Reference Mark Sensor - Flywheel Pin Troubles

[Bradical]

Some may remember, I have been having intermittent issues with my 924S dying while driving. I posted here earlier because I found that the connections at the Speed/Ref sensors were damaged and I redid them. This rectified the problem for a few weeks until it started happening again.

Last weekend the car died after warming up for about 5 minutes and would not restart, no tach bounce. Every other time the car has restarted immediately or after a minute or two.

Using an oscilloscope I found that I had a good signal from the speed sensor a VERY weak signal from the reference sensor, which reminded me- I know the PO had issues with the TDC pin in the flywheel causing no-start about a year before I took ownership of it (we AutoX'd together).

I have done literally everything but pull this flywheel as of yet- new sensors, rebuilt the entire DME wiring harness, new ignition switches, multiple DME relays, DMEs and speed/ref sensors. The entire fuel system is new. The pump runs with the DME relay bypassed... I am just telling you this so no one copy and pastes the typical no-start troubleshooting procedures because I have gone through ALL of it and I know that my issue lies with the reference mark sensor. I read the resistance of the sensor through pins 25 & 26 at the DME connector but I am reading less than a 0.5v peak during cranking. The speed sensor reads the same resistance and shows a perfect 2.5v peak. I have checked, double-checked, set and re-set the sensor gap.

Anyone had a similar experience and can shed some light on this situation? I was able to see the reference mark set screw through the OT mark view hole and it's definitely THERE, but it looked a little rough, like it was stripped. Thoughts?

[kdjones2000]

[9FF]

Originally Posted by Bradical

Some may remember, I have been having intermittent issues with my 924S dying while driving. I posted here earlier because I found that the connections at the Speed/Ref sensors were damaged and I redid them. This rectified the problem for a few weeks until it started happening again...

What did you do when you "redid them"?
How intermittent was it before you "redid them"?
Was it exactly the same intermittent issue when it came back a few weeks after you "redid them"?
When you replaced the DME harness did that include the complete wiring to the sensors?
After you replaced the DME harness was the problem exactly the same?
Do you have an aftermarket flywheel fitted, like a fidanza?

[Bradical]

Originally Posted by 9FF

1.) What did you do when you "redid them"?
2.) How intermittent was it before you "redid them"?
3.) Was it exactly the same intermittent issue when it came back a few weeks after you "redid them"?
4.) When you replaced the DME harness did that include the complete wiring to the sensors?
5.) After you replaced the DME harness was the problem exactly the same?
6.) Do you have an aftermarket flywheel fitted, like a fidanza?

I numbered your questions so I can respond to them in order.



1.) I cut back the speed and reference sensor wires and installed replacement connectors. I also re-gapped the sensor bracket and installed 2 new sensors. I also installed a new DME relay.

2.) Completely random. I have had the car shut off a few feet after leaving the driveway, on the freeway at speed after having been on the road for more than an hour, and just a few seconds after startup.

3.) After replacing the connectors, I ran in an AutoX without issue. Then, on the way home the car cut off and restarted after 2 minutes. This was the first time it had happened in weeks.

4.) I made a two hour drive, slept, then woke up and started warming up the car. It ran for 10 minutes, died and would not restart. No tach bounce. Of course I tried everything I could- swapped out the DME relay again, jiggled the sensors connectors, tapped the DME and relay, shook the connections at the DME. Nothing- had to get towed. SO, when I got home I pulled the entire DME harness from the car and replaced the reference sensor and fuel injection harnesses with the ones from Lindsay Racing, so complete replacement all the way back to the DME.

5.) Still no start, no tach bounce. Continuity and signal on 8 & 27 for RPM, continuity on 25 and 26 (1100 ohms) but really weak signal for reference mark. Before anyone asks, YES I did verify that the sensors were plugged into the correct connectors- I checked continuity across pins 8 and 27 at the DME and unplugged the speed sensor (closest to the transaxle) and observed the meter read OL, then reconnected the plug and observed the meter report 1.1kohms. I followed the same process for the reference sensor, a la pins 25 and 26 and the sensor closest to the engine and made the same observation in that case.

