Help with Motronic no start

[88911coupe]

I réalise there a a million threads on this subject...I have many of them printed however I need some experts to help me determine what my current situation is. Quick back story, started having really rare non starts, replaced DME twice but so rare can't really tell what is problem. HOWEVER, today it would not start at all and I had a list of things to check when this happened, so hopefully my results will help some experts point me to the issue.
Car cold, not started several days.
Turns over great (new hi torque starter)
Relay "clicks" at on position
I checked ICV valve and vibrates in "on" position
Fuel bypass trick starts fuel pump, fuel pump hums away but car won't startporch911

With Ignition in "on" position I have power at pins # 30, 85b, 86 and 87. NO power on 87b and 85
Ignition in "start" same as "on" BUT now have power on 87B as well.
From what I have read this is as it should be...correct?
Does this help narrow down the problem?
I tested the TDC and Speed sensors and they both tested exactly as they should (at least as far as I can tell). This was done accessing the harness to the DME. 8 and 27; 25 and 26, both showed about 900 ohms. While cranking I got about 1.6 volt on 8/27 and under .03 volt 25/26, if i read my notes correctly.
I've tapped on the DME and the relay and no effect. I'm sort of glad the problem has occurred in the driveway and its no longer intermittent.
What should I check next while it's still acting up?
THANKS!

[DRACO A5OG]

Try to find a Pelican with a 3.2 and put your DME ( the ECM/PCU ) in their car to see if it will start. DO NOT put their DME in yours.

Do you smell fuel at tail pipe?

[darrin]

base problem is either fuel or spark (since you're still breathing when around the car, seems safe to rule out air . . .) -- your testing above doesn't indicate whether you've pulled a plug to see if it's sparking when you turn the engine over -- that's where I'd start with next steps, as it should help narrow down where to go next (i.e., if no spark, troubleshoot why you're not getting a spark OR if spark, troubleshoot why you're having a fuel issue)

Pulling a spark plug will also tell you what's going on in the combustion chamber (i.e. if it's wet and smells like gas, good chance you're getting fuel)

[rusnak]

shoot a couple of quick full bursts of starting fluid in the intake, then turn the key. If it coughs, then you have spark. If it starts, then you should check fuel pressure, change fuel filter, and start looking for possible vacuum leaks. You might also test the CHT sensor.

[Chuck.H]

Good chance you have a cracked solder joint in the DME computer...

I've found if you unbolt the DME computer (I'm assuming you mean DME relay in your description) bring it out from under the seat and shake it lightly while trying to start the car, it's a good test for cracked solder joints.

The most common cracked joints are the ones going to the ignition coil, because that transistor stands up off the circuit board, putting extra strain on the solder.

Chuck.H
'89 TurboLookTarga, 393k miles

[88911coupe]

OK thanks for the suggestions. Pretty sure this is "electronic" and not related to fuel pressure or filter since the issue is completely an of/off situation. Runs perfectly, when it runs, and won't start at all when it happens. Prior to today it would fire up, and run, perfectly after a few minutes. I was hoping the findings today would help focus on what the problem is, such as Chuck's comment that it's most likely with the DME ECU. I did have some bad solder joints fixed a few months ago so maybe another one has broken loose? Does the results on which pins the Relay have power point to anything?
I'll be doing a spark test after work.

[darrin]

good luck and keep us posted -- not sure the on/off situation clearly points to spark, as it would seem to me that if the power to the fuel injectors were to cut out, you'd run into the same problem (on/off situation), as with no power there'd be no fuel squirting into the engine.

[scarceller]

Attach a spark plug directly to the coil and ground the plug. Does the test spark plug have spark while cranking?

Your DME Relay voltages look fine. If you get power on Pin 87B while cranking then the 2 internal relays are fine, both the DME main relay and the FP relay have activated as they should with key in 'START'

I also recommend you test resistance of each injector if one shorts out all 6 go out because they are wired in parallel. The other trick is unplug 3 injectors and see if she catches, then try the other 3. Just unplug cyl1-3 then put them back and unplug 4-6 YES these motors will run on 3 cyls.

[88911coupe]

OK, just did spark test but not the way Scarceller suggested, I'll try that next. However what I did do was pull the #1 spark plug, hook it back up to its wire and then wrapped a heavy copper wire around its threads and then connected to the ground point on the intake. Tried to start and no apparent spark that I could see. I DID notice a faint fuel smell in the tailpipe after I tried this.

