Power to ignition coil - can I test with multimeter?

[CharlesJones]

Hi,

I am investigating a no spark issue on my 1984 3.2 Carrera. I have tested to see if there is power going to the coil by putting a multimeter between the positive and negative terminals of the coil with the engine cranking. This reads zero volts, but I have since read somewhere that you can't use a multimeter to test for power to the coil since the power comes as a pulse from the DME, and thus the multimeter will always read zero volts.

Is this true, or does my test indeed show that no power is going to the coil?

Many thanks,
Charles

[304065]

Charles,

At best, that test is inconclusive. At worst, you can be shorting out the coil driver on the DME board.

Why don't you start with the secondary side of the coil? Are you getting sparks at the plugs? Why not pull a plug wire, get a test plug, the kind that attaches with an alligator clip, and see if the DME is sending a signal to the primary, therefore inducing a current in the secondary and giving you a spark? That is also more valid as a diagnostic test.

Good luck!

[rusnak]

Yup.

[2.70Racer]

Charles,
12 volts should be present at pin 15 of the coil when the ignition is in the on position. Test for this with your black meter lead to a good ground, not pin 1 of the coil.
Pin 1 of the coil goes to the DME. The DME takes pin 1 to ground momentarily to fire the coil.
A simple test of the coil is to remove the wire at pin 1 of the coil, connect a test wire that will reach to a convenient ground (I think you folks call it earth), with 12 volts at pin 15, tap the jumper test wire to ground, any sparkplug connected to the output of the coil should then fire.
The DME does not send any kind of "pulse" to the coil, it just takes pin 1 of the coil momentarily to ground.

The reference sensor and/or the speed sensor tell the DME when to fire the coil.
If these sensors are original to the car they should be replaced, 28 years on the job, time to retire them.
Could also be the DME relay located next to the DME under the seat. A new relay here is good preventive maintenance.
Good Luck,

[ischmitz]

Have you verified that you get fuel? If you get fuel pulses but no spark pulses you could have an issue with the DME (ignition stage output). If both fuel and spark pulses are absent then the sensors could be suspect.

You can use an LED test light to check for fuel pulses on the injectors and pulses on the coil. Better yet an oscilloscope. A DVM is not good because it is too slow.

[CharlesJones]

Thanks very much for all you input and suggestions. I can see I wasn't testing power to the coil at all then - I just really hope I haven't damaged anything in doing so.

Thanks for the explaination of how the DME fires the coil Doug. I will try testing for voltage to pin 15 like you suggest, and if its getting power I will connect a test plug to the output of the coil to see if that is giving a spark. At the moment I have tried a test plug (i.e. spare spark plug grounded against the inlet manifold) in one of the HT leads and am not seeing any spark there. If I getting power to the coil but no spark from the output of the coil, I will try testing the coil like you suggest. If the coil is good perhaps I should just replace those speed sensors - it sounds like that would be good preventative maintenance even if they are not culprit anyway.

Yeah, I tried a spare DME relay but that didn't make any difference. I also tested the fuel pump by touching 12 volts to the fuel pump fuse, and I could hear it spin. I tested for voltage at this fuse with a multimeter during cranking, and can see its getting about 9 and a half volts (not sure why it isn't 12 volts).

I have read that the fuel pump should only operate during cranking, and not with just the ignition key switched to on, but before this problem I am sure I used to hear something whirr when I switched the ignition key to the on position prior to cranking. Now when I switch it to on I don't hear anything. I feel the ICV valve buzzing if I put my hand on it with the ignition on, by the way - not sure what that tells me, but I know it is expected behaviour.

I haven't yet verified that I am getting fuel. I can't smell any fuel at all in the engine bay, so I am guessing I probably am not. With an LED test light do I plug one of the injector wires into it to check for a pulse? Sounds like a very useful bit of equipment - I will try and purchase one tomorrow.

Thanks again for all your input. I have a good few more diagnostic tests to try - I will report back what I find.

[CharlesJones]

Hi there,

I just wanted to give you an update on my progress. Today I tested the voltage to the coil using my multimeter correctly this time, and tested between pin 15 (red wire) and earth, and then pin 1 (black wire) and earth. I found that I got 12 volts on both sides whilst cranking, so my previous test was clearly completely flawed!

I checked the resistance across the pins on the connectors of the flywheel speed and reference sensors, and found I had just over 1000 ohms across pins 2 and 3 for both sensors, so they look good. I then went back into the car and reached under the seat to put my hand on the DME relay to see if I could feel it click - I read somewhere that you should be able to feel it click twice when you switch on the ignition. However, I had already tried swapping out this relay with a spare, but that didn't make any difference, so I had swapped it back again. Anyway, I held the relay with my hand under the seat and twisted the ignition key and "boom", the car started right up! It scared the living daylights out of me. When you're concentrating on feeling for a tiny relay click and not expecting the engine to fire it gives you quite a shock when it does!

