david@st ives

Geoff

Thanks for all your help, I cannot do any more this evening .

I am really stressed out , I have spent hours with the scope, flat batteries , lighting problems in my garage , freezing winds , nagging from the wife , working out how to post pictures, work , you name it ive got it.

I am just having my first beer since new years eve, so i am going to chill out and finish reading my book.

I just wish i had paid attention to some of the earlier posts, but when i could only spare the occassional hour on this problem somehow the advice gets lost on me.,this has been going back to March now , so i dont want it to be 12 months off the road.

Dave

Geoffrey

Loren,

The ECU needs to know which edge to trigger on. A sine wave is EITHER rising or falling and you must tell the ECU which it is. If you have a falling wave and you set the ECU to detect a rising wave or vice versa, your ignition timing retards as the engine RPM goes up.

"I am really stressed out , I have spent hours with the scope, flat batteries , lighting problems in my garage , freezing winds , nagging from the wife , working out how to post pictures, work , you name it ive got it."

Think of all of the learning you had today.

ABQ Slim

"Think of all of the learning you had today."

But at what price? Dave, I empathize, and commiserate. I've followed this thread for months, hoping for a satisfactory ending. Certainly, the techs and mechs and elecs who have tried to help on this forum are the best there are. They've probably forgotten more than I'll ever know.

But I know a thing or two about a budget, and how sometimes a monetary budget (and trying to stick to it) can wreak havoc with one's sanity budget.

"I am really stressed out , I have spent hours with the scope, flat batteries , lighting problems in my garage , freezing winds , nagging from the wife , working out how to post pictures, work , you name it ive got it."

If there's any way you can get the car to a trained professional, or pay to bring the trained professional to you, perhaps you'll end up in a better place. Even a temporary (seemingly exorbitant) expenditure now may save you sleepless nights, etc., and get you and your P-car back on the road sooner.

My Porsche prayers are with you, amigo. Best of luck!

Tony

david@st ives

Thanks for the support guys.

I did bring a Porsche mech to my house last week and he told me it was the hall sensor causing the problem and that as a starting point was to change the distributers.

I asked on here about the hall sensor and everyone was unanimous it wasnt that, the local auto electrition will not touch it because he has not got the fault reader and said it could cost thousands to detect manually.

My nearest OPC is over 100 miles away and that would cost me £200 ish for diagnosis 2 trips on a trailer , whatever that costs plus the hire of a trailer also my time in doing these trips.
All this equates to a lot of money, maybe over £500 before a part is even replaced add that onto what i have already spent , coils , dme test, dme relay replacement, not to mention the scope and leads.

I feel i am so close to the problem , there is only one place now the problem can be and that is between the ecu and the coils via the ignition modules.

With the help of these guys i am sure the problem will be cracked this week.

Dave

Geoffrey

You can test the hall sensor with the oscilloscope. It will provide a 5 volt square wave with one pulse per 2 engine revolutions. In fact, you can put the distributor hall sensor on channel 1 and the crank mag sensor on channel 2 and watch them. I probably have a picture of what it looks like on my scope if you need one.

The engine will run with a failed distributor sensor.

david@st ives

Well here we go first picture ignition off

second picture engine cranking.

I just used the tip part of the probe without the earth clip on no 1 wire just as the wire comes out of the DME

Red rooster

David,
You have the scope probe ground lead connected to the DME/car ground ?

If you have some jump leads, connect one to the battery ground terminal and then connect the scope probe ground lead to that jump lead.
Or if you have a ground terminal on the front of the scope connect the jump lead to that .

With ignition off you should see a straight , horizontal line on the scope, with the probe on pin 1 .
Anything else is a measurement problem.

Geoff

david@st ives

Geoff

Assuming the end white wires are no 5 pin.

I have found a signal at no 4 pin white wire on the ignition module nearest the rear bumber.

No signal detected at the 2 white wires at the other pins.

I will post the pictures

Red rooster

David,
DME pin 1 goes to ignition modules pin 5 ( white wire ) .OK , lets see that signal.

Geoff

david@st ives

Geoff what goes to pin 4 ?

Red rooster

David,
Ignition module Pin 1 Coil drive - Green/Red - Coil 1

Pin 2 Ground - Brown

Pin 3 input signal shield -- DME pin 2

Pin 4 +12 supply - Red/White - also Coil 15

Pin 5 input signal - White -- DME pin 1



Have you got a good signal on DME pin1 / Module pin 5 ?

Lets go through the system in order and then find where there is a problem ?

Geoff

david@st ives

Geoff

I have got a signal at pin 1 dme end

I am having trouble with my photo hosting.

When i put the ignition on i got a signal , then cranking i got a bigger signal.

I will go and test the other wires

david@st ives

pin 1 ok 12 volts

pin 2 earth ok

pin 3 signal similr to pin 5 but weaker

pin 4 ok 12 volts

pin 5 signal with ignition on turning approx double strength whilst cranking

Dave

Lorenfb

"The ECU needs to know which edge to trigger on"

Wrong! We're talking about a 964 and NOT a '78-'83 911SC. No guessing allowed.
The reference signal is a smooth sine wave with a missing wave and without "edges".
It appears to have to an edge when using a A/D digital type scope versus an analog
scope because of the digital's sampling rate. This should be apparent from the signal
posted by Dave's analog scope. Also, check here for more waveform info:
www.systemsc.com/waveforms.htm

You need to open the 964 DME and:
1. make a schematic of the circuits
2. understand all the I.C.s and how they function
3. make a test bench for running the DME
4. use a scope to analyze all the signals & waveforms
5. trace the reference signal from pin 47 thru ALL circuits
to the up interrupt input pin 12

Again, it's NOT edge triggered. If it were edge triggered, then ALL pulses that form the
reference signal have "edges", thus preventing discrimination. The "missing tooth" approach
is used on many Bosch based ECMs, e.g. Porsche, BMW, & Mercedes.

Bottomline: Let's not guess at how systems function!

Geoffrey

Loren, I will agree to disagree with you on this one. A sine wave is either rising or falling and the ECU needs to know which it is otherwise the timing of the crank pulse will be incorrect. As the engine increases in speed, the period changes and it matters which direction the wave is traveling as it passes 0 volts. It is easy to demonstrate with an aftermarket ECU and a timing light. You can't fake the facts and the facts show that if you have a falling wave and the ECU is setup to detect a rising wave (or vice versa) then as the RPM increases the ignition timing retards. You can try it by running it on your emulator with a crank reference sensor connected backwards and watch the ignition output pulse.

It does not matter which edge you trigger off of when you have a digital square wave, but it does on a magnetic generated sine wave.