Will corrupted firmware sometime mimic the symptoms of failing heads?

[Teruel]

Will corrupted firmware sometimes mimic the symptoms of failing heads or bad blocks?

For example, sometimes when you go to image a drive, bad blocks appear everywhere on the hdd but when you fix the
firmware, they disappear?

[Nick_TS]

Hello,

Could you tell me model number of the drive you mentioned?

[DataMedics]

Sometimes FW issues can make a drive extremely slow such as with the WD Slow Responding issue or Seagate issues with the media cache. When this is happening the drive will tend to timeout before successfully reading the sector making it appear as a bad sector. Other issues such as translator problems can make all sectors beyond a certain LBA appear as bad because the drive isn't accessing the correct sectors. Each sector has a unique identifier stored in the drives servo (maybe not technically the servo but not accessible data you can read) that's used to verify the right one was read. The translator is basically a bunch of shifts where it skips reading a block of bad sectors. So if the translator is corrupted, it'll be trying to read sectors which even if they aren't bad are the wrong sectors with the wrong identifiers. So it'll return an error on every sector beyond where the translator is correct.

Just this week I had a case that really threw me for a loop. The drive was a Hitachi which seemed to have one weak head, all others read just fine. The weak head could read some sectors, but was very slow and hit a lot of bad sectors, so I disabled it and read the other heads out completely. After I went back to read the weak head and it wouldn't read at all, so I assumed it had failed. Replaced the heads, and it still wouldn't read. Finally I initialized the translator in RAM, and it worked just fine. 100% of the sectors were read in the end without a single bad block. The only thing I can imagine would explain what happened is that the translator went bad between when I finished imaging the three good heads and started imaging the one bad one. I can't imagine that a corrupted translator could affect only one head, but either way case solved.