[Levanime]

Originally Posted by denzelx

I don't understand the underlying memory concept of this chip.

So there are 3 different logical memory components:
- The D-Flash
- The EEPROM
- The P-Flash

As far as I understand, the EEPROM is simply emulated by a part of the d-flash? This means physically, that the d-flash and the eeprom are somewhat the same?
However if we interpret this chunk of memory as d-flash (which for some reason is not faulty at all), we then can transform this data into eeprom-format, write it back to memory and everything is fixed. I don't understand the logic behind this at all. What is exactly faulty in a bricked frm?

you are right.

what messes up is partition table, data itself (in DFlash) is intact, therefore EEE is recoverable

[Fx323i]

Hi guys, I have read through the whole topic from the start to the end and back and I need to ask the more experienced guys one thing:

I was using xProg v5.72, China clone. It always gave me "No BDM connection" error and could not read anything. After a while (like an hour) it gave me "Unknown command, software error" error. So I knew the Atmega64 firmware inside was shot somehow and I did a reflash to v5.51 software and I think v2.4 firmware.

Then I tried to read the data from FRM3, MC9S12XE384 and the D-Flash was empty. Seems like something messed up the data. What I did was that I found the correct, exact part number D-Flash, EEE and P-Flash files. The P-flash I write and read without any problems. But when I first erase D-Flash (telling it that EEPROM has 0 sectors), then write it, then erase the EEE (telling it the EEPROM partition has 16 sectors), then write it with a proper file.

After then I try to read EEE and it reads just fine, but once I read D-flash, it is empty. I know this corresponds to the information given in the window where I choose the number of sectors and it seems to be legit, but I still would like to ask you guys if this is normal. I am trying to repair the FRM module for a friend, this is my first time, and the car is 150 kms away from me. It would not be nice to drive 300 kms for nothing, so I am rather asking.

EDIT: I believe, after reading a lot, that the described situation is correct although the data size of EEPROM and Dflash are way different. As EEPROM is emulated by DFLASH, basically it only has one memory or the other (if talking about all sectors used and not mentioning the PFlash) and the Dflash can only be read as a whole memory in the case the EEPROM is corrupted in its partition. Am I right?

[Levanime]

Yes ur right

[Fx323i]

Thank you,
your reply helped me a lot to actually understand what is going on with the memory. Although I went through reading a lot I needed to ask because I am no expert in this area. Understanding the matter feels to me the best way to be able to repair these kinds of things, because just following some tutorials is a job for many people, but I do not feel like it.

So once more, thank you very much!

[tomvleeuwen]

Some programmers (like Orange5) allow you to read D-flash while the partition table is not corrupt. Hence you were able to find the D-flash dump from a good unit.

Also, it might seem that D-Flash is empty if you only look at the first sector, since it is possible that the first sector is empty but there is some data further down in the file. I wondered once for a long time why it was empty, so just a warning for everyone trying to recover via D-flash.

But since you have overwritten the data this information is useless to you...

[Fx323i]

You are totally right. I got several spare defective FRM3 modules from the dealer's shelf to have more background in testing, comparing the data etc.

Now I can see that the xProg version 5.72 of the China clone would have been able to read the dflash when the partition is corrupt. But, within the first try to read the dflash, the Chinese software I have in v5.72 deletes the xProg's serial number and then every attempt to read ends with "Waiting certificate" status bar.

Previously I tested it with v5.51 and I am able to write with no problem and read while the memory is not corrupt. This version is also by far the most stable according to all the information I found in other forums like MHH, where I can see many of you contribute, too.

Today I will reflash the Atmega64 in my xProg again and test with v5.55 and if not successful, I will go above. The information about working versions of xProg might be useful for someone else.

[Levanime]

make sure internet connection is off
or xprog files are read only

[Fx323i]

I had internet connection off and LAN/WLAN drivers disabled, it happened anyway.

