1985 911 starting problem

[Sid99]

I'm new to the forum. Thanks to all who write here.
I have a 1985 911 Carrera that failed to start when it was quite cold. I have the car in a garage and start it and sometimes drive it every week or two, normally. It turns over well but does not hit at all, and has not since the first failure to do so some weeks ago now.
With the key on, a voltmeter reads about 12.6 on each post of the coil, but, when I check for a spark out of the distributor, nothing. Although I have not always smelled fuel, I did the last try before writing this post.
From my reading it seems it may be the DME, but would welcome other possibilities. I do not have a friend who owns a 911 to try a DME from another car.
What things may be at play here, please, that I can try to pinpoint the problem? Maybe other tests I could run"?
Thanks for any help anyone might be able to offer.
Sid

[ischmitz]

Sid,

First the basics:

Have you confirmed you have power to the DME? Turn the key to ON and check that the silver canister on top of the engine (ICV) is vibrating. If not maybe an aftermarket alarm is killing power to the DME. Then, while you crank make sure the fuel pump runs (have voltmeter on the fuse and watch through windshield)

Yes it could be the DME but can also be as simple as a bad DME relay. These are cheap (around 25$ from our host) and you should carry a spare. They are known to go bad.

Then, if cars sit long the injectors get stuck shut and it takes some gentle tapping to make them work again. But that would still give you a spark, just no fuel.

The other component that will kill spark and fuel are the sensors on the flywheel. There are two and they do age and fail. You can Ohm them out for a quick check. Do a search for DME flywheel sensors. Lots of posts. You should see something like 900Ohm between the active pins if I remember correctly. If one is open the engine won't start.

Good luck,
Ingo

[Sid99]

Thanks for the reply. To eliminate that possibility, I have some time ago checked that the fuse for the fuel pump is getting power while cranking.
It may seem like a basic question, but where is the best place to check for power getting to the DME?
Thanks, again.
Sid

[Sid99]

Ingo,
Also, let me make sure that your comment, "Turn the key to ON and check that the silver canister on top of the engine (ICV) is vibrating," applies to the after-market alarm possibility, instead of the DME. The car has no after-market alarm.
Thanks.
Sid

[6771911esses]

DME relay under driver's seat. 9 out of 10 times.

[Sid99]

Had the relay out earlier but can't find a wiring diagram to see which pin to check to make sure there is power to it.
Do you have such a diagram?
Sid

[ischmitz]

Sid,

When the key gets turned to ON the first stage of the DME relay turns on. That powers the DME itself, the injectors and the coil and makes the ICV vibrate. So that is an easy check for whether the DME gets power. A vibrating ICV tells you the DME is alive, its chip is working and the internal CPU is processing code. If you don't get power the most common issues are an aftermarket or factory alarm system or a really bad DME or a bad GND or DME relay.

Then, once you crank the second stage of the DME relay comes on and powers the fuel pump and the O2 sensor heater. This click can be felt when touching the DME relay while starting to crank. Unfortunately, the DME relay turns on even if the flywheel sensors are bad. So you really need to get a NOD light onto an injector connector or an in-line spark tester to tell whether you get spark and fuel pulses during cranking. If not the common issues are the flywheel sensors or the DME itself.


Cheers,
Ingo

[ischmitz]

Originally Posted by Sid99

Ingo,
Also, let me make sure that your comment, "Turn the key to ON and check that the silver canister on top of the engine (ICV) is vibrating," applies to the after-market alarm possibility, instead of the DME. The car has no after-market alarm.
Thanks.
Sid

What I am saying is: it has to vibrate with they key in ON - else you are not getting power to the DME. The alarm is one of many possible reasons.