[ischmitz]

Hi Buck,

sorry it took me a little longer to chime in here. After reading all the suggestions I believe you have gotten very good information from Sal.

If your spark test is conclusive and you did get the DME reworked not too long ago I would focus on the sensors before suspecting the DME again. The reason I am saying this is as follow:

- no spark can be caused either by poor solder or non-working sensors (or sensor evaluation circuit inside the DME).

- if yours is intermittent this more or less points away from the sensor electronic. So it leaves solder or sensor.

- the "faint fuel smell" at the tail pipe is not conclusive. You really want to get a NOID light and check for fuel pulses. If they are also absent this shifts focus onto the FW sensors. The NOID light kits go for under 20$ on evil bay and really are a great help in diagnosing DME issues.

- If you do see fuel pulses it's either solder or something is wrong in the ignition portion of the DME.

Going forward, your test with the voltmeter in AC on the speed sensor gives it more or less a clean bill of health if you see an AC during cranking. For the reference sensor the test is a little more difficult and your result is not very conclusive since there is only one pulse per FW rotation. A better test is to use an oscilloscope to make sure the pulse if produced since 0.03V is not much to go by.

If you have access to another DME do the swap test. Either your DME into someone's car or someones DME into your car. One quick test for less than 20 seconds wont cause any damage either way.

If you don't have any access to another DME or other car contact me and we can work something out.

Ingo

[88911coupe]

Thanks Ingo, I just sent you a PM before I saw this reply. I did the spark plug test and didn't appear to get spark. I also put a test light on the terminals of the coil and saw power on both with ignition in the "on" position...not sure what this is telling me, maybe I need to actually use a meter for more specific info? I'll see about getting a NOID light next.
Thanks for the help!

[ischmitz]

Power on both sides of coil is normal with ignition on but not cranking or running. Think of the DME is the points. It closes path to GND momentarily to charge coil and then a spark is produced when DME opens path to GND again.

Any access to an oscilloscope or other DME?


Sent from my iPhone

[scarceller]

One more visual clue, while cranking does the tach needle bounce? If it does not then I suspect the ref or speed sensor.

[mysocal911]

Originally Posted by scarceller

Attach a spark plug directly to the coil and ground the plug. Does the test spark plug have spark while cranking?

This is the proper way to check for a spark, i.e. to eliminate the distributor and its
wiring. Checking for a spark is ALWAYS the beginning step in evaluating a no-start
condition as it's the easiest test to do! Next, one always checks for the necessary
powers to the control unit. After that, the sensors are checked.

[88911coupe]

Dang it..went out to do some more tests and the darn thing stared...hopefully it will act up again soon and I can do some more checking. I was looking for NOID lights and found one that says it has connections for Bosch and several others...I assume that's all I need to look for, correct?
Just to note...earlier today I'd pulled the wire from the coil off and cleaned it up, then re attached it and it didn't start. It did start just now so not sure if that may have had something to do with the issue.

[sugarwood]

Originally Posted by scarceller

One more visual clue, while cranking does the tach needle bounce? If it does not then I suspect the ref or speed sensor.

My tach does not bounce when cranking, and my car is running perfectly.

[88911coupe]

OK...problem back and tried to test with a loaner NOID light from AutoZone. There is one "test light" with BOSCH PFI on it but the pins seem a little narrow for the connector on my car. I can get them to make contact but there is a connector (GM I think) that looks like it would be more appropriate. I can get the pins to touch the female receptors in the injector wire and had my wife try to start the car. I got nothing on the NOID light. I can smell faint fuel at the tailpipe but have heard that does not necessarily mean the injectors are working. Should I try on of the other NOID lights or was my test sufficient to confirm no injector pulse?
ok went ahead and tried the GM PFI which fits perfectly and it indicates no power to injector.

hanks

[ischmitz]

Try plugging the NOID light in both ways. If still nothing there's a good chance your reference sensor is bad.


Sent from my iPhone

[scarceller]

I mentioned this before already, be sure you don't have one injector shorted out.

[88911coupe]

Thanks for the suggestions, Sal, thanks for that suggestion, have not tried that yet. What resistance amount am I looking for? Maybe its in the Bentley manual, I'll check.
Really appreciate the suggestions! Really sucks to be scared to take the car anywhere...