Wouldn't SCANDISK.EXE mark any faulty areas? Meaning you can use the areas that work?

I almost forgot about that. I might have to give it a try. 🤣

EDIT: For some reason, this disk I'm working with won't even take a format. I'm trying to format it to FAT32 using the Linux tool Gparted so I can scandisk it later, but it's taking seemingly forever. 😜 Should I give up at this point, or should I keep trying to format it?

EDIT2: I've decided that I'm better off taking this hard drive apart and stealing the magnets from it. 😜 I still have another drive to screw around with though that I know at least somewhat works.

I'd use HDAT2 to run a thorough surface scan, if you haven't done that already. Once you know where the problematic areas are, you might be able to avoid them. Partition the drive such that the bad regions won't be used. Bad sectors should be marked once you've scanned for them, but more of them will probably keep showing up nearby.

The most picky test seems to be the one in HDAT2 that it calls the "most powerful test". I've seen it make remarks about questionable sectors which ultimately "passed" but took some retries before doing so. Such sectors are probably only going to get worse in time, so they'd be good to avoid. That test is slow though and it gets the hard drive quite hot - enough that I'd prefer to put a fan over the drive while testing, if it doesn't have one already.

Sometimes you're lucky and the bad spots are clustered. Other times the damage is scattered everywhere across the disk.

should listen the HDD carefully for 'clicky' sound,
such sound usually indicates physicaly bad-sectors, the HDD wont last long as it usualy affect the 'needle' that reading the plate, you WILL see the bad sectors counts increases,

if its only magnetic damages, stuff like spinrite (commercial porgram) might help somewhat,

but if its hdd controler damages ...

shamino wrote:

I'd use HDAT2 to run a thorough surface scan, if you haven't done that already. Once you know where the problematic areas are, you might be able to avoid them. Partition the drive such that the bad regions won't be used. Bad sectors should be marked once you've scanned for them, but more of them will probably keep showing up nearby.

The most picky test seems to be the one in HDAT2 that it calls the "most powerful test". I've seen it make remarks about questionable sectors which ultimately "passed" but took some retries before doing so. Such sectors are probably only going to get worse in time, so they'd be good to avoid. That test is slow though and it gets the hard drive quite hot - enough that I'd prefer to put a fan over the drive while testing, if it doesn't have one already.

Sometimes you're lucky and the bad spots are clustered. Other times the damage is scattered everywhere across the disk.

That sounds like a good tool. I'll have to check it out.

I suggest listening carefully (or even touching the drive's outer casing) to see if it stops spinning on its own after you power it up.

The one bad WD drive I have actually spins down completely a minute or two after I start it up, and I get the impression that trying to run any kind of software on it at that point is futile. I keep hoping the Internet elves will some day find some ingenious solution to make it go that will not cost me thousands of dollars.

Damn, I just fried my spare 60GB laptop drive. 🤣 I knew the cheap adapter I used it with would do something bad eventually. 😜 I plugged the thing in, and only afterwards did I notice that I had missed the connection by one row of pins, causing power to surge through the drive's electronics and kill it for good.

I guess it's back to messing around with my unformattable drive and seeing how that turns out. 😜

It takes a long time to do a full drive format, and even longer when trying to format bad sectors. It could take many hours. I would recommend doing another surface scan after formatting completes, just to be sure.

Well, I ran HDAT2 on the "unformattable" drive (which actually turns out to be an 80GB 😳), and other than some bad sectors near the beginning of the drive, it actually appears to be fine. I'm going to try giving it another format.

EDIT: So far, so good. Running HDAT2 in it's most intensive mode seems to have straightened this drive out. 😁 I'm currently formatting the drive with Ranish right as I type this.

EDIT2: W00T! The format was successful! I only formatted half of the hard drive, but now I'm going to get an install of 98 going on it and see just how well it works. Ranish does a sector scan as it formats so I don't really have to worry about running Scandisk again. 🤣

@Shamino: Thanks for suggesting HDAT2. 😁 I've been looking for a tool like this for a long time.

I never heard of HDAT2 until now. Is it new?

TESTDISK used to be the hawtness.

If you've over-voltaged your hard drive you can get some life out of it by simply removing the TVS diode (Transitional Voltage Suppressor). Most drives that people think are 'fried' actually just have blown diode.

If you can solder, it would be best to add a new one as you don't get any second chances with it gone!!

Interesting reading here from Seagate but the info translates across all drives:

http://forums.seagate.com/t5/Barracuda-XT-Bar … -FAQ/m-p/118908

How very intriguing! I had not heard of this before. (Previously, people only proposed the elaborate idea of ripping off the control board and replacing it with one from an identical drive.)

I hate it that there may be a $5 fix for a problem that a data recovery company will charge thousands of dollars to solve.

That's exactly why I started digging around for info!

I found this out when I accidentally plugged the wrong adapter into my external drive that contained all my backups.

A data recovery company quoted me £150 to take a look and the recovery would have been £1200!!

A replacement diode cost me 35 Pence (approx 55 Cents) from Maplins!!!!

I recently had a 2.5" 250GB WD drive "die". It just sat there silent when powered up. Out of curiosity I gave it a finger flick and it came to life. Bad connection or some sort of motor issue?

I had a IBM Deathstar years ago that would just sit there clicking and not spin up. A tap to the case got it going too.

I also tried the "freezer trick", which some folks enthusiastically endorse, but had no success.

Machine_1760 wrote:

That's exactly why I started digging around for info! […]
Show full quote

That's exactly why I started digging around for info!

I found this out when I accidentally plugged the wrong adapter into my external drive that contained all my backups.

A data recovery company quoted me £150 to take a look and the recovery would have been £1200!!

A replacement diode cost me 35 Pence (approx 55 Cents) from Maplins!!!!

So was your drive spinning up at all? Did you have to solder the new diode in place?

As the TVS diode is directly in series with the 12V rail it will prevent the drive powering up if there has been an overvoltage event. Test your diode with a multimeter, there should be no continuity between the diode legs if the diode is intact. The diode will clamp small surges and recover but sustained (longer than about 250ms)surges will destroy it.

The drive can run with the TVS diode removed - just snip it off. Be aware that if there is another overvoltage then the drive will fail permanently. Without the diode even the smallest power surge is enough to kill it!

It is recommended that you only run the drive for as long as it takes to get your data off. Otherwise you will need to solder another diode in place. You can get them from RS cheaply, or, take one from another scrap drive if you have one.

In my case I replaced the diode and the drive has run for about another year o far with no further issue.

If the diode is in series, wouldn't you need to replace it with a jumper?

Nope, the failed diode acts like a jumper to ground. You have to have new one or remove it.

The diode is in series only after it has failed.

Machine_1760 wrote:

As the TVS diode is directly in series with the 12V rail it will prevent the drive powering up if there has been an overvoltage event.

So, you're saying your drive in particular wasn't spinning up at all? My external drive definitely spins up, and I definitely have a rapidly flashing LED, but like I said, the power cuts out pretty quickly. I guess the only way to be sure is to break out the ol' multimeter.