specsalot

My 951 has had an intermittent problem for some time where it stalls while idling in traffic. Wigglying the connector wires for the reference mark and speed sensor was the cure for some time. Most times the car would restart and run without having to wiggle the wires. Ultimately, I cut back the DME harness and installed new connectors. This seemed to help for a while.

Then I got stranded in the car. It would start, but only idle. Any attempt to raise RPM would result in the car cutting off at about 2000 RPM. Never behaved like that before. So I decided to "kill the problem" by replacing what was probably (reference mark / speed sensor issues) a wiring issue. I have the full shop manual for this car and wiring diagrams. I've worked through a diagnostic process without luck so far, this is what I've done:

A. I checked, cleaned and tightened the ground points which give issue (MP VI, VIII, IX and X). I also checked the electronic connectors referenced in the manual.

B. I initially read an open circuit through the crank reference sensor. So I decided to bite the wiring bullet. Purchased / retrofitted (3) harness repair kits from Lindsey Racing. Rational - I had the intake off several years ago and re-lagged the wiring harness. Every connector boot I ever pulled back found wires with flaking insulation. I decided it was time to "kill the wiring issues" once and for all. I ordered and install the 3 kits shown below.



C. I did numerous point to point checks across the newly installed wiring to verify pathways between reference mark and speed sensors and the Motronic DME plug [pins 25 & 26 for mark reference sensor] [pins 8 & 27 for speed sensor]

D. I measured resistances of both sensors directly through the wiring harness and found both to be ~ 945 ohms. I also checked and verified that neither signal path showed any grounding.

E. I've checked, cleaned, retightened the key ground points on the engine block and flywheel housing.

F. I did the clutch several years ago. While the bell housing was off, I reset the gap / bracket for the speed sensor. I decided to verify the gap using procedure on Clarks Garage (gluing a 0.8 mm washer onto the speed sensor and resetting gap). One thing I did notice when I reset the sensor gap is that I moved the bracket significantly lower from the position I set several years ago. I fully recognize that this doesn't make sense.

G. I've verified B+ at terminal 15 of the ignition coil. I've also verified a complete circuit between terminal 1 of the coil all the way to Motronic DME plug terminal 1. I've replaced the coil, coil wire, plug wires, and plugs. The new coil and old coil produce some spark to bite you but not enough to light up the plugs. Intermittantly firing the coil totally independent of the DME (wired directly off the battery - intermittent energizing of primary circuit) fails to deliver meaningful spark on the secondary.

H. I drew the conclusion that perhaps the Motronic DME developed a problem. The symptoms of the car when I got stranded represented something new for me. I pulled the Motronic DME and sent it off to Specialized ECU Repairs. They benched it said it was not operating correctly and needed repairs. They did a rebuild ($300) and I now have it back. With the repaired Motronic installed the car still produces insufficient spark.

I've started to re-trace my steps a bit with the following observations:

I've noticed is that the reference mark sensor seems to generate an inconsistent AC voltage (using a multimeter - not an oscilloscope) while cranking the engine. It varies between 0.07 and 0.4 VAC. The speed sensor delivers a consistent 2 VAC. I've swapped sensor's physical and still get the same results. To take potential sensor issues off the table completely, I've just ordered replacement sensors from Pelican.

I've noticed that the battery terminal voltage dips to ~ 11.2 VDC while cranking. I'm wondering if this is enough of a voltage depression to be affecting coil performance? At this point, I hope I can rule out issues with the DME as a source of insufficient spark.

Am I correct in assuming that B+ at coil terminal 15 takes out any potential alarm / immobilizer functions which can interfere with spark generation?

The car builds fuel pressure and the injectors are delivering fuel to the intake manifold when I crank the engine. I've flushed enough gas down the intake during test cranks, that I did pull the plugs and "re-oil" the cylinders. I know from experience that this can impact an engines ability to start. My view is that this is definitely a spark issue. As an optimist, I've so far refused to pull the fuel pump fuse while I've test cranked to diagnose the spark problem.

Any ideas on what I am missing here?

Thanks in advance,

Paul (aka specsalot)

lart951

I would suggest posting this on the 944 Turbo Forum

https://rennlist.com/forums/944-turb...bo-s-forum-72/

specsalot

Thanks - Will post there as well

konakat

Sounds like DME to me. One of the sensors only gets a pulse once per revolution, so I wouldn't think the bad AC voltage on one sensor to be an issue. It definitely sounds like a sensor problem, so I think swapping your DME is the next step. The DME has to convert that analog pulse to a digital pulse, so that circuit needs to function correctly for everything to work.

mytrplseven

I don't know where in Florida you are, but good luck in finding someone to swap out a DME with you to diagnose your problem. I've made a couple of posts on both forums to have someone with an '87 N/A to help by letting me use their DME for a road check on my stumbling accel issue, but no one ever responds.

konakat

If you have access to a scope you can probe the dme to find out if that signal is the issue. I can help walk you through it if you aren't comfortable with reading schematics.

specsalot

Thanks all - lart951 suggested that I repost this to the 944 Turbo forum, so I have reposted there as well (same title). In that thread so far:

Chris White pointed out that initial symptoms pointed directly at a problem with the TPS (throttle position switch) or its wiring. My plan is to check the circuit from DME / KLR plugs to the TPS connector. And also to measure / verify TPS operation. Prior to this issue the car had some high idle issues which (I'm learning) point possibly toward TPS issues. Chris also pointed out that sensors (reference mark / rpm) are probably not the issue, suggesting the culprit may be wet plugs. Sensors are already on order and will be changed this week after they arrive. I will have the plugs cleaned before next attempt to fire it up.