6.) No, it's a factory flywheel. I remember having a conversation with the PO before I had any knowledge that I would one day buy the car from him where he told me he had a month-long no start escapade where he found that the flywheel reference stud was damaged during his clutch replacement. I pulled the clutch today and found that the bolts for the throwout bearing guide tube were almost falling out and had scored the face of the throwout bearing, and 2 of the flywheel bolts came out by hand...

On top of that, the reference mark stud was pretty rough on the end and appeared to have been stripped. I was able to drill and extract it FAIRLY easily (this in itself tells me it isn't the factory stud and wasn't installed with Loctite 270) and I had already ordered a couple last weekend when I suspected I might be driven to this point. I am glad that I am doing this even if it ends up being something else entirely if not to discover the fact that the PO has obviously reused or not torqued the flywheel bolts correctly.

[thomasryan]

When you say you rebuilt the wiring harnes, does that mean new ends or new wires?

Are you scoping it at the sensor or at the DME?


Edit: sorry just read your response. Can you put an ohm meter at the DME and manipulate the refrence wire to see if the reading changes. (fixed a knock code on another car with a fresh coax wire recently)

[9FF]

First off I was thinking you had a wiring problem as these sort of intermittent problems are normally always electrical related. I know you replaced the sensor connectors by cutting back to good wiring but the wiring in the looms on our cars may have also degraded somewhat over the years, that's why I asked if you replaced the loom. The sensor wiring may be difficult to measure resistances over if the actual wiring may be suspect, probably causing a higher resistance. I've seen it on the fan resistor wiring but I'm sure it's elsewhere too especially as the resistances you are measuring are quite low.

It seems like you are now digging deeper into this by accessing the flywheel and the stud. Certainly seems like fitting a new stud, if the end is damaged, is the way to go. The stud may also be loose and is backing in or out while the engine is running. It was originally held in by loctite though that may have degaraded, idk if that's common. If the problem persists I would change out the complete wiring from sensors to DME next.

Finally, I know you have swapped out the DME, I presume with another used one. If the used one was a known working one that's fine. However our DME's, as they go through heat cycles, do suffer from solder dry joints, especially on the two internal power circuits. It's a known problem and will cause intermittent no-start problems like you are experiencing. I would suggest you resolder the DME even if it's not confirmed as the problem now, just a wyit as you go through resolving the issue. I would not be surprised if both your old and new DME had dry solder joints, I have resoldered all my (3) car DME's when one failed, they all looked dry. GL

[Bradical]

Originally Posted by 9FF

It seems like you are now digging deeper into this by accessing the flywheel and the stud. Certainly seems like fitting a new stud, if the end is damaged, is the way to go. The stud may also be loose and is backing in or out while the engine is running. It was originally held in by loctite though that may have degaraded, idk if that's common. If the problem persists I would change out the complete wiring from sensors to DME next.

Yeah, I actually pulled the entire DME harness from the car, stripped off every single bit of wire loom and exposed every wire, replaced the speed/ref sensor harness and the fuel injector harness and reinsulated the whole harness with high quality heat resistant loom insulation and reinstalled the harness. I actually didnt find anything suspect beyond the original breaks in the insulation at the connectors I fixed weeks ago. BUT, I am glad I did it, its really helped my peace-of-mind and it looks great, too!

I have 3 DME's on hand right now (I think I literally keep 3 of almost every part), perhaps today I will flow some solder on one of them while I am waiting for my reference sensor studs and new throwout bearing to ship. Thanks for your insight, it helps to discuss it with people who know what is going on.

[Pauld_94S2]

Before you go too crazy with this problem, would you be able to swap the two sensors since they are the same part?

In my understanding of what you have stated, you have two sensors with the same resistance, as measured from the DME computer, but with unequal pulse outputs.

