I think I screwed up my (951) DME, s there anybody here who can help me out/

[wild man]

The thread is here;

Troubleshooting the DME - no ignition trigger pulses.

[Joe Bob]

Try Ingo, 3.6@cox.net

[rick-l]

Having no ignition pulses at the micro controller but having fuel injection pulses there does not seem possible. Are you sure you have the right schematic?

[wild man]

Wow, these threads move down very quickly in this forum. I don't post very often here because I don't have a 911, but I know there are memebers here who are very knowledgeable about DME's.

Rick, I'm pretty sure I've got correct schematics, for a late Motronic v3.1 turbo DME (0 261 200 088). They came off of clark's garage, with the names 944t-dme-a.jpg (analog board), and 944tdme-b.jpg (digital board), and when tracing the circuitry, everything seems to match up, even the 2764 (28-pin) EPROM. And the "systemc" site says that no ignition pulses IS the most common failure mode for that unit. But I can verify that the injector pulses are good, because I can smell fuel whenever I crank it, but there is no firing of the plugs.

I would really like to hear from someone that has gone through this, and what the component was, that was found to be the culprit. The systemc site does not provide that information.

[benp911]

go to rennlist and pm TT he is the 951 computer man. He wrote the vitesse chips.

[rick-l]

Originally Posted by wild man

But I can verify that the injector pulses are good, because I can smell fuel whenever I crank it, but there is no firing of the plugs.

Since you have it apart, sitting in the front seat, why didn't you put your scope on the FI pin?

[scottb]

Try Loren at Systems Consulting...Automotive Electronics. Love him or hate him...Loren knows his stuff when it comes to Porsche DME.

HTH.

[wild man]

Originally Posted by rick-l

Since you have it apart, sitting in the front seat, why didn't you put your scope on the FI pin?

I actually did, and the waveform looks exactly like the waveform shown in the book. I was just saying that the smell of fuel while cranking it has verified what I was seeing on the scope.

Scottb, I WAS on the systemc site, as stated in my last post, but he does not offer much (or should I say, ANY) info regarding possible failed components inside of the DME. He is basically saying that if it is not outputting any ignition pulses, the entire DME should be replaced. But what I am looking to do is replace only the component within the DME that is causing it to fail to output the ignition pulses. I saw a post from a while back from someone that was looking to replace the S100 chip (F12438-03 - on the digital board), which is apparently a custom chip, that I have no idea what it is or does. But the poster did not say what his symptoms were. What are the odds that THAT might be the offending chip?

[wild man]

Originally Posted by benp911

go to rennlist and pm TT he is the 951 computer man. He wrote the vitesse chips.

Unfortunately, I am not a registered member on that site, so obviously I would not be able to PM him. Slowly, but surely, I am making some progress on my own though. At this point in time, I have determined that everything on the digital board is operating correctly, narrowing the problem down to the analog board. I am not as good with analog stuff as I am with digital, but I think I'll have it licked within the next couple days. I will post my findings.

[rick-l]

I thought you said there was no ignition timing at the port of the micro controller?

Originally Posted by wild man

I've got a situation where the DME isn't putting out any ignition trigger pulses, causing a no-start situation.

I was able to trace the lack of trigger pulses through the analog and digital boards, all the way back to the output port of the 8051 microcontroller.

How about the cold solder joint at the big darlington?

[wild man]

Wasn't aware of it. Now that I just purchased a temp controlled soldering station, I'll re-flow all 6 joints (using Ag4 solder) at both the transistor and circuit board. Thanks for the heads-up on that.

There is a chance that I may have had the scope on the wrong setting, or something, when looking at the signal on the micro-controller chip, so maybe it is there. I did a swap of logic boards with another DME, so I know for sure that the problem is definitely on the I/O board.

[benp911]

The Rennlist 951 board is huge and thats where you will find your answers. The RL 911 board is tiny compared to Pelican. You want 911 info then this is the place, but Rennlist is the place for 951 info. It is free to post just like pelican. Many of us frequent both boards because they both have lots to offer just some boards have their strengths. When I had my 951's yes four of them!! I lived on Rennlist. Now with my 911 I get my information here. If I still had a 993 then I would spend more time on rennlist.
enjoy and good luck

[ischmitz]

If the 8051 doesn't generate ignition pulses on pin 2 the flywheel decoder IC could be damaged. It doesn't "see" the reference (TDC) signal and hense no spark signals are generated. The 8051 needs to get at least one pulse from the TDC sensor to start generating a train of spark signals based on the signal of the speed sensor. Check pin 12 and 13 on the 8051 for the speed and reference signals while cranking.

