william_b_noble

73.8 ohms is very suspicious. substitute a 12v turn signal light bulb for the coil, e.g. put the bulb across green and black. If the bulb is on steady, you have found your problem, the drive transistor for the coil would be bad.

note, I haven't checked - one wire has +12, the other goes to the DME, whichever the one that goes to the DME, that's the one I'm talking about that goes to the drive transistor for the coil. certainly with the DME disconnected, it should read as an open circuit, with the DME connected you should see the emitter/collector junction of the transistor, and the transistor should be OFF, and so the resistance should be high.

PerryB

Mr. Noble,

Thanks for your input. Sorry about not getting back with you, I've had a busy work week.

A light between the coil terminal wires doesn't light at all. The coil does get power, but the other wire that goes to DME terminal #1 must not be grounded. I am unwilling to test the ground continuity through the DME as I've heard that even the small current used by an ohm meter will damage the DME.

This is NOT what my oscilloscope showed. When testing the DME output to the KLR my scope showed that the DME was grounding the coil, it just never alternated that ground to fire the coil.

The 73.8 ohms was between the positive wire of the coil (un-hooked from the coil) and ground. I assumed that the was because of some other item that gets it's +12V from the same circuit that gives +12V to everything that goes on when the key is turned to the "on" positision.

Perry

Joe Baumbach

Not sure if this has been suggested.

You can remove the KLR from the equation by disconnecting it and bridging pins 9 and 16 to feed back the ignition signal directly from the DME to the coil (just like the NA 944s do). This is suggested for diagnosis in the factory service manual.

PerryB

Thank you Mr. Baumbch,

Yes, I have tried that. It made no difference. On that same page it also says that the DME should send a square wave signal to the KLR. I tested that using my oscilloscope and the DME didn't produce that signal.

Perry

PerryB

I am still looking at the oil level warning light when I turn the key on and crank it. Is there some lock out for the ignition if the oil level is low?

Perry

VirginiaF1

Hi Perry,
Epic Masterclass Thread.. sorry you are it's subject.

On page 70 of my 1988 owner's manual the Oil Level Warning System is detailed.
When my '88 had a rear cam cork gasket fail, this light illuminated and engine did not crank.
Yours seems to crank but not start?... Fuel pump fail?

Best wishes, Mikey

PerryB

VirginiaF1,

Thanks for your reply. Epic Masterclass Thread, yes. I have been feeling like a rat in a maze with this problem, or at least the subject in some strange psychology experiment on fustration.

My '87 owner's manual does not mention the oil level warning light at all. I have disconnected the wire from the esnding unit on the sump and the dash light stays lit. I also grounded the wire and the light stays on. I would really like some definitive yes or no on whether Porsche ties this warning light into the ignition.

My car cranks strongly, but is getting no spark. I do get gas now, I did not originally. I'm still thinking that it's my computer, even though it's been back to Specialized ECU twice. I have done every possible test that I can think of. The DME still gets all the correct information from the sensors, etc., but refuses to fire the coil.

When I originally had this problem (way back in February!) the car would neither spark nor fire the injectors. Now that I've had Specialized rebuild the DME, I have fuel but still no spark. That fact points to a computer problem. All the harness connections at the DME are good. The last time I sent the computer out to Specialized they told me that they put my computer into a running car and it worked fine.

At this point I either need some other explanation as to what might be causing this problem (other than the zillion thiings that I have tested already) or I need someone from the forum to put my DME into their 951 and film the results. I have already had Specsalot do that test before the lastest return to Specialized. I am hesitant to contact him again and impose on his good nature and his time in this matter. If anyone is willing to help please contact me on the forum or via PM.

I spent some time over this holiday weekend looking over the car for any other possible items to test. There are none that I can think of. I am willing to entertain anyone's suggestions on other possible courses of action.

Perry

Marajit

They did the same to me four times.

Did they promise they would fix your computer?

No.

You're on your own.

Sucker.

https://rennlist.com/forums/924-931-...urus-long.html

VirginiaF1

Yikes..

PerryB

"Yikes"--I agree!

VirginiaF1 and Marajit,

I read all of Marajit's thread. Not good news. I have nothing to add. I talked with the tech at Specialized earlier this summer and they were willing to retest my computer at no cost to me other that shipping it out to them. I should point out that originally they sent me an exchange computer. My original computer served as a core.

Unfortunately, I have no way to test the computer directly, so I have no way to know if Specialized has done a compentent job or if the computer is malfunctioning causing all this disaster. Over the years that I have been working on cars I have tried to do as much work myself as possible. This is partly for financial reasons but also because I know that I care more about my cars than any mechanic. For them, my car is just another job, one of thousands they will do in a lifetime of work. If something needs extra attention or time or even a complete re-do, I can put in that time because I'm not concerned with making a profit everytime the hood gets opened.

If Specialized really is not giving me a good computer I'm going to have to go to one of their competitors, which means I may start this whole process over again with another company, only to trust that THIS time the computer people will do a good job.

I still can't help but think that something I did during the clutch job caused this and if I could just figure out what, I'd solve this problem and get back on the road.

