In the OLS that was attached earlier there are the following that I would assume does just that:

"3D126A Codiermatrix für Codierzelle Getriebeart"
    Translation: "Coding Matrix for Coding Cell Transmission Type"

"3D1269 Attribut-Byte für Codierzelle Getriebeart"
    Translation: "Attribute Byte for Coding Cell Transmission Type"

"3D1485 Attribut-Byte für Codierzelle Getriebespreizung/-übersetzung"
    Translation: "Attribute Byte for Coding Cell Transmission Spread/Ratio"

"3D1486 Codiermatrix für Codierzelle Getriebespreizung/-übersetzung"
    Translation: "Coding Matrix for Coding Cell Transmission Spread/Ratio"

Nice find, it correlates with the ECU calibrations

If it matters, in the DS1 we change the power class at run time in ECU RAM to 0 and it is an architectural decision we cannot easily back out of. As far as I know it hasn't given us any issues so far across S6 and RS6/7, but thought I should mention it as you did.

Thanks for the tip,

the concept of powerclass and all of this is still new to me, so I will run my thoughts by you to see if they seem feasible.

On the ecu side my original plan was to change powerclass and/or the scalars projectls listed on the ecu side while retaining the S6 file.

Since you stated powerclass is set to 0 on ds1 for s6 and rs7, I imagine the powerclass mismatch must not be an issue?

An alternative to my plan was to flash RS7 bin via ds1, with the assumption that RS models are already equipped on the ecu side for Al551 input. Does this seem like a reasonable workaround?

So rs7 bin with ecu immo off + tcu immo off = should work maybe?

If not, I will look into coding the AL551 1st by changing ShrVar_xEngTyp_C on the tcu side to match S6 engine

if thats not enough I will look into trying to get the A6 Transmission to read like an RS transmision if possible, maybe cloning from RS TCM (both al551 might work? maybe?)

of course I will test and see, but just wanted to make sure there are no obvious roadblocks that would standout to someone with experience on this

We haven't heard of any problems by always having the Power Class set to 0 on original cars, chassis swaps, engine swaps, transmission swaps.

We've had people with ZF swaps ask us to change the ECU to RS7 and we can do that on request. You then update the DS1 and flash the ECU. My only experience of flashing all sorts of non standard versions to MED17 has been to do it through "OBD on the bench", with immo off. If you already have flash access to your ECU with a working immobiliser or the ECU has been bench flashed to immo off, you can flash the immo off file through OBD with the version change, and then ensure that every flash thereafter through OBD retains the immo off setting (which is in the xdf).

The flash of the ECU doesn't change the (emulated EEPROM) coding and I don't have info on what may also need to be changed there.

Bit of a jigsaw to piece together, lots of us have our own little niche on what we change, but I've basically swerved EEPROM and coding except to store ethanol content and map switch in it. As long as we know or contain the effects of coding on the way things like ignition tables are looked up, we don't have to do anything with coding/EEPROM when we flash or calibrate.

You can just go right ahead and flash through OBD. It is nice to have bench flashing available for recovery, but when working with a DS1 I would let it do its thing. After you update and get the RS7 software, it will force a full flash to ensure all the segments fit together. After that it will just do the 30 second calibration flashes. Just keep the immo off flag (in the xdf) chain intact always if you need that, break it once and you'll have to bench the immo off to get OBD flash access again.

Update on the Swap:

I got the transmission in.

I got a bench read of the a6 transmission flash and eeprom. I also got some help from ProjectLS for the AL551 immo and some help identifying stuff from artemesia.

After repinning the transmission connector, VCDS was able to read ecu and TCU fine. The only code was on the TCM ECU/TCU incompatible fault

I flashed the modified TCM immo off file without issues.

For ECU side I did a full flash with the DS1 RS7 Bin with IMMO Off. For some reason this completely wiped my ecu long coding in VCDS and some new codes appeared. Including an immobilizer fault. Even though immo has been off on ecu side and should be on TCM as well.

I will attach the autoscan. I am still digging, I have a couple ideas but I wanted to post here in case anyone else could chime in.

I do suspect maybe the issue is on ECU side and that maybe the full flash of the RS7 bin from DS1 was not enough, or immo is not completely off, may have to be adapted or flashed with ODIS on the ECU side if it comes down to it I'm guessing.

the flash file for tcm I was using is FL_4G0927158BD_006 from C7 RS6. I did not find a tcm file for C7 RS7 but I figured this would work fine with ECU SW 4G0906560B__0008

I have not tried to start the car as I dont want to lock the ecu. But the immo fault popped up after flashing the ECU file.

I also noticed that it is saying the TCM cannot be reached now. I repinned my transmission connector to match the A6 transmission. (Connection type 1 in the attached PIC) I am wondering if maybe since I flashed an RS7 TCM file I must also switch the pinout to another connection type? Just an idea.


My other friend who is doing this swap locked his ecu from immo at this step. I think he tried copying some VCDS coding to fill in the wiped coding on the ECU but it did not work so I will go another route


* I also attached the stock a6 bench read NNT code transmission in case anyone needs it or wants to take a look. I also attached bench read of TCM after flashing immo off patch and RS SW

Speculation only, but perhaps the ECU erases the long coding as a plausibility/safety feature in the event of a fault? Unlike on BMWs, we completely ignore the coding as part of the flash, only the fingerprint is written because it is a pre-requisite for flashing. We do flash all the OBD flashable segments in a full flash.