964 ignition failure

[TomUban]

My '90 964 failed to start a week after it was running fine. It cranks over fine and I've verified/tested that the fuel pump is running, the DME relay is good (I tried bypassing it too), a known good replacement DME did not help.

I don't see any spark when either looking for it to jump a gap or using an inductively coupled timing light. It would seem that this narrows the problem down to one of the following:

1. coil
2. hall effect sensor
3. ignition module

Does this sound right? Could it be something else? Is there a way to test each of these without replacements?

--tnx
--tom

[964sw]

Hi

Had the same problem on my 964 -91, Tested every thing for weeks.
Then i replased the Hall engin speed sensor that gives the signal to the Dme unit. The sensor i in stainless steel and the holder for it in aluminum and there was so mutch corision that the sensor was pressed to gether and the signal was not rigt. new sensor and holder and the car was running again.

steven

[38D]

My guess would be the sensor.

[ischmitz]

The fuel pump will only run if the DME and the sensor are working. So if you are sure the fuel pump came on during cranking your sensor is fine. And if you swapped the DME with a known good one chances are your fault is downstream from the DME. However, I just repaired a 964 DME that had a bad output for the igniton signal. So the DME can fail.

The 964 has one ignition channel inside the DME that is routed to both ignitors. From the ignitors onwards every is there in pairs (ignitors, coils, distributors, cables, plugs). Even if one circtuit goes bad the car will still run.

If both circuits are dead it seems unlikely that both coils or both igntitors died at the same time. My guess is either the harness from the DME to the ignitors (remember, one signal from the DME triggers both ignitors) or the GND connection for the ignitors or the +12 volts for the coils is bad.

Ingo

[964sw]

In my case the sensor gave a signal strong anuf to trigger the injectiom system
but not the ignitionsystem and i also had fuel pressure.

steven

[ischmitz]

Steven, the 964 DME is a computer with analog I/O stages. It uses a 8051 derivative CPU that "decides" when to turn on the fuel pump and the spark signal. If the sensor produces any signal the CPU will trigger the fuel pump to run and at the same time produce sparks.

Maybe in you case the sensor signal was very intermittent, it triggered the fuel pump but did not produce steady spark signals to make the engine catch and run properly. Once the CPU sees any type of signal on the sensor input it turns on the fuel pump for at least a half second. If you signal faded in and out the fuel pump could have come on while there were only random spark signals.

By the end of the day I would hook up and oscilloscope to the sensor and check whether I get the steady 2 volt p-p signal with missing tooth dip before shelling out $$$ on a hunch.

Cheers,
Ingo

[964sw]

Hi Ingo ! Just what I did. Traded a osilloscop on ebay for 50$ and then in the sensor
but i dont think everbody like to go that way and or knows how to use one

steven

[ischmitz]

Good point Steven, I think you did the right thing. I was just cautioning folks to go out there and start throwing parts at a problem. This usually burns through a lot of money fast if you don't have these items laying around. I have heard so many horror stories where the entire ignition/DME was replaced just on a hunch......

It seems like you had your bases covered and tracked down your issue. Way to go.

Cheers,
Ingo

[TomUban]

Here is the current status:

I hooked an oscilloscope to to the DME connector to look at the speed sensor output and it has a strong signal when the engine is being cranked over.

I hooked up a hammer, but as the DME had been disconnected from the battery, any stored faults had already been lost. The hammer does not appear to be useful unless the engine has run (and stored a fault) or is running.

I checked that there is 12v on across each coil.

I tried disconnecting one ignition module at a time in case a failed one was taking out the DME signal to the other. This did not help.

Can a failed coil cause both to misbehave?

Perhaps the signal from the DME to the ignition modules is not making it. Any suggestions as to how to look at this signal and what to expect would be appreciated.

[MURF]

a failed coil can put secondary ignition voltage into the dme.it will still run on 1 ignition till the voltage damages the computer...we had several cars that would cut out and bad ignition wires or coils were the cause.search for posts on this issue on the site..ecu doctor.com.will check out your dme for a $50 fee. either a bad coil,ign module or bad ignition wires will do this..replace both components upper and lower..the excessive voltage will cause the dme to shut down and reboot.

[ischmitz]

The hammer should be able to test the injectors and all other sensors and actuators of the engine. If any of the sensors is bad the hammer will let you know. So it is very useful IMHO. Try if you can actuate each injector.

The only item you cannot check/trigger with the hammer on a 964 is the ignition. If the output from the DME is damaged from overvoltage (bad coil, bad ignition wires) the DME needs repair. I recently had a 964 DME on the bench with that exact failure symptom: The engine would not start but I had strong fuel smell at the tail pipe after 30 seconds of cranking.

It spark output stage was damaged. The output is a simple trigger signal to both ignitors. You should be able to measure it at pin 1 of the 55 pin connector.

To test if one of the ignitors is bad disconnect one at a time and try if the engine starts. Also, do you smell fuel when cranking? If yes, it is likely a bad DME, if no you have another issue.

Finally, if your DME requires repair and send me a PM.

Ingo

[ischmitz]

Another thing: The engine will run w/o the hall sensor. The DME will throw an error code and retard the ignition by 6 degrees. This will reduce power but won't harm anything.

What MURF said is spot-on: Bad ignition wires or a bad coil will cause the DME to reset frequently (engine cuts out). While the DME has internal protection against overvoltage it can only take so much and eventually burns out. This leaves the DME dead.

