As anyone who has been reading my rebuild thread knows, after a year of engine work I am now stumped by the fact I have no spark. So to diagnose the problem I thought we could all put our heads together and talk about exactly what the minimum requirements are for a DME to fire the plugs.

As I see it if these requirements are met, regardless of all other sensors in the car, there will be spark when you turn over the engine:

-The DME must be functional (I dont know any way to test this except put it in a different car)

-The DME must be getting battery power, (At DME connector Pins 24-18 have ~12 Volts with key off)

-The DME must be getting switched ignition power from the DME relay (At DME connector pins 24-37 have ~12 Volts with key on) This check ensures your DME relay ok.)

-The flywheel reference sensor must be functional (At DME connector pins 47-48 speed ref sensor input shows a signal > 3v p-p)

-The distributor hall effect sensor must be working. (I dont know how to check this)

-The distributor must be turning

-The coils must be ok

Then in my mind, there will be spark.

???


Kirk

 

" DME must be functional (I dont know any way to test this except put it in a different car)"
Put your scope on the line out to the switchs , crank over the engine .
See a flat line or does it wave at you ?

" hall effect sensor must be working. (I dont know how to check this)"
When you scope and crank , does it wave at you ?

When it cranks , does it crank fast ?

 

Kirk, this isn't exactly what you are asking, but if you could verify if you are getting injector firing it could tell you that the dme is at least partially working. If you aren't then you are looking for something more than just a spark loss. If you are getting gas then I'd go right to the coils and check for power and the trigger to them.

This probably isnt useful, but perhaps something I wrote might help in some way. Good luck!

 

To test coils
I think if I jumper pin 18 (+ power) at the DME cable to Pin 1, each time I remove power the coil should spark. ??

Also for the hall effect if I apply +12v to (+) and gnd the (-) I should see a wave form between 0 and gnd while cranking the engine? correct?

I also thought that I should check for a pulse at one of the injectors to see whether it is a spark only problem or if there is nothing coming from the DME.

These spark computers are tough and I can't think of any reason why mine would have failed. The fact I have no reading from it on the hammer is suspect, but I believe someone else recently had the same problem, and the DME still ran the car fine.

Kirk

 

There has been some disagreement on this area, but I don't think the hall effect sensor is needed to run the car. Again this is just from memory, but I think it is just used by the car to tell which cylinder to retard in case of knock.

I can verify this tonight if someone doesnt confirm.

 

The Hall sensor in the distributor tells the ECU which cycle the engine is on so it can operate in sequential mode. To test the sensor you can use an oscilloscope to see if it has a square wave output. Alternatively, you can check for power and ground on two outer pins and with the opening or window of the trigger wheel at the sensor you should read 0v on the signal line and when the trigger wheel is solid at the sensor you should be getting some type of voltage, usually about 5v since the ECU holds the input high.

To test the crank sensor you need an oscilloscope to see the mag wave.

If you ground the DME Ignition output line and release it, it should fire both coils.

The ignitor needs to have big power and ground.

 

So Geoffrey. w/o the hall sensor... Will the car run?

 

I cannot recall on a stock DME whether it will or not. With the MoTeC, it will run without it, however, it may be running on the wrong cycle so the injectors, while operating in a sequential mode, they won't be timed properly. This is noticable at low RPM, part throttle where the injector duty cycles are low and the AFRs are off slightly due to the loss of efficiency from the fuel sitting on the back of a closed intake valve. At WOT, there really isn't any noticable difference.

Why doesn't someone pull their distributor connector and see?

 

Thanks for the explanation. I will tonight if noone confirms before then. Not near my car atm.

 

the engine will run with the hall sensor unplugged

 

kirk

i suggest you re-check the dist timing if you haven't done so already

 

"-The distributor hall effect sensor must be working. (I dont know how to check this)"

Not needed to run! Only for sequential injection & knock control (really not needed for this but the
way Porsche/Bosch designed the Porsche ECMs).

 

"I think if I jumper pin 18 (+ power) at the DME cable to Pin 1, each time I remove power the coil should spark. ??"
Why do that ? If you can put the scope on the dme to ignition switch line and watch for a wave , that would be quick and safe . Doing jumpers can cause problems .
If you have a dual trace , you can watch the dme side and the switch side of the cable .
If you have 3 traces you can watch the dme side and the two switch ends .
Scopes are safe , jumpers are sometimes not so safe .

"i suggest you re-check the dist timing if you haven't done so already"
The spark timing is done by the computer .
If the dist timing was 180 off , it would still have a spark . A spark at the wrong time , but a spark .

 

Well I had my friend take my DME down to a local garage, whose owner has the only other 964 I know of, a 1990 C2. He swapped my DME in and his car wouldn't start. I guess thats about the end of the diagnosis. DME is dead. (Owner owed me a favour because I always lend him my hammer)

Now I have to find a used DME. Preferably one of the later 964 series ones. I guess I'll post a thread and see if anyone has one they're willing to let go for cheap.

More $$$

Kirk