The 8051 uses interupt lines 0 and 1 to evaluate the speed and refernce signals. Both need to be triggered at least once after a power-cycle.

I am surprised that a wrong EPROM in itself would cause that sort of damage. More likely it might have been caused by ESD when handling the DME. Early ICs don't have a very sophisticated ESD protection and can be very sensitive to electrostatic discharge.

Ingo

[wild man]

I'm really impressed with you guys here on the 911 forum, with how willing you are to help a 951 guy like me out. Especially going out on a limb to recommend another site, which Wayne probably wouldn't be too thrilled about. But it does look like it would be to my benefit to become a regged renner.

It sounds like you are da man, ischmitz. I don't know how much of a software person you are, but maybe it will help if I provide some more detailed information on exactly what I did to cause the fault.

In deciding to become a DIY-tuner, I purchased a chip programmer. The first thing I did when I got it, was to read all of the porsche chips that I have, for the purpose of making backup copies of the bins, including the APE MAF chips that were in the car. Where I screwed up, was in believing that the APE DME chip was a 16K chip, rather than 8K. When I tried to peel off the sticker to verify the size, it started to tear, so I stopped.

In order to verify that the programmer was working correctly, I burned a copy of the DME chip onto a 16K EPROM, thinking that was the correct size. (I now have several 8K EEPROM's to work with) and installed it. I know it is possible to use a 16K chip, if the 8K of data is properly scaled onto it, but that isn't how things went down. Anyways, the car would not start. So with my fingers crossed (not literally, only mentally) I put the original APE chip back in, hoping it was going to start and run, but unforunately, it did not. At that point, I realized that I had probably screwed up. So I then grabbed a heat gun, which I should have done in the first place, and successfully peeled off the sticker to reveal that it was an 8K chip. Now I knew for sure that I had screwed up.

After some initial troubleshooting, where I thought there was no pulses coming out of the 8051, I hatched a plan to determine if it was a problem on the logic board, or not. I have another Motronic v3.1 DME (type 077) that I pulled out of my 87 924S, which I determined to be "almost" fully compatible with the 89 951(S) DME, which is type 088. After adding 4 pins to the EPROM socket and adding a (removable plug type) B700 jumper to the 87 logic board, I was able to make it 100% compatible with the 89 one. All of the differences are on the I/O boards.

The I/O board differences that I have picked up on so far are:

R250 - Present on 088 board, NOT present on 077 board.
C501 - Present on 088 board, NOT present on 077 board.
R525 - Present on 077 board, NOT present on 088 board.
R524 - Present on 077 board, NOT present on 088 board.
R519 - Present on 077 board, NOT present on 088 board.
R511 - Present on 088 board, NOT present on 077 board.
T820 - Present on 088 board, NOT present on 077 board.
D501 - Present on 088 board, NOT present on 077 board.
R251 - Present on 088 board, NOT present on 077 board.
R822 - Present on 088 board, NOT present on 077 board.
R823 - Present on 088 board, NOT present on 077 board.
R826 - Present on 088 board, NOT present on 077 board.
R513 - Present on 088 board, NOT present on 077 board.
R501 - Present on 088 board, NOT present on 077 board.
R502 - Present on 088 board, NOT present on 077 board.
Most of the 8 of what we used to call "select" resistors (I think they are calling them "equalize" resistors) are of some what different values, which is to be expected. And at one point, I believe I saw what was a diode on one of the boards, and was a resistor on the other, but I'm so bleary-eyed now, that I can't seem to find it.

Anyways, I dropped the 951 EPROM into the 077 DME, attached the connector, and the car started and ran. It didn't run great (somewhat missy), but it did run. Next step was to swap the logic and I/O boards. The result was that the car started and ran with the 077 I/O board with the 088 logic board, but not visa-versa. Therefore, I can definatively conclude, that the fault is on the 088 I/O board. With that, I must also conclude that the flywheel decoder IC (S100, on the logic board) is NOT the damaged component. For me to be able to do powered-up troubleshooting, I need to sit outside of the passenger side of the car, which is ont in a garage. There is mostly rain in the forecast here for the next few days, so I' not sure when I'll be able to get back out there on it.

With all of the info that I have just presented, do you have any other ideas, on what I should be checking out? Thanks,

-Steve

[ischmitz]

Steve,

I didn't read all the way down your previous post - this is why I missed your second statement that you might have had the scope on the wrong setting. So strike my comment with the flywheel decoder IC.