Dougs951S validated that the dip in battery terminal voltage during cranking (down to ~ 11.2V) should not be enough to prevent the car from generating sufficient spark to fire up.

My DME has been sent off and rebuilt by Specialized ECU repair. They resoldered defective solder joints, replaced a defective memory module and also a defective voltage comparator. This shop has a good reputation and I have to believe I can rule out DME issues after the rebuild.

Getting access to a scope (through work) is not out of the realm of possibilities, but it will take some real string pulling which I'm hoping to avoid. I have to do the rest of the leg work before I consider going down that alley.

Beyond checking out the TPS, I plan to make a long list straight off the wiring diagrams and verify external wiring (DME/KLR) one pin at a time to rule out wiring harness issues.

The DME relay is operating so this rules out any issues with the alarm control module. The Alarm Relay (G18) socket has been vacant since I've owned this car. The former owner said that the alarm system was a PITA and that it was best to leave it alone.

I'll keep everyone update as I work through this process. Thanks for the input.

specsalot

Update - 2/1/2013 - Drawing a conclusion that I've got an ground cable problem. NEW OEM coil not deliveribng much spark. But coil ground (coil terminal 1 voltage to B-) seems to pulse implying DME is alive and well. I recently repaired the lagging on the ground cable and noted that there was some evidence of internal corrosion within the cable.

More parts on order from Pelican (and others) to do other maintenance while the car is out of service (timing belt, balance belt, reseal throttle body, battery cable replacements).

More to follow ....

specsalot

COMMENT - I've always loved the lines of these cars. It's challenges like the last few weeks that test ownership commitment. For the most part this car has delivered on the goods. In vehicle terms its a bit of a senior citizen (like it's owner). Some days I don't start up as well as I used to either

Marajit

OK, that made me laugh and throw up a little bit in my mouth.

specsalot

Hopefully your previous meal remained tasty during the rerun - sorry to hear my post causes such an upset.

I appreciate you taking the time to post. I'm hoping you can take the time to elaborate? I'm assuming your not a fan of Specialized ECU Repair.

I'm still struggling with a weak spark issue, but at this point it may be a grounding issue. A trial and error process of elimination for sure. I'm ready to call BS on Specialized ECU Repair. With a 5 year warranty on their work, I can send the DME back anytime if I can't solve this issue.

fwb42

If you think you have ground issues, look for post about Ice Shark cables. I put a set on my car and eliminated any electrical issue that may develop. This is a very high quality electrical fix and worth the $$.

specsalot

Thanks fwb42 - I read several Ice Shark cable postings. Looks like great kit. I made a decision to just go back with new Porsche OEM. I'm also replacing the starter / alternator harness to eliminate that as a source of future troubles. Getting the cables out is a bit of work, but I removed both B+ and B- to inspect them when I got the car in 2007. They were fine than, but its been 6+ years. Time to eliminate current (or future) problems. It's hard to point directly at this as a cause of low spark, but I can say after a bit of coil testing (independent of DME ground path) that spark seems to have a problem finding a ground path on the engine. With some effort I've been able to the coil to intermittently fire 1/4" spark from the coil, but only at solid independent ground - not at engine block grounds. Pretty strange (as electricity can some times be).

mytrplseven

The guys at Specialized offered to take a DME off the shelf and plug it into my car to isolate whether that's at the core of my ongoing stumbling accel issues. That's customer service, and I haven't even bought anything yet.

Where in Florida are you?

specsalot

I'm up in Gainesville. I'm trying to run down the bases before I go back to Specialized. I opened my DME for a look see before I sent it off. No visible issues like water intrusion, burning, etc. That said I didn't go to the level of inspecting solder joints with a magnifying glass. They found some cracked solder joints and some kind of problem with a defective memory module.

Squinting at the schematic (TY to Clarks Garage), one of the two boards is appears to process analog signals in and out. The bulk of intelligent processing occurs on the other (digital) board. For those here who are knowledgeable, this is square 1 info. For those of us traversing the learning curve this is interesting stuff. Spark is a generated via a conduction pathway from DME terminal 1 to a DME ground (not sure which DME pin yet). The reference mark and RPM sensors are passed directly to the digital side of the DME. Beyond that it's all greek to me. The fact that I'm getting decent injection (fuel side) implies the spark side should also be working. It suggests the computer is working.

Next step (beyond waiting for parts to arrive) is to try to make some preliminary measurement of current flow in the active coil circuit during crank up. This would confirm coil circuit integrity.

What I've read is that stumbling acceleration could be related to TPS (throttle position switch), AFM condition / calibration, or potentially the O2 sensor. These problems are probably a bit more tricky to diagnose. If you can get CH 28 of the shop manual, it walks through all the steps to totally diagnose the vehicle. If you would like, I can scan it and e-mail your a copy?