Quote:
http://cgi.ebay.com/ws/eBayISAPI.dll...m=120127163992 Quote:
|
Quote:
Quote:
Charl |
Quote:
|
Quote:
You may also be able to find a +12V adaptor for your other laptop for less than $80. CompUSA just closed a bunch of stores across the US and probably had some (but it's a bit late in the game for them). This kind of thing frequently turns up in the clearance bins at the office stores. Also try eBay and your local Craigslist -- people kill their laptops and sell (or even give away) the still-perfectly-good power adapters and other accessories they no longer need. I've been looking for a free laptop to use for this kind of thing. Doesn't matter if it has an old, slow processor or a battery that no longer holds a charge. It just needs enough horsepower to run a comm program. I actually want this to hack TiVos, but it would work to talk to ECUs too. |
Quote:
http://cgi.ebay.com/ws/eBayISAPI.dll...m=110131694986 |
Quote:
There is no wire connected to pin 2 (receive data) or pin 8 (clear to send) of the PC serial port. It looks to me like they are not using the UART in the normal manner. They seem to be sending and receiving the data one bit at a time by directly accessing the handshaking lines. :eek: |
You mean like this thing I found in the basement with my dad's old cables? :rolleyes:
http://img523.imageshack.us/img523/4...rialqs2.th.jpg Great, what's step two? :D EDIT: The laptop I have I actually got from Vikash for testing his Scan Tool software. But he gave it to me without the brick and with a dead battery (thus why I got it), and I haven't found an adapter for an OmniBook 4000CT anywhere. |
More on the problem with some 92's
I wanted to put the "92 problem" into context and explain how it affects the VWRX select monitor software. I may have given people the incorrect impression that this method of extracting data from the ECU doesn't work on 92's at all.
There is a problem on the three 92 cars that have so far been tried, but as Charl rightly pointed out, we don't have a big enough sample group to conclude that all 92's have this problem. Nor are we likely to get one. Ever. Even if the problem does affect all 92's, it is not insurmountable and, depending on your application, you may hardly notice it. Kevin's VWRX software is the best software currently available and I assume that most people who build one of these interfaces will be running it. (At least until I learn windows programming and produce a better one, specially for the SVX ;)) This software does not validate the data coming back from the ECU as far as I can tell. It doesn't care if the address bytes in the reply is corrupted and it displays whatever data comes back for the user regardless of whether is it sensible. The upshot of this is that you can run Kevin's software even if your car has the same problem as mine. If the data from the ECU is getting corrupted then you will see the gauges on the screen fluctuating a lot. It's returning several values a second and unless you are doing some detailed diagnostics, it probably doesn't matter too much to you if some of them are incorrect. You can usually pick out the correct information from the values that are obviously incorrect. For my application, I need valid data to examine, to see how the ECU software works and to find the parameter addresses. That is why the corrupt data returned by my 92 ECU has been a problem for me. But now I have what looks like a workable solution. I hope that sets a few minds at rest. As I see it, the main stumbling block in this project now is the lack of good data logging software available. I would like to write some, but I don't know how to write windows software and it will take me some time to learn it to a good enough standard. Phil. |
Quote:
|
Quote:
Why? Well, if you don't know whether the device you want to communicate with uses standard RS-232 bit rates or word formats, this method gives an assembly language program complete control over the format and timing. The monitor code was first written to work with the PC's parallel port in a similar way (flipping bits "the hard way"). This circuit may have just been a cheap and easy way to use a serial port while re-using most of the previous code. The original code and parallel port circuit may have simply been a cheap and easy way to talk to the ECU without having to build an RS-232–to–TTL converter. |
Quote:
Sometimes the data quality is not too bad. Like on Friday night I could pick out one good response every dozen or so bad ones. I could work with that, and so could the VWRX software. On Saturday it was the same. Using the Friday night technique, I managed to copy the first 32K of the ECU address space to disk. It took nine hours though and flattened my battery. Unfortunately the most interesting part is the last 16K which I didn't get.:rolleyes: But on Sunday Morning, after recharging the battery, it was futile. There was no good data at all coming back from the ECU. Just pages and pages of random crap. No software could possibly do anything with that. Sunday evening it was a little better. I managed to get about 4K of good data from over the space of 3 hours but then it went back to random crap again. |
Just a thought.
What about running the engine for a while and then shutting it off and then dumping the memory? Your problem with the ECU lends a liitle more belief to the 92's having that extra 50 horsepower:D It appears Subaru didn't want anyone getting inside the ECU. |
Quote:
I had a crazy idea the other day that it was to do with the battery voltage. It was working (to an extent) until the battery went flat. After I recharged the battery it no longer worked. So I left the headlights on for a while to drop the voltage a little. Didn't seem to make any difference. I downloaded the firmware from my 92 TCU the other day. There are quite a few differences from the 94 TCU - although they are based on the same program. I can't help wondering if they made some changes on the later cars to try and stop the gearbox from self destructing. :rolleyes: I haven't got far enough with understanding the code to say what functionality has changed. It does raise the question of whether a 92 box would last longer if it was controlled by a 94 TCU. I'm also thinking that it must be possible to reprogram the TCU the way that LAN does for the ECU. There is a space provided on the board for an eprom socket. It's got the same number of pins as the socket on the ECU. Maybe LAN's memory adapter board would plug right in there. If so, it might be possible in future to reprogram the TCU for different shift points which could be useful for those who have raised the rev limiter on their ECU. |
The 94 and up have a speed limiter in one of the computers. Which is it? If it's the transmission computer, that might be where the difference is.
|
Quote:
Hopefully one day I'll understand enough about the TCU to say what the differences are. |
All times are GMT -6. The time now is 11:31 AM. |
Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
© 2001-2015 SVX World Network
(208)-906-1122