Ingo

[TomUban]

This all sounds like good information. I did try disconnecting each of the ignition module connectors one at a time while trying to start the engine. It sounds like I need to put my scope onto the ignition module connector to see if the DME is sending the signal and if the signal is not present there, then check it at the DME
connection, which may be more difficult to do while the connector is attached to the DME. As I have a spare DME, and that spare did not allow the engine to start, it would seem that the problem must be elsewhere. At this point, I'm guessing it is something to do with the wire harness. Are there any other connectors between the DME and the ignition modules?

[TomUban]

I checked the signal from the DME to the ignition modules and there was nothing there, so I removed the cover from my spare DME in order to allow me to attach my oscilloscope to the signals while the DME is attached to the wire harness. In this way, I observed that while the speed sensor signal looked good at the connector, once the sensor was attached to the DME the signal evaporated to nearly nothing.
As the manual says this signal should be about 3v peak to peak, but doesn't say if this is connected or not connected, I'm hoping that the sensor is the problem and I'm going to replace it. I'll followup after this has happened.

Thanks for everyone's help and suggestions so far...

[ischmitz]

Originally Posted by TomUban

I checked the signal from the DME to the ignition modules and there was nothing there, so I removed the cover from my spare DME in order to allow me to attach my oscilloscope to the signals while the DME is attached to the wire harness. In this way, I observed that while the speed sensor signal looked good at the connector, once the sensor was attached to the DME the signal evaporated to nearly nothing.
As the manual says this signal should be about 3v peak to peak, but doesn't say if this is connected or not connected, I'm hoping that the sensor is the problem and I'm going to replace it. I'll followup after this has happened.

Thanks for everyone's help and suggestions so far...

The input impedance of the DME should be a couple of 10kOhm. So it could be the sensor going bad. You can actually open the 55 pin connector and gain access to pins that way while it is connected to the DME. Have you checked with the second DME as well if the sensor signal gets small when connecting the DME? Did you get fuel injector signals?

Ingo

[TomUban]

I did not check the speed sensor signal as connected to both DMEs, just the one I opened. Both DMEs were known to be working before this recent trouble, so I'm thinking that the sensor is bad. Not sure what a good sensor should read, but this one is 524 ohms. I did verify that the sensor cable from the DME to the connector in the engine compartment is good.

Ingo, I see that you have a 3.6 in a '74. Which transmission are you running?

[TomUban]

Ops, forgot to mention that I did not try to check if the hammer actuated the injectors, but I did see that the throttle switches (e.g. idle) were being sensed.

[ischmitz]

Hm, I really don't know what to tell you (regarding the sensor). I would run out and measure mine but I am 3000 miles away from home. Your reading of 500 Ohm sounds about right when I would guess. The 3.2 sensors are around 1kohm. The input stage is typically medium impedance, around 30 - 40 kOhm so it doesn't bog the signal down too much and is not too sensitive to ESD. The sensor is a simple coil on a soft iron core. Hard to believe that the signal goes away when you connect it.

The bottom line is that if you don't get fuel smell at the tail pipe after cranking for about a minute then it could be the sensor (or the harness for the sensor signal). It uses a Coax wire. Maybe it has shorted out. Measure the resistance between the two pins of the harness connector for the reference sensor with the DME connected and ignition OFF. Then disconnect the DME and measure again. With the DME disconnected it should be in the MOhm range.

Are you sure that both DMEs were fine. It would be rare for both to have died at the same time but then again you never know.

There is no other connection between pin 1 of the 55-pin connector and the ignitors. Both ignitors are driven off the same signal. Just for kicks, what do you measure at pin 1 of the DME (or the corresponding pin of the connector to the ignitors) with the ignition turned on? It should be +5 Volt if I remember correctly.

I run the stock 7:31 magnesion 915 with my 3.6. I put in the one-piece bearing retainer and the later-style differential cover. So far I put about 30,000 miles on the tranny and had no issues whatsoever. Took it to the track a couple of times but it is mostly driven on the streets. It's an absolute blast.

Ingo

[ischmitz]

To be more specific the reference sensor has to two leads (1) and (2) that connect to its internal coil and (3) that is connected to the housing as GND. The signals go to pin 47 and 48 of the DME connector via shielded COAX wires. I would measure the resistance between those pins with the DME disconnected. Then connect the DME and disconnect the reference sensor. Measure resistance between pin 1 and pin 2 at the harness connector to get the input impedance of the DME. Then measure resistance 1 - 3 and 2 - 3 to confirm there is no short to GND.

[TomUban]

If I disconnect the DME and measure the impedance between the three pins on the connector in the engine bay, they are all open. If I connect the DME and do the same, I see 64K ohms between pins 1 - 2 and 1 - 3, and zero between 2 - 3. Note that I may have pins 1 and 3 reversed as I did not look for a reference mark.

I believe that the original sensor part number is 964.606.217.03, but when I called Stoddard, they say that the old one is no longer available as it has been superceded by the 993.606.217.01 part, which has a different connector, etc. The part I removed from my car looks as if the sensor may have been spliced with the original cable, although in a very professional manner, so maybe it had been replaced at some point in the past. I think that I will open up the splice to see if I am correct and also check if there is some sort of problem in the connections there before ordering a new sensor. Surprisingly, the new sensor is only $45, so won't be the biggest error I've ever made if it doesn't turn out to be the problem.