If you have Linux floating around somewhere, you could also try ddrescue to make an image of the failing drive, but it sounds like it's too late now. I did that on my old HP laptop when I replaced its failing drive (which had been showing hundreds of bad sectors for 10 years before I finally replaced it).

GuillermoXT wrote on 2024-11-03, 00:03:

My 20-month-old son hit my second wireless mouse, which I use when I'm sitting on the couch and want to watch series, on the table several times, causing a microswitch in it to break.

I had a couple of the original Microsoft wireless mice - (1st and 2nd edition I think) - great mice, but the "left click" button on both wore out
a lot faster than any other mice I've had.

Not a standard thing (at least not one I recognized at the time) - I was able to put one of the little square board-mount switches on a couple of
wire standoffs and it worked well - but the 1st one I did then needed quite a bit more pressure on the left button.

Later I fixed the 2nd edition, I did the same thing but found a much lighter press button.. It needs only a little more pressure then it did, and I
still use it quite often. The 1st edition still works too but the difference in left/right button pressure make it a bit weird to use...
I've not bothered changing the button again, I just keep it on my workbench for the few times I need a mouse when working on PC hardware.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

You can get those buttons with taller nubbies on too. What can happen also is the post moulded into the case button that hits the switch button can get burred and compressed shorter than it should be, so you get an identical replacement switch and it still doesn't click because the mechanics don't reach far enough any more. I have put in little shims of plastic before to build the post back up. I guess in this age of 3D printers one could make them little clip on "shoes" in nylon or something hardwearing and replace those as required.

Spent the evening continuing to work on the 486 motherboard I had previously been repairing. Last time I had repaired a trace with a temporary bodge wire. The board otherwise seemed to work, so I wanted to implement a more permanent fix.

After removing the keyboard socket and sanding down a bunch of the corrosion, I installed a replacement trace using some enamelled wire. I also noticed that the via near C15 was clogged with corrosion, so I cleaned that out too.

After soldering a new keyboard socket, I tried booting the system thinking it would be fully repaired. Much to my surprise, it failed to POST. The only thing that was working was the reset line wasn't stuck on, but other than that it seemed completely DOA.

I proceeded to spend the next hour trying to figure out what had gone wrong. I triple checked my repair work, testing for shorts, continuity, voltages, and inspecting for other damage I might have caused. Everything seemed fine. I even double checked the processor and RAM in another board to make sure they still worked.

After getting nowhere, I resigned myself to the fact that maybe something else on the board was faulty and I'd have to spend more time diagnosing it. On a whim, I figured I might try a different processor completely. That's when I checked the switch that controls the FSB speed.

Oops.

Turns out that stupid switch was set incorrectly. The "3" position should have been set to on. It must have gotten nudged into the off position at some point. This resulted in an invalid setting and the reason the board wouldn't POST.

After fixing the switch, everything booted up as normal.

This makes twice in the past week I thought I screwed up a repair only to discover the solution was a simple settings issue.

Will I learn my lesson after this? (Probably not.) 😅

BitWrangler wrote on 2024-11-04, 02:31:

You can get those buttons with taller nubbies on too. What can happen also is the post moulded into the case button that hits the switch button can get burred and compressed shorter than it should be, so you get an identical replacement switch and it still doesn't click because the mechanics don't reach far enough any more. I have put in little shims of plastic before to build the post back up. I guess in this age of 3D printers one could make them little clip on "shoes" in nylon or something hardwearing and replace those as required.

I this case the problem was electrical contact - both mice got so "noisy" that one click would often register as twp (or three).
It's been years since I looked inside them, but IIRC the button housing was a taller-thin thing I'd not seen with a bigger set of contacts
inside than I expected... little square board-mount device was a simple fix!

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

The ones I have pried open have a little metal disk that's a flattened dimple, kinda like how you could make a clicker out of an old tin, and that would be the positive action click and one contact. After many many cycles, those fracture across and the action gets glitchy.

edit: hah, didn't read all the way through. Yah, that above is the tiny square ones, the rectangular ones are actually microswitches usually, they have bronze leaf springs in, those tend to get glitchy due to not being self cleaning at only 5V and build up some oxide. They can be rejuvenated by switching something higher voltage with an inrush current a few times, which clears them up for a bit. However they can also get contact fatigue, or stick at the actuator thing when it is full of worn plastic powder and galls.

Shponglefan wrote on 2024-11-04, 02:34:

Spent the evening continuing to work on the 486 motherboard I had previously been repairing. Last time I had repaired a trace w […]
Show full quote

Spent the evening continuing to work on the 486 motherboard I had previously been repairing. Last time I had repaired a trace with a temporary bodge wire. The board otherwise seemed to work, so I wanted to implement a more permanent fix.

The attachment 4DMU HL3-L4-VB motherboard bodge wire.jpg is no longer available

After removing the keyboard socket and sanding down a bunch of the corrosion, I installed a replacement trace using some enamelled wire. I also noticed that the via near C15 was clogged with corrosion, so I cleaned that out too.

