hwh wrote on 2020-06-24, 21:03:

Other stuff, I don't know, drivers mysteriously uninstalling themselves.

I recently had an amazing experience: the GPU in my retro system insisted on using IRQ5 and when I manually assigned IRQ5 to Legacy ISA in the BIOS setup, Nvidia drivers just vanished from Windows 98 SE. I had to reinstall them.

Weirdest troubleshooting experiences? Motherboards not being able to boot into DOS/From floppy/From HDD/what have ya until cache is disabled (usually internal).
I have no idea what causes this, but iv had it happen to me on multiple motherboards, multiple socket types.

I don't even know if im just seeing things with this, probably happens after i poke the bios unintentionally in wrong places, as the minute i clear the bios or load "default settings" in the menu after a restart, it usually fixes itself and boots fine after that.

Fuggin spooky :0

hwh wrote on 2020-06-24, 21:03:

Mostly things like drives that will only work in some really arbitrary cable twisting or jumper combination. Stuff like, these two drives will only work if one is set to CS, the other to slave, and they're on separate channels,

Ah spent many a night doing the same, the joys of cannibalised hardware eh?
Pretty tame but my PPro board that only likes 40 pin IDE cables. I've got earlier PC's that don't care less but this M/B just wont detect hard drives properly with 80 pin cables.
(and of course I'd thrown out just about all 40 pin cables earlier that year, Lucky I'd missed one inside a case)

Full credit for the speaker fix, not sure I'd of put 2 and 2 together on that one

Mine was with a system I built with an Asrock K7VT6.
Sometimes it was fine other times it would blue screen before windows was loaded.
Long story short, the memory I was using was not on the QVL and I had the bios memory timings set to auto.
Every time it was powered on, it would set random timings for the memory. Setting them manually fixed it.

hwh wrote on 2020-06-24, 21:03:

I was amazed the first time I witnessed a Macintosh refuse to boot without a fresh battery, but that kind of thing seems commonplace now.

I will never understand the Macintosh II/IIx/IIfx needing TWO 3.6v "1/2 AA" batteries just to power on. Some other Macs just won't output video without a battery (and can be tricked into doing so) but only the "big box" Mac IIs seem to need batteries to work at all.

Had a computer used for automotive service manual reader back in the day (bit later), rebooting repeatedly, checked the outputs including power good, but P.G. is bouncing in time with rebooting. Manual is keyed to a computer and very special designed around special chassis, looks like metal box than a PC, based on a intel baby AT pentium motherboard but power supply also is custom designed for the case too and it does not have power good signal built in.

A P.G. is from a small external board mounted to the exterior of the power supply with a special application precision timer IC (6 pin IC, not 555 timer) to go off (go high in specific delay) when signal is turned on which was the 5V from one of outputs when turned on.

A small capacitor that formed timing system of that IC had one leg fell off when removed. Another capacitor fixed that.

Cheers,

The strangest troubleshoot I made was purely accidental. When I was a teenager, I broke my gamecube's audio by inserting a audio jack in the video output (I don't know what I had in mind, but it was pretty stupid). So there was a spark, the gamecube went to black screen, and after rebooting it, no more sound.

Sad, I was about to give up on that poor gamecube. I was going to end its misery by having fun at making more sparks with its innards. So yeah I opened it, took the motherboard off and used a screwdriver to connect several pins together to make sparks.

However through this, a miracle happened : by shorting two pins, the sound came back !!
It was amazing !!

These two pins, among many others were actually the back of a connector which led to the gamecube's "psu". So as a pretty crude and simple fix, I took a tinfoil and wrapped it up into the connector to make contact.

To this day my gamecube still works perfecty with no issues, and sound works again.

Later on, I discovered why exactly sound stopped working and by what miracle it came back to life.
See this post : http://forums.modretro.com/index.php?threads/ … 752/#post-47537
And this picture https://www.gc-forever.com/wiki/index.php?tit … ile:GCPower.jpg

So basically the only component using 12V in the cube is the sound chip. So by making that stupid thing in the connector, I basically fried the +12V power line.

Then by poking around in the gamecube, I fortunately shorted what used to be the +12V power line and +5V which restored power to the sound chip once again and made it work like new.

But still, thinking about how randomly I fixed this just blows my mind.

Ooh boy I've got a dozen but I'll post here a (shortened down) version of my most "weird" encounter:

It begun when I wanted to pop in a P233MMX into my Shuttle HOT-555A board (rather affectuonously known as "Shittle Hot garbage"), set the jumpers correctly and booted it first try, or so I thought...

It would post fine but would flake out when trying to start Windows, popping Scanreg in a constant loop.

Bemused, I attempted to reinstall Windows, no dice. Got a Suwin.exe error, tried with w95, same deal.

Then I get a HDD error, immediately I swap the 20GB WD IDE HDD I had there at the time (gah why did I smash it to bits... My own stupidity will never cease to amaze me), put a SCSI IBM DDRS HDD (aka "Ultrastar 9ES"), with a Adaptec AHA-2940U2W, same punishment...

So I popped back my P133, set the jumpers accordingly to the manual, and it works... Though later on the WD drive started acting dicey so I left the 9ES HDD in there.

Now, if the PSU is dying, I'm fucked. If the 233MMX I got is a dud, I'm... Well, that blows. I'll maybe get one of those industrial ATX PSUs that still bear a -5V rail and a ATX2AT harness (I'm kinda bummed as I wanted to keep this PC as original as possible).

I even tried putting a fan on the voltage regulators, running the mobo outside of the case (never again, it's a plaie sans nom to tear down, and in the process I managed to give myself a few small cuts on my fingers, which were 20x more painful than papercuts, but since old PC cases sometimes require a blood sacrifice...)

