Shreddoc wrote on 2022-04-02, 02:33:

I did the heatsink chopping-up thing, once. Never again.

Far too much metal dust anxiety and cleanup to be worthwhile. Everything within a 5m radius of the workspace (including own skin + clothes + hair) afterwards has some quantity of microscopic metal shards (aka Magic Shorting & Random Fault Dust) randomly all over it. Virtually gotta walk through a decontamination shower before you're safe to go within a metre of anything electronic afterwards!

All over a measly $1 bit of aluminium.

It's not that bad with a manual hacksaw, well tensioned sharp blade, fine tooth.

Shreddoc wrote on 2022-04-02, 02:33:

I did the heatsink chopping-up thing, once. Never again.

Far too much metal dust anxiety and cleanup to be worthwhile. Everything within a 5m radius of the workspace (including own skin + clothes + hair) afterwards has some quantity of microscopic metal shards (aka Magic Shorting & Random Fault Dust) randomly all over it. Virtually gotta walk through a decontamination shower before you're safe to go within a metre of anything electronic afterwards!

All over a measly $1 bit of aluminium.

That sounds pretty terrible 🤣

Btw, did it at least work? 😋

BitWrangler wrote on 2022-04-02, 14:29:

Shreddoc wrote on 2022-04-02, 02:33:

I did the heatsink chopping-up thing, once. Never again.

Far too much metal dust anxiety and cleanup to be worthwhile. Everything within a 5m radius of the workspace (including own skin + clothes + hair) afterwards has some quantity of microscopic metal shards (aka Magic Shorting & Random Fault Dust) randomly all over it. Virtually gotta walk through a decontamination shower before you're safe to go within a metre of anything electronic afterwards!

All over a measly $1 bit of aluminium.

It's not that bad with a manual hacksaw, well tensioned sharp blade, fine tooth.

Probably so. I don't recommend doing it with a dremel or angle grinder. In that context, invisible or barely-visible metallic particles are being flung around in far greater quantities and coverage than many realise. I just don't rate the risk, personally.

Tetrium wrote on 2022-04-02, 15:38:

Shreddoc wrote on 2022-04-02, 02:33:

I did the heatsink chopping-up thing, once. Never again.

Far too much metal dust anxiety and cleanup to be worthwhile. Everything within a 5m radius of the workspace (including own skin + clothes + hair) afterwards has some quantity of microscopic metal shards (aka Magic Shorting & Random Fault Dust) randomly all over it. Virtually gotta walk through a decontamination shower before you're safe to go within a metre of anything electronic afterwards!

All over a measly $1 bit of aluminium.

That sounds pretty terrible 🤣

Btw, did it at least work? 😋

Nope! 😀 It could have worked, but after cutting I had second thoughts, and decided not to put a hard-to-obtain $$$ board at even a tiny risk by following up a metalwork session of invisible flung particles, with a play-with-the-computer-innards session, all over a $2 heatsink that I could pick up in town the next day. Possible? entirely! Economical, or worthwhile? I say hell no.

Solo761 wrote on 2022-03-31, 22:01:

Welp... I found out why I had issues with VT6421A PCI SATA/PATA controller cards not getting detected on boot on my P233. Story […]
Show full quote

Welp... I found out why I had issues with VT6421A PCI SATA/PATA controller cards not getting detected on boot on my P233. Story time...

I bought two a long time ago to try and mod XBox360 drive and they laid around since. They worked fine then (about time when Core2Duo were mainstream) but for some reason they don't now. They have ROM chips so it should have it's own BIOS on boot time, and yet nothing shows up beside regular BIOS boot...

I thought maybe it's something to do with too old PCI standard on my QDI Titanium IB+ so I tried them on one Athlon X2 I had nearby. It booted to XP and I got prompted with "found new hardware". OK, so they do work. Great, lets try flashing them with custom (or just updated) firmware I found here http://toastytech.com/files/w95stuff.html.
But flasher tool said "Unknown LPC" and returned strange strings for flash chip ID (FFh/FFh)... Google wasn't that helpfull so I was puzzled a bit and took off sticker that was on ROM chip to see which one is it. It was CAT28F010N. Haven't encountered this one before, but it looked like standard 128kB flash ROM so it should work...

Card.jpg

And then I noticed one small detail that explained all of this... ROM chip isn't soldered!

ROMChip01.jpg ROMChip02.jpg

On each side only two pins are soldered to hold it in place! I tried checking few of these unsoldered pins for continuity and yup, no continuity, they're not even touching the pads below... So although these cards have ROM chips they're not used and they behave the same as variants without firmware. That's why there's no drive check on boot, but it gets "noticed" in windows.

Well... now I have new project. Desolder chip on one and pop it in standalone programmer to see if they're real or just dummy chips. Nothing would surprise me now... If they're real I'll flash firmware into it and solder it back on to see if that does something.

The more you know...

Well, saga continues. If someone's interested in how it went...