The attachment 4DMU HL3-L4-VB motherboard enamelled wire.jpg is no longer available

After soldering a new keyboard socket, I tried booting the system thinking it would be fully repaired. Much to my surprise, it failed to POST. The only thing that was working was the reset line wasn't stuck on, but other than that it seemed completely DOA.

The attachment 4DMU HL3-L4-VB motherboard no POST.jpg is no longer available

I proceeded to spend the next hour trying to figure out what had gone wrong. I triple checked my repair work, testing for shorts, continuity, voltages, and inspecting for other damage I might have caused. Everything seemed fine. I even double checked the processor and RAM in another board to make sure they still worked.

After getting nowhere, I resigned myself to the fact that maybe something else on the board was faulty and I'd have to spend more time diagnosing it. On a whim, I figured I might try a different processor completely. That's when I checked the switch that controls the FSB speed.

The attachment Stupid Switch.jpg is no longer available

Oops.

Turns out that stupid switch was set incorrectly. The "3" position should have been set to on. It must have gotten nudged into the off position at some point. This resulted in an invalid setting and the reason the board wouldn't POST.

After fixing the switch, everything booted up as normal.

The attachment 4DMU HL3-L4-VB motherboard successful boot.jpg is no longer available

This makes twice in the past week I thought I screwed up a repair only to discover the solution was a simple settings issue.

Will I learn my lesson after this? (Probably not.) 😅

Re: What retro activity did you get up to today?
nothing else to add...

BitWrangler wrote on 2024-11-04, 04:01:

The ones I have pried open have a little metal disk that's a flattened dimple, kinda like how you could make a clicker out of an old tin, and that would be the positive action click and one contact. After many many cycles, those fracture across and the action gets glitchy.

edit: hah, didn't read all the way through. Yah, that above is the tiny square ones, the rectangular ones are actually microswitches usually, they have bronze leaf springs in, those tend to get glitchy due to not being self cleaning at only 5V and build up some oxide. They can be rejuvenated by switching something higher voltage with an inrush current a few times, which clears them up for a bit. However they can also get contact fatigue, or stick at the actuator thing when it is full of worn plastic powder and galls.

These weren't normal looking microswitches, the mice themselves are fairly old, and actually have screws on the bottom under "feet" for disassembly...
you don't have to pry them apart. I think at least one of them still has the original switch on the right-click position (I think I might have
swapped both in one to make them more consistent .. I suppose I could take them apart and get a pic of the switch if you want to see it.

Dave ::: https://dunfield.themindfactory.com ::: "Daves Old Computers"->Personal

Damn, I am having too much fun with the alligators still, keep forgetting this is a support operation and I'm supposed to swing back to that 286 Packard Bell..

Anyway, got tired of chasing the weirdness, now starting to suspect something to do with the CPU which is a late SL part, and maybe it doesn't get optimally configured, and maybe something I have run put it "right", then running faster.

So I took a detour and threw together a speaker enclosure in about 30 mins... I wanted to mount the test speaker somehow so it wasn't just rattling around on the bench. So, this happened. Vitamin bottle upended on that light fitting I wrenched off the ceiling the other week. Which I think is a little too large for the base for this really. Will have to keep an eye out for something solidish but a bit smaller. As it is like that though I am quite pleased with the sound, it was that little modem speaker I was impressed with before, so it took that up at least 13/37ths of a notch. Interesting thing, with the bottle kind feeding down into the fitting like that, it doesn't really "thump" the bench on sharp lows, but lightly tap it, and you can just about feel it... hyper low end rumble pack 🤣 .. Anyway, that's interesting enough I wanna find a 4" cone somewhere and stick it in a listerine bottle or similar, with another flared base thing, thinking I might lift that one up somehow, maybe skirt it like 240 degrees, with the hole pointed forward.... IDK if that's worth doing to run off the speaker header, might need an LPT DAC and small amp.

Today I cable managed my windows 98 PC (CPU is out as I'm about to volt mod it)

Maybe I'm too old, but I prefer to listen music on a portable players instead of a smartphones. Firstly, the quality in a smartphone may be inferior, secondly, convenience - the touch screen is not very suitable for control.

At the moment, I use a rather rare example of e.Digital MXP 100 around 2000, with a chip from Texas Instruments and with Compact Flash or MicroDrive as a storage. It even has voice control, you press a button, name a song and it starts playing, cool even for 2000!

I bought it second-hand without a battery, but at least with a charging cable. All ports here are proprietary, no micro or mini USB. I picked the only USB data cable that fits the shape, from a Casio camera. I picked the new battery based on voltage and amperes, I had to change the connector, fortunately I cut off the connectors from old junk earlier and collected them in a box, one matched.

