ssokolow wrote on 2024-08-18, 09:06:

I haven't encountered that exact combination before, but it's commonplace in the vintage mac community, where the drives are oft […]
Show full quote

iraito wrote on 2024-08-15, 15:59:

I did and now the system is complete.
Everything works but the damn CD drive, it can read CD audios and I can listen to Redbook tracks but most other CDs can't be read, for some reason it can read some games but it's unreliable.

I haven't encountered that exact combination before, but it's commonplace in the vintage mac community, where the drives are often more proprietary, for the optical drives to have flaky compatibility because, even if you don't use them, laser diodes also slowly go bad with age and one symptom of that process in progress is that, for example, they struggle more with burned discs than factory-pressed ones.

I'd try grabbing a $30 SATA DVD rewriter and one of the appropriate variant of those $3 Aliexpress SATA-PATA converter boards and sticking a new optical drive in it. (Both prices in Canadian Dollars, so I assume they'd be 70% of that amount in USD.)

(The rate they go bad varies. I've got a hand-me-down HP prebuilt from 2012 and its drive is already more or less dead and I've also got a Power Mac G4 Quicksilver 2002 where the replacement drive I put into it was an actual PATA DVD rewriter I pulled from a PC in the closet.)

I saw suggestions a while ago that the failing laser was a bit of a myth, at least while drives still worked at all, just clinging on to audio capability say. Instead it was proposed that there was a capacitor aging out in the read head drive circuits, possibly ultimately resulting in laser death but while "something" happens, repairable. I don't think I saw full detail at the time, and was meaning to dig it out when I "got serious" about needing some of the more ancient optical drives to be working, but that has not happened yet.

Working on my DOS6.22 build, trying to figure out why no matter what drive I use, it keeps telling me it's write-protected.

Posting here to try and figure that out.

Enjoying some Knight Rider while doing it. Got the full series super cheap on DVD a while back, just now getting around to watching. 😀

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

That's fantastic 😀 Posting it somewhere to ask what it is would be fun I think.

Trying to keep old computers working is tough work, I just dusted off my small open testing PC with the K6-2: Re: What retro activity did you get up to today?
Upon first turning it on the cd-rom drive which is a laptop type was just spinning up and spinning down constantly, saying "drive not ready" when trying to read.
I've got lots of optical drives but they're in the storage place right now so rather than replace it I took the drive apart, then cleaned & re-seated all the flex cable connectors. That seems to have helped and now it can read CDs again so it doesn't need replacing.

Some time ago I got another Diamond Monster 3D II 8MB card with a plan to add 4 more MB to it, to match the other same card that's already been upgraded, then try SLI out 😀
Earlier in the day I had finished up fixing a Voodoo 1 card which was sold as faulty but I got it working. This Diamond Voodoo 2 card was sold as working but not tested so I was very disappointed to load up a game on this new (to me) Voodoo 2 and all the textures were flashing like crazy and very similar to this:

Even when the Mojo diagnostic data showed that the FBI was there with full RAM and both TMUs were reporting and showing 2MB RAM each, which is correct.

Argh, a new problem. Pressing on bits of the card doesn't make much difference and because its textures that are flashing while the geometry is good, I know it's one of the TMUs that's got the problem. Turned off one of the TMUs using this in autoexec.bat: "SET SSTV2_NUM_TMUS=1"
Started up GLQuake which was a horrendous mess before and success, with just 1x TMU enabled the card is working properly. vswitchzero's videos are good on this so I know which TMU is which, the TMU near the VGA port is the secondary one - which makes sense if you look at the Voodoo 2 as a Voodoo 1 card that grew an extra TMU in the largest remaining area of PCB space. The second one got disabled and that fixed the card so I turn my attention there.

Instead of checking stuff the hard way, I get some titanium tweezers and run them along all the TMU's QFP pins that hook to the memory, some of them were indeed loose:

This is not kind to the legs of the QFP chip but the method is certainly quick.

Pulled out the microscope again, the day after packing it away once more. Soldered those loose pins back into place and checked that nearby pins were soldered down well which is far easier through the scope. Tested the card again and all games look great now, the speed boost from V1 to V2 is quite impressive.
I had some weirdness trying to get Unreal to display on this card in Glide where it would just show a black screen even though all other games in Windows were fine. For some reason setting the Glide resolution in unreal.ini manually to 640x480 instead of 512x384 made it display. I'm glad that was all it was.
Now this Voodoo 2 works properly I'm planning to put the extra RAM on it soon 😀 Then make an SLI cable

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

The piece of art 🤣 made me think: why was there no slotkets designed for Socket A CPU on Slot A motherboards? Wikipedia has only a short description:

Wikipedia wrote:

Slotkets were never introduced to take advantage of the AMD Athlon processors' transition from the Slot A form factor to the Socket A form factor.

