Alexraptor wrote on 2025-06-01, 12:57:

As I've had it explained to me, it's caused by a chemical reaction between the plastics in the device and cable, resulting from […]
Show full quote

devius wrote on 2025-06-01, 10:48:

That looks more like hot-melt glue.

Ydee wrote on 2025-06-01, 11:01:

It looks like a thermal deformity from touching a hot object.

As I've had it explained to me, it's caused by a chemical reaction between the plastics in the device and cable, resulting from prolonged contact. There are marks like this all over the back side of the keyboard, wherever the cable was been pressing against the keyboard.

It was factory sealed anyway.

yes, those blemishes are caused by contact, the "rubber" of the cable and the plastic. I have two or three keyboards that suffered these injuries, like the Dell I'm using right now.
With a razor blade, you remove the ridge, then with a metal sponge, you scrape and smooth it, and that's it.
that s what I did
If it's on the back, good luck. It's bad if it's on the top. Although in this case, it tends to happen on the edges.

Heh. I always figured it was a simple friction melt from the usually sloppy handing of them in bulk.

I figured it wasn't that different to a band saw, or wire cutting machine.

Bought 4 IDE to dual CF card adapters. Of which 4 had bent pins on CF card connector and some even on the IDE connector. -sigh-

So I made a claim on AliExpress, requesting a 50% refund, and straightened the pins myself. Seems to be working now ...

H3nrik V! wrote on 2025-06-02, 06:54:

Bought 4 IDE to dual CF card adapters.

This exact model is ok for onboard IDE controllers with Cable-Select capabilities (e.g. Socket7 boards).
To be used with older PCs through multi-i/o cards (i.e. IDE/floppy/LPT/com) you have to cut the ground trace around the pin 28 on card's IDE header.
Those old i/o cards expose ISA bus ALE signal directly into the IDE cable, and shorting the ALE on the CF adapter disrupts the ISA bus logic.

weedeewee wrote on 2025-06-01, 13:15:

That's exactly what it is.

The plasticizer leeching out of the cable and deforming the rigid plastic shell.

I've never seen anything like this before. So the "plasticizer" is what -- alien slime?
I have the exact same keyboard, and other than yellowing, nothing like this has happened to her.

Ydee wrote on 2025-06-02, 08:58:

I've never seen anything like this before. So the "plasticizer" is what -- alien slime?

A plasticizer is a volatile chemicals, added to other material to alter its properties. With time those chemicals either completely evaporate, causing material to go hard as rock, or it continue slowly reacting with the hosting material, causing its disintegration. You know how the rubber rollers in cassette decks or some rubber bumpers in old HDDs turns into slimy goo. Or like power cables laying in the styrofoam packing materials (printers/monitors/etc) often gets that styrofoam etched onto the cable, covering it with hard to remove residues, or carving into the cable itself. Just because they touch, and because chemicals aren't completely inert.

H3nrik V! wrote on 2025-06-02, 06:54:

Bought 4 IDE to dual CF card adapters. Of which 4 had bent pins on CF card connector and some even on the IDE connector. -sigh-

So I made a claim on AliExpress, requesting a 50% refund, and straightened the pins myself. Seems to be working now ...

I tried these cheap CF card adapter and they break easy.
I use the ones from "Startech" they cost more but are worth it.
My PC hardware is expensive and I don't want to ruin it with cheaply made components.

vstrakh wrote on 2025-06-02, 09:13:

Ydee wrote on 2025-06-02, 08:58:

I've never seen anything like this before. So the "plasticizer" is what -- alien slime?

A plasticizer is a volatile chemicals, added to other material to alter its properties. With time those chemicals either completely evaporate, causing material to go hard as rock, or it continue slowly reacting with the hosting material, causing its disintegration. You know how the rubber rollers in cassette decks or some rubber bumpers in old HDDs turns into slimy goo. Or like power cables laying in the styrofoam packing materials (printers/monitors/etc) often gets that styrofoam etched onto the cable, covering it with hard to remove residues, or carving into the cable itself. Just because they touch, and because chemicals aren't completely inert.

Thank you for a detailed explanation - I know that some types of rubber decay, swell or melt over time (for example belts from DVD/CD drives). But I've never seen cable softeners react so aggressively with plastic. In addition, the damage is above the F4 key, but the cable is insulated from the body of the keyboard by an LDPE? or PVC? bag. Weird.
I've had the same keyboard for several years, a cable wrapped around her body, and I don't see any such signs of damage, even on the cable.

