec.

I can't figure out how to check the resistance on the triggering wire, though. See this post.

http://www.renntech.org/forums/index.php?showtopic=26802

I have a brand new brain, with the orginal/same software that came from the factory on it. Can't see anything wrong in the harnessing. Unless I take the next step of simply replacing all the wiring harnesses out of the DME, I'm back to stumped.

[Topless]

Eric,

I feel for you man. These are the troubleshooting issues that make my brain hurt. I do think you are very very close. All indications point to DME/sensor communication issues. You said you had your tech burn in fresh DME software? What about the DME hardware? Those little brains don't care for heat much. If the problem is heat to the harness it will usually fail at the connectors and not inside the bundle. Harness chaff or rodent damage is usually easy to spot. I think you have ruled everything else out. What is still missing???

[ericinboca]

Eric,

I feel for you man. These are the troubleshooting issues that make my brain hurt. I do think you are very very close. All indications point to DME/sensor communication issues. You said you had your tech burn in fresh DME software? What about the DME hardware? Those little brains don't care for heat much. If the problem is heat to the harness it will usually fail at the connectors and not inside the bundle. Harness chaff or rodent damage is usually easy to spot. I think you have ruled everything else out. What is still missing???

New hardware too. So, I think we can eliminate the DME as a problem.

I can't find any visible damage to any of the wiring.

Are you suggesting the heat can damage the connector modules? the part that plugs into the brain? I don't see any damage to them, but maybe something inside of them is out of whack. I think the only thing to do with those is replace the full harness.

I'm confused by the new codes popping up, and the diagnostics on them, and how they may tie into all of the this. The solenoid valve passes the test, and the voltage reading to the valve is fine too. I get lost, though, on checking the resistance between III/1 and the solenoid valve pin. I read that to mean pin 1 on the third plug module, and valve connector pin 2. But there is nothing in pin 1 on the third plug module.

And I can't find the connector X59. I found in some of my manuals a reference to it being a white 21 pin connector located at the left front side of the engine compartment, but I sure don't see anything like it. And when I trace the wire back from the solenoid connector, it simply goes into a bigger wire bundle.

[ericinboca]

The heat damage.

http://photos.streetfire.net/photo/IMG2280jpg_1690766.htm

[Topless]

If you eliminate the DME the only thing left is heat damage to the harness or DME power source. This is very likely considering what was going on inside your trunk. Any other possibilities? The process of elimination works great as long as we don't overlook anything along the way.

[ericinboca]

If you eliminate the DME the only thing left is heat damage to the harness or DME power source. This is very likely considering what was going on inside your trunk. Any other possibilities? The process of elimination works great as long as we don't overlook anything along the way.

anything you can tell me about a discreet power source to the DME??

Figured out connector X59 - it is the white 21 pin connector in the trunk next to the black one and the pink/purple one.

[ericinboca]

Some progress - I got it to start and idle, AND I am no longer getting 02 sensors codes.

I spent hours looking at, feeling and tracing the wiring bundles. Since I had all the heat damage in the rear trunk, I spent a lot of time back there, unwrapping them, tracing individual wires, disconnecting and inspecting the X connectors, etc.

I did find the wires that control the tail lights, the license plate light, the rear spoiler motors and the rear trunk latch to be damaged. I don't think these were shorting or grounding anything out, but I did finally find something I could fix. There was no change, BTW, after replacing this damage.

After cleaning and inspecting connectors for the umpteenth time, I tried to start again. It turned over and acted like it wanted to run, but would shut down like it was not getting enough fuel. So I gave it some accelerator when it first fired up - it "misfired" through it and caught and I held it at 1.8K to 2K RPMs for a bit, and once warm, it idled fine. I also shut it down a couple of times, and it would refire fine. We'll see if it refires once cool.

I let it run with Durametric connected. Like I said above, no 02 sensor codes. And using actual values, I can see the voltage on both sensors now fluctuating properly. Only change has been the new DME. I think, knock on wood, one issue is solved.

BUT, and it is a big but, I still get the hydraulic solenoid valve lifter errors for both banks; an error I was not getting before the new DME. I clear it, and it comes back. I am also getting a PSM/ABS error, saying "5281 version coding."

[ericinboca]

More progress, I think.

At last report, you will recall I had starting and idling problems, a PSM/ABS failure and hydraulic solenoid valve errors in both banks.

I spent most of yesterday checking continuity of wires, checking connectors, looking at wire bundles, etc. Again. I still did not find anything.

In searching renntech on the ABS/PSM failure, I ran across the possibility of a bad MAF. Using Mike Focke's trick of trying to run the car with the MAF disconnected, it started up immediately. With the MAF connected, it struggles to start and idle.

So, my conclusion is the starting, idling problem till warm and the ABS/PSM are MAF related. I'm hoping Tuesday or Wednesday to have a new MAF in hand.

I still get, though, P0026 and P0028 errors in both banks. I'm unsure is I am checking the right thing.

[ericinboca]

Update

This is nearly a closed issue, hopefully it will be completely closed after a track test on June 6.

We have concluded the DME was "going south", and failing as it heated up. The hotter it got, the worse it got. A failing DME caused the power loss issues, and since the very thing that is supposed to tell you what is wrong, was itself in bad shape, we never got codes. This was an educated guess, made for all the right reasons, and while I cannot confirm it was the root of the original problems until the car is back at the track, it all adds up.

So, we replaced the DME, and the new box was flashed/created with the software from the old box. But, for some reason, that did not work correctly, much to the puzzlement of the DME guy and the enormously knowledgeable local indie (who are separate entities but work togehter - the DME guy does the DME work for the indie).

The car would start, it would run, but we got PSM codes and valve lifter codes, when in fact we could find nothing wrong with either. Further, a hook up to PIWIS wouldn't show critical DME info that should have been there.

A couple of calls from the indie to his contacts at Porsche got some key info - i think he called them I pass codes - and with his PIWIS, he and the DME guy where able to rebuild the sofware on the box.

All looks good right now. Fingers crossed.