So now I am hoping that I don't have an intermittant problem, or worse still that the car doesn't die somewhere when I am far from home and not start again! I am thinking that maybe I just had a bad connection due to moisture or something - the car has been stood outside for a little while. Maybe my disconnecting and reconnecting the speed sensors helped, or even my holding the DME relay under the seat maybe bodged a connection to make it work. Like I say though, I did swap out the DME relay only the other day, so my money is on the flywheel sensor connectors.

What I'll do is use the car and see if the problem occurs again. If it does I will wiggle those sensor connectors and see if that helps, and if not I will wiggle the DME relay and try that. If that doesn't work I will call a tow truck!

I will update this thread again when my problem resurfaces (it seems inevitable that it will). Thanks again for all your help and support - you really have all been a terrific help.

Best regards

[rusnak]

Charles:

You feel for the second "click" on the DME relay to check for the signal from the ref sensor to the DME. This is a lazy way to do it, but in a "no start" scenario it usually works so you can move on to the next item.

If it dies again, then I'd suspect a faulty cable, or faulty ground.

[CharlesJones]

Thanks for the explaination. I really wish there was a write-up somewhere which explained all this stuff - I don't even understand what the ref sensor is for. In fact, I don't understand what both the ref sensor and the speed sensor are for, or why there are the two of them when they are identical sensors!

Yeah, I suspect a faulty cable - maybe it was my disconnecting and reconnecting the sensor cables which wiggled it and got it working again.

[rusnak]

The ref sensor tells the DME when to fire the plugs and open the injectors. The speed sensor tells the DME what RPM the engine is turning. You can troubleshoot a bad ref sensor by plugging in a good one and pulsing it against a metal object like the fender while cranking the motor over with a remote starter button.

[ischmitz]

The two most likely culprits for intermittent operation on the 3.2 are the DME and the DME relay. Both are known to suffer from issues that "come and go"

If you have the next no-start issue make sure you have a known good DME relay in your glove box to exclude it as root cause. I am a little concerned that you previously swapped the DME relay and it did not solve the issue. Assuming that you installed a new DME relay and still had the no-start and then went back to the old relay only to find the car fired right up tells me both relays are O.K. and you most likely have something else going on. Just put the new relay in and make sure it runs on both. Then put the new relay away as the spare.

The sensors are more prone to failing under temperature. Cable issues are not very comon. To make matters worse there is some complex logic to the sensors:

You need the reference sensor to function at least for one pulse to get the engine started. After that the DME "occasionally" looks for the reference pulse to make sure timing is synched to TDC of the engine. The speed sensor signal is required at all times and that is what tells the DME when to fire the spark and injectors. An intermittent reference sensor is very hard to diagnose. As someone else mentioned there were some instances where an aftermarket flywheel didn't have the right material pin installed to trigger the reference sensor. To diagnose you can plug a spare sensor into the harness and wave a metal object in front of it. One pulse is enough to transition the DME into 'run' mode.

If either sensor is not present (disconnected, damaged) the engine will never start or even sputter. Not a single spark event will be issued and a single injection pulse will be issued. However, once the engine is running you could unplug the reference sensor and it would run on for a little bit.

If the engine doesn't start but the exhaust starts reeking of unburned fuel after several minutes of cranking you most likely have a bad DME (failed ignition stage). That failure can be very intermittent. One day it runs perfectly fine and the next it is dead as a doornail. Then all of a sudden it works again only to fail in the most inopportune moment. Sometimes it is just cutting out.

The only way to properly diagnose that is to disassemble the DME and inspect with a high-powered microscope for hairline cracks in the PCB.

Good luck,
Ingo

[dshepp806]

Complexity aside,.....make sure the FEMALE side of the DME relay does not have any issues as to connectiviity........

Otherwise: the DME may require some investigation.

BEST!

Doyle

[Lorenfb]

"After that the DME "occasionally" looks for the reference pulse to make sure timing is synched to TDC of the engine."

Not really! Once the engine starts the ref sensor is no longer needed. It can be left
disconnected until the next time the engine needs to start. This DME/Motronic
functionally exists both in the earlier BMW DME and the later similar BMW DME.
Disconnecting the ref sensor while running can help evaluate the speed sensor.

[ischmitz]

Loren, I disassembled most of the code for the Bosch DME's (Porsche, BMW and a few others) and the reference sensor signal will be used during engine running conditions. There is a sort plausibility logic in the code that checks signal integrity of the speed sensor and "resets" TDC location under certain conditions (jitter).