First off Bosch was somewhat methodical in numbering the components in these DMEs. Component numbers starting with 2 are in the power supply block, component numbers with 5 are in the ignition block and numbers with 8 are in the Lambda block. This will give you some ideas of what the differences are between the two I/O boards. Note that components on the digital board around the A/D converter could be different between DME versions as well. This is especially true comparing air-cooled versions to water-cooled versions.

From what you describe you seem to have a damaged ignition stage on your analog board. With an o-scope you should see a square pulse coming from the 8051 for the ignition signal. The analog board holds the driver circuit that switches the coil. It is essentially a low-side switch built with discrete components. This is more or less straight-forward analog electronics. Admitted - it is a somewhat complex pulse-shaping stage with current-sensing and a Darlington output. But with an o-scope you should be able to determine why it isn't working.

From what you describe when you put the 8k EPROM into the DME I still wonder how this could have caused damage the analog board. If the ignition driver circuit was working properly (including its current limiter) it should not have caused any damage even if the 8051 output was constantly on for several seconds. So I'd venture to guess there either was prior damage or it is a coincidence. If on the other hand the current limiter wasn't working you could have simply overloaded the Darlington transistor. It is the component in the TO3-like housing.

I have put many chips into these DME and even if you accidentally put it in the wrong way the DME normally doesn't get damaged in the process. It simply doesn't do a thing because the code is not running at all. As soon as you swap in a good chip it starts working again though. And while I did that I had the coil connected. So it should have gotten damaged but never did. The same happens if I miss-write an EPROM or forget the 700 jumper. I do work on these DMEs for quite some time now and even built myself a DME test station

[wild man]

Wow, that's some really interesting info on the Bosch methodology. You really know your stuff, when it comes to these boxes. I haven't had very much time in the last few days to work on it, as I've been pretty busy with other things, but I hope to be able to put some time into it tomorrow, now that I'm armed with more info.

I don't know if it will help , but I will try to recount what I recalled seeing when I was in there with the scope. I started out at pin 1 of the connector, working my way back through the analog and digital boards, to the 8051. What I recall seeing, was an initial spike when it was powered up (ignition turned on), but only flat-line while cranking. I don't recall exactly where the timebase was set on the scope, but I believe the spike is somewhere between maybe 10 and 100ms in width. I don't recall seeing the power-up spike at the 8051 ignition pulse output port pin, but If I have the opportunity, I will go back through everything tomorrow.

[ischmitz]

Steve,

right now I am sitting in a meeting in Tokyo so I don't have access to my stuff. I don't remember offhand how the ignition pulse looks like through the stages. At the 8051 side there should be a 0 - 5V (logic level) signal with a square shape. Not sure about its length but it will be in the lower milliseconds. If I remember correctly the outout is normally high (5V) and goes to low (0V) when an ignition event is triggered.

Now that you have ruled out the logic board simply follow that pulse from the 8051 pin 2 to the analog board. Eventually, it needs to show up at the base of the large transistor. There is one more inverter on the logic board before it gets to the analog board. The analog board receives a normally zero voltage signal that goes to 5V for a small period of time when the spark is triggered.

Ingo

[wild man]

Hey ingo, I finally was able to spend a little bit of time on the DME this morning. I seem to stuck on something, and here's what I got:

First let me state that I am comparing signals on the 088 (turbo) board, with signals on the 077 (non-turbo & no schematic) board, which is the one that IS producing ignition pulses.

After comparing signals on comparator S800, I am seeing that pin 6, the - input that feeds (via pin 1) the pulse shaper circuit remains at around +12 volts on the 077 board, and pin 7 (the + input for pin 1) has 30ms square pulses on it, while the 088 board does not. But when I try to trace the signal back to S800 pin 3, which looks like what is feeding pin 7 via some resistors, diodes, and caps, it doesn't look like there is any signal there that is changing in amplitude, so I'm not really sure where the pulses on pin 7 are coming from. Could you advise me on what is going on there?

[rick-l]

Since he is sleeping....
Pin 6 should have 2 volts on it was that a typo?
pin 7 is fed from S704 pin 10 on the other board. Is the signal there?
that comes from port 1 pin 1 of the 8051.

[ischmitz]

finally woke up (I am in Japan right now...)

Rick is right: The ignition signal should come from the digital board to R525 and R507 on tee analog board. Pin 6 of S800 needs to remain at 2V (defined by the voltage divider). If the comperator draws current through its input via pin 6 that is bad sign and tells you it is blown.

S800 is an example of an IC that by its number is identified as being part of the lambda function block but it is shared between the lambda and the ignition stage.

Cheers,
Ingo