I managed to find with difficulty, having sorted through dead links, one live one with an application something like Music Explorer, since simply placing files on the card will not give anything, the player requires the creation of technical files with a list of your files, the manufacturer decided to complicate the user's life. Well, as expected, the application is glitchy, it sees the player and is even ready to change files on it, only immediately after the change it crashes and the maximum it gives is to create an empty file or folder. And this is independent of Windows (tried in Win98, 2000, XP - all the same, probably some libraries are missing). I had to write a PHP script that creates a ready-made technical file with certain bytes from a list of files and only then can you play files downloaded independently to the card.

The sound quality is very good, no worse than the slim player from Sony, large buttons, in short, in the end I am happy and consider the quest completed.

ErroneousHyphen wrote on 2024-11-05, 12:38:

The attachment IMG_20241105_233236896_HDR.jpg is no longer available

Today I cable managed my windows 98 PC (CPU is out as I'm about to volt mod it)

LGA socket aint no win98 PC 😁

smtkr wrote on 2024-11-06, 00:39:

ErroneousHyphen wrote on 2024-11-05, 12:38:

The attachment IMG_20241105_233236896_HDR.jpg is no longer available

Today I cable managed my windows 98 PC (CPU is out as I'm about to volt mod it)

LGA socket aint no win98 PC 😁

hey Pentium 4s were still in the Windows 98 era... just... its just mine is a very very late Pentium 4 😜

Successfully imaged time pc backup/restore cd which uses Norton ghost. Was worried couldn't image it because of bit rot as it had always thrown errors when trying to image. Used alcohol 120% and saved as MDS then successfully burnt using raw+dao

ErroneousHyphen wrote on 2024-11-06, 02:24:

smtkr wrote on 2024-11-06, 00:39:

ErroneousHyphen wrote on 2024-11-05, 12:38:

The attachment IMG_20241105_233236896_HDR.jpg is no longer available

Today I cable managed my windows 98 PC (CPU is out as I'm about to volt mod it)

LGA socket aint no win98 PC 😁

hey Pentium 4s were still in the Windows 98 era... just... its just mine is a very very late Pentium 4 😜

hey, my ASUS P5B supports Win98 with official drivers ... of course the Q6700 works better with XP ... dual boot to the rescue 😂

Wouldn't be the first to do a 775 based 98 build 🤣
I have a 775i65PE waiting for a similar treatment, got almost everything needed.

smtkr wrote on 2024-11-06, 00:39:

LGA socket aint no win98 PC 😁

LGA775 is the *best* Win 98 PC. 😜

BitWrangler wrote on 2024-11-05, 04:15:

Damn, I am having too much fun with the alligators still, keep forgetting this is a support operation and I'm supposed to swing back to that 286 Packard Bell..

Anyway, got tired of chasing the weirdness, now starting to suspect something to do with the CPU which is a late SL part, and maybe it doesn't get optimally configured, and maybe something I have run put it "right", then running faster.

So I took a detour and threw together a speaker enclosure in about 30 mins... I wanted to mount the test speaker somehow so it wasn't just rattling around on the bench. So, this happened. Vitamin bottle upended on that light fitting I wrenched off the ceiling the other week. Which I think is a little too large for the base for this really. Will have to keep an eye out for something solidish but a bit smaller. As it is like that though I am quite pleased with the sound, it was that little modem speaker I was impressed with before, so it took that up at least 13/37ths of a notch. Interesting thing, with the bottle kind feeding down into the fitting like that, it doesn't really "thump" the bench on sharp lows, but lightly tap it, and you can just about feel it... hyper low end rumble pack 🤣 .. Anyway, that's interesting enough I wanna find a 4" cone somewhere and stick it in a listerine bottle or similar, with another flared base thing, thinking I might lift that one up somehow, maybe skirt it like 240 degrees, with the hole pointed forward.... IDK if that's worth doing to run off the speaker header, might need an LPT DAC and small amp.

👏👏👏😆

Not PC related and only slightly retro -

Today and for the past few days I've been playing around with running RISC OS 5 on the Raspberry Pi.

I did the initial setup on a Pi 3A+ for easy WiFi access to download everything, before transferring the SD card to it's intended home - a Pi Zero 1.3, where I don't need any networking.

Installed ADFFS and a bunch of Acorn Archimedes games.

Had loads of stability issues with it however, random games freezing the whole system at random and multiple power on/off reboots required. After a bit of trial and error, I downgraded ADFFS 2 versions lower and it seems to have improved stability a bit. Further testing required.

The Pi is a little stopgap until I ever decide to acquire a real Acorn Archimedes or later RISC PC in the future.

The 30 minute speaker is now the 35 minute speaker, as it took a few mins to carefully carve a hole in the bottom of a fruit cup container... Thus it is now leaner and meaner... it seems just wide enough to keep it from being tippier than say an air blower or can of contact cleaner, so doesn't appear to need weight at the moment.

Don't think I changed the acoustics a heck of a lot, was noticing a mod file sounded like a shitty 56k MP3, but then reminded myself it was probably made from samples done on a 90s homebrew 8 bit sampler and over processed to get rid of hiss and distortion, so yeah, sounds as good as the input lets it.