H3nrik V! wrote on 2026-09-21, 07:03:

Briliant tool! Had it been there before I designed my own, I'd definetely use that. […]
Show full quote

Lopo wrote on 2026-09-20, 08:31:

Today's retro activity was more CAD than retro: I got tired of CPUs rattling around in boxes, so I wrote a small generator for s […]
Show full quote

Today's retro activity was more CAD than retro: I got tired of CPUs rattling around in boxes, so I wrote a small generator for stackable 3D-printable chip trays.

Chip Tray Generator (runs in the browser, single HTML file, source on GitHub, MIT)

What it does:

• PGA/LGA CPUs from 386 (PGA132) through 486, Socket 5/7, 370, A, 423, 478, 754/939, AM2/AM3 up to LGA 775, plus QFP/PLCC and DIP chips in channels.
• Pick a package, rows x columns, walls, tolerance; live 3D preview; export OpenSCAD (exact, parametric, engraved label) or STL.
• Trays stack: rim on top, matching recess underneath. With "fixed outer format" trays for different sockets get the same footprint, so a 386 tray stacks on a Socket 7 tray.
• The chip never rests on its pins: ceramic PGAs sit on a pedestal inside the pin-free centre, FC-PGA 370 on an edge ledge (Intel puts capacitors on the underside), QFP/PLCC on a pad under the body, DIPs on a centre rail like in a tube.

Package sizes and pin-free centres come from the datasheets (Intel 231630, 241997, 245264, AMD 25175, 21850, JEDEC outlines), not guesswork. What is NOT done yet: I haven't printed every variant, so if you print one and the tolerance or pedestal is off, tell me and I'll fix the defaults.

Briliant tool! Had it been there before I designed my own, I'd definetely use that.

I found a little glitch though

Using 178x130 milimeters as outer limits (to fit the trays I've designed myself), selecting DIP-20 and above the pockets gets outside of the tray.

Thanks, good catch — fixed and live now. With a fixed outer format the channel length was never checked against the tray, so a DIP-24/40 channel with 4 chips (135 / 218 mm) was placed into a 130 mm tray and the rails poked out through the walls. Now the chips-per-channel count drops automatically until a channel fits (the label shows what it was limited to), and there's a new "Channels along the width" option that rotates the channels 90° — at 178×130 that gets DIP-40 to 3 chips per channel instead of 2. And if nothing fits at all, the tray is drawn empty with a warning instead of cutting through the wall.

For anyone interested in the tray generator: I've opened a dedicated thread for it, so fixes and print reports don't clutter this one — Chip Tray Generator — stackable 3D-printable trays. The DIP channel overflow reported above is fixed there.

dormcat wrote on 2026-09-20, 19:09:

Cleaned up the degraded thermal tape between the heatsink and the main chip of my no-brand Voodoo Banshee reference board.

The attachment Banshee.jpg is no longer available

Will need a new thermal pad or tape. Plus a new 4×4 cm fan.

Think they originally had heatsink, no fan? Your board also has holes for nylon plugs so might be able to avoid the tape.

I became frustrated trying to uncover the origins of my -4ns 128bit down-clocked Radeon 7500, which carried both Manli stickers and HIS VBIOS but was claimed by neither Manli or HIS.

Going back to the PCB components, had I misread something? The actual integration and procurement looks a lot more performant than the downclocked firmware was aiming for, and that's the hook that gets me digging because I like it when good work is properly attributed.

The only permanent identifier is DH in the silkscreen. Now, I've seen DH on various cards from Butterfly and others. After extensive sleuthing I think this particular card is more than likely HIS-specified, Manli-integrated, with parts DH-procured. And, as stickers and VBIOS can be changed the only permanent identification is DH.

But who is DH? Probably a company called Double Hero; Double Hero have been a subsidiary of PC Partner since 1994 and their role is sourcing parts from Chinese factories to satisfy the demand set by PC Partner's partners.

To be continued..

I watched a VHS tape on my commodore 1702 monitor like I used to in the 90s.

I burned an Audio CD, 'nough said.

EduBat wrote on 2026-09-21, 18:37:

I burned an Audio CD, 'nough said.

My last car had a cd player and i burned a lot of cds to listen to music in it. Now its FLAC files on usb though

Law212 wrote on 2026-09-21, 17:17:

I watched a VHS tape on my commodore 1702 monitor like I used to in the 90s.

At one point I had a Betamax hooked up to a 1024S in my "shop" as it was for a year or two, sharing it with an Amiga of course. Didn't have a lot of media for it though. Didn't see video as crisp until DVD on early noughts TV though.

I took out my Cambridge soundworks four point out of storage and connected to turtle beach montego aureal vortex only to find out it's a 2.1 card only

gmaverick2k wrote on 2026-09-21, 20:10:

I took out my Cambridge soundworks four point out of storage and connected to turtle beach montego aureal vortex only to find out it's a 2.1 card only

I could be wrong but the Turtle Beach card kinda looks like a 2.1 only design.

Granted idk if my MX300 is any different in terms of what it can do.

Small note: is your card using a AU8820? I seem to find the design used both variants. (8820 and 8830)

EduBat wrote on 2026-09-21, 18:37:

I burned an Audio CD, 'nough said.

Maybe I should get back to that, just for the hell of it. I have spindles of CD-R discs and nothing to do with them.

I got the "untested" Slot-1 box I grabbed for practically nothing from VCFMW working. All it needed was a new CMOS battery and the drives configured in BIOS. I'm sure this will be useful to somebody...

