FRporscheman

The car is an '87 951. It cranks nicely but doesn't start, and I immediately noticed the tach does not bounce while cranking. The stock boost gauge comes up to 1 bar with the ignition turned on (which, I guess, means the DME is good), and the tach will give one bounce right when I stop cranking. My first thought was speed and reference sensors, so I replaced both with used ones I had. One of them I confirmed to be good, it came from a running car, and it reads about 1020 ohms. The other one is unknown condition, and reads about 450 ohms.

The PO mentioned that the car has issues starting in wet weather. Also, according to him, the engine would stutter a little when the engine bay got wet from rain.

The old sensors were hard to take out. The speed sensor (rear) came out with just a little difficulty, and when I looked at it, it looked like something had hit it a little bit on the end and slightly deformed it. The reference sensor (front) was mangled and would not come out. I had to unbolt the bracket, break the top of the sensor off, and push the sensor down and out. The EMI shield was also slightly deformed so I cut off about 1/2" with a hacksaw and gave it a little chamfer with a file. I don't think lack of the EMI shield will prevent starting, will it? I gapped the sensors hastily... I'll go back and gap it properly.

EDIT I found this and it answered my wiring questions. Now, it must be a bad sensor or missing/damaged trigger.

quinnfiske

I had a similar problem. It turned out to be the DME. The driver circuit that sends a pulse to the coil failed, so no spark was formed. I would check for spark at the plugs.

In the course of solving the problem, I replaced the sensors with new ones as well as the harness from the bracket to the DME. Lindsey makes an excellent replacement harness and comes with very good instructions. Autobahn Extremist also has one, but I have never seen any comments on theirs.

F18Rep

I think I would plug them in and then test for the 1020/450 from the DME end - I bet your reference is grounded to the shield somewhere. The chassis side of the sensor disconnect is always suspicious.

Also, I think I saw someone post that you could actually wave a screw driver near the sensor and see a pulse on a meter - never tried that myself though. ...Bruce ;]

odb812

+1 on the DME side of the speed and reference sensor connection. I had the same symptoms you described and I pulled the boots back on the DME side of the connection at the back of the intake manifold and the wires were frayed.

BTW I'll get back to you about the parts list.

FRporscheman

OK, new problem: Now the boost gauge is dead at all times. I admit a moment of negligence in forgetting to disconnect the battery when I unplugged the DME. Did I fry it? I have 1 spare.... I'll try it out.

I replaced the 2 sensors, but it wouldn't start, probably because of the DME.

I'll check the pins at the DME plug with an ohmmeter. Also I might try checking with a voltmeter while cranking, I might see the reading change/jump, indicating the sensor and trigger are doing their jobs.

Take your time Todd.

Thanks guys!

MC951

I doubt that you fried the DME. I've unplugged it more than once without disconnecting the battery with no harm.

Ronin-951

Takes a fairly high end meter to read those pulses, most use an O-Scope. A 'Fluke' 87 is fast enough and has a Bar display making life easy,@$350.

MooreBoost

had this same problem a couple days ago on one of my cars. i went down to the local junk yard and cut some crank sensor connectors off a 80's bmw 325. spliced those connectors on and adjusted the sensors and it started right up.

Bri Bro

The boost gauge signal comes from the KLR. I am betting you disconnected the connector that goes over to the dash that supplies the tach and boost signal when you were testing the DME. It is a small flat connector.

FRporscheman

I forgot to put the DME relay back in after testing it in my other car - that explains the boost gauge.

But even with the DME relay in, I still have no tach bounce! I installed new sensors, I adjusted the gap at the bracket, and I confirmed the ohms of the sensors through the DME plug (confirming the harness and connectors are good... right?). I even tried a spare DME and no difference.

I thought, maybe the trigger setscrew is messed up. I turned the engine until the screw became visible through the ref sensor hole, and the damn thing is mangled. It looks like part of it sheared off, and what nub is left is only protruding from the flywheel by about 1.5-2 threads. I measured the distance from the tip of the screw to the top of the sensor bracket, and it's about 6mm longer than the length of a sensor.....

Is it possible to access this setscrew without doing a clutch job?

I had an idea, where I shim up the speed sensor to also be 6mm away from the ring gear, and regap the sensor bracket. Would that work?

I used my DMM to try to see voltage from the sensors, and I saw erratic readings (obviously) but the speed sensor gave me up to 60mv and the ref sensor gave me much less, like 1-2mv.

FRporscheman

Why are the first 3 tags of this thread "928, autobahn, extremist"???

FRporscheman

These are the ideas I currently have:

- if the flywheel stud is supposed to be sticking out by 5mm and it's only got about 2mm left, then shim up the rear sensor by that much and lower the bracket.
- drill/tap a new hole just next to the current messed up stud, and use a new stud. I figure a couple of degrees before shouldn't make too much of a difference, until I pull the engine later in the year to freshen it up.
- Stick a 5mm-high magnet to the flywheel next to the current messed up stud, again, a couple of degrees before. Would a magnet go flying off at high rpm? It's a strong Nd magnet.

Unless someone knows of a way to remove a mangled setscrew with the flywheel still in the car...

Bri Bro

I know the magnet will not work and the shimming may not. Can't you get at it through the starter port?

FRporscheman

I have not tried, but I know I won't be able to reach it from there, the hole isn't deep enough / with enough clearance for a tool.

I didn't think the magnet would work - figured it would produce the wrong B field inside the sensor coil.

What if I epoxy a 5mm tall nut to the flywheel?

Bri Bro

You have the problem of which pole to use and it would affect the sensor way before it got to it. It also would come loose at some inopportune time like in the middle of rush hour traffic.

Putting a nut on the stub might work but that small point sets the engine timing reference. The gear teeth (speed sensor) count how many degrees from the reference. I would find a way to replace it. If you can see it, you have a chance of replacing it. I do believe it has loctite on the treads so heat would help. A soldering gun should do the trick. Have you though about a drilling a small access hole in the bell housing?