Vida CEM swapping

[dikidera]

I did a bunch of TCM adaptation dumps before and after adding Lubeguard Platinum + drain and fill of 2.3l of Ravenol T-IV fluid. It appears that a power cycle is required after driving to commit the adaptations to the EEPROM, luckily I have a base adaptation which I can use to restore them . I was unsure of the requirement for committing but it appears this is required.

In any case I am seeing the changed adaptations now, they are not little. I shall continue to observe and drive the car to find a sweetspot for dumping them again and then overwriting on each drive cycle to get my proper driving experience.

Obviously the biggest problem is the sweetspot will get overwritten unless the conditions are reverse engineered so we can alter them. As I've said before, not always does the gearbox adapt to how it should be.

[xanthefin]

I did a bunch of TCM adaptation dumps before and after adding Lubeguard Platinum + drain and fill of 2.3l of Ravenol T-IV fluid. It appears that a power cycle is required after driving to commit the adaptations to the EEPROM, luckily I have a base adaptation which I can use to restore them . I was unsure of the requirement for committing but it appears this is required.

In any case I am seeing the changed adaptations now, they are not little. I shall continue to observe and drive the car to find a sweetspot for dumping them again and then overwriting on each drive cycle to get my proper driving experience.

Obviously the biggest problem is the sweetspot will get overwritten unless the conditions are reverse engineered so we can alter them. As I've said before, not always does the gearbox adapt to how it should be.

Are you using Indicator mode to show with light if Adaption has been made (if its supported on your TCM)?

I loved the idea to make flappy pedals to change manually "gears" on my 850 though i would need find the culprit for speed limit in TCM which prevents manual changing at higher speeds.


My TCM has some commands which i need still decrypt what they really do.

btw What you were earlier looking for manual shifting i guess you look at AW55 GT with Lever unit but it does not example have manually command block functionality of automatic gear changing like 4T65EV-GT does... or you can bring that functionality from the firmware?

[dikidera]

These adaptations happen all the time while driving. It is not necessary to delete them. I have not deleted mine, but the experience worsens after initial drain and fill.

My tcm has the firmware with manual shifting completely removed from the firmware. I basically need a tcm with aw55 + geartronic firmware(actual tcm so I can install directly).

I've discovered some maps, but without checksum algorithm + ability to refresh I cannot edit them. At the very least I can improve my driving experience.

[xanthefin]

These adaptations happen all the time while driving. It is not necessary to delete them. I have not deleted mine, but the experience worsens after initial drain and fill.

My tcm has the firmware with manual shifting completely removed from the firmware. I basically need a tcm with aw55 + geartronic firmware(actual tcm so I can install directly).

I've discovered some maps, but without checksum algorithm + ability to refresh I cannot edit them. At the very least I can improve my driving experience.

You know checksum location?

[dikidera]

Not yet. Last time I looked at it, it was very unique in the way it worked. But I decided to focus on other things so I left it as an exercise at a later time.

[xanthefin]

Not yet. Last time I looked at it, it was very unique in the way it worked. But I decided to focus on other things so I left it as an exercise at a later time.

Have you tried just asking it from TCM? A5A601 you in theory get 4 byte checksum. Search it where it is then.

[dikidera]

Exactly, but the idea is to derive the algorithm in C or Python so you can have it before reflashing TCM.

The TCM exposes two variables via DHA. Expected Checksum and Calculated Checksum. The Expected Checksum is read from EEPROM. Or at least has something to do with EEPROM. I have not studied the code much more.

[xanthefin]

Exactly, but the idea is to derive the algorithm in C or Python so you can have it before reflashing TCM.

The TCM exposes two variables via DHA. Expected Checksum and Calculated Checksum. The Expected Checksum is read from EEPROM. Or at least has something to do with EEPROM. I have not studied the code much more.

Yes TCM Calculated checksum is by A5A701 and replacing old checksum with that in EEPROM. I thought being lazier for starters by using TCM itself calculate it.

[dikidera]

Oh, that seems like quite the helping hand . I am off to write a memory logger. There are some control variables I am very curious to know what they do.

And gears or something of the sort have the following values 0x101, 0x102, 0x103, 0x105, 0x10D, 0x207, 0x208 0x20A, 0x20B, 0x20C and so forth. A few control variables like FFFF88F0, 88F2, 88F4(I think) have these values, but I am not sure what they represent. Is it the next gear? Is it the current gear? Is it a bit pattern for gear + locked/unlocked torque converter? So many questions and very few answers. However h'F0,gbr(will vary for each firmware) is quite important here.

[prometey1982]

Oh, that seems like quite the helping hand . I am off to write a memory logger. There are some control variables I am very curious to know what they do.

And gears or something of the sort have the following values 0x101, 0x102, 0x103, 0x105, 0x10D, 0x207, 0x208 0x20A, 0x20B, 0x20C and so forth. A few control variables like FFFF88F0, 88F2, 88F4(I think) have these values, but I am not sure what they represent. Is it the next gear? Is it the current gear? Is it a bit pattern for gear + locked/unlocked torque converter? So many questions and very few answers. However h'F0,gbr(will vary for each firmware) is quite important here.

Here is the script for adding xrefs to gbr offsets. With xrefs code analysis is much easy.