No spark tried everything - Ingo even loaned me his DME??

[4flyboy]

Hi guys,

My friend and mechanic, Alex and I, are at wits end. We took a step back this evening to think and ask the forum for help .

My 1984 3.2 Carrera Turbo-Look has been sidelined for some time with an intermittent No Spark situation.

Alex is a BMW guy with extensive electrical and Motec experience, he is patient, thorough and as mystified as I with this infuriating issue.

Oh, and Alex has an oscilloscope.

We've read most every post on this subject and even enlisted Ingo Schmidt in a limited way.

Ingo has bench tested my DME (twice) and even Express mailed me his test mule DME so that we could eliminate the computer as the culprit. In fairness we haven't been able to pick his brain, he's been traveling and even picked up his cell in Europe when I called.

Thanks Ingo!

Alex will chime in because I'm an illiterate electrical enthusiast but here's my take on what we've done:

1. DME bench tested and loaner to no avail
2. New coil
3. New Magnecors
4. New fuel pump - it runs and gets signal
5. New fuel pump relay
6. Check all sensors for continuity
7. Check ignition switch
8. Oxygen sensor plugged/unplugged
9. The fuel whatchamacallit vibrates as it should
10. New battery - the old one gave out during this process
11. Two new speed/reference sensors, the gap is perfect, the signal is good on O scope
12. Reversed the speed and reference sensors
13. Checked the flywheel sensor is in place

What led up to this?

I had some work done to replace the rear seat pans, cutting, grinding and welding.

Car wouldn't start at the shop, I reset and cleaned the referece/speed sensors and the car started right up. Got it home and it quit on me.

Two new speed/reference sensors, car started and ran for 50 miles and quit again.

Now no spark and not even an attempt to fire.....when we turn on the ignition there is a WEAK almost imperceptible SPARK that goes out as soon as the key is advanced to engage the starter.

I'm at a loss guys, any insight or suggestions gratefully acknowleged, nothing will be scoffed at.

Thank-you, Scott

Delray Beach Florida

Here's a picture for color...Ref Sensor

[al525i]

Hi everyone!
I'm Alex, and I've been trying to help work through the no-start on Scott's car.

I have extensive motronic experience but most of it has been on similar vintage BMW's (same systems really)

Here's my take on whats happening, or not happening.

We have a no spark, no fuel situation. when the ignition is turned on, the motronic does an initialization and pulses a single weak spark. The Idle control valve vibrates when the key is on so I know the DME is "working"

When the ignition is turned to start the engine spins over at normal cranking speed and the DME engages the fuel pump. It does not, however, pulse the fuel injectors (either batch) and does not attempt to create a spark. There is also no movement on the tachometer.

This seemed like a speed or reference sensor problem so I performed a resistance test on the sensors. They were within spec. Next I broke out the oscilloscope. The speed sensor waveform looked great but there was some noise in the reference sensor output. I removed the sensors and the retaining bracket and spun the crankshaft around to visuallly inspect the flywheel teeth and the reference pin. Everything appears to be in order there and there is no debris on the flywheel to cause erroneous signals. The retaining bracket was reinstalled and the gap adjusted to factory spec (.8mm) I replaced the reference sensor with a new one and the noisy output went away but it still wouldn't start. I tested power and ground inputs and outputs to the dme and everything tests fine.

At this point I tried to rule out other possible failures. I disconnected the dme and probed all the wires to verify that all the sensors were getting viable readings to the dme connector. (this was done with a fluke 77 digital meter or two channel oscilloscope where applicable) The only thing that didn't look perfect was the air flow meter, which had a few bad spots when doing a door sweep. That situation has been remedied and the sweep shows a smooth, linear output curve with resistances as specified in the manual.

Seemingly, the DME was the next logical item so it was sent to Ingo (who is an amazing individual and an asset to this community!) for testing and rehab. He found nothign wrong with it and sent it back with a clean bill of health and a twin that he knows to be good as well.

Still no start.

At this point I have tested each and every wire in the harness from the DME end all the way out to the connector end for continuity, short to ground, and short to other wires in the harness. I found a weak connection at the speed sensor chassis side connector in the engine compartment but that has been repaired with no change in situation.

[pmax]

Probably worth while testing the known good DME again in a known good car to see if it has been damaged.

[steely]

When Scott worked the seat pans, was the DME removed?
Just a guess, recessed DME pins on the harness connector?
If the FP is on, the DME must be seeing the sensor input, so maybe its the coil & injector output not getting to harness? (although 2 failures is unusual)
Can you get a breakout box to vfy?

[ant7]

Ignition switch ?
Check with a multimeter' [on the DC range] that you have the correct voltage on the primary side of the ignition coil, when you turn the engine over!
Hope this helps.
A...

[DanielDudley]

And you checked grounds ? Tried starter fluid as a diagnostic ? New distributor caps can be bad right out of the box if they are dropped. So can coils.

It sounds like you tested the continuity of every wire in the harness. This still sounds like a continuity issue.

