Several of my old (under 2GB) WD drives have stopped working also, and I imagine the few remaining will do, all the Conner drives I have no longer work also. Quantums and Seagates I have all function and don't seem to have any bad sectors developed on them.

As with everything: eventually all will fail.
Designed to last five years and reaching over 20-30 years is quite an accomplishment.

Most of my old drives (wd, Seagate, quantum, IBM, Samsung, Conner and so on) still work without a single hassle. But reliability? No.

Some failed drives still lay around, but there's nothing bound to one or two brands.

I already prepared myself on replacing hard drives with CF adapters. Not the real deal, but a second best. At least for the time CF cards will be around.

The problem lies within the aging mechanics.
Rubber becomes brittle, lubrication goes away. If some one had a 'clean room', refurbishing would easily be possible.
Unfortunately, that's rarely the case.
And thus, tiny dust particles would crash into the platters like meteors from space!

Edit: This reminds me of the energy savings nonsense.
From a personal experience, lamps that are left turned on, last much longer than those that are switched on/off regulary.

That's because of a change in temperature, which causes damage in the material over time.
Like a piece of wire that's bent back and forth, so to say.

Things like this makes you wonder if an old 386, with CRT monitor+screensaver, left turned on for 20 years might still be running. 🙂

Well, unfortunately the coating on the cathode will still slowly evaporate, leaving the crt with little to no brightness. The same happened in the past with vacuum tubes in radios. The cathode wears out when the heater is on and sooner or later you had to replace the tubes in your radio. Fortunately that was very easy back in the day, they were always socketed.

Thanks for message, one people confirm my experience with old WDC Harddisk (only old WDC with round shape hard case )

Quantum Fireball 1.7GB is actually main HDD of my 486 PC with homemade Vision968 card. Fireball runs fine.
I have open one WDC HDD, it is lost, there is no lost anymore. It spins up , head doing only period "klack" . Time wasted.

Yes, i make thougt about CF cards, i works sometime with industrial PC -> i collect CF card from scrap. 😉

Greetings
matt

I don't think one can say anything certain about HDD reliability unless they have tested large samples of the same model with known data of its power-on hours, which is the unit used by manufacturers to specify the HDD life expectancy.

In particular, for old drives that do not support SMART there is no way of knowing about the past usage of an HDD, so when you get a pre-owned drive it may be on its last legs due to running 24/7 for several years, or some drive from a grandma's house that only saw a few hours of use in its lifetime. And usually there's no way to know about that.

This is one of the reasons old SCSI drives seem to fail more often than IDE drives, but it's not surprising considering most SCSI drives have probably been running non-stop for several years on harsh server/workstation environments.

So, I don't think the drive brand/model is going to tell you much about how reliable they are (unless we're talking about known problematic models, such as the early 00's Deskstars). In my experience with old HDD pulls, I have only had one WD drive fail (2.1GB) while several Seagate and Maxtor drives from the same era were dead or near dead with faulty sectors. But again, I don't think that means anything really because I don't know the past history of these drives.

Make sure you RTFM. I threw out a bunch of old drives, that I now realise were probably OK, because they used non-standard configuration settings, and I was unaware of this. Years later I read the docs and realized my mistake.

I've handled quite a bit of old drives over the last years and I agree with the last post, drives I thought were dead were actually ok but needed a different configuration. For example some drives would refuse to work with a "slave" drive on the same cable but would work on their own, regardless of the jumper setting. Other drives that hadn't been powered for years actually needed a good shake to unstuck the internals, etc.

Western Digital Caviar drives around 1996 had a firmware issue that the firmware accepted the command "execute drive diagnostics" even before the spinup cycle was complete. Upon execution of that command, the voice coil started moving the heads from the park position into the drive. If the drive was not up to speed, this could cause head crashes that damage the inner tracks and cause wear on the heads. The drive firmware had the excuse that "it did not indicate the ready-for-command state" while it was spinning up, so no conforming BIOS should send the "execute drive diagnostics" command at that time.

At the same time, the Award BIOS had a bug that caused the "execute drive diagnostics" to be sent to IDE drives even if their status was not perfectly ready.

This situation resulted in updates to both the drive firmware (don't accept any commands during spinup) and Award BIOSes (wait for drive ready until a command is sent). One of the two fixes is enough to mitigate the issue. The private PC of my family lost a Caviar 2.5GB hard drive due to this flaw combination. IIRC it was replaced on warranty, though.

I don't see why, other than maybe congealed lubricant, they would all fail like that. I therefore suspect they can be "revived" if not reliable.

That said, I haven't done much checking on old hard drives. I just verified the failure of some I had issues with. One's a 91 WD which sometimes likes to be recognized, other times not. One's a WD Caviar (2850) which I am scrapping. It does run, it is recognized, and it will write and read data. It's just really really intermittent and slow. And therefore worthless.

I also switched on my oldest one just to see what happens. It's an 83 Computer Memories (CMI). Look them up if you like :p No idea if it "works."

Unfortunately I haven't had much of my hardware out in years.

On some HDDs, you can actually log-in via serial connection for diagnostics purposes.
Especially on Conner/Seagate models.

http://www.os2museum.com/wp/seagate-serial-talk/