Mike

If any of you have been following my tales, you'd know my 951 died on the road. Getting no spark was the symptom - there's no pulsing ignition signal from the DME to the coil.

I sent my DME to DEVEK Performance who were gracious enough to swap it into a working 951 and they confirmed my box was dead. I looked around a bit for used DME controller but gave up.

So I opened up mine and poked around. Tracing the ignition pin (pin 1) back I found the big transistor mounted 90deg to the board. This transistor is the driver for the non-existent ignition signal.

The board is marked BCE so it's a bipolar device - likely an NPN from looking at the circuit. I removed it from the circuit (by desoldering the skywire thingus) and checked it out - the Base-Emitter junction was OK (and confirmed it was a NPN device) but the Base-Collector was blown. So I went to the local high-priced TV repair parts joint and picked up a two-pack of 2N3055 devices for $2.83 + tax. These are NPN devices in a similar package to the blown part.

I sky-wired the part in for a quick check and the old gal fired right up. Woo-hoo!

I then carefully removed the old device physically (it's held in with rivets) and replaced it with the 2N3055. One problem was that the new device's case (collector) did not have its own pin (they assume you'll use a connection to the case for that) so I had to scrape the surface and solder a pigtail to it - which I then tied to pin1 of the big connector. Some care needs to be taken here as the case of the device must be insulated from the aluminum plate that holds it at the 90deg angle. This is the same situation as the original part - it had a mica insulator and two shoulder washers that must be re-used.

Now to life test it. This makes me a bit nervous as it is a single-point failure (what isn't, I suppose?) and might leave me stranded. So short trips are in store!

I was beginning to think I'd miss all the warm weather in upstate NY - it only lasts a few weeks

I've enclosed a few shots.

Here's the bad part with its cryptic numbering:




Here's a close-up of the replacement part after install. You can see how I soldered a lead to the case (collector) and ran it over to pin 1 on the big connector. Remember that the case pin was missing on this device. BTW, pin 1 of the big connector goes directly to the ignition coil.

Adrian

Dear Mike,
If these rather inexpensive 3055s last more than a few weeks I would be surprised. However you never know your luck in the big city. For more faith and reliability I recommend (maybe wait until this one trashes itself) you go a purchase the slightly more expensive Motorola or RCA versions.
Single point 3055 failures in these type of units is much more common than is reported. Plus well done. Most people cannot be bothered to do what you have,
Ciao,
Adrian
911C4

PS: You might be able to make a little business out of this.

Mike

Yes, the "character" of this device concerns me. I was lucky when the original died - I was pulling into a parking lot. Would have hated it happening while I was passing someone. This new part leaves me concerned about that. I intend to find a better part ASAP.

What I'd really like is to find the "correct" part. The marking on the original is clearly an OEM part number that means nothing in the real world.

Is the original really a 2N3055 or some equivalent? IOW, if I get my hands on a "real" Mot or RCA 2N3055 - will that be "good enough"?

Mike

From the "other" 951 forum
http://forums.rennlist.com/scripts/r...&f=15&t=000777

Mike

I've replaced the 2N3055 device with the more suitable BU941Z
http://forums.rennlist.com/scripts/r...&f=15&t=000777

icat

Great troubleshooting Mike! I have my own computer firm and we often just toss parts we think are bad due to the low cost of replacements. But when your dealing with a Porsche - everythings expensive! Keep the board posted on how your repair holds up.

ALS951

Mike,

How were you able to diagnose the problem? What type/kind of equipment did you use?

TIA
Al

Mike

Hi Al,

When the car first died, I discovered that there was no spark at the output of the coil. I swapped coils - still no spark. All along there seemed to be a signal present on the switched side of the coil's primary (using a scope) but I couldn't really tell if the signal was to spec.

I swapped my DME into another 944 turbo (actually, DEVEK Performance did that for me) and confirmed the DME was hosed. The DME essentially does two things - outputs the ignition pulses to the coil and outputs the injector signals based on a bunch of input signals. I was pretty sure I had fuel. My guess was that whatever was the output driver component for the ignition signal was the culprit.