Could it be that Slot A was very short-lived and was replaced / succeeded by Socket A, while Slot 1 had much more market share and coexisted with Socket 370 for two more years?

First Slot A CPU announced: 1999/06/23
Last Slot A and first Socket A CPU announced: 2000/06/05

First Slot 1 CPU announced: 1997/05/07
First Socket 370 CPU announced: 1998/08 (date unspecified)
Last Slot 1 CPU announced: 2000/07/31 (even outlived Slot A)

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

🤣 great work, but is the slot PCB actually a Slot-A or a Slot-1?

Thermalwrong wrote on 2024-08-21, 01:46:

That's fantastic :) Posting it somewhere to ask what it is would be fun I think. […]
Show full quote

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

That's fantastic 😀 Posting it somewhere to ask what it is would be fun I think.

Trying to keep old computers working is tough work, I just dusted off my small open testing PC with the K6-2: Re: What retro activity did you get up to today?
Upon first turning it on the cd-rom drive which is a laptop type was just spinning up and spinning down constantly, saying "drive not ready" when trying to read.
I've got lots of optical drives but they're in the storage place right now so rather than replace it I took the drive apart, then cleaned & re-seated all the flex cable connectors. That seems to have helped and now it can read CDs again so it doesn't need replacing.

Some time ago I got another Diamond Monster 3D II 8MB card with a plan to add 4 more MB to it, to match the other same card that's already been upgraded, then try SLI out 😀
Earlier in the day I had finished up fixing a Voodoo 1 card which was sold as faulty but I got it working. This Diamond Voodoo 2 card was sold as working but not tested so I was very disappointed to load up a game on this new (to me) Voodoo 2 and all the textures were flashing like crazy and very similar to this:

The attachment 3dfx party mode.JPG is no longer available

Even when the Mojo diagnostic data showed that the FBI was there with full RAM and both TMUs were reporting and showing 2MB RAM each, which is correct.

Argh, a new problem. Pressing on bits of the card doesn't make much difference and because its textures that are flashing while the geometry is good, I know it's one of the TMUs that's got the problem. Turned off one of the TMUs using this in autoexec.bat: "SET SSTV2_NUM_TMUS=1"
Started up GLQuake which was a horrendous mess before and success, with just 1x TMU enabled the card is working properly. vswitchzero's videos are good on this so I know which TMU is which, the TMU near the VGA port is the secondary one - which makes sense if you look at the Voodoo 2 as a Voodoo 1 card that grew an extra TMU in the largest remaining area of PCB space. The second one got disabled and that fixed the card so I turn my attention there.

Instead of checking stuff the hard way, I get some titanium tweezers and run them along all the TMU's QFP pins that hook to the memory, some of them were indeed loose:

The attachment IMG_4246 (Custom).JPG is no longer available

The attachment IMG_4245 (Custom).JPG is no longer available

This is not kind to the legs of the QFP chip but the method is certainly quick.

Pulled out the microscope again, the day after packing it away once more. Soldered those loose pins back into place and checked that nearby pins were soldered down well which is far easier through the scope. Tested the card again and all games look great now, the speed boost from V1 to V2 is quite impressive.
I had some weirdness trying to get Unreal to display on this card in Glide where it would just show a black screen even though all other games in Windows were fine. For some reason setting the Glide resolution in unreal.ini manually to 640x480 instead of 512x384 made it display. I'm glad that was all it was.
Now this Voodoo 2 works properly I'm planning to put the extra RAM on it soon 😀 Then make an SLI cable

It might sounds weird but a retro build becomes reliable after 1 year, for some reason something will always not behave correctly after 1 month from the finished build, then comes the fixing, 3 more months and you start noticing things that can be improved, so comes the enhancing part, 7 months later you check again for issues and if nothing comes up the build is stable.

I'm having fun

dormcat wrote on 2024-08-21, 03:12:

Could it be that Slot A was very short-lived and was replaced / succeeded by Socket A, while Slot 1 had much more market share a […]
Show full quote

Wikipedia wrote:

Slotkets were never introduced to take advantage of the AMD Athlon processors' transition from the Slot A form factor to the Socket A form factor.