Switch them and see if the low output follows the sensor or stays with the location.
If the problem follows the sensor, then looks like a defective sensor.
If the problem stays the same, then you may have an incorrect gap between the sensor and the screw.

And this is based on the fact that you have verified the loom wiring is correct.

[Bradical]

Originally Posted by Pauld_94S2

Before you go too crazy with this problem, would you be able to swap the two sensors since they are the same part?

In my understanding of what you have stated, you have two sensors with the same resistance, as measured from the DME computer, but with unequal pulse outputs.

Switch them and see if the low output follows the sensor or stays with the location.
If the problem follows the sensor, then looks like a defective sensor.
If the problem stays the same, then you may have an incorrect gap between the sensor and the screw.

And this is based on the fact that you have verified the loom wiring is correct.

Absolutely, yes. The pulse output from the reference position was the same across different sensors unfortunately- the low output stays with the location so to speak.

I will hopefully have the car reassembled this weekend and I will update everyone on what happens. While I had everything picked apart, just to reassure myself I did jump out the pin on the DME plug that grounds the DME relay and the relay engaged the fuel pump and I built pressure at the rail. This issue is certainly the DME not having the qualifying triggers to start- testing it that way seems to me to include every related component and wire excluding the motronic computer. There's simply no where else to take this but to the flywheel pin at this point! (I hope! Ha!)

[Bradical]

Jeeze, so the odyssey continues. I set a factory pin in the flywheel at 5mm and reinstalled it, still no start. I proceeded then to completely dismantle my dash in search of an electrical issue. Obviously when I reinstalled the bell housing I took great care to re-check and re-set (this makes 4 I think) the sensor clearance.

Here's what I have SO far.

1.) Sine wave on the speed sensor
2.) Start signal from the ignition switch to pin 4 (~10vdc)
3.) Activation of the 1st coil on the DME relay
4.) Power returning to the DME on pins 18 & 35 from the DME relay

What's not happening.

1.) Activation of ground to 85b from pin 20 (if I jump out pin 20 to ground, the fuel pump runs)
2.) Ignition pulses the ignition. coil from pin 1
3.) Signals to the fuel injectors
4.) TACH BOUNCE! I still don't have tach bounce.

Note: the stuff the DME computer does when it confirms the engine is being started, or in other words, the stuff that it won't do if it doesn't see speed/ref & start signals.


Oye vey, so I have purchased my very own oscilloscope so I can check the reference mark signal for myself and come to my own conclusion because this has to be, somehow, the source of my troubles and if not I need to definitively rule it out. At this point I would be 100% convinced that I had a bad DME computer, HOWEVER, I have tested 3 known good DME computers on this car to absolutely no avail.

On a good note however, man am I lucky that I ended up pulling this flywheel. The PO forgot to torque 2 of the flywheel bolts (or reused old ones, resulting in their loosening), AND the bolts holding the throwout bearing guide tube to the clutch bell housing- resulting in some damage to the throwout bearing- something I wouldn't have found until catastrophic failure... so there's that. I am also taking the opportunity to go through the central electric system as I was insuring I had good signals back and forth between the DME, DME relay and the fuel pump.

[kdjones2000]

Do you have the EMI sleeve, and if so the spacer ring required for the speed sensor?

[Bradical]

Originally Posted by kdjones2000

Do you have the EMI sleeve, and if so the spacer ring required for the speed sensor?

I do have the EMI sleeve, no spacer on the speed sensor. Wow, could that be it? That would make the reference sensor too high. Holy hell, I forgot about that- I mean its crazy that it was running without it for a year or more and now it wont, but I am willing to try anything at this point. I am going to order that thing and try it out!

Thanks! This is what I was hoping for- somebody to mention something I hadnt thought of!

[Bradical]

Well, the spacer for the speed sensor isnt the trick. You only need the spacer if you are running an early bracket without the machined recess for the EMI sleeve.