Never mind, I managed to get a setup of xProg v5.55, which works fine for this purpouse. I just had to use a different EEPROM and FLASH files for ATMEGA64 in xProg, because those for v5.51 did not work with this version. Had to mess with it a little bit as I needed to convert them from bin to hex format, but finally I made it and now it works.

So the first version working for reading dFlash with corrupt partition is v5.55.

[Levanime]

good job,

I also have 5.55 on xprog, but I use it rarely.

[xBMW13]

I can't seem to get my xprog 5.55 clone to read the FRM - I soldered the connection points, but I get the error "Can't access BDM adapter". Is there a way I can test whether the xprog is working correctly in the first place? I'm not a programmer, but can follow instructions and technologically literate..

I have tried several of the recommended solutions outlined earlier in the thread (make sure connection points are contacting correctly, with and without the BDM adapter, use a serial cable, external power source, etc).

I now have a few questions after trying to get the xprog to work for the past 2 weeks on and off.

1. If you're using an external power source, would you still need to plug the xprog into the usb slot on the computer if you are using a serial data cable? When I try to do it without the USB cable, it states "the device is silent".

2. Is the VCC point on the FRM suppose to be contacted by the VPP wire on the xprog? There's no VCC wire.

3. Aside from meeting the minimum specs required for the xprog, could a computer be too old to work properly with the xprog? I'm using a very, very old computer that runs XP, and maybe that is the problem.

4. Is there a link/place where I can buy a xprog that has been proven to work by users here? I live in Canada.

[Fx323i]

Vcc = Vddr on xprog. Mentioned in the first post. On my xprog, there even is a Vcc on the BDM board. Which revision is your BDM board?

Seems like your problem is caused by power to the FRM3.

[xBMW13]

Originally Posted by Fx323i

Vcc = Vddr on xprog. Mentioned in the first post. On my xprog, there even is a Vcc on the BDM board. Which revision is your BDM board?

Seems like your problem is caused by power to the FRM3.

I have uploaded photos of my BDM adapter to imgur found here (

View post on imgur.com

). The revision appears to be Rev.C 2009.04.06.

That makes more sense now - so the VDD(+5v) is what VCC would be then, correct? and for the GND connection, which one would it be since there is 3 GNDs? I used the one on the left side under EXTAL.

[Levanime]

Bdm adapter is not needed for frm3

[xBMW13]

Originally Posted by Levanime

Bdm adapter is not needed for frm3

Yes, I have read that some users require it and others do not - I will try again later tonight without and without the BDM adapter to see if any works. Does it matter which GND pin is used?

[xBMW13]

Originally Posted by Levanime

Bdm adapter is not needed for frm3

Okay, I managed to get the xprog working thanks to Levanime's and Fx323i's advice and info.

I read and backed up the P-flash file, erased the EEE Prom, and flashed a new EEE Prom file. I then installed the FRM back into my vehicle and some light functions were back. I used NCS Expert to read my VIN from the CAS file and FA_write to the NFRM, then SG_CODIEREN the NFRM with a blank .MAN file. The NFRM VIN is still incorrect at this point.

Now I entered INPA to clear all errors, and this seemed to have restored most of the light functions and is throwing ZERO errors as far as I know. However, when I enter WINKFP to update the ZUSB, I can't find my number (9240529) hence I can't update/progress further.

Do I need to update the ZUSB? and if so, how do I go about this since I can't find my number and there doesn't appear to be anything wrong with my car anymore aside from the VIN being different on the NFRM still. Thanks a lot to everyone here for the information and help!!

Edit: The auto window can't close for some reason, it'll go up and bounce back down to the middle. I can manually close it, so it's still a win for now. Hahah!

Edit2: Found out that I can re-initialize the window to probably get it back to normal. Will have to try that the next time the vehicle is here, since it's not mine, to fix that.