Ingo

[Sid99]

I did find, I think, what you called the ICV (chart in the manual I believe, if it stand for idle control valve), on top of the engine, and, except the color is tarnished, it is silver. AND, with the key on, it did vibrate.
I guess with that checked, the problem is likely to be the relay, sensors on the flywheel, or one of the weird connections corroded somewhere. Since is has been started and driven on a somewhat regular basis, it seems unlikely to be a corroded connection somewhere, although I understand that could happen at any time.
Do you have a wiring diagram or clue how to check the relay?
Thanks.
Sid

[ischmitz]

Next, ohm out the two flywheel sensors. Disconnect the 35-pin DME connector and measure inside the connector. Measure resistance between pin 25 and pin 26 (reference sensor) and resistance between pin 8 and pin 27 (engine speed sensor). Both should show around 900 ohm. You can also crank and should see about 1.5V AC with the meter connected to pin 8 and pin 27.

[Sid99]

Thanks so much. Will do some more work on tracking with this new information when I can get into my "car clothes," hopefully later this evening.
Sid

[hayday]

Has it been cold weather every time you've tried to start the car?

[Sid99]

No, not cold for all tries. I suppose the cold could have caused some failure if something close, but has been in the 50s and 60s sometimes.
Sid

[hayday]

is because it might be the cylinder head temp sensor. However, as said here already, I'd look at the DME relay, then the reference sensor. This sort of thing is often something simple. One day I shut my 84 off in the driveway, went out an hour later and it would crank but wouldn't start. It was the reference sensor.

[kidrock]

We need to make sure to not overthink this.

First thing for the 3.2: *always* keep a spare DME relay. I don't believe I've ever read a thread on how to "test" this relay. Whenever the 3.2 has a starting issue, we replace the relay. If the starting issue continues, then we move on to the basic analysis: fuel and spark.

Fuel: I prefer to introduce ether or carb cleaner to the intake. If the car immediately starts for a second or two then dies, then your car is starved for fuel and your work starts there.

Spark: I prefer to pull a plug and wire, and use a jumper wire from the base of the plug to a good ground on the engine (away from any fuel supply). Having a friend turn the key, I observe the spark. No spark? Then follow Ingo's instructions by the letter.

My guess is that you have a bad reference sensor. Just my .02 cents, FWIW

[Sid99]

Thanks for the helpful comments.
The reference sensor is one thing I haven't checked. Did try ether in the intake but still did not hit. After that checked for spark and got none.
With the weather, looks like Saturday is the day to continue. I have hope that it is narrowed down now enough to get it running.
Sid

[Sid99]

Found a relay to borrow, that came off a car that was running, and still engine does hot hit. I double checked for spark; none. And, fuel; ether still did not make it hit.
Seems reference sensors are next to check out.
Sid
Thanks again for all the tips. . . .

[kidrock]

Probably the ref sensor. If and when you're ready to change it out, do the speed sensor as well....they're both the same. Be careful to:

1) not move the sensor position bracket when removing and installing the sensors. Not a major deal, but you'll avoid having to space the sensors properly upon installation.

2) make sure to not mix up the connectors to the harness in the engine area, and

3) buy the identical but considerably less expensive BMW part (BMW part number 12-14-1-708-619-M14) and save some big $$$$

[Sid99]

I just checked the impedance and found 966 on the pins for the reference sensor, but no reading for the engine speed sensor pins. I then checked the voltage on the two sets of pins and got nothing on either set.
Given the comment by someone from another thread on the speed sensor sending a signal to the fuel pump to continue after cranking is stopped, and since I smelled fuel pretty strong after cranking to check voltage, I let the car sit a while and tried again. Still does not hit.
After all the other checks, it seems the problem is one or both sensors. Just wanted to see if that sounds right. Any input at this point would be appreciated.
Thanks again for all the tips.
Sid

[ischmitz]

yes, sounds like at least one of the sensors is bad, maybe both from your measurements.

However, for what it's worth, with bad sensors there won't be any fuel pulses. So even with the fuel pump running during cranking you no fuel is introduced to the engine. Maybe you simply smelled old fuel and richness in the exhaust.

Get the new sensors and let us know how things go from there,
Ingo