[gomezoneill]

If the problem started after the welding then that's what started your problem, go from there. Possibly a melted wire inside a harness? Also that connection to the red sensor plug looks suspect in the photo. Sounds like a reference sensor issue though.

[faverymi]

I would start testing every wire on the main harness

That seems to me likely the cause

[DanielDudley]

Caig Laboratories DeoxIT D5 Spray Contact Cleaner, 5% solution, 5 oz. | D5S-6 (D5S6) | Caig Laboratories

[DanielDudley]

Originally Posted by faverymi

I would start testing every wire on the main harness

That seems to me likely the cause

Check every wire for continuity, and check every wire against every other wire to eliminate crossed wire shorts.

FWIW, I have seen wires test fine when unplugged, and lose connection when the wire is moved back into its customary position, due to an internal break being flexed into making contact.

[al525i]

Originally Posted by steely

When Scott worked the seat pans, was the DME removed?
Just a guess, recessed DME pins on the harness connector?
If the FP is on, the DME must be seeing the sensor input, so maybe its the coil & injector output not getting to harness? (although 2 failures is unusual)
Can you get a breakout box to vfy?

I currently have the DME connector shell open so I can back probe the connetors with everything assembled. I've checked all the terminals and there are none recessed or damaged. I've verified the integrity of the coil and injector wiring. I disconnected every injector in case there was one pulling the circuit dead and still have no pulse on a noid light.

Originally Posted by ant7

Ignition switch ?
Check with a multimeter' [on the DC range] that you have the correct voltage on the primary side of the ignition coil, when you turn the engine over!
Hope this helps.
A...

the hot side of the coil gets battery voltage when the ignition is in run or start and it doesn't drop out

Originally Posted by DanielDudley

And you checked grounds ? Tried starter fluid as a diagnostic ? New distributor caps can be bad right out of the box if they are dropped. So can coils.

I've checked all grounds and even tried adding jumper wire grounds to see if there was just a weak ground possibly. Starter fluid isn't going to help here but yes that was one of my first checks trying to see if it was fuel or spark related. (then i found out its both)

There's no output from the coil to a tester, there's no spark to even make the distributor part of the problem. I have no DME output to the coil, even so I tried a known good coil with the exact same results.

It sounds like you tested the continuity of every wire in the harness. This still sounds like a continuity issue.

Thats what I keep coming back to but I've run end to end continuity tests and tests for short to ground and short to other wires of every wire in the harness while wiggling, moving, and jiggling them to no avail.

Originally Posted by gomezoneill

If the problem started after the welding then that's what started your problem, go from there. Possibly a melted wire inside a harness? Also that connection to the red sensor plug looks suspect in the photo. Sounds like a reference sensor issue though.

The connection to the speed sensor has been replaced. The picture Scott posted was before. I found that and thought it was going to be the solution.. no dice.

I thought the same about a melted wire inside the harness but I can't find anywhere the harness looks even slightly damaged. Also the electrical testing points to there not being an issue there. I'm almost to the point of pulling the entire harness and opening it up to physically inspect every wire

Originally Posted by DanielDudley

Check every wire for continuity, and check every wire against every other wire to eliminate crossed wire shorts.

FWIW, I have seen wires test fine when unplugged, and lose connection when the wire is moved back into its customary position, due to an internal break being flexed into making contact.

Thats been done... twice now. The second go round I was more vigorous with the shaking flexing and jiggling to try and find an issue.

[al525i]

Originally Posted by steely

When Scott worked the seat pans, was the DME removed?
Just a guess, recessed DME pins on the harness connector?
If the FP is on, the DME must be seeing the sensor input, so maybe its the coil & injector output not getting to harness? (although 2 failures is unusual)
Can you get a breakout box to vfy?

I currently have the DME connector shell open so I can back probe the connetors with everything assembled. I've checked all the terminals and there are none recessed or damaged. I've verified the integrity of the coil and injector wiring. I disconnected every injector in case there was one pulling the circuit dead and still have no pulse on a noid light.

Originally Posted by ant7

Ignition switch ?
Check with a multimeter' [on the DC range] that you have the correct voltage on the primary side of the ignition coil, when you turn the engine over!
Hope this helps.
A...

the hot side of the coil gets battery voltage when the ignition is in run or start and it doesn't drop out

Originally Posted by DanielDudley

And you checked grounds ? Tried starter fluid as a diagnostic ? New distributor caps can be bad right out of the box if they are dropped. So can coils.

I've checked all grounds and even tried adding jumper wire grounds to see if there was just a weak ground possibly. Starter fluid isn't going to help here but yes that was one of my first checks trying to see if it was fuel or spark related. (then i found out its both)

There's no output from the coil to a tester, there's no spark to even make the distributor part of the problem. I have no DME output to the coil, even so I tried a known good coil with the exact same results.

It sounds like you tested the continuity of every wire in the harness. This still sounds like a continuity issue.

Thats what I keep coming back to but I've run end to end continuity tests and tests for short to ground and short to other wires of every wire in the harness while wiggling, moving, and jiggling them to no avail.

