On Sat, 26 Sep 2026 20:13:15 -0400, Spalls Hurgenson <
spallshurgenson@gmail.com> said this thing:
Awww, my WinXP PC just died.
The WinXP box is a heavily modified "Republic of Gamers" gaming PC I >scrounged from the kerbside some years back. I rebuilt it, and have
used it to play mid-noughties 9x and XP-era games that sometimes have >problems with more modern hardware and operating systems. Although
mostly it just runs nifty screensavers in the background, 'cause I
like the pretty pictures. ;-)
Still, I'm not really surprised it's given up the ghost. It's an old
machine and none of its parts were new when I built it. The real
question is whether or not I intend to fix the PC or junk it entirely.
I don't really have any spare PSUs in storage, and I'm not sure
spending $50 (or $100, or whatever god knows if PSU prices have gone
up like everything else computer) to fix a PC that barely gets any
real use.
Final Update:
So I finally got my hands on a spare PSU and tinkered with the dead
PC.
Through the aid of a colleague, I have a (mostly) new PSU that
definitely works. But the PC is still dead. It turns out it wasn't the
old PSU that had died after all. I plugged in the new one and no got
no response.
Now, my second guess as to what might have gone wrong was bad
capacitors. The PC does date back to the mid-2000s and that was not a
good era for caps. But a quick visual scan didn't reveal anything
wrong; none of the caps had leaked or even were apparently bulging,
and there was no fishy smell. Maybe, I thought, it was something
simpler? Perhaps the power-switch broke (or the cable had gotten
disconnected?)
But no, nothing so easy. I replaced the power-switch and still had no
luck. At that point I realized I'd have to get elbows deep and take
the whole PC apart so I could investigate the components more closely.
Maybe some card had gone bad? Perhaps there was a short somewhere?
I've seen some motherboards refuse to boot if their CMOS battery was
dead. There was a whole host of possibilities.
But, long story short (something difficult for me!) it turns out my
second guess was correct. It was, of course, the capacitors. It was
difficult to see while the motherboard was in the chassis, but once I
got it out, there was some minor but obvious bulging on a number of
capacitors, including more than a few clustered around the connectors
for the power-cabling. In truth, the board wasn't entirely dead
--there was enough juice running through it that its onboard LEDs
could (barely) flicker-- but some half-hearted tests with a multimeter
seemed to indicate that not enough voltage was getting through to
power the CPU.
I guess I could try and replace the capacitors... but I won't. My
soldering skills are barely passable, there's a lot of capacitors, and
almost all of them are in extremely inconvenient places tucked tightly
between internal connectors. Even if I could get to them, my luck with replacing caps is about 30%, so I have high doubts that -with so many capacitors to replace- I'd end up with a working machine.
The board is dead, and I'm not Jesusing it back to life.
Shame. I liked that machine and -tinkering with it- I realized the
board was actually quite capable. But now it's just junk. I'm not all
that surprised it died, though; it was a curbside-find, and is 20
years old. Even if it hadn't been built during the capacitor-plague
years, it's still somewhat amazing it worked as long as it did.
On the plus side: now I have TWO spare PSUs! (the new one, and the old
one that wasn't really dead after all). So some good came out of it.
--- PyGate Linux v1.5.19
* Origin: Dragon's Lair, PyGate NNTP<>Fido Gate (3:633/10)