Ydee wrote on 2025-06-02, 11:34:

vstrakh wrote on 2025-06-02, 09:13:

Ydee wrote on 2025-06-02, 08:58:

I've never seen anything like this before. So the "plasticizer" is what -- alien slime?

A plasticizer is a volatile chemicals, added to other material to alter its properties. With time those chemicals either completely evaporate, causing material to go hard as rock, or it continue slowly reacting with the hosting material, causing its disintegration. You know how the rubber rollers in cassette decks or some rubber bumpers in old HDDs turns into slimy goo. Or like power cables laying in the styrofoam packing materials (printers/monitors/etc) often gets that styrofoam etched onto the cable, covering it with hard to remove residues, or carving into the cable itself. Just because they touch, and because chemicals aren't completely inert.

Thank you for a detailed explanation - I know that some types of rubber decay, swell or melt over time (for example belts from DVD/CD drives). But I've never seen cable softeners react so aggressively with plastic. In addition, the damage is above the F4 key, but the cable is insulated from the body of the keyboard by an LDPE? or PVC? bag. Weird.
I've had the same keyboard for several years, a cable wrapped around her body, and I don't see any such signs of damage, even on the cable.

I think contact surface is key.

There's very little of the cable actually touching the plastics when one just casually wraps the cord around the keyboard, just a few light pressure points here and there. But when you have everything packed tightly inside a box for 25+ years... well. 🤣

Alexraptor wrote on 2025-06-02, 12:40:

I think contact surface is key.

There's very little of the cable actually touching the plastics when one just casually wraps the cord around the keyboard, just a few light pressure points here and there. But when you have everything packed tightly inside a box for 25+ years... well. 🤣

I agree, but where is contact between cable and keyboard body here?:

Just picked up these two.

I just bought two new old stock GeForce 5700LE 128MB 128-bit.

Intel486dx33 wrote on 2025-06-02, 10:46:

I tried these cheap CF card adapter and they break easy. I use the ones from "Startech" they cost more but are worth it. My PC […]
Show full quote

H3nrik V! wrote on 2025-06-02, 06:54:

Bought 4 IDE to dual CF card adapters. Of which 4 had bent pins on CF card connector and some even on the IDE connector. -sigh-

So I made a claim on AliExpress, requesting a 50% refund, and straightened the pins myself. Seems to be working now ...

I tried these cheap CF card adapter and they break easy.
I use the ones from "Startech" they cost more but are worth it.
My PC hardware is expensive and I don't want to ruin it with cheaply made components.

Haven't found any other dual ones, I like that. Soldering is one of my better skills, and since the connectors are the only actual components, I just give all the solder joints a freshing up ..

Nicolas 2000 wrote on 2025-06-01, 17:01:

... ati rage 128 pro ultra (so that would be a 16mb version of the Pro)...

What do you call the 32MB version then?

Nicolas 2000 wrote on 2025-06-01, 17:01:

I found a 512mb pc133 sdram unit and an ati rage 128 pro ultra (so that would be a 16mb version of the Pro) on the flea market today. Nice upgrade from the 384mb and rage 128 non-pro that I used to run in my W98 machine.

No, Rage 128 Ultra is to Rage 128 Pro the same that TNT2 M64 is to TNT2 Pro. It's the same GPU, but linked to 64 bit memory. Rage 128 (non pro) should be faster in most games than Rage 128 Ultra.

Ydee wrote on 2025-06-02, 12:59:

Alexraptor wrote on 2025-06-02, 12:40:

I think contact surface is key.

There's very little of the cable actually touching the plastics when one just casually wraps the cord around the keyboard, just a few light pressure points here and there. But when you have everything packed tightly inside a box for 25+ years... well. 🤣

I agree, but where is contact between cable and keyboard body here?:

I would just assume it leeched through the protective plastic bag as well, which probably has some plasticizers of it's own.

Thermalwrong wrote on 2025-05-05, 03:06:

Ah wow, haven't seen one of those for a while - I knew I'd seen it before and I have indeed but mine is branded WTC POP3254: Re: […]
Show full quote

kinetix wrote on 2025-05-05, 02:07:

yes, there are several similar, but in this one the extra slot is shorter. surely all those use a similar if not the same BIOS. […]
Show full quote

Horun wrote on 2025-05-04, 17:00:

That looks like this: https://theretroweb.com/motherboards/s/artek- … ems-inc-opti-25
and a bunch other too, and there is a bios 😀

yes, there are several similar, but in this one the extra slot is shorter. surely all those use a similar if not the same BIOS.
I'm working on this board. I had to extract a couple of resistors and diodes, the external battery pins, and two ICs., and two broken capacitors. I had a really long hard time with the chips because I don't have a desoldering gun, and the oxidation on the top made it hard to melt the tin. Applying enough heat and time to melt the tin so I could extract it with the manual solder sucker was also a limitation (I used a wet-cooling trick on the ICs), the small holes and ground plane didn't helped either . I'm rebuilding some broken lines, but beyond this zone in the board, there don't seem to be any more problems. I looked with a camera and didn't see any oxidation. I'll take some time before continuing.
I still need to get another 386 CPU; I only have a "486" DLC/40 in another board.

