I don't think it would be possible to set the value low enough without firmware modification. I played around with this (the read not the write) on the few drives I had that supported it, and the values are limited to how low you can set it. Plus the values are based on 100ms increments, and 100ms is plenty of time for a successful read or write. You would need to be able to set it to 0, which is the default value which tells the drive to use its default recovery timeout.

_________________
http://www.hddsuperclone.com
Home of HDDSuperClone

If you have an old WD drive that is supported by WDTLER, then dump SA module 02 before and after changing the TLER setting. I suspect that module 02 is where this setting is stored. If so, then make the same changes to module 02 in unsupported drives and see what happens. I would check whether these changes affect the defaults for the ATA ERC command.

_________________
A backup a day keeps DR away.

Sorry I have not responded, I have been away from the forum for a couple days, and just checked back in. Glad I could have stimulated some discussion.

I could be wrong, as I am trying to learn this stuff by watching all of these discussions, but isn't Module 02 in the SA? One of the issues I am having with this drive is that I cannot read the SA with WDMarvel, because it does not become busy. I would be willing to try performing that Module 02 Fix, but alas, WDMarvel and HDDSuperTool both cannot get to the SA to perform any functions.

Let me better lay out what we have tried.

Brought the drive in because it was not recognized in customers machine.
Hooked it up to RapidSpar, and it was recognized and tested fine.
We performed a clone immediately, and while RS cloned it fairly quickly, About 70% of the files in the image were unrecoverable.
To get it off of our primary diagnostics machine, we put it on the DR machine in my office and tried to access the files on the drive with DMDE. DMDE saw the files and recovered files that were not recoverable on the RS.

It plugged along for just short of 2 months, but recovered the main 60GB of the files the customer needed asap. Decided I would try a clone with both DMDE and HDDSuperClone, but both reported between 180 to 250 days for completion at an average rate of 30k/sec.

Pulled the firmware off using my eeprom reader to see if I could find anything wrong with it.

So it appears that the files and structure are intact, but it is just responding really slowly. I dont want to have it running any more than needed at this point, because I dont know what is wrong with it.

I also considered this, but was not near my machine when I did so. I can certainly give it a shot when I get back to the shop.

Well that is unfortunate. I would like to thank everyone for their insight and assistance in helping me try to figure this out.