Originally Posted by gomezoneill

If the problem started after the welding then that's what started your problem, go from there. Possibly a melted wire inside a harness? Also that connection to the red sensor plug looks suspect in the photo. Sounds like a reference sensor issue though.

The connection to the speed sensor has been replaced. The picture Scott posted was before. I found that and thought it was going to be the solution.. no dice.

I thought the same about a melted wire inside the harness but I can't find anywhere the harness looks even slightly damaged. Also the electrical testing points to there not being an issue there. I'm almost to the point of pulling the entire harness and opening it up to physically inspect every wire

Originally Posted by DanielDudley

Check every wire for continuity, and check every wire against every other wire to eliminate crossed wire shorts.

FWIW, I have seen wires test fine when unplugged, and lose connection when the wire is moved back into its customary position, due to an internal break being flexed into making contact.

Thats been done... twice now. The second go round I was more vigorous with the shaking flexing and jiggling to try and find an issue.

[T77911S]

here is the thing with a MM.
you can have good readings but the connection/wire may still not be able to carry the current needed to operate the load.
prime example:
I see guys measuring 12 volts at a brake light but the bulb will not light. they find a corroded or bad connection that if they had used a test light they would have found it sooner.

the welding and cutting caught my eye because maybe it damaged some wiring. yes I saw you tested all the connections.

here is what I would do since you have already done a ll the right things and it still does not work.
I would test the wiring with voltage. apply 12v to each pin of every connector. then use a test light at the DME connector to check it.
you can test the ground side of wires by connecting a test light to 12v and check for a good ground.

a light bulb requires more current to light it. a MM can measure 12v even with questionable connections.

[Joe Bob]

I had an issue with DME input power from the fuse box. It ended being a fuse with worn ends. Inconsistent power. I ended up replacing the box with a modern system with blade fuses.

[mysocal911]

Use this as a troubleshooting guide:

:
Simple tests to localize a 911 3.2 no-start problem:

1. Remove the coil wire from the distributor cap and place the end about 10mm
from the engine case.

2. Remove the #1 injector connector and insert a noid light, buy one at
AutoZone/Reillys or Pep Boys.

3. Have someone crank the engine while you monitor both the coil wire & noid light.

Tests results:

A. Both immediate spark and noid upon cranking then;
Both speed & ref sensors, powers/grounds O.K., and DME ECM O.K. - Check for fuel
pressure and/or spray carb cleaner into air filter and re-do steps 1-3 above.
Check distributor cap next. The temp sensor can be bypassed by jumping the
temp connector (white) with a paper clip, which simulates a warm engine.

B. If neither noid nor spark then; check sensors (1k ohms, speed > 1.0VAC,
ref >.10VAC), DME ECM powers (1/18/35) & grds (5/16/17), or DME relay

C. If noid but no spark then; DME ECM a problem, or power to coil from ignition
switch (if already replaced the coil), also check coil wire for resistance.
Once it starts, remove the ref sensor connector, the engine should keep running
and accel perfectly (If not, then a bad speed sensor). Next, try reversing the
speed & ref sensors on the engine and then the two connectors on top,
i.e. To check for a possible weak sensor.

D. If spark but no noid then; power connection to injectors & ECM (pins 14/15) or
bad DME Relay.

Note: The temp sensor, AFM, and altitude sensor affect the fuel and are essentially
eliminated as problem sources for basic starting by using carb cleaner (A above).

The above is a starting point and NOT an exhaustive/in-detail troubleshooting guide.

When checking power voltages, use a test light (incandescent bulb type) to load the
circuit better with the voltmeter connected.

[mysocal911]

Originally Posted by 4flyboy

Now no spark and not even an attempt to fire.....when we turn on the ignition there is a WEAK almost imperceptible SPARK that goes out as soon as the key is advanced to engage the starter.

You haven't focused on the other half of a no-start, i.e. the injectors. If the injectors
are working then all the inputs are O.K. and the effort needs to focus from pin one
of the DME ECM out, i.e. since the ECM has been tested O.K.

Check the power on the coil (using a test light) while cranking. One side (ignition switch #15) should
stay at 12 volts (no flashing of the test light like the DME ECM pin 1 side).

[rick-l]

Originally Posted by mysocal911

You haven't focused on the other half of a no-start, i.e. the injectors.

Originally Posted by al525i

When the ignition is turned to start the engine spins over at normal cranking speed and the DME engages the fuel pump. It does not, however, pulse the fuel injectors (either batch) and does not attempt to create a spark. There is also no movement on the tachometer.

It sounds like he knows what he is doing

You did verify the correct sensor signals are at the right pins at the DME when you put a scope on them?

It almost has to be the wire harness.

[rick-l]

Just a thought..... If any of the 4 sensor wires were shorted to chassis the signal would still look good on your isolated scope.

[4flyboy]

Thanks for all the responses, gives us fresh eyes on it fellas.

I'm sure Alex will chime in here again.

Scott

[4flyboy]

Guys, Alex attempted to respond to some of your suggestions this morning and he received a message indicating that his response would be reviewed by a Moderator.

Anyone know what the problem would be?

He is new to this forum....

In any case we sure appreciate the help.

Thanks Scott