Just spent the last two evenings wondering why every single ATX motherboard I tested would power off after 2-3 seconds.

Turns out I was using a latching toggle button for the power button. Talk about a dumb mistake. 😮

Combine two broken floppy drives into one working drive, also I’ve changed the LED to yellow.

Shponglefan wrote on 2024-01-19, 18:07:

Just spent the last two evenings wondering why every single ATX motherboard I tested would power off after 2-3 seconds.

Turns out I was using a latching toggle button for the power button. Talk about a dumb mistake. 😮

It should be possible to make them nonlatching if the switch can be taken apart. Usually there's a simple mechanism with a little metal bit going in a predefined groove and taking that metal bit off makes the switch no longer latch.

Tiido wrote on 2024-01-19, 21:15:

It should be possible to make them nonlatching if the switch can be taken apart. Usually there's a simple mechanism with a little metal bit going in a predefined groove and taking that metal bit off makes the switch no longer latch.

Oh, I have momentary buttons as well. I usually use the latching toggle buttons as turbo buttons when testing older AT boards.

I just happened to grab the wrong button to use for power on these ATX boards. It took me two whole days to realize it. 😁

Oh, hahaha. I have had a similar issue but it was because the switch got stuck inside while the button on top did come out, it took a while to figure out why the machine went off not long after powering 🤣

Package from China finally arrived

Fixed a purple GF2 MX400 with quite a bunch of missing SMD caps.

Totally not flexing my SMD soldering 🤣, I feel like it's far from looking good. But it works, and that's all that matters.

Not a good day for retro for me I guess. I got this untested Compaq motherboard, which I connected to a regular PSU power supply. When I connected the ATX cable to motherboard something exploded on the motherboard and I could also smell burning electronics. It's a shame. I could not locate the damage. Does Compaq have different wiring on ATX PSU connectors or what?

Turbo -> wrote on 2024-01-20, 15:09:

Not a good day for retro for me I guess. I got this untested Compaq motherboard, which I connected to a regular PSU power supply. When I connected the ATX cable to motherboard something exploded on the motherboard and I could also smell burning electronics. It's a shame. I could not locate the damage. Does Compaq have different wiring on ATX PSU connectors or what?

It looks like a cap and two diodes, and maybe the mosfet.

I've only heard of Dell using proprietary power supplies. Is the damage at the top right corner? Those components are easily accessible for repair.

Kahenraz wrote on 2024-01-20, 18:30:

I've only heard of Dell using proprietary power supplies. Is the damage at the top right corner? Those components are easily accessible for repair.

It seems Compaq did too, as I don't recall seeing a 24 pin ATX lead on a Socket 370 mobo.

Ahh, very true. I associate that green shroud on AGP slots as a "Pentium 4" era thing. But is that still too early? When was 24 pin ATX introduced?

IIRC 775 was the earliest I know of 24 pin ATX, Excluding ASRock's funky boards.

Just for fun, I hooked up my Win9x test rig to a 42" LCD TV using a DVI to HDMI adapter. I made sure to set the scaling to "Aspect Ratio" in the Nvidia driver panel. The end result is pretty decent, as long as you use normal TV viewing distance (2-3 meters). Output from the sound card goes to the TV's RCA inputs. I also connected a Logitech RumblePad 2 for a true couch gaming experience. 😀

Currently having some fun with Final Fantasy 7 on that setup. Looks and plays great.

Today I sorted some of my stuff, I’ll probably get rid of my mainboard/CPU combos from C2D and up to get some space back.

My goal was to only have SS7/370/A/754/939 systems and nothing more. Well now I have some 486 and other stuff I absolutely don’t need 😂

Tried running 3dMark2000 on the GF2 MX400 above. Getting a "3dMark2000 requires a DX7 GPU" error unfortunately.

Test machine - Gigabyte 6VX7B-4X, Coppermine P3 700/100FSB, 80GB Maxtor, GF2 MX400, 384MB RAM, Realtek NIC, Win2k Pro SP4.

Working on the EPIA-800 Win95 machine.

Got the network up and running for file transfers, and couldn't help but install Netscape Navigator 4 and see how many "errors" it would take to load Vogons. Did it in IE as well, just for kicks. 😀

About a dozen errors later...

Captured directly from Win95 using Print Screen and PSP 5. 😜

Good times!

Find an SSH client and an X server and you can run modern Firefox and Chrome over the network with X11 Forwarding on Windows 95.

This worked for me in Windows 98. I believe that Cygwin works with Windows 95 as well. There are caveats though.

https://msfn.org/board/topic/183781-x-forward … eue-processing/

https://msfn.org/board/topic/183777-how-to-de … n-windows-9xme/

Having way too much with this EPIA-800 today. 😜

Was doing a lot of playing around in DOS, thought it would be fun to run some old school DOS benchmarks.

Now I'm getting ready to dump a ton more games on it, mostly for Windows 95.

dominusprog wrote on 2024-01-19, 19:53:

Combine two broken floppy drives into one working drive, also I’ve changed the LED to yellow.

IMG_20240119_205556.jpg
IMG_20240119_205356.jpg

Congratulations on making one working one 😀 I've done similar and if you managed that while keeping the calibration then that's excellent - even impressive if you had to recalibrate after. At this point I've done motor swaps and other stuff on Citizen W1D drives to get 1 working one from 2 duds - always worth it.

The past few days things have settled down and I've been able to print a much more suitable Compact Flash card holder since I now have much more than 10, this design screws into the bottom of a wood shelf and uses VHB tape on the front to attach while keeping the space above fully usable. This design holds 25 cards and is the width limit of my printer - it sucked having to re-do it as well since it takes 4 hours to print and the first print I forgot to put in the tolerances so no cards could fit.

Previous designs I've done for a PCMCIA card holder were just 1 row holding six cards but they weren't staggered like this - the 45 degree angle means I can get to each row and column easily, it prints very nicely too since no overhangs.

The very dirty Tecra 750DVD is all cleaned up now and apart from the smashed LCD panel (which I have a spare for), it's working great. The DVD drive didn't want to read discs though and it's very unique to this laptop model. I spent some hours with the drive in bits and eventually got it functioning by reflowing the flex cable connectors for the drive sled and laser mechanism. A USB > laptop IDE converter + a bench PSU were a great help in getting it working.
Also this laptop got upgraded to a CF card instead of the 5GB hard drive which was making some crazy clunking noises and the bearing noise drove me crazy. So much nicer with general operation now being silent.

I'm glad I was able to get the DVD drive working - it's a custom physical size so a replacement drive just isn't an option, this laptop has an integrated DVD decoder so dropping to a CD-ROM drive would be a real loss.

Actually using USB-C Power Delivery is such a huge convenience - cables that can request their specific voltage mean I don't have to use the bench PSU to power every single thing any more:

Not pictured but I even made a USB-C Power Delivery 'dongle' for my Toshiba Portege 3000 series and Libretto 100CT laptops - they use a really tough to source tiny 2-pin connector and I hacked one up by melting a suitably sized JST connector. Now my Portege laptops are portable again even with dud batteries since power banks can source enough power to pretend to be the DC power brick 😀