[kdjones2000]

I have that spacer for the bracket on my '88 Turbo, so I don't know where you are getting your information from.

[djnolan]

Originally Posted by thomasryan

When you say you rebuilt the wiring harnes, does that mean new ends or new wires?

Lindsey racing sells partial replacement harnesses for the injector harness portion and the sensor harness portion. I have to agree that the wiring and insulation inside the looms could be toast and either shorting in certain conditions or loosing connectivity intermittently.

Also 944online will test your DME, assumedely that is a load test.

[Bradical]

Originally Posted by kdjones2000

I have that spacer for the bracket on my '88 Turbo, so I don't know where you are getting your information from.

There are 2 brackets. 1 has a machined recession for the EMI sleeve to fit into that does not effect the height of the reference sensor, and the other bracket does not have a machined recession, which causes the EMI sleeve to make the reference sensor sit higher (the thickness of the shoulder on the EMI sleeve) and thus require the spacer to be installed on the speed sensor to compensate.

Originally Posted by djnolan

Lindsey racing sells partial replacement harnesses for the injector harness portion and the sensor harness portion. I have to agree that the wiring and insulation inside the looms could be toast and either shorting in certain conditions or loosing connectivity intermittently.

Also 944online will test your DME, assumedely that is a load test.

Thats what I did- and I have 3 DMEs and an oscilloscope- what I am seeing is a strong sine wave from the speed sensor and a weak (0.5v peak) signal from the reference sensor.

What i tried yesterday was cutting a thinner washer for aligning the reference sensor. I cut one that was 0.7mm (instead of 0.8mm, I am using a 0.0001 micrometer to hit within a few 10 thousandths of this measurement) and realigned the bracket. The car started an ran, but scored the face of the speed sensor. Again- I had set the reference pin on the flywheel to 5mm, so everything should be working fine... I need to just pull the engine and check this relationship on an engine stand. Its bizaree that 0.1mm closer would cause damage to the sensor, especially when the factory manual says you should be between .5mm and .8mm. Very odd.

To be totally honest, the fact that the car actually ran was pretty vindicating- it at least proved that all my testing has been right. I am really not sure how or why this issue has come up, but again I do know that the PO wrestled with a problem with the reference sensor pin which he thought he resolved but was one of the influencing factors as to why he sold me the car in the first place.

[kdjones2000]

Hmmmm...

Sounds like if you PUT THE SPACER WASHER on the speed sensor, and gapped things properly, then everything should work as it should.....

If you are worried, the do the above, and slowly hand turn the engine and see if the reference sensor and pin touch.

[v2rocket_aka944]

are all the grounds hooked up at the block/bellhousing?

[Bradical]

Originally Posted by kdjones2000

Hmmmm...

Sounds like if you PUT THE SPACER WASHER on the speed sensor, and gapped things properly, then everything should work as it should.....

If you are worried, the do the above, and slowly hand turn the engine and see if the reference sensor and pin touch.

Oh I tried that. I was convinced it would work. The reference pin hit the sensor- also the bracket adjusted all the way down, so I am not sure if the speed sensor was even remotely close enough.

[Bradical]

Originally Posted by v2rocket_aka944

are all the grounds hooked up at the block/bellhousing?

Yes sir.

There is something not right about the relationship between the positions of the sensors themselves, the ring gear and the pin. I am just going to pull the bellhousing again and possibly the engine itself and check all the alignments and gaps with a feeler gauge. Here's why I am so confident.

Here's the sine wave I am seeing for the speed sensor. It's peak is nearly 8.5v, meaning that it's as close to the flywheel as it could possibly be without touching. In reality this signal can be as low as 2.5v, which a variance in 0.3mm spacing.

Imgur: The most awesome images on the Internet

The reference sensor, on the other hand. The positive flank is only ~0.5v, about 25% of the intensity the DME needs to see to fire.

I feel like the best way to go about fixing this is to dismantle the car and do it right.