aging HP C-drive bites the dust, and then...

Thanks Skynet. Believe it or not I abbreviated that a little. Black-screen started immediately after my trying to use the ancient recovery DVD. But at that point it was also doing a much quicker boot-fail cycle, in which the fans barely get up to speed before it would give it up and then try again.

After unplugging all external devices and finding the same rapid cycling persisted, based on some old recollection of that behavior relating to RAM failure, I unseated the RAM sticks and then re-seated them one at a time. It didn't short-cycle until I got to the last RAM stick. Not really buying that, though...so to test that too-easy nonsense diagnosis, I then swapped the 3rd and 4th RAM sticks. The problem stuck with the SLOT, not the RAM. I tried some other RAM configurations, and eventually it stopped short-cycling with all four sticks installed, so that was yet another mystery that never was solved. They're 8GB sticks. Is there a minimum I should work with just for diagnostic purposes, or is this a wrong tree to be barking up to begin with?

Oh, and to be clear, no, I didn't pull that EPS cable - at that point. I did at one point start to pull the mobo out, the next day, thinking I'd like to see the back, but in so doing, I realized that in the crappy light, I'd reconnected old HD and DVD drives to wrong SATA ports. I'm still not 100% sure these are right, but I've got HD (or new SSD) connected to SATA0, and DVD to SATA2 now. Also extracted the vid card and went back to the original mobo DVI connection to my monitor.

Well, here's the post-mortem and resurrection report. I don't know how it will seem to the rest of you, but to me this was wildly unexpected.

A few scattered sessions of returning to this ever-deteriorating desktop, and I decided to pull the mobo out, see if there was some nasty problem on the back side of it - whatever it might be - and to clean the dust out of the catch-points like the CPU fan, etc. I've had some problems with carpenter ants, and wondered if they'd nested somewhere in the PC, but really hadn't seen enough of them around it to think that was it.

And it wasn't. Still, I spent an hour or so cleaning up the dust, using compressed air and swabbing here and there, and reassembling. I was rewarded with the exact same symptoms, maybe worse. I'd only reinstalled the CPU fan, and it would spin a few times, then stop - no longer cycling up for ten seconds regularly as it had been previously. Started to seem like a PS problem. The PS LED, I then noticed, would be green until I hit the power-button, then dim or go out. Finding a popular replacement on Amazon for less than $25, I decided to go for it.

New PS fit perfectly, and...the fan jumped to life as hoped. I decided to get back to the 'last known good config' (remember this began with MANY repetitions of what, by all accounts, was my C-drive failing, attempting to repair itself, getting to the same 91% figure, then repeating, ad nauseum) so I hooked up the original drive.

And that's the configuration from which I'm typing this out. It's been running and rebooted once for updates, for more than 12 hours now.

So...a failing PS can somehow simulate a failing C-drive, through many, many cycles, reaching the identical "91%" self-repair?
Then begin a shorter cyclic-failure, lasting a couple of minutes, again repeatable seemingly forever?
Then begin a set-your-watch-by-it shorter cycling lasting like 9 seconds?
Then start cycling in about 2-second intervals, where the fan never even gets up to speed before the cycle repeats?

Who knew? Not me, that's for sure.

--Dave

These things being what they are, turns out a weekend wasn't a long enough shakedown period. I shut it down Monday morning, came back to it Tuesday morning, and it booted only to the point where it wanted me to enter F1 (to ignore that the case fan went down while I was cleaning it - bad connections at entry point, not a big deal since I've got the case open and it's got other fans). But unlike the recent boots over the weekend, it did nothing when I hit F1 this time. I powered down, and (mutha&@#%) wouldn't you know: it started doing that same two-second attempt at startup that it was doing before I replaced the PS, barely spinning up the CPU fan, then stopping, repeating ad nauseum.

All I can guess at is that there's something on the mobo that's easily disturbed, eg by my thorough cleaning effort, that was "fixed" for those couple of days by my various pokings and proddings - and now it's back to its former ways. Any diagnostic methods would be appreciated.

I'm thinking of literally going at the mobo with a chopstick or some such to see if I can find a cold-solder joint or whatever.