Desoldered EEPROMs, tested them on my programmer(s), one works, other is not getting detected... Flashed the firmware into working one, soldered it back onto the board and still nothing. No boot screen during motherboard boot, new hardware detected in windows, same as before. Tried to flash it with VT6421(a) flash tool and again same message, "LPC Flash Not Exist"... this made me a bit curious and it seems I learned something new.
It seems that these chips are wrong type for these cards. Chips are 32 pin, but at most third of them are connected to something, others look like just pads, and no visible vias near them. It seems that these card use so called chips that have so called LPC mode, short for "Low Pin Count" mode. That's why flasher reported "LCD Flash Not Exist". These CAT28F010N are not LPC type...

Now it's a matter of pride (and stubbornness) , time to get some LPC chips and see if that will work 🧐

Picked up another laptop - one I'm quite familiar with....as I've had 2 other ones (a color one and a monochrome one.

This was the most anemic one I've had so far: 25MHz 486 SX, 8MB RAM, Casio 10" DTSN Monochrome LCD with a scratch in it, but I got to hand it to them, NanTan could actually make a great machine when they wanted to and this seems to be the ticket. Took it apart to upgrade to a DX2-66 and see what I could do to move from the DTSN screen to a color Active Matrix or something similar. So that upgrade might become a future project - once I'm done getting the M/75's touch panel done, which I'll probably wait until winter to carry out. Right now it's kicking but with the spare DX2 I have laying around in it with full SoundBlaster sound. Actually, the CASIO LCD is kind of surprisingly good for such a cheaply made panel. I also put an 80GB Drive with FreeDOS on it in there.

Windows 3.1's birthday coming up this Wednesday I started messing around with a 386. I'd like to connect it to my network, but the out of the 4 ISA networking cards I have I only have one 3Com which I want to keep for a more premium 486 build I'm planning. All the rest is SMC and I can't find there original driver. Do anyone know where to get it?

Continuing The Great Cataloging Of It all, taking approximately 2 forevers. Just did a bunch of laptop power adapters and desktop PSUs, as well as finally breaking out a box from the last Liqmat major haul (that I know of) and getting that stuff photo'd and listed.

I had an XT PSU from Kingspao that someone had tried and failed to disassemble and gave up and threw out. As far as I can tell they just removed the fan screws and let the fan fall down inside. SO I fixed that (the warranty sticker hadn't been breached), and, well, I haven't tested ripple yet but the voltages are correct (enough). I have a few failed AT PSUs I need to break down and check but I need to finish my catalog and get some more stuff out of here. The funny thing is I got rid of a monitor and a couple of computers and somehow it seems *more* crowded!

I mentioned in the bought-it thread I got 5.25" and 3.5" rolodex-style disk boxes, which I've been looking for for a long time but haven't found except for very expensive ebay listings, these were cheap. Goodwill may not have computers anymore, but every so often...

Kadotus reminds me I have a Socket 7 I think needs recapping - I need to test some more CPUs on it, because it could just be that my Celeron 566 and P3 1000 are both dead - not sure if that's worse or better than the mobo being bad.

Ordered a pile of capacitors to recap my Abit BH6 and Apoen AX34-U and learned that an overworked power supply could not be fixed by adding another diode to the regulator.

;

RandomStranger wrote on 2022-04-03, 05:50:

Windows 3.1's birthday coming up this Wednesday I started messing around with a 386. I'd like to connect it to my network, but the out of the 4 ISA networking cards I have I only have one 3Com which I want to keep for a more premium 486 build I'm planning. All the rest is SMC and I can't find there original driver. Do anyone know where to get it?

That looks like an SMC 8416T or a variant thereof .

Try this

EDIT :
This one looks newer .

Source : https://web.archive.org/web/20030423053444/ht … brary/ETHER.EXE

Yep, thats an SMC EtherEz 8416T. Here is the Data Sheet and Users Guide. There are some more drivers here.

I was using one for my Windows 95 network booting experiments a few weeks back - they seem like nice enough cards.

Thank you both!

worked on my new 386 😀
Re: Bought these (retro) hardware today

The included 60MB hard drive was defect so I had to use a new 40GB drive.
In the bios it was setup like this :

To be safe I wanted to setup the new 40GB HDD as a 127MB drive :

So I used Seatool for DOS on another computer to fix the drive capacity to the exact amount ( in LBA ) corresponding to C/H/S 1023/15/17.
Plugged in the laptop, Bios set on Drive type 42 and installed MSDOS 6.22 + Windows 3.1 flawlessly.

It seems that even if the drive is setup with a larger size ( like 520MB) it's working as well, even if the Bios is setup for a smaller drive. I'm not sure why and how it's possible. But I didn't want to face bugs and issues so I preferred to stick with 127MB. For this computer it's good enough 😀

Spent a sunday afternoon writing Yet Another DOS Int33 Mouse Driver.
I wrote it in C because why not and because I guess there's no other open source mouse driver in C.