Edit3: My hardware is the exact same as Xerix, and I'm experience the same WINKFP problem (ZUSB is the same as well since I originally flashed EEE Prom from the original post). Would you guys recommend me to follow through with the solutions you guys suggested for him/her to update the ZUSB (use files from the one suggested by levanime for Xerix, then repeat above steps, update using ZUSB 9383048, that should appear)? I do not have any problems that I know of thus far.

[Fx323i]

1. It is quite a pity you did not read through the whole topic. The best way to repair FRM module seems to be reading dFlash memory. When you try to read it, xprog tells you the partition is correct and then you choose the size of eeeprom of 0 sectors - then you read full dflash. A few posts back there is a link to this page. There you could have downloaded the dflash to eee convertor, convert the data and write back as eee.

http://tlvps.tomvanleeuwen.nl/frm/

Anyway, if you can not find your version of frm, I suggest you firstly try to load the WinKFP with the correct new SP Daten files. BMW Coding Tool utility is quite helpful regarding this.

[xBMW13]

Originally Posted by Fx323i

1. It is quite a pity you did not read through the whole topic. The best way to repair FRM module seems to be reading dFlash memory. When you try to read it, xprog tells you the partition is correct and then you choose the size of eeeprom of 0 sectors - then you read full dflash. A few posts back there is a link to this page. There you could have downloaded the dflash to eee convertor, convert the data and write back as eee.

http://tlvps.tomvanleeuwen.nl/frm/

Anyway, if you can not find your version of frm, I suggest you firstly try to load the WinKFP with the correct new SP Daten files. BMW Coding Tool utility is quite helpful regarding this.

Yes, I understand that now. I initially tried reading the dFlash too, but was getting a corruption error and used 16 sectors instead of 0 as you have stated (I thought it would be similar number of sectors as EEE Prom).

As for updating WinKFP with new SP Daten files, I double checked that I had the latest files installed (60.1) and that they were written/load into winKFP via BMW Coding Tool.

Now I have some specific questions regarding the flashing process so I can learn and get some things clarified.

1. Can the original dFlash be extracted again if it was never erased in the first place or does erasing/writing EEE Prom automatically overwrites the dFlash as well?

2. When doing the "old" method of writing EEE Prom with a dump instead of extracting the original EEE Prom, do you have to write the dFlash, and pFlash from the dump folder as well?

3. When you extract the EEE Prom from the dFlash, it does not require coding. Would any NCS Expert coding done on the FRM (such as turn signal blink count) remain intact still?


When I did my flashing, I essentially just erased the EEE Prom, then wrote the new file but did not write either pFlash or dFlash. (I did back up my original pFlash since I was able to read this with error)

[Levanime]

I dont have time to read all posts,

if you have overwritten eeprom then your original DFlash (on which eee is based) is lost.

Now, don't overthink it. If I'm not sure what you wrote there, but, basicly you should write eee only from same type of frm3 (look not at ZB but whats written on label (lets say e90 AHL XE, or E70 VDM) then find eee from same type of frm and just write it, thats it.

after that you will need to code it according to FA with ncsexpert.

if VIN number will cause you a problem you can change it with tool32, two jobs
c_fg_schrieben and c_fg_auftrag, you can see syntax in one of tool32 windows

basicly its 17 digits of vin + checksum, (you can see checksum in ncsexpert after reading FA.

whatever is written is in aif does not really mater, thats for informational purposes only.

[xBMW13]

Fair enough, everything seems to be working fine with no errors that I know of.

I will leave the ZUSB and not update it then, if it's not needed. Thanks for the help and clarification!

[Levanime]

if it aint broken, dont fix it

[n54ride]

Originally Posted by Levanime

here you go

https://dl.dropboxusercontent.com/u/...0e92%20e93.zip

Hello, I need 6135 9249082 frm dump to recover mine, someone has posted it 3 years ago but the link has expired.
a dealer tried to help but it made worse as he wrote some different dumps and it's not working at all, I need pflash, dflash and eeprom I believe to restore it.

thank you very much!

[Fx323i]

The requested file is in the google drive link in the first post of this thread