If anyone can think of anything for me to test...

Perry

VirginiaF1

Whereas Specsalot graciously installed your DME, can anyone lend you their's to install into yours and perhaps eliminate it?

PerryB

VirginiaF1,

That would be great, but I would worry that by placing their computer into my car it would be somehow damaged. It is possible that there is something wrong with my car that is damaging the DME. I couldn't guarantee that they would get back their computer in working order.

Perry

Marajit

Perry, I can't for the life of me understand how one can screw up the electronics doing a clutch job.
Gotta ask.
Did you resurface your flywheel?

Did you install it with the pin facing straight down?

PerryB

I broke the speed sensor removing it from the car. I ended up replacing both (speed and reference) as my first step to finding out why I had neither spark nor fuel. I was convinced that the no start had something to do with my replacing these sensors (wrong gap or defective wiring, etc.)

As I got more into it, I located and used an oscilloscope to test the sensors. Both tested good. After running through the complete official Porsche diagnosis flowchart for a no start on these cars, I eliminated every variable except the computer. I then sent out the DME to Specialized with no success. I then sent out the KLR, with the same result. Specialized charged me (I believe) $350 each time. Each time that I have re-examined the problem, everything points back to the computer.

I did have the flywheel resurfaced and balanced with the new pressure plate. The pins were installed. The back of the crank has a roll pin installed so that the flywheel can only go on one way. Also, since the speed and reference sensors both give the correct output, I know that they are "reading" the pins.

I have sent the DME back to Specialized twice for testing. The first time they assured me that the DME was good. The second time they put it into a running 951 and it started and ran correctly. Each time I had to take their word for the validity of their tests, and the honesty of their reporting.

Don't hesitate to ask these questions. I'm hoping that someone will ask something that will solve this problem.

Perry

specsalot

Perry - I'm sorry for not replying sooner. Stopped thinking about working on cars during August heat in FL. Thinking cap back on ☺. To my knowledge there is nothing about the oil level sensor reporting low oil that will prevent a car from starting.

Reading through what is available in various sources, it appears that the crank reference sensor signal is responsible for coil firing and the speed sensor signal (along with other inputs) determines injection quantity.

Both of these sensors have embedded magnets and generate AC voltage when iron / steel interact with magnetic fields of the probes. The speed sensor picks up ring gear teeth. The crank reference sensor detects the passing of a steel pin embedded in the fly wheel.

The reference sensor sits in a sleeve which helps prevent the probe from picking up other metallic interactions. Perhaps the output voltage level of the reference sensor is not falling low enough for the computer to recognize that the fly wheel pin is no longer under the sensor. In general the more narrow the gap, the higher the generated voltage. Perhaps adjusting the gaps further open might change things.

Also, the bracket which holds the sensors is held in place by two bolts. One bolt hole is elongated (this is where adjustments are made) the other hole is where the bracket pivots. The bracket actually pivots on a "tensioning sleeve". Perhaps your "sleeve" is deformed or rusted away. This would upset the relative geometry of the two sensors. For a correct setting of the ring gear clearance, you could have a crank reference sensor gap that is too small. Here again - I'm just grasping at straws. Something that could explain why your car won't run.

You seem to be clearly stating the issue - No square wave output from the DME. The question is why. You seemed to start getting fuel injection after you split the common grounding of the sensor repair harness. I believe your crank reference sensor may be producing voltage all the time because it's being influenced by the ring gear. Or perhaps your DME is just picking up some kind of signal cross talk being induced into the sensor leads between the sensor and the DME unit.

If you want to send me your electronics to try in my car once again, please do.

Best Regards,
Paul

For new readers here is a brief summation with relevant post numbers:

Post 1 No start (no spark, not tach bounce, no fuel injection) after a clutch job Dec 2013. Replaced both speed and reference mark sensors and their harness wiring. DME computer replaced with rebuilt unit. Sensors gapped per recommendations. Sensor voltages measured within range using an oscilloscope and VOM. DME/KLR test plan shows no square wave output from DME to KLR.

Post 33. 39. 40- Electronics seem to self excite without cranking the engine. Coil randomly fires, tach bounces, and injection occurs during key off - key on sequences. Situation also occurs when B+ is supplied directly to injectors. During these self excitation events the DME did send a square wave to the KLR. Window Relay gets hot during this testing.

Post 67 – Electronics run is another vehicle – No Spark – No Tack Bounce – However – Fuel Injection was occurring.

Post 70 Full Electronics (DME + KLR) Sent to Specialized ECU for Diagnostics – Results – Rebuilt DME was damaged by KLR – KLR & DME repaired.

Post 85 – Crank / Reference repair harness re-configured with independent grounds (per OEM wiring).

Post 86 Full Electronics (DME + KLR) sent to Specialize ECU – Electronics bench tested good and were able to start a test car. Electronics re-installed in car. No spark, but fuel injection was taking place.

Post 90 – Speed and Reference sensor tested with oscilloscope – Test same as all previous tests. No square wave being sent from DME to KLR.

Post 107 – B+ circuit to Terminal 15 of coil shows 73.8 ohms to ground resistance with coil disconnected.