Toddimus

... build my own tester for my car.

After searching, reading and hemming and hawing, I've come to the conclusion that I'm going to make my own diagnostic tool for my car.

Everybody talks about getting / making a circuit board to talk to their OBD1 964 or Euro/Early 993 (non OBDII) car's diagnostics. Between the group buys of bare/assembled circuit boards and home brew versions (like I made for my 964), I got to thinking... Why shouldn't I be able to do it for my 1997 US 993? Upon a first reading of most threads, I was lead to believe that I was screwed. That it wouldn't work for my "late" model 993.

Then I read this...
He knows what he's talking about. JasonAndreas always pipes up with the crucial tidbits of info. You deserve more recognition than you get around here!!! Thanks again for past, present and, (hopefully) future help with these cars.

As Steven Colbert might say, ...

Moving on.

Ok, back to the original question. Why shouldn't I build my own tester for my car? Here's my options as I see them:

• Buy a PST2 / PIWIS / Hammer = Financially ridiculous. Thanks Porsche for your tremendously underwhelming generosity with your "proprietary code"
  
• Buy Durametric's solution. -- ~$250, still somewhat expensive, and they don't have as many features as I'd want (i.e. alarm, actual values, etc)
  
• Buy a Kragen/Ebay general OBDII tester -- ~$100, less expensive yet, but still expensive enough. Probably not going to get alarm, actual values, etc in this either.
  
• Make my own and use Doug B's software -- priceless! Well, if you pardon the reference to the TV commercial, it is priceless if I'm seeing it the right way.

Here's my rationale. Please let me know if I'm tripping out and missing something important.

With a home-made circuit (now mostly built on my kitchen counter), and Doug's software, I will be able to get more info than I could with the Durametric software, for very little money expense. Plenty of time expense, but that's free when I do the work.

Again, back to the original question. Why shouldn't I build my own tester for my car? JasonAndreas said it will work, and I get more functionality out of it than I can with anything else that's obtainable to me. I may still need to buy a generic tester to get the engine/smog readiness codes, but I'll still come out having saved some money. And I can't help but learn more things about the car and electronics. It seems like a win-win situation. I may even be able to find/buy another software package that can use this interface to cover the missing generic OBDII info. I need to do more homework to verify this.

I'm thinking about taking it a step further and adding a second tier to my project box to add a knock counter numeric indicator. I'd use some sort of microprocessor or commercially available purpose built event counter, and wire it up to some sort of display that I could see from the driver's seat. I might even wire up some sort of an indicator light circuit in the box to light up an un-used warning light in the clock to alert me of repeated knocking. The circuit would have some dwell time to the indicator signal after a knock is sensed to make it easier for me to recognize. More of a geek project for sure, but it wouldn't hurt to know if the DME thinks the engine is knocking. Since I don't have a Tiptronic, I'd use the indicator that looks like a Tiptronic shifter **** to remind me that I need to downshift. It would be like the opposite of those annoying upshift lights I've seen in older VWs. You see that light every upshift, but I'd never want to see mine. Upshift light = useless Downshift light = potentially useful

That's my Friday evening stream of consciousness post. Thanks for bearing with me, and please offer any input. I hope nobody bursts my bubble though and tells me it won't work!

Mike J

No reason at all. I used to write code for imbedded processors and I often wondered about writing some code that would decode all the fancy Porsche codes...I think it would be neat to take a protocol analyser while a PST2 is talking to a 993 and decode all the handshaking and command/response codes. You could then write effectively a PST2 with the Porsche specific codes.

But it all take time....and I don't have much these days..hopefully someone (like yourself?) would be willing to take this on...

Cheers,

Mike

Toddimus

I'd love to reverse engineer that thing. Problem is, I'm much more of a mechanical guy with some hardware and software experience. I am getting pretty good at LabView, so I've thought of trying to do an interface using it as a programming language. But why re-invent the wheel when Doug B has done most of the work for us? I'd like to work with him or somebody else to make a scantool.cfg and trouble codes.txt file that works for the 1996-1998 cars. Has anybody done this, or worked on it?


I just did my "alpha" test, and IT WORKS!!! (for the most part). Using the 993 scantool.cfg file for the 1995 cars found on Rennlist, my home-brew circuit (based on Ray's) and a project box, I think I was able to communicate with ALL of the systems on the car, except the SRS Airbag. At least I know I was communicating with something, and part numbers showed up. No error codes were registered, and some of the actual values/inputs didn't seem right. I was able to cycle through the climate control output tests successfully. I heard the servo motors and fans and solenoids doing their thing each time I hit the space bar during the test. That's always so cool to be able to press a key on a laptop and have your car do something. I guess I'm a geek, but it's pretty fun being a geek, and it comes in handy sometimes too.

Tomorrow, I'll do a much more thorough test, but for now it looks pretty good. I'll also post some screenshots of scantool4 and some photos of my handiwork. Stay tuned.

It's time for me to tune into Formula 1 at Fuji in Japan.

JasonAndreas

I'm pretty positive that the ISO9141 interbyte transfer time was changed for the MY97+ airbag controller so there is a good chance that it won't work. But, I haven't personally tried it so I can't say with 100% certainty. It might also just be an issue with the wrong control unit ID being sent out. Was the engine running or was the engine off but with the ignition turned to power-on when you tried? Did you tie the two K-lines together?

Toddimus

Since it was 11pm and I have LPMM exhaust, I decided to be nice to my neighbors. The ignition was on, engine not running. I didn't try tying the K and K' lines together. I'll give that a shot today.

Toddimus

And my interface box


I did try connecting the K and K' lines to connect to the airbag, as well as having the car started. No luck there.

There are definite bugs going on here. The actual values don't seem to be correct at all. Also, I tried to induce an error on for the DME (by disconnecting an O2 sensor plug), and on the alarm (by having the glove box open during arming). Neither error showed up on the scantool software.

Also, when I try to get the inputs or actual values (I can't remember which), the scantool application seems to lock up. I tried both circuit toggle switches, and one of them freed up the scantool application. Sorry, bad explanation of the symptoms and "cure", but I have more testing to do.

Cheers