"It is easy to demonstrate with an aftermarket ECU and a timing light."

Right, some aftermarket ECUs may use an edge triggered approach but NOT the Bosch DME.

You obviously have not tested the 964 DME and don't understand how it functions, but have guessed
about how it must function. There are NO edges to trigger on, i.e. all the "edges" are the same. You need to get a better scope.

Bottomline: You're guessing again! Get a better scope so you can "see" really what's happening.

Note: If the wires (47 & 48) are reversed, the signal is the SAME, i.e. each tooth generates
the same signal no matter what the polarity of the wires.

 

Dave,
Sorry for the delay. Had to go to Walmart !

DME pin 1 / Module pin 5 should be a horizontal straight line trace with just the ignition on .

When you crank that signal should be the same horiz. straight line with square pulses going upwards. The width of the pulse should be much less than the gap between pulses.

___I---I______________________I---I_________ etc.

Thats about as close as I can get to give you the idea . Make sure that probe ground is connected to the car ground - battery -ve or a bunch of brown wires bolted to the body .

How does that look ?

Geoff

 

I think my $4000 Tektronics scope works just fine.

 

"I think my $4000 Tektronics scope works just fine."

Not based on the waveform posted. Dave's scope has a MUCH better image of the true signal.

 

Loren/Geoffrey ,
I think you are both right ! The sine wave crank sensor signal is passed through a level comparator where it is converted into 5v amplitude "square "
pulses with edges generated for counting etc.
The input comparator has a level of hysteresis for input signal noise suppresion. Hence the requirement for the signal to be above a certain amplitude .
On my motor simulator I tried reversing the input wave form before switch on ,while monitoring ignition timing etc. There was a hesitation while the system readjusted and then values reverted to the expected.

All the best

Geoff

 

"The sine wave crank sensor signal is passed through a level comparator"

Correct!

"On my motor simulator I tried reversing the input wave form before switch on ,while monitoring ignition timing etc. There was a hesitation while the system readjusted and then values reverted to the expected."

Correct again! Thus the polarity doesn't matter and it's NOT edge triggered, but only "looks" for the
missing tooth.

"Hence the requirement for the signal to be above a certain amplitude ."

Correct again! Thus, the missing tooth provides a reduced amplitude.

 

Loren,
The conversion of sine to square wave does imply edge triggering by eliminating sine wave amplitude from the equation.
If I move the missing teeth position ,the DME timing moves with it . No different to moving the flywheel on the crank .
Therefore the input edge after the missing teeth is a time reference for DME calculations .

Anyway , back to Davids no start !!

All the best


Geoff

 

"When you crank that signal should be the same horiz. straight line with square pulses going upwards. The width of the pulse should be much less than the gap between pulses."

Dont' guess. The ignition pulse is 0/5 volts whose pulse width varies with the desired dwell time
set by the microprocessor and NOT the tooth gap.

"The conversion of sine to square wave does imply edge triggering by eliminating sine wave amplitude from the equation."

No it doesn't! Edge triggering implies that the edge defines relative timing phase/position.
The DME doesn't sense ANY edge, just a tooth (a pulse/zero crossing). The comparator
just looks for "zero crossing" (the pulse) above a reference voltage which allows for discrimination
of the missing tooth. The rate of the pulses (teeth) and NOT any edge determines the RPMs
and the missing tooth (NO edge) determines the TDC point.

"Therefore the input edge after the missing teeth is a time reference for DME calculations ."

It's NOT the edge, but the pulse.

Come on, now you're guesssing too!

 

Geoff

I have packed away for the evening now, I have no lighting in the garage and my battery is flat again apart from that its freezing out there.

I will try again tommorrow about 4.00 pm uk time.

I am sure i have had that trace on my scope on pin 1, the other trace i described is with the var /volts turned more or less to max.

Dave

 

"Dont' guess. The ignition pulse is 0/5 volts whose pulse width varies with the desired dwell time
set by the microprocessor and NOT the tooth gap."

I take it the pulse is 0.5 volts is that half a volt.

If so i have got that voltage reading

 

David,
No , the pulse is 0 volts to 5 volts - 5 volts high .

Catch you tomorrow.

Geoff

 

Loren,
We are trying to lead a guy, who doesnt understand using a scope , along a path using the least technical terminology possible.
This is not a proper occaision for anyone try to demonstrate their respective technical prowess !
If it helps , I graduated in Electronic Engineering 1967 and have been involved with Defense / Automotive development ever since.
I know very well just how Motronic operates but see little point in this constant nitpicking .

You are stating that the time frame of crank sensor edges is irrelevant. That is nonsense. The missing tooth pulse is defined by its start and finish edges , as are all subsequent timing pulses .
As for the Pin 1 drive pulse , would you like the map address for the control algorithm ? or any other function in a 964 DME ??

None of this helps David resolve his problem at all , which should be our primary objective.I hope you agree.

Geoff

 

"which should be our primary objective"
How about , ONLY , objective ?

I know I'm nobody to tell anyone anything but ,
all this about edge , not edge , front / back side , bite me , etc , not really at all important in this car repair , is it ?

Well worth a new thread maybe ?

 

Indycam,
I stand corrected ! You are 100% right.
Being from the UK, I understand Davids predicement very well ,hence the help.

Lets hope that tomorrow we can get that motor running!

All the best

Geoff

 

"You are stating that the time frame of crank sensor edges is irrelevant."

The edge in itself is irrelevant for signal generation here, i.e. it's the relative voltage level of the
sine wave that provides the pulse. The 911SC ignition (and the early 928, Mercedes Benz)
which uses the edge dV/dT to generate an igntion signal. There the polarity of the edge
affected the triggering time, thus the ignition timing.

Yes, the thread does get off the main direction occasionally by mis-information being posted,
but that's one of the purposes of a forum, i.e. to allow the correct info to "surface".

The scope's settings to check the Pin 1 ignition signal are:

1. Voltage - 2 volts/cm
2. Time Base - 5 ms/cm

Note: If your probe is X10, then the volts setting is .20 volts/cm.
The scope's triggering is set to automatic.

As the engine is cranked, multiple pulses should appear. Also, check Pin 6
which is the tach signal that's basically the same (but 0 to 12 volts vs 0 to 5)
as the Pin 1 spark trigger signal. If Pin 6 has a signal, but no Pin 1, then:

1. bad Pin 1
2. shorted wiring
3. shorted ignition module input

"all this about edge , not edge , front / back side , bite me , etc , not really at all important in this car repair , is it "

But it is necessary, especially for those ignitions, e.g. 911SC, that it's critical.