Could it be that Slot A was very short-lived and was replaced / succeeded by Socket A, while Slot 1 had much more market share and coexisted with Socket 370 for two more years?

First Slot A CPU announced: 1999/06/23
Last Slot A and first Socket A CPU announced: 2000/06/05

First Slot 1 CPU announced: 1997/05/07
First Socket 370 CPU announced: 1998/08 (date unspecified)
Last Slot 1 CPU announced: 2000/07/31 (even outlived Slot A)

Yes! I think this is exactly why.

The prevalence of Slot 1 boards today vs Slot A pretty much explains how short lived the platform was.

dormcat wrote on 2024-08-21, 03:12:

The piece of art 🤣 made me think: why was there no slotkets designed for Socket A CPU on Slot A motherboards? Wikipedia has only a short description:

There is one single mention of a Socket A slotket (german production) that I have come across:
https://tweakers.net/nieuws/11711/slot-a-slot … et-adapter.html
According to another article around that time there was a batch of 250 pieces issued.

Ozzuneoj wrote on 2024-08-21, 06:01:

dormcat wrote on 2024-08-21, 03:12:

Could it be that Slot A was very short-lived and was replaced / succeeded by Socket A, while Slot 1 had much more market share a […]
Show full quote

Wikipedia wrote:

Slotkets were never introduced to take advantage of the AMD Athlon processors' transition from the Slot A form factor to the Socket A form factor.

Could it be that Slot A was very short-lived and was replaced / succeeded by Socket A, while Slot 1 had much more market share and coexisted with Socket 370 for two more years?
First Slot A CPU announced: 1999/06/23
Last Slot A and first Socket A CPU announced: 2000/06/05
First Slot 1 CPU announced: 1997/05/07
First Socket 370 CPU announced: 1998/08 (date unspecified)
Last Slot 1 CPU announced: 2000/07/31 (even outlived Slot A)

Yes! I think this is exactly why.
The prevalence of Slot 1 boards today vs Slot A pretty much explains how short lived the platform was.

There seems to be also a technical issue in that Slot A and Socket A are not as compatible as the equivalent Pentium cpu's. You can read about it here (scroll down to may 5th, 2023) https://www.hardwareluxx.de/community/threads … 584473/page-289

An first retro to Gameboy game launch party for Israel indursty in a Tel Aviv mall:
https://youtu.be/nloZGAHQUtw?si=ZLSEBP4xAS9Egyxe
https://youtu.be/xBTk28t0Q-A?si=LsOARHj4t692whif

I finally finished the last piece of another modded case, i left this one unfinished for like 8 months because it was one hell of a problem i created for myself, cutting a perfect hole in a case with a lot of curves? not ideal if you want a form fitting mask for it.
Given the hardware (athlon XP 2800+\audigy\9800 XXL) i went for colors present in the wallpapers of the time for ATI, i just went for a classier look, no branding just a slick design (i might add the ATI logo in red on the side panel)
3D scanning the case helped a lot for the creation of the filtered mask.

Postman5 wrote on 2024-08-21, 05:52:

I'm having fun

Holy carp 🤣. That is some dedication ! At first I thought "what dirty pins" and then I saw what actually was going on hahaha

Postman5 wrote on 2024-08-21, 05:52:

I'm having fun

kudos!

Postman5 wrote on 2024-08-21, 05:52:

I'm having fun

I'm afraid to ask how long that took!

H3nrik V! wrote on 2024-08-21, 04:14:

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

🤣 great work, but is the slot PCB actually a Slot-A or a Slot-1?

They use the same slot - slot A id the exact reverse of slot 1 physically, so they can use the same slots on motherboards.

Postman5 wrote on 2024-08-21, 05:52:

I'm having fun

What kind of black voodoo magic are thou practicing there?

They want to keep making frags:)

debs3759 wrote on 2024-08-21, 15:39:

H3nrik V! wrote on 2024-08-21, 04:14:

Major Jackyl wrote on 2024-08-20, 21:55:

Oh my word. I wasn't up to this, but my brother made this abomination (for shiz and gigglez). I almost died of laughter when I actually looked at it closely. At first glance, nothing unusual, but well, a picture speaks a thousand words, 🤣

🤣 great work, but is the slot PCB actually a Slot-A or a Slot-1?

They use the same slot - slot A id the exact reverse of slot 1 physically, so they can use the same slots on motherboards.

That's kind of ironic 😀