Ah wow, haven't seen one of those for a while - I knew I'd seen it before and I have indeed but mine is branded WTC POP3254: Re: What retro activity did you get up to today?
Perhaps the tracing there could be useful to you 😀 There must have been a reference design because there are 5+ boards all with the same layout.
For clearing out those oxidised pins / vias, scratching them first with maybe a ceramic tool can help. I've also found that rosin flux does a better job with corroded solder than modern no-clean fluxes do.

And the later follow up where I found that pulling cleaning and reinstalling the two chips by that 8-bit ISA slot resolved my DMA problems: Re: What retro activity did you get up to today?

Good luck with the fixes, getting all those boards for that price each now-a-days is rather rare. This Opti 386 board is quite decently quick even though it doesn't have cache.

Last weekend I ran the final checks on the repaired lines and put all the components back. I replaced another resistor.
Today I ran the first test...
BANG! !!
A replaced tantalum capacitor blew (it was one salvaged from other boards, maybe I even put it in backwards). I put in another one and everything was fine.
First I measured the main clock in the socket... and there was no signal. After a closer inspection, I noticed the crystal had been replaced at some point, but it had been installed backwards!! (I'm not the only idiot) so I corrected this problem and now I had 33 MHz in the socket.
But when I measured the clock in the ISA, there was no clock either!!!
Turns out the crystal was missing. I had looked at the empty space before, but the vias were filled, and I thought it didn't need the component. But now looking more closely, I noticed they weren't perfect, so it must have been touched too. I checked out pictures of similar boards and saw that crystal at 14.31818MHz. So I installed one and got a 4MHz clock signal on the ISA . Shouldn't it be around 8MHz? Could it be due to the BIOS configuration and the divider. Perhaps without initial configuration (no BIOS or CPU installed) it boots at minimum speed. However, I remember when I picked up this and other boards that were on top of each other, a 25MHz crystal fell to the floor. Perhaps that's the one I really should have installed? (Although all the photos of similar boards show a 14.318MHz one,) the manual for another similar board shows that it has a bus clock divider of /2, /3, and /4, so with 14.318MHz it would have 7.16, 4.77, and 3.5MHz, and with a 25MHz crystal it would be 12.5, 8.3, and 6.25 MHz.
Anyway, I have several faster crystals to run some tests, if the board finally works. In the photos, I see up to 50MHz for the main clock, and I have one. I'll also try some possibly compatible BIOSes (this one was missing the BIOS chip).
Any opinions or advice on this?

Addendum: Today seems to be the day of the blown tantalums. Another 386 motherboard I restored months ago, and when I was giving it a few tweaks, one of them blew. 🤣. It was on the 12V rail. The rest of the capacitors there are 16V, too close, so I'll replace them all with 25V ones.

kinetix wrote on Yesterday, 01:32:

Last weekend I ran the final checks on the repaired lines and put all the components back together. I replaced another resistor. […]
Show full quote

Thermalwrong wrote on 2025-05-05, 03:06:

Ah wow, haven't seen one of those for a while - I knew I'd seen it before and I have indeed but mine is branded WTC POP3254: Re: […]
Show full quote

kinetix wrote on 2025-05-05, 02:07:

yes, there are several similar, but in this one the extra slot is shorter. surely all those use a similar if not the same BIOS.
I'm working on this board. I had to extract a couple of resistors and diodes, the external battery pins, and two ICs., and two broken capacitors. I had a really long hard time with the chips because I don't have a desoldering gun, and the oxidation on the top made it hard to melt the tin. Applying enough heat and time to melt the tin so I could extract it with the manual solder sucker was also a limitation (I used a wet-cooling trick on the ICs), the small holes and ground plane didn't helped either . I'm rebuilding some broken lines, but beyond this zone in the board, there don't seem to be any more problems. I looked with a camera and didn't see any oxidation. I'll take some time before continuing.
I still need to get another 386 CPU; I only have a "486" DLC/40 in another board.

