Re: [opendiag] Vauxhall Vectra (1997) ecu communication

From: Bülent Özen <ozenelekt@...>
Date: Sat Oct 11 2003 - 13:07:54 CEST

Hi Ian

you have discovered the KW82 with keywords 0x52 and 0x80 . you send to ecu the invers of KB2 0x7F . it is correct.
The baud rate is almost always 4800. KW82 is a derivate of KW71 but runs mostly like KW1281
After the init you must communicate like kw1281. when you receive a byte , send inverted out. till you receive 0x03 <ETX>
than you can begin to send data and you receive the inverted back. dont forget to increment the block counter ( begin with 1) .

Sequence so far (PC set to 4800 baud)
PC 0xE0 YOU SEND THE ECU ADDRESS
ECU 0x55 0x52 0x80 YOU RECEIVE SYNC BYTE , KW1 AND KW2
PC 0x7F (after 25mS) YOU SEND INVERS OF KW2

ECU 90569797 0696 0696 3897 013 3 IT IS THE FIRST IDENT BLOC BUT YOU MUST SEND BYTEWISE COMPLEMENT BACK

PC 0x7F (after 25mS) YOU SEND FALSE DATA

more info you can find in KW71 documents.

best regards

Bülent
  

---
-- Original Message ----- 
From: ianmacnor 
To: opendiag@yahoogroups.com 
Sent: Saturday, October 11, 2003 12:19 AM
Subject: [opendiag] Vauxhall Vectra (1997) ecu communication
Sirs
I have been trying to communicate with my Vauxhall Vectra (X18XE 
engine with Simtec 56.5 ECU) using my PC and a "Jeff Noxon" 
interface. I have searched everywhere for some hints or 
information on this vehicle and have found absolutely nothing. I have 
studied and tried ODB2, KWP2000, 10400 baud, handshake bytes etc. - 
all to no avail. So, experiment I did.
Wake-up was the first hurdle. I tried sending every character from 
0x00 to 0xFF at 5 baud 8/N/1 and looked for a response at 10400 baud. 
I got a response only from 0xE0 but could make no sense of the 
characters I was receiving. After examining the bit pattern I changed 
the baud rate to 4800 and received back 0x55 0x52 0x80 which looked 
sensible, although not understood.
What then to send back to the ECU? After a delay of 25mS, I tried 
every character and checked for a response. 0x7F worked and the ECU 
then continuously sent out
90569797 0696 0696 3897 013 3 which looked like it might be some sort 
of identification together with date codes (the car was built in late 
1997). I responded with 0x7F again and received 
0x06 0xF0 0x55 0x10 0x32 0x10 0x01 0x9D which again looked 
intelligible.
0x06 - 6 bytes to follow
0xF0 0x55 0x10 0x32 0x10 0x01 - the 6 bytes
0x9D - a checksum (equal to sum of all the other bytes)
I am now in the position of many other opendiag members in trying to 
work out the meaning of these codes as they are unrecognisable as 
KWP2000 codes. Is 0xF0 the "tester" address and 0x55 the ECU address? 
What does 0x10 0x32 0x10 0x01 mean? What do I respond with?
Sequence so far (PC set to 4800 baud)
PC 0xE0 (at 5 baud using bit-bang method)
ECU 0x55 0x52 0x80 (this is repeated once more if I don't respond)
PC 0x7F (after 25mS)
ECU 90569797 0696 0696 3897 013 3
PC 0x7F (after 25mS)
ECU 0x06 0xF0 0x55 0x10 0x32 0x10 0x01 0x9D
?
If I don't make the PC respond after 25mS then the ECU continues to 
output its data continuously.
Regards
Ian McFarlane
Yahoo! Groups Sponsor 
ADVERTISEMENT
             
       
       
To unsubscribe from this group, send an email to:
opendiag-unsubscribe@egroups.com
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service. 
[Non-text portions of this message have been removed]
 

Received on Sat Oct 11 04:08:08 2003