Page 1 of 1

PF530 (2C84C54B)

Posted: 15.09.16, 14:45
by lcoughey
After ECC correction:

Chip 0
- dump 0 - 15.97GB of 16GB Bad
- dump 1 - 15.97GB of 16GB Bad

Chip 1
- dump 2 - 2.57GB of 16GB Bad
- dump 3 - 5.41GB of 16GB Bad

Thus far, we have been unable to read anything more.

Any suggestions on getting these chips to read better? Set to slow read hasn't made a difference. I will probably apply heat later, but am not optimistic that it will help dig out of this big hole.

Also, is it worth trying for a RAW recovery to see what I can get?

Re: PF530 (2C84C54B)

Posted: 16.09.16, 14:41
by Roman_TS

Re: PF530 (2C84C54B)

Posted: 16.09.16, 14:51
by lcoughey
Tried all three yesterday.

1. ECC - is how I got it to show the numbers above
2. Readout with read retry read not a single sector more
3. Readout with Voltage control and switched-off read retry read not a single sector more

Amazingly, I've been able to get the broken file structure and recover 7.28GB (2,901 Files) of good healthy files, with file and folder name, from the very broken dumps.

Re: PF530 (2C84C54B)

Posted: 16.09.16, 15:12
by Roman_TS
It might be mistake. Do not confuse those things. Your Flash drive might be filled by data on 15%, and all other 85% would be empty. If your CPU is AU669x or SSS (those one which are crypt not only DA but also SA with ECC), there might be situation when empty ranges would be marked as "uncorrected" - because there is no data in sectors, so there is nothing to fix.

You didn't tell that your controller is AU699x, and that's why you get almost perfect result with folder structure and good files.

BEFORE STARTING ANY RECOVERY PROCESS, DON'T FORGET TO ASK YOUR CUSTOMER HOW MUCH DATA SHOULD BE ON FLASH! :)

Because looks like in your case, total number of all files were aprox 8GB ;)

Re: PF530 (2C84C54B)

Posted: 16.09.16, 15:20
by lcoughey
If it were not for the fact that the file structure shows 80% of the folders (ironically enough, the most current) are empty, I'd think you were correct. But it is good to know of the issues with AU699x controller and how they store data on the NAND.