Ah wow, haven't seen one of those for a while - I knew I'd seen it before and I have indeed but mine is branded WTC POP3254: Re: What retro activity did you get up to today?
Perhaps the tracing there could be useful to you 😀 There must have been a reference design because there are 5+ boards all with the same layout.
For clearing out those oxidised pins / vias, scratching them first with maybe a ceramic tool can help. I've also found that rosin flux does a better job with corroded solder than modern no-clean fluxes do.

And the later follow up where I found that pulling cleaning and reinstalling the two chips by that 8-bit ISA slot resolved my DMA problems: Re: What retro activity did you get up to today?

Good luck with the fixes, getting all those boards for that price each now-a-days is rather rare. This Opti 386 board is quite decently quick even though it doesn't have cache.

Last weekend I ran the final checks on the repaired lines and put all the components back together. I replaced another resistor.
Today I ran the first test. BANG!
A replaced tantalum capacitor blew (it was one salvaged from other boards, maybe I even put it in backwards). I put in another one and everything was fine.
First I measured the main clock in the socket... and there was no signal. After a closer inspection, I noticed the crystal had been replaced at some point, but it had been installed backwards!! (I'm not the only idiot) so I corrected this problem and now I had about 32 MHz in the socket.
But when I measured the clock in the ISA, there was no clock either!!!
Turns out the crystal was missing. I had looked at the empty space before, but the vias were filled, and I thought I didn't need the component. But now looking more closely, I noticed they weren't perfect, so it must have been touched too. I checked out pictures of similar boards and saw that crystal at 14.31818MHz. So I installed one and got a 4MHz clock signal on the ISA . Shouldn't it be around 8MHz? Could it be due to the BIOS configuration and the divider. Perhaps without initial configuration (no BIOS or CPU installed) it boots at minimum speed. However, I remember when I picked up this and other boards that were on top of each other, a 25MHz crystal fell to the floor. Perhaps that's the one I really should have installed? (Although all the photos of similar boards show a 14.318MHz one,) the manual for another similar board shows that it has a bus clock divider of /2, /3, and /4, so with 14.318MHz it would have 7.16, 4.77, and 3.5MHz, and with a 25MHz crystal it would be 12.5, 8.3, and 6.25 MHz.
Anyway, I have several faster crystals to run some tests, if the board finally works. In the photos, I see up to 50MHz for the main clock, and I have one. I'll also try some possibly compatible BIOSes (this one was missing the BIOS chip).
Any opinions or advice on this?

Addendum: Today seems to be the day of the blown tantalums. Another 386 motherboard I restored months ago, and when I was giving it a few tweaks, one of them blew. 🤣

I think that the chipset doesn't know what to do without a BIOS so defaults to the lowest value possible for ISA bus speed. It should be a register setting that the BIOS would normally configure.

I haven't set up my board in a while and can't easily check it, but it's definitely 14.318mhz that goes there, my working board has that and I apparently replaced it while fixing my board - see Re: What retro activity did you get up to today?
I don't know whether the bus clock is derived from that 14.318mhz crystal - page 4 of the document here covers it in detail: https://theretroweb.com/chipsets/366#docs

For the BIOS, try out the one from the Biostar MB-1325PD on it, it's a late version but worked the best on mine. With the original old BIOS my board had I couldn't even install 1MB SIMMs.

Managed to find an AOpen AX6F (rev 2.1) in box (even got the manual!).
That's right, a very early Slot 1 MB with the FX chipset and thus no AGP och SDRAM.
Works perfectly and looks like its almost unused.
I think I'll combine it with a 233MHz and another of my recent finds, a Diamond Viper V330 PCI.

RaverX wrote on 2025-06-02, 18:49:

Nicolas 2000 wrote on 2025-06-01, 17:01:

I found a 512mb pc133 sdram unit and an ati rage 128 pro ultra (so that would be a 16mb version of the Pro) on the flea market today. Nice upgrade from the 384mb and rage 128 non-pro that I used to run in my W98 machine.

No, Rage 128 Ultra is to Rage 128 Pro the same that TNT2 M64 is to TNT2 Pro. It's the same GPU, but linked to 64 bit memory. Rage 128 (non pro) should be faster in most games than Rage 128 Ultra.

I'm confused now 😀.

My old card said "rage 128" on the card.

My new card says "rage 128 pro" on the card. Dxdiag says it's 16mb.

Which of the two is the better card?