What's the longest-running HDD you have???

[LongRunner]

NOTES:
Seagate, WD, and Quantum drives count in hours.
Maxtor drives count in "minutes" (actually units of 64 seconds).
Fujitsu drives count in seconds and at least some of them have a threshold for the runtime counter.
Samsung drives count in units of 30 seconds.
I'm not sure about Toshiba but I know they only rate their laptop drives for 20,000 hours.
If you have a ball-bearing drive as well as a fluid-dynamic bearing drive that has been running for even longer, mention both.
Also, if your longest-running drive has bad sectors, also mention the longest-runner that doesn't have bad sectors. (This means any type of bad sector reported by S.M.A.R.T., not just uncorrectables.)

Of my drives:
Longest runner - a Fujitsu MPE3064AT which I mentioned here.
Longest runner that doesn't fail S.M.A.R.T. - an ST3120026A that I occasionally use (34,358 hours and counting). No bad sectors, no CRC errors, no write errors. By the way, you'd hardly believe what model of power supply was in the PC I got it from.

[LongRunner]

Toshiba's PR department will hate me for posting this.

[c_hegge]

you'd hardly believe what model of power supply was in the PC I got it from.

Antec SP or SL?

[LongRunner]

Yep. Surprisingly, none of the caps popped.

[c_hegge]

I had an SP-350 come my way fairly recently with no bad caps, although I think it was only ever used like once per week as it didn't have a spec of dust in it. I replaced it as a precaution, as it was starting to whine slightly.

[Pentium]

I saw a Quantum Fireball come in at work manufactured in 2000 with ~77,000 hours on it. I was depressed when the customer requested we smash the drive

Second, a WD drive made in 1999 that had 65,000 hours on it and under 300 power cycles. Talk about leaving your computer on all the time huh?

[LongRunner]

This might explain the weird S.M.A.R.T. values on Seagate drives.

The Antec PSU is actually a True330, but that's still of the Fuhjyyu-ridden generation. It's rather dusty inside. At least the fans are both dual ball-bearing. I haven't plugged it in. +3.3V and +5V have CapXon GL; the other rails have Fuhjyyu TN. Primary caps are Panasonic but only rated at 85°C (which isn't usually a big problem but considering those PSUs' reputation for running hot, anything's possible…).

I think the failures of Nichicon VR on Intel motherboards were less to do with the temperature rating and more to do with quality problems (IIRC, Nichicon VZ has been reported to have the same problem).

[Pentium]

You would have gotten high ripple in that system even if the caps were good because they're general purpose caps on the 12V and 5VSB (fuyhjyyu TN) they have high ESR for a PSU. Most I have seen for output filtering are >0.08 ESR

[Wester547]

I don't think there's anything actually wrong with the VR series, they're just rated at 85*C, so they're somewhat ill fitting for motherboards... any motherboard that runs hot will cook them in less than 4-5 years of use. The capacitors of the 1990s that had quarternary ammonium salt problems were no longer being produced as of 1999, so the VZ series with date codes in the early 2000s shouldn't have that problem. I don't think the VR series is bad.... I've seen plenty of (well cooled) Intel motherboards with them last much longer than motherboards with crap like KZG, or the Teapo/OST/G-Luxon/etc motherboard failures. But KZG is truly awful. I see as many KZGs fail on motherboards as I do CapXons in ATX SMPS units.... why they're still being used and how they even survived torture testings (KZGs) mystifies me. They ruin good motherboards the same way CapXons ruin good power supplies.

That said, the only place for 85*C capacitors is bypassing/decoupling/coupling (all those tasks only involve passing a signal which doesn't pass much ripple through capacitors at all), and in the voltage doubler, and maybe in audio applications, like a sound card, anything that doesn't need low ESR or anything that isn't too stressful, and anything that won't heat up the capacitor, like so. So I think temperature rating is definitely the reason (remember that 85*C capacitors only have 1/4th the life of 105*C capacitors, and the VR series is only rated for 2,000 hours at 85*C, which is equal, in lifespan, to 500 hours at 105*C), not the electrolyte, aluminimum foil, cans, etc. On that note, though, unfortunately, what Intel did do on a number of boards is put VRs right near the linear regulators for the AGP slots, and those DID get hot, and eventually those capacitors would be shot because of that. Nice going Intel...

The Seagate read/write errors don't bother me, but those would be a problem on any other brand of drives, with the exception of Hitachi, I believe. The longest running drive I have is a WD84AA. 32,000 power hours, 7,000 power cycles, no problems at all.... yep, I don't run my computers 24/7 and I'm a power cycle person (though not excessively).

[c_hegge]

The only Nichicons to have had problems since 2000 were the HM and HN series, but that was fixed in 2005