Being in C it is more suited for a modern/virtualized PC so I added VirtualBox integration to it.
https://user-images.githubusercontent.com/123 … eb242551422.mp4

Plus Windows 3.x integration so that it works on DOS windows too.
https://user-images.githubusercontent.com/123 … 50e6c0e71dc.mp4

And despite the bloat it's still like 8KB of memory, less than MS Mouse 9 😀

That's great! Can you be tempted to add VMware integration as well?

Tried installing Raspberry Pi Desktop (Debian Buster) on an old Atom laptop, again. Installation went through but then it entered hibernation EVERY 30 SECONDS! The problem seems to be quite common on different Linux builds and there's no easy way to fix. 🙄 Guess I have to stick with Windows XP.

Removed, cleaned and lubricated the fans on my Voodoo5.
Tested everything on my ASUS-P3BF sitting on my home-made test bench.

That homemade test bench is adorable. How did you decide to raise the board up from the base? Did you add rubber bumpers or standoffs?

Kahenraz wrote on 2022-04-04, 17:51:

That homemade test bench is adorable. How did you decide to raise the board up from the base? Did you add rubber bumpers or standoffs?

The base is made from 4 plynths (similar to the legs) that form a hollow square base. A thin layer of OSB board was then added as the base.
The legs are made long enough to elevate the base, with each leg having a felt pad (or rubber if you prefer) to stop it from scratching any surface.
Big thanks to my dad for that idea 😀

Decided as part of distracting myself so I don't finish The Great Catalog I would start "nailing down" systems to get them kind of off my mind. Had a radically wrong video card and no NIC in the Dell R450, so I swapped in a Creative TNT2 "Pro" (identification info is a little irregular but it seems faster than a regular TNT2) and a NIC and Win98 SE (found some fun incompatibilities with partition utilities and had to do several reboots to get it cleared so Win98 could repartition), just got USB working and got the driver packages for some components and the SBLive (Audigy drivers), lots more to go but I'm satisfied that, barring some radical failure in testing the hardware will work out okay. Stole the Maxtor 15Gb for my "small drives" box and used a 40Gb in it for now.

USB 1.0 sure is slow.

Have plans for some specific games from back in the day, themes, etc. Will not finish all that for a while but it's usable now instead of sitting there incomplete. Probably next up the AST Advantage P100 with the funky case or maybe slapping the DX4-100 in that pristine Gateway tower.

I will try to get un-distracted tomorrow, as I have something to put on eBay and a couple more things to put on FB Marketplace. I'm starting to get a pile together for a giveaway, no idea when I'll be able to get that rolling, but it won't be all at once so maybe that will start soon.

Despite having thoroughly recycled the *actually* impossible to shift or irreparably broken stuff, I keep finding new stuff for the pile (like the Zumax power supply - whose case I might actually re-use for a project, though as a PSU it's a no-go.

I’ve been on the hunt for a while for an excellent PCI or PCI-Express graphics card to pair with Windows 95 (both Retail and OSR2) and my Ivy Bridge overkill system. I haven’t had any luck with PCI-Express cards from either Nvidia or ATI because their Win 9x drivers just don’t work properly with Windows 95.

I want a card with DirectX 7 and 8 compliance as well as the ability to display at 1080p widescreen.

I found a lot of references to the Radeon 9200 and 9250 PCI cards working with Windows 95 and the Catalyst 3.4 drivers. I’d hate to buy yet another PCI graphics card without knowing for sure if it would work for the task (I’ve gotten quite a few PCI cards over the years that I discovered didn’t meet my needs). So, as I sat down to think about how much I wanted to spend on yet another PCI card, it hit me. I remembered that I had a Mobility Radeon 9000 (AGP, 64MB) in my IBM Thinkpad T42.

I fired up the Thinkpad and installed Windows 95 OSR2 onto its hard drive. While most of the computer’s hardware won’t work properly with Windows 95, Windows detected the graphics correctly using the Catalyst 3.4 drivers. I initially got a blank screen on the laptop’s LCD, but the display began to work for me once I attached my widescreen monitor to the external VGA port. This Mobility Radeon 9000 only seems to work properly with Windows 95 when I have an external monitor attached during bootup (it will display on both the LCD and external monitor simultaneously). Weird!

I then installed DirectX 8 and some benchmarks. I have to say, I’m impressed. A 1024x768x32 resolution got over 60 FPS at high details in 3DMark 2000. I think the AGP version of the Radeon 9000 can definitely game with Windows 95. I didn’t have any stability issues either.

3Dmark also offered to test at 1080p. It got a respectable 60 FPS average at low details in 1080p (32-bit), but averaged slightly less than 30 FPS at high details. Still, I got widescreen resolution that way. The Catalyst drivers don’t offer widescreen resolutions to Windows by default, but I enabled them by adding “1920,1080” registry keys to the 8, 16, and 32-bit color depth keys with defaults values for “60” Hz. That worked to add a widescreen choice to the Windows 95 display options.

I don’t know how much of a difference to expect between the PCI variant and the AGP one in the Thinkpad, but I feel a lot more confident about buying a PCI Radeon 9200 or 9250 knowing that the Windows 95 drivers work pretty darn well.