Better safe than sorry...its always best to have a backup of your flash and NVM EEPROM. Never know when yours will get corrupted or physically damaged. Just be carefully opening it and putting BDM on.

ok, thanks for input. i will buying a bdm next friday then. 
cant wait to get started! just need to figure out where everything is now.
for a typical "stage1" remap i assume its mostly upping the boost and maybe some slight timing changes???
my car is still completely stock, as i still have till may 2013 or another 10k to go.
i may put a new air filter in. and maybe a cheapo ebay downpipe. but thats it while its in warranty.

Does anybody know if the MT and DSG files are the same?  A friend's B box died, but the only ori I see is for MT - his car is DSG.

Can somebody sanity check my thoughts here?

So, with the checksum DTC set, I was forced to start digging to figure out how to un-set the DTC.  I didn't have an original e2p, so I took my spare ECU, flashed a working e2p and the fixed file.

I booted the ECU, cleared the DTC's, let the DTC's get set (it was on the bench), and then powered off the ECU and dumped the e2p.

Then I intentionally broke the file, flashed it back, booted the ECU, and verified the DTC was set.

Then I compared the two files.  Broken on the left, no DTC on the right:

00000102: F0 C4
00000103: 3D 67
00000104: 57 20
00000105: BF CE
0000011F: B5 DF
00000902: F0 C4
00000903: 3D 67
00000904: 57 20
00000905: BF CE
0000091F: B5 DF

So I figured I could fix this - but the checksum needs to be fixed on those pages... That's why 0000011F and 0000091F were modified and presumably at the end of the page.

I remembered Setzi posted about ME7 checksums, so I found his guide here:

http://nefariousmotorsports.com/forum/index.php?topic=362.0title=

Obviously the MED9 e2p is bigger, but why would they change anything?  I assumed it was divided into 128 pages of 16 bytes per page.

Since I don't know what the page info table is like, I just made some assumptions.

Based on Setzi's notes, I followed:


Boom.  A sample check of each page showed the last two bytes were correct for the above checksum routine.  MED9's eeprom checksum works just like ME7.

The interesting thing is my BDM would never flash e2p files that have been modified, during my immo-off quest.  It would always fail the verification.

I used the above method as a test to see if I could fix a file... and sure enough - it flashed!

Now to get this DTC unset...

Observation:

e2p's that don't flash via BDM either contain invalid checksums or if they are the 4096 byte variety don't have the first 2048 bytes matching the last 2048 bytes.

I got immo off working on my development ECU today...

My BDM will not flash e2p's that have been modified unless the checksum for that page is corrected.  It also won't flash an e2p that has a mismatch between the first and second half of the flash.

Clone running 3.1.0.1255