Brawndo wrote on 2026-09-22, 01:47:

EduBat wrote on 2026-09-21, 18:37:

I burned an Audio CD, 'nough said.

Maybe I should get back to that, just for the hell of it. I have spindles of CD-R discs and nothing to do with them.

I use 'em for my PlayStation 2. The original games stay in a box and this way I can use things like undub patches without having to mess with shady pirate loaders.

PcBytes wrote on 2026-09-21, 22:56:

I could be wrong but the Turtle Beach card kinda looks like a 2.1 only design. […]
Show full quote

gmaverick2k wrote on 2026-09-21, 20:10:

I took out my Cambridge soundworks four point out of storage and connected to turtle beach montego aureal vortex only to find out it's a 2.1 card only

I could be wrong but the Turtle Beach card kinda looks like a 2.1 only design.

Granted idk if my MX300 is any different in terms of what it can do.

Small note: is your card using a AU8820? I seem to find the design used both variants. (8820 and 8830)

I own an SQ2500 as well but I'll have to go digging around looking for it

PcBytes wrote on 2026-09-21, 22:56:

I seem to find the design used both variants.

There are three variants.
https://www.amazon.com/-/zh_TW/LYSB00UQK73TU- … S/dp/B00UQK73TU

Hello,

I am trying to make Companion of Xanth working with both AWE64PnP voice and Roland MT-32 Music. I use SoftMPU but the MIDI music stops each time speech begins.
SET BLASTER=A220 I5 D1 H5 P330 T6 & SOFTMPU /SB:220 /IRQ:5 /MPU:330.

DataPro wrote on 2026-09-22, 14:20:

Hello,

I am trying to make Companion of Xanth working with both AWE64PnP voice and Roland MT-32 Music. I use SoftMPU but the MIDI music stops each time speech begins.
SET BLASTER=A220 I5 D1 H5 P330 T6 & SOFTMPU /SB:220 /IRQ:5 /MPU:330.

Does the game have a configuration executable to change hardware settings?

MattRocks wrote on 2026-09-22, 11:30:

PcBytes wrote on 2026-09-21, 22:56:

I seem to find the design used both variants.

There are three variants.
https://www.amazon.com/-/zh_TW/LYSB00UQK73TU- … S/dp/B00UQK73TU

Only 2 official (Turtle Beach made.) ones.
8810's design from that link is not a official product, rather a chinese clone (which also raises the question whether that 8810 is actually genuine, as the Chinese have forged Vortex chips in the past.).

For reference, this is what 8810 cards look like: https://theretroweb.com/expansioncards/s/aure … ortex-advantage

Turtle Beach has also never used the 8810 with the specific PCB used on Montegos - only 8820 and 8830.

The PCB itself has been (somewhat) reused by Aureal themselves with SQ1500's release, but not by Turtle Beach, and also has different 3.5 jacks used (including a S/PDIF jack not present in either of the previous designs card itself has been based on - both Aureal's own 8820 and Turtle Beach used the rather generic 3 jack + gameport design for the original 8820 (and TB's 8830), somewhat similar to some older ES1373-based SoundBlasters.)

My Rialto research got a little exciting yesterday and this morning.

I now know with a good amount of certainty that rialto performs independent reads of a 18 byte config block before the normal GPU startup happens. Furthermore, the behavior is consistent between X700, X1950GT and HD3450. I traced all SPI bootup traffic from these cards and was able to determine that they behave the same.

It appears that Rialto is reading a fixed address in the EEPROM that has a pointer to the 18 byte configuration block. That block can be stored "anywhere" (well, not anywhere, there are some limits) - relocating it to a different position and changing the pointer will maintain the card operational. This is awesome. I found as well that the config block is stored in different places in the VBIOS, while X700 and X1950 differ - the HD series all appear to have it in the same place. Further between generations, the 18 byte block differs only in 3 bytes - the rest is entirely the same.
So now the idea is to figure out if the X700 would work with the same config block from the X1950 and HD3450.

For a nerd like me this is pretty exciting 😁

Almost lost my sanity with SiS730 and a PowerVR Kyro (3D Prophet 4000XT).

In the end I admitted defeat and swapped in a GF4 MX440.
The Kyro simply WON'T work with about anything that runs a 462 Athlon, for whatever reason. (yes, it does that on multiple boards. KT266 refuses to even POST with it, nForce 2 and SiS 730S share the same driver issues.)

Slapped it on a GB BX2000 with a slotket'd 750MHz P3. No crashes, on 512MB RAM.

been messing around trying to get build engine games running on both my dos and 98 machines. had an itch to play shadow warrior and redneck rampage. dos is too slow with the pentium 166 mmx and dont want to drop the redundant k6-2+ 500 in there as its meant to be a dos machine entirely. moved onto my 98 machine and in pure dos mode with the vortex 2 the system kept crashing. used google ai to suggest which to use or if i can combo my different dos pci soundcards, ymf744, solo-1 and vortex 2. i had given up on the solo-1 because it wasnt behaving in the past. in pure dos mode, the solo-1 would also crash in redneck rampage. there's no way to get around TDMA mode on socket A even though i tried to force it with /D switch in config.sys. google ai suggested using both the solo-1 and vortex 2 with sc-55 to cover all bases. solo-1 would be configured to deal with dos emulation within win98 (redneck and shadow warrior work great) and vortex 2 with soundblaster emulation disabled to do windows sounds and gameport. so output of solo-1 as input into vortex 2. vortex 2 with cd in and mpu-401 duty. problem solved