Couldnt really find a section for this..

[The1]

when i get home ill reflash one of the chips, ill go and drive it to the next km and then get another read and see if we can figure something more out then possibly.

[The1]

ok got it working now.
448982KM

Code: Select all

C0B5 E113 FF00 DF20
F20D DD22 6E91 BE41
F20D DD22 6E91 BE41
F20D DD22 6E91 BE41

not sure what the first line is yet but the number lines, each segements need to add upto FF, so DD+22=FF then you need to recalc the checksum at the start then it will work.

[festy]

does decreasing the number work too?

[VL400]

Nice work fellas

[The1]

Thanks VL, no cigar yet though, These worked for simplicity ill just copy the line, the header i used was

C0B5 E113 FF00 DF20

Code: Select all

448,982
F20D DD22 6E91 BE41

348,982
FC03 D32C 6E91 BE41

248,982
F40B DB24 6E91 BE41

Though a Random pick think i have it right, didn't work
1523km
6C93 FF00 57A8 3BC4

[The1]

tried a few others which worked, using same working

8,982KM
82km
10km
1km

now back to these

Code: Select all

1530 no work
A45B FF00 57A8 F30C
1529 no work
3CC3 FF00 57A8 6B94
1528 work
BC43 FF00 57A8 EB14
1527 no work
4CB3 FF00 57A8 1BE4
1526 no work
CC33 FF00 57A8 9B64
1525 no work
0CF3 FF00 57A8 5BA4
1524 no work
8C73 FF00 57A8 5B24
1523 no work
6C93 FF00 57A8 3BC4
1522 work
EC13 FF00 57A8 BB44
1521 no work
2CD3 FF00 57A8 7B84
1520 works
AC53 FF00 57A8 FB04

Some checksum related thing to do with the first line?

[festy]

odo.jpg.jpg (9.64 KiB) Viewed 4697 times

[The1]

haha yeh not to many people would have looked at that one in there car.

[festy]

These clusters normally read about 2% slow don't they, but I recall seeing a reference to some sort of 'police mode' that increases accuracy?

I was just thinking, word 3 (DF20) looks like it's self-checksumming (df^20=ff) but I've got mine set to ff04 and it's still happy - so that would mean word 3 was 16 bits of data, not 8 bits + checksum.
DF20 in native decimal is 1275, or 102% x 1250 (ppk) so 7402 (04E2 native, decimal 1250) should give the higher accuracy mode if I'm right.
As I don't have any method of triggering vss pulses just at the moment, it's only a theory for now...

Edit: on second thought, the dash would show 2% slow, not fast - so disregard this post

[festy]

Started putting together a basic UI for my programmer on the way to work today, calculating the values by hand and typing into hyperterm gets old pretty fast
Lots more to do (only calculates the checksums at the moment), but the plan is to have it plug in to the dash's programming port to save unsoldering ICs...

dashtool01.JPG (13.49 KiB) Viewed 4659 times