Nice read ! The oddest troubleshoot for me that I remember was a 10 year old single core P4 soc 478 back in 2014. The machine would quickly reboot at random. Tore it apart to basic parts and it ran fine, but still did all the regular clean, reseat, etc and testing things using a few of my known good parts and it all came out good. Checked PSU and it was within tolerances too, swapped it out just to make sure. When adding everything back in it suddenly started quick random reboots again. Ok this machine had one HD on primary and two CD burners on secondary IDE, one master - one slave. Once those CDR were connected it went reboot crazy so checked each as a single and no issue BUT when master+slaved together it caused something to reset that mobo. Odd part of the story: both old CDR (2002-2003 era) work perfect as single drives still today in other machines but for some reason could not get them to master - slave each other no matter which was which on that motherboard. And yes that motherboard still works just fine...

I used to work at a computer repair store around 1999. One day I got a computer that wouldn't respond to the keyboard.

I kept pressing ctrl-alt-del and nothing would happen. So I say "this thing keeps locking up" and guy next to me says "this things keeps rebooting" then we looked at each other's screens. We got our keyboard wires mixed up.

kingcake wrote on 2020-06-25, 03:09:

I used to work at a computer repair store around 1999. One day I got a computer that wouldn't respond to the keyboard.

I kept pressing ctrl-alt-del and nothing would happen. So I say "this thing keeps locking up" and guy next to me says "this things keeps rebooting" then we looked at each other's screens. We got our keyboard wires mixed up.

🤣 now that is funny ! Thanks !

kingcake wrote on 2020-06-25, 03:09:

I used to work at a computer repair store around 1999. One day I got a computer that wouldn't respond to the keyboard.

I kept pressing ctrl-alt-del and nothing would happen. So I say "this thing keeps locking up" and guy next to me says "this things keeps rebooting" then we looked at each other's screens. We got our keyboard wires mixed up.

*laughs*.

Cheers,

I built a cheap PC for somebody that ran into issues after some months - it would fail to boot. So I went around and had a look and discovered no HDD being detected by the BIOS. Then, it miraculously reappeared and all was well. I ran diagnostics and everything checked out ok.

Figuring it may have possibly just been a loose cable I left it be. But a couple of weeks later, same issue again. Eventually ended up practically re-building the PC with a new different motherboard, new different CPU & new different HDD.

Then a couple of months later - same issue back again. This time there was nothing left to change except the case itself, so I did and what do you know - it cured it permanently!

I kept the case for testing purposes. I couldn't see anything physically wrong with it, nothing that could be shorting anything etc. The case had only 1 quirk - a plastic funnel installed over the side panel vent near to the CPU. I removed this funnel and decided to push it by installing a Socket 478 P4 3.0GHz setup. Never gave me any problems and was used for years afterwards.

I can only guess that plastic funnel was causing overheating of some sort.

Troubleshooting a friend's computer not showing anything on the screen, which ended up with us buying a new motherboard and swapping it out. In the end, problem was the DVI connection on the monitor side had come loose.

Ha, been there. I remember panicking for 30 minutes because I thought I had killed an expensive video card. After reseating/cleaning/inspecting it a dozen times, I noticed the monitor connector was loose.

(DisplayPort with its auto-locking mechanism can be annoying, but it does at least put an end to this problem.)

This week actually. Elderly friend lost all of her pictures and was going insane trying to find them saying she thought she may have had a virus. Well Onedrive being annoying as sin, I had uninstalled it previously. Welp, after noticing Onedrive BACK ON AND BEING ANNOYING AGAIN!!! I figured out that Onedrive had deleted her picture folder on her desktop and buried them in a special Onedrive folder upon reinstalling it's self. So went through the settings and set Onedrive to never turn on and such instead of uninstalling it this time, moved everything back onto her desktop folder back where she could find it and all is well in the world. Thanks Microsoft.

OneDrive is pure fucking evil.

It's weird how often I miss obvious signs of malfunctions: touching pins, broken traces, burned parts, cracked plastics, missing jumpers.
"I've checked that thing at least 50 times, how could I miss that?!"
Happens more than I'd like to admit.

appiah4 wrote on 2020-06-25, 07:58:

OneDrive is pure fucking evil.

OneDrive is okay; Microsoft cramming it down our throats sideways at every update is evil. 😀

When I briefly worked at a janky computer repair shop in the early 2000s, I was tasked to sort out someone's gaming rig that wouldn't boot. Nothing visibly wrong with it, but the motherboard was screaming that the processor was missing. Well it's clearly under the heat sink, so let's look at it. I pulled the cooling block, and noticed that the processor was crooked. Removed it, and most of the pins were mashed flat. As I didn't have a spare processor of its socket type handy, and I was left at the store with no assistance (as usual), I grabbed the only edged tool I could find, a carpet knife, and set about straightening the pins. None broke off, thank heaven for small miracles, and I popped it back into the socket and re-applied thermal paste. Once everything was back where it belonged, I fired it up, and it booted. Went into BIOS to check the voltage values, temperature, and everything else, and it was all stable. Mission complete. My corrupt boss charged the customer for a whole new processor. Not so much weird, as it was more infuriating at every step.

The only accidental good fortune I've had was when I decided to restore my Goldstar 3DO years after ordering a replacement CDROM laser assembly not made specifically for the 3DO. Right after I disassembled the system and the drive tray to replace the laser, I randomly happened upon a page talking about replacing laser assemblies, warning that there is a solder bridge that must be opened before installation. It's soldered to prevent ESD during shipping, but apparently this isn't common knowledge. I removed the solder bridge, installed the assembly, and the system plays discs without issue now.