I opened the DME up, followed pin 1 (the ignition signal) back to the part I eventually replaced. Understanding electronics helps at this point. The PC board was marked "BCE" telling me the part was a bipolar device. I isolated the device from the board (I just carefully unsoldered the leads) and checked the junctions with a DVM - found one was blown, the other junction indicating an NPN device.

I swapped in a lame substitute (the 2N3055) and the car fired right up (didn't run great @ ~4500RPM). With Wilk's help, I tracked down a more suitable part - the BU931 or BU941 series. These are Darlington devices which are actually two transistors in series - one driving the other yielding more gain. The are specifically made for driving ignition coils.

ALS951

Mike,

Wow, thanks for the reply, mine went out a while back and it was doing the same thing yours did. I did find out it was the DME by swapping one out of my brother's 951. But was trying to figure out a way to repair the old one. Any suggestions as to where I might be able to fix it. Considering I'm not much of a electronic guy.

TIA

Al 87 951 Zermatt Silver/Burgandy

Dash01

Did you ever find a suitable replacement ignition driver for your DME? Source?

Please advise, and thanks ver much.

StoogeMoe

If you know electronics, then you know you must choose the correct substitute based on all the parameters like collector current and so on. This trans probably gets hot, so make sure it is heat sinked properly or it will just fry again. That is if it even has a heat sink. I never saw it. If it does, make sure you use the heat conductive paste so the heat gets transferred to the sink.

I wouldn't worry about single points of failure. The DME is full of them. This transistor should last a long time. Good job on the diagnosis.

buckshot

a good general purpose darlington power transistor (npn)
is the nte- 263 most electronics stores should be able to get them in a day or two.
its rated at 10 amps, can't see how you could blow one of these in the dme.
we use this part in cash registers to drive some hi speed stepping motors.
just think, a p-car with a cash register part???
have fun!
buckshot wi

Danno

I think you want to get one rated for at least 20amps to have some safety margin. The coil's primary is around 1.0-1.5ohms and some people put in aftermarket coils with 0.6-0.7ohm primaries. So it'd be a good idea to get a higher-rated part.

Dash01

Per the thread from several years ago, a BU941Z transistor worked for Mike, IIRC.

If Mike (or anyone else who has successfully replaced his Porsche 951 DME ignition driver) would chime in here and tell us where he got the successful replacement, and how much it cost, and most importantly, how he correctly diagnosed the DME fault, I'd appreciate it enormously.

Dougs951S

Bringing this thread back to life. After fixing my front end, I went to take my car out for a test drive and it wouldnt start. This is the first time since buying it that it has ever failed to start. I swapped in a known good DME relay, and I have tach bounce. Car has fuel, but no spark when tested directly off the coil wire. I WAS able to get the coil to spark by running another ground to the coil and tapping it against the frame of the car, proving the ignition switch and coil are good. Also tested for continuity and confirmed the wiring from the DME to the coil is good. Jumped pins 9 and 16 on the klr and still no spark. The DME is not providing the ground pulse to the coil. I checked the boards over pretty carefully under strong light and magnification and nothing really seemed amiss accept for a small unrelated cap C890 that looked like it might have popped. I went ahead and got a replacement cap but I dont think it has anything to do with my starting issue. I know the large silver NPN transistor T504 is the ignition coil driver, and the cause of many related issues in our DME's. The solder joints all looked ok but I went ahead and resoldered the three leads on the board as well as the 3 skywire connections. The transistor its self looked really good, but im wondering if it, and not a solder joint failed? I have not put the board back yet to test it. Whats the best way to test if that particular transistor failed and if so, what is the best way to remove and attach a replacement, since it looks riveted?

Thanks

Edit: Damn if it wasn't just a cold solder joint. It just cranked up on the first attempt. Time will tell now how well those joints hold.