@Br*ste1 One possible solution is to use a modified procedure of the one I posted.
Basically, you could try:
1. backing up the unique part of your Reader using dd. I don't remember the exact command line. @Rupor can you give the command to use.
2. Backup a good PRS-T1 completely.
3. Restore the backup to the faulty one followed immediately by
4. a restore of the unique raw data block at the beginning copied in the first step.
When I had checked my PRS-T1 for bad blocks it found some here and there but my procedure made my Reader behave "as new".