The FI2438-03 is armed every time the MCU "writes" to a certain address space in external memory. Once armed it will trigger a non-maskable interupt at TDC.

The fact that the engine will continue to run when you disconnect the reference sensor might not cause an issue during idle. But I wouldn't take the engine through the rev ranges if I were you.....

[Lorenfb]

"The fact that the engine will continue to run when you disconnect the reference sensor might not cause an issue during idle. But I wouldn't take the engine through the rev ranges if I were you....."

Really, kinda short sighted, right? This has been a know fact for the last 15+ years to many who have
troubleshot both the BMW & 911 3.2. Only when the speed sensor is weak, sensitive to noise, or the ring
gear is bad, there'll be a problem. Those that have another view point have not worked extensively
on problematic BMWs & 911 3.2s in the field to test this or don't fully understand the earlier DMEs.

Bottom line: Try it, you may learn something that'll help you troubleshoot in the future.

[CharlesJones]

Thanks for all the responses guys. This is all a very interesting discussion. So just to recap to make sure I am understanding this right - the DME just takes a cue from the reference sensor at start-up to locate TDC, and then (possibly - under discussion) checks this occasionally after start-up, whereas the speed sensor is used continuous to track the speed of the engine so the DME knows when to ground the coil for spark and when to trigger the injectors?

Ingo - thanks for the diagnostics tips. Indeed with my no start issue the engine was completely dead - it would crank of course but there was no attempt at firing at all, and not even a hint of unburnt fuel smell, so I guess the injectors weren't doing anything either. What I will do next time it fails is fiddle with the sensor cables and plugs, try it again, and if that doesn't work then fiddle with the DME relay, and then try it again. When it started yesterday I had actually physically touched both of these things prior to the start, so I can't isolate either one.

I really hope its not the DME itself. If fiddling with both the cables and the DME relay doesn't have any effect the next time the problem resurfaces (of course, they might have nothing to do with the problem - if might just have been coincidence) then perhaps I will just send off the DME to a company to have it checked over. Its not something I would have the confidence to attack with a magnifying glass and soldering iron!

Thanks again for all your responses. I will update the thread if I have any more news.

[Chuck.H]

Hi Charles -

The DME is a very simple 8051 based computer - if you're careful to avoid static build-up, it's easy to open and inspect. The coil drivers are the components that stand up on heat sinks above the circuit board. Look at the underside of the board where these components attach and you might see the telltale signs of a ring around the middle of the solder joint. You may even be able to wiggle the component lead in the middle of the solder joint. Using a low power soldering iron will allow you to touch these connections for a few seconds to 'reflow' the solder joints.

BTW, your problem was exactly as mine when I had an intermittant loss of spark...

Hope this helps,
Chuck.H
'89 TurboLookTarga, 333k miles

[CharlesJones]

Hi Chuck,

Thanks for the tip - that is very valuable post. Indeed this could well be my problem - if anything I did got the car to start it could be that my taking hold of the DME relay knocked the DME enough to wiggle such a bad connection as you describe. If your problem had identical symptons then it seems likely. If the problem reoccurs ("if"? I should just say "when"!) I will take a look at the DME like you suggest.

That's impressive that your car has done a third of a million miles by the way! It doesn't surprise me though - these cars are just so beautifully made that I could imagine them going forever with the correct care and maintenance!

[healy2442]

Hi Charles
I was searching the internet for a resolution on the exact same problem you are
having with yours!
My Car is also a 1984 3.2 carrera and isnt starting i have replaced both sensors, DME Relay,
and coil the car ran fine for months until here i am now back with the same problem
no spark from the coil when i am cranking but when i turn on the ignition there is
a spark from the spark plug momentarily confusing! I am going trough all wiring again looking for a bad earth somewhere hopefully.

Its incredible to think that i have the same problem as you and im in Cork

Regards
Brian

[CharlesJones]

Hi Brian,

That's a coincidence alright! It must be something to do with the Irish weather!

When I was searching for similar problems I found a thread where someone had the same sympton you describe - just a single spark when they turned the ignition on, but none when cranking. Unfortunately though it was one of those threads where the poster never added a concluding post, so I don't know what they found to be the cause of the issue in their case. Maybe you should take a look at the DME like Chuck suggests above - that's one thing I will try if my problem resurfaces and cable-wiggling of the sensors doesn't cure it.

You have an excellent taste in cars by the way! If you should find yourself over this neck woods please give us a shout. Maybe we could hook up over a coffee for a chat.

What do you do about road tax incidentially? Do you pay the 1600 euro a year or whatever it is?! I am so much looking forward to 2014!