drchambers

I have a 1995 C4 with 79K miles. My intermittent no-start and stumble at 3K RPM has turned into a permanent no-start. There's no spark.

DME Relay

• Pins 1 & 3 are hot
• Pin 6 gets ground cranking and 5 is hot
• Pin 8 no ground and no 12V at pin 7

DME Plug

• Pins 18 & 27 have 12V, nothing at 37
• Pins 48 & 49 show about 2.9VAC when cranking
• Pins 45 & 14 show 2.8 Kohm

Ignition Module Plug

• Pins 1 & 6 have 12V
• Pins 2 & 7 only have .11VDC measured from pin 4

I pulled the cover off the DME and found nothing obvious. I touched up some solder joints on the power transistors but everything looked pretty good. Fuel pump runs when jumped. I recently replaced the crank sensor and thought that had fixed the no-start problem. It was a cheap replacement, so I wouldn’t completely rule it out. The car has a TPC supercharger installed by TPC in 2002.

Does the DME only need to see the crank sensor and temp sensor to generate a start signal? Crank sensor signal too weak?

IainM

When you say DME, do you mean the relay or the ECU?
I ask because there’s no transistors in the DME relay.
have you tried a new DME relay?

drchambers

The relay's good. It isn't getting a ground signal to turn on the pump, which seems consistent with no trigger input to the ignition module.

But to answer your question, I did try my spare DME relay as a first step.

jay@EZimmoblock.com

- Check again...If PIN 5 on the DME relay is hot, Pin 37 on the DME ECU plug should also be hot - they're wired together.
- No GND on DME relay PIN 8 is what happens when the ECU doesn't detect engine rotation.
- 2.8K on the temp sensor is fine, but I don't believe it's required for start.

systemsc

The temp sensor resistance value is never a condition for a no-start, i.e. there will always be spanks and injector pulses irrespective of the temp sensor value.

drchambers

Good to know. I was going off of the graphic in the shop manual under "DME Diagnosis" which has an "x" at "Engine will not start" and "Engine Temperature Sensor 2".

systemsc

The Bosch 993 DME ECM is very reliable and rarely if ever fails. Problems do occur, on the 88 pin Bosch ECMs, when soldered in chips are removed and replaced for re-flashing ("performance tuning").

drchambers

I went out to confirm voltage on pin 37, and nothing, key on or off. I didn't think to crank it. I do have 12V on pin 36 key off.

I plug the connector back into the DME and turn the key and it starts!!! I then pulled the crank sensor just to do a visual in case I screwed up the install and it looks fine.

I'd generally be the last person to blame the ECU, but I don't know what else to conclude??? Grounds seem OK. Holds just under 14V when running.

jay@EZimmoblock.com

When you turn the ignition on, the ECU turns the first stage of the DME relay on, putting 12 v out on DME relay 5 which is wired to ECU 37, so those 2 pins should always read the same voltage (provided the relay is installed). When/if rotation is sensed, the ECU then turns on the DME relay second stage to power the fuel pump.

Sounds like your "interrmittent" may have decided to go away for a while.....

Moose

May be random but my clutch switch (at the pedal) went out causing intermittent start problems. Since you are looking, may be one thing to check.

systemsc

The clutch switch only causes a no-crank mode. The starter relay (fuse box) can always be bypassed, if there's a no-crank mode.

drchambers

I put the original crank sensor back in and I get a more reasonable 3.5VAC at the DME plug. The problems persist though. Intermittent no-start and intermittent problems under load at roughly 3K RPM. It either stumbles and bucks, acts like it hit a rev limiter, or sometimes runs great. The MAF is OK, the throttle sensor runs through a range and the fuel pump is good, so I would guess the no-start and stumble have the same cause.

I think the only signal that the TPC computer intercepts is the crank sensor, but that seems to be OK when it doesn't start. I don't have a scope, but the VAC seems good. The circuit board is conformally coated so a visual inspection didn't help. The no-start originally seemed heat related, but now also happens cold.

I keep coming back to the DME but the solder joints look OK and Loren's comment about 993 DME reliability makes me question that the DME is the problem.

Any suggestions?

jay@EZimmoblock.com

Just to clarify: when you say "no-start" you mean it cranks but doesn't fire, right?
Getting back to your original post:

The ECU has to ground DME relay Pin 8 to get power to the fuel pump through the second stage of the DME relay. If that's not happening, you have a problem.
The ECU grounds DME relay 8 when ignition is ON and it detects rotation. No rotation(or rotation signal) = no start.
How did you measure for gnd on DME relay pin 8? Relay in?

The fact that the original crank sensor is outputting a different level than the one you took out might just be differences in the sensor and/or installed gap. There's an acceptable min/max range, just not sure what those limits are. Since you have a problem with both sensors I'd stay with the original for now.

systemsc

You need to check the DME ECM spark outputs (pins 25 & 31) to the coils driver module. Since you don't have a scope, use your VOM set to VAC while cranking.
Then using a test light, check each coil's primary for the spark module output signals while cranking. You also need to use a noid light to check for injector signals.
If you have no spark but an injector pulse, the problem is in DME ECM.

jay@EZimmoblock.com

If that's still the case, THAT's the problem.
Maybe attach temporary pigtail wires or a switch across DME relay contacts 3 and 7 and next time it doesn't start or stumbles, make the connection.
You checked the raw signal from the crank sensor but you should also verify that the ECU is detecting that signal by looking at the ECU "Tn" output signal on the OBD port pin # 14. If not present there, the ECU isn't seeing the sensor and won't turn on the fuel pump.