VAG-COM/VCDS EVERYTHING

[heshamzn]

@heshamzn

It is necessary to solder 4 wires coresponding those 4 holes to your pcb. You need to power up your cable with 12v from a power supply or direct from your car-obd2.

How to power it up with my car obds ?? i think i will make china happy cuz i don't understand this thing att all

[DKa0S]

I'll! Maybe ABS control module fault, ECU reflashed, faulty or brake light switch faulty. I read that brake switch could cause that problems He replaces them and we will try again!

[macwol01]

Even cheaper version of the cable you refer to:
http://www.obd2cartool.com/vagcom-16...version-p-1266

Basically it's the same just older version but you can always reflash to Kolimer's.

These cables fully support VW Golf 6 (5K).
These cables support K2 with both jumpers (J1 and J5) removed).

What is the jumper setup for a 2014 Seat Leon F5 184 TDI mate ?

My cable is the same as this so after reading 384 pages not come acroos this snipet of info before.

[Tib82]

Hi Guys,

First of all I wish you a Happy New Year!

I checked Kolimer's v7.x packages and it's not clear to me why they have EEWriteLang and HWType files with different file size or data structure. What are the differences between them?
Eg:
EEWriteLang in v7.0 = 927,0 kB
EEWriteLang in v7.1 or v7.2 = 929,0 kB

[kaalis]

eewrite autodetect vid/pid checkbox, hwtype detects bootloader status.
Anyone can see those changes.

What is the jumper setup for a 2014 Seat Leon F5 184 TDI mate ?

Keep both jumpers removed and you will be fine.

[C3H8]

Elektrik,
at least I have problems to follow you.

My suggestion would be:
Let's focus on 32 combinations for input pins 2,3,6,7,8. These pins select the routing of other inputs (11,19,17,5,14) to outputs (18,16,12,13,15).
If you provide the result (which output is following a toggle on which input?) for every of these 32 combinations on pin 2-8 we have the logic.

A GAL 16V8 have some restrictions: pins 1 - 11 can only be an input, pins 12-19 can be both.

Happy new year!

Performed scan of 0x46 GAL, performed reconstruction using MATLAB, got 31 different cases, as it can be seen, CLK pin and pins >9 does not influence on output logic. GAL logic can be reconstructed with this truth table.

Attachment 583453

But it's incorrect, since I have filtered out rows by selecting only unique XOR'ed differences of input-output, which hides output pins changes. So I am leaving capture and formatted (raw) input combinations (187498 occurrences, including changing input pins) as an attachment. Every integer represents bit sequence in such manner: PIN1,2,3...9,11,12,...19. According to schematics, PIN 17 and PIN 19 should be inputs, not sure about OE pin.

[Elektrik1]

C3H8, Happy new year!

I have already dumped 0x44 and tested if everything corresponds, it fits fine your logic, also shows how different bits (XOR) corresponds to ones that are produced by you decompiled logic. The hardest part is to perform Karnaugh maps on many inputs (>8).

sshot-2019-01-02-01-47-08-1.jpg

[Elektrik1]

Actually, for 0x46 logic is very similar, but missing 9-th pin changes equations.

sshot-2019-01-02-03-09-32-1.jpg

[Elektrik1]

Also, C3H8, you decompiled logic is very similar to 'original' 0x44 version but produces a bit different output, four bits are different.
sshot-2019-01-02-03-33-40-1.jpg

[C3H8]

Elektrik,
regarding the jedec file: Please check your pin 11: your jedec file above (called 0x44) doesn't have the negation on this pin (if I decompiled it correctly since I had to count and copy all the bit positions manually). Please double check (in red) with
nc1=1 i2=2 i3=3 i4=4 i5=5 i6=6 i7=7 i8=8 i9=9 GND=10 !i11=11
o12=12 o13=13 f14=14 o15=15 o16=16 f17=17 o18=18 f19=19 VCC=20

equations

!o18 = !i5 & !i9
# !i8 & i11
!o16 = !i4 & !i9
# !i7 & i11
!o15 = !i9
!o13 = !f14
# !i2 & !f17 & !i6
# !i3 & !f19 & !i6
!o12 = !i2 & !f17
# !i3 & !f19

[C3H8]

Elektrik,
Please check: if this is correct than pin 5 have no influence to the logic. But this pin is connected to the FTDI and Atmel needs to see this pin always on RX0! On the other hand pin 4 have impact to the logic, but there is nothing connected in schematic.

Actually, for 0x46 logic is very similar, but missing 9-th pin changes equations.

sshot-2019-01-02-03-09-32-1.jpg

[Elektrik1]

Ah, my bad for 11 pin got it correctly in excel thou'.

Tried to improvise:
sshot-2019-01-02-14-07-53-1.jpg

Got green light, but not connects to any ECU, test passes
20190102_140933.jpg

Attaching excel, captured 0x46 logs, and MATLAB code used for analysis and unique combinations extraction

[C3H8]

Elektrik,
with your current logic the ECU can see all the traffic between PC and Atmel. This traffic is on pin 5 (TX TFDI) and a copy to pin pin 15 (RX Atmel). But it is also always on pin 16 and 18 (TX K, TX L). There is something missing in the first 2 lines to keep away all the communication between PC and Atmel, like in this example in red:
!o18 = !i5 & !ix # !i8 & i11

[Elektrik1]

Hehe, actually found the issue, one of wire got broken, so incorrect truth table was generated. Fixed, and got different result.
20190103_022111.jpg

sshot-2019-01-03-02-22-13-1.jpg

Filled in formulas for this table, and basically got hardware GAL delete performed on software (), and surprisingly it works fine on simple KLine radio, but perhaps won't work on LLine protocols and so on since a lot of cummunication crap is sent to K&L Lines

sshot-2019-01-03-02-24-34-1.jpg

So it seems I have filtered too much combinations leaving only unique combinations (1024 -> 31 combination), which are clearly not enough.

[Elektrik1]

I am leaving a final excel file here, formulas needs to fixed to match all greens on C3H8 logic test columns. We're very close guys, let's finish it!