Yeah, all the 32X issues are because of those ribbons... I did the nuclear option in mine and got rid of them not so long ago. It was a pain in the rear but there will not be any more of that pain again 🤣.
EDIT: of course the image order is reversed 🤣

Tiido wrote on 2024-11-27, 14:24:

Yeah, all the 32X issues are because of those ribbons... I did the nuclear option in mine and got rid of them not so long ago. It was a pain in the rear but there will not be any more of that pain again 🤣.
EDIT: of course the image order is reversed 🤣

I like it! I don't think I'm going to attempt it, but I like it 🤣.

An easy fix 😁.

Tested my ASUS P5A-B today after the PS/2 to DIN adapter arrived. Worked out great, no issues at all, no leaking caps, nothing ... out of the box from the bay. Great find.

TheChexWarrior wrote on 2024-11-23, 20:50:

2 activities: Bought Zire Palm PocketPC and Retroid 4 Pro.

The second bought this huge Lego Inovation System set and its epic. With the CDs, wow how much '98 was awesome in tech and gaming.

I have a update with the Creatures Extreme Pack and CD.

Yesterday, i have another pass at try to repair the AIT4 drive that i got on local market.
I have two of them, one will not power on and one does power on but won't accept tape.

With the one that power on, I opened it and check the behavior, in the past, I observed that it load tape but the head drum is constantly spin up then down.
I thought that it due to the head dirty but when i take a closer look, it happen when the drive engage the pinch roller to the capstan motor and there are no movement.
When i tried to manually spin the capstan , it seems to stuck. I tried with the other drive and it spin freely. So i have a drive that have defect mainboard but good mechanical and a drive that have working mainboard but defect mechanical.
So i do the most reasonable thing we should do: i take the working motherboard out and installed it on to the other drive. And it working perfectly! It even pass the read/write test.

So now i won't feel that i wastesd my money when i tried to source some AIT tape.

Then today, i take my external SDT-9000 DAT drive out for a yearly exercise and it greet me with the usual Sony quirk:
- dirty contact of detection switch -> clean it by dripping some IPA into the sw and work it out.
- drive suddenly decide to eat my tape -> had to open the enclosure and take out the drive then manually eject the tape and clean everything up.
- pinch roller harden -> replace it with the spare
- broken tape -> luckily, it happen on the beginning so i keep the clear transparent part and splice the remaining tape to it, lost around 1 minute of song but it no big deal.
now the drive is happy playing audio tape

I started getting graphical glitches on an Athlon/WinXP system with a x1950 PRO AGP, opened up the case to have a look and saw an orange LED on the back of the card that I hadn't ever seen before.

This dual-Athlon system if anyone is interested... (link to vogons post)

Assuming it was a kind of warning light, and being quite sleepy, I absently and very gently touched it with the tip of my finger.... and could hear my skin sizzling as I moved my hand away!

Yeah turns out that was a SMD cap absolutely destroying itself.

Waiting for it to cool down first, I removed the cap with tweezers (it was basically attached by ashes-- no soldering iron required), cleaned the area, and the card now works perfectly fine.

I suppose I will take the risk on EMI regs, and now I have a neat little diamond burned out of my fingerprint...

Adding a PS/2 mouse port on a Micronics JX30 (09-00183-xx) (Gateway 2000 4DX2-66V MB).

By default not supported but provisioned on the MB. Added a header, some "dremel time" to resolve clearance issue with the KBD connector, populate W7. And final but not least, the BIOS from this post:

Re: Larger Hard Drives on a Gateway 2000 4DX2-66V

Cyfrifiadur wrote on 2024-11-30, 10:41:

I started getting graphical glitches on an Athlon/WinXP system with a x1950 PRO AGP, opened up the case to have a look and saw a […]
Show full quote

I started getting graphical glitches on an Athlon/WinXP system with a x1950 PRO AGP, opened up the case to have a look and saw an orange LED on the back of the card that I hadn't ever seen before.

This dual-Athlon system if anyone is interested... (link to vogons post)

Assuming it was a kind of warning light, and being quite sleepy, I absently and very gently touched it with the tip of my finger.... and could hear my skin sizzling as I moved my hand away!

Yeah turns out that was a SMD cap absolutely destroying itself.

Waiting for it to cool down first, I removed the cap with tweezers (it was basically attached by ashes-- no soldering iron required), cleaned the area, and the card now works perfectly fine.

I suppose I will take the risk on EMI regs, and now I have a neat little diamond burned out of my fingerprint...

I love that dual amd build!
re cap ... are you gonna replace it or just leave that as is ?

Well, I've made an interesting discovery today while testing and refurbing an old Radeon 9700 Pro that I picked up in excellent condition.

I gave it a quick test in my Abit NF7-S 2.0 (Nforce 2 Ultra + Athlon XP) system. After a minute or so the PCB behind the GPU die was definitely hotter than the heatsink, so I knew the contact was pretty poor. I shut it down, but there weren't any artifacts up to that point (in Windows XP, but without drivers installed yet), so I knew it was at least worth improving the cooling. When I took the cooler offer, I saw that someone had already been in here, and that the stock thermal putty\crust (whetever it was...) on the heatsink had already been carved away in a square around the die and there was a tiny dot of what looked like Arctic Silver in the center... obviously making terrible contact with the cooler. There also still seemed to be a bit of a "rim" of old thermal crud around the die, so I carefully scraped that down to level it out.

To avoid having to remove the notorious big shim around the die (sounds like a nightmare), I sandwiched a thin 0.3mm copper shim between the die and the heatsink with thermal paste. I also installed some new clips since they were a bit less fiddly than the originals. During testing the heatsink was basically the same temp as the back of the PCB so the copper shim definitely did the job.

However, after I installed drivers the system would just lose display signal when loading Windows. BUMMER. I tried a few things and basically was ready to call it dead. Then, I read some posts online about this happening all the time back in the day... especially on higher-end Athlon XP systems, even some speficially mentioning the NF7-S 2.0! In some cases a different power supply fixed it. Rather than mess with all that, I just dropped it into another system I have, running an EpoX 9NDA3J (Nforce 3 Ultra) and an Athlon 64 X2 4200+. The card seems to be working perfectly fine on the second system! I haven't tested and heavy 3D applications yet, but it is working with whatever drivers the system found to install by default, and I ran the brief Directx 9.0C dxdiag 3D acceleration tests with no issues at all.

So, if you've got a 9700 Pro that seems dead, definitely try it in another system. I guess these things were quite finicky back in the day. It is either due to a high demand on the voltage rails, chipset\BIOS quirks or possibly even drivers. If I am able to narrow it down any further, I will post back here again. I will say, both systems are running the same Seasonic 550HT 550W 80Plus White with 30A on the +5v rail and multiple 12v rails, so I'm not sure if the power supply would be to blame. If it is, then maybe +5v rail is the culprit. The board is running a 1700+ Thoroughbred B core at 2Ghz, so it shouldn't be too power hungry, but the A64 is probably more balanced toward 12v.

I thought for sure the thing was dead, so I am quite pleased. 😀

Ozzuneoj wrote on 2024-12-01, 01:02:

Well, I've made an interesting discovery today while testing and refurbing an old Radeon 9700 Pro that I picked up in excellent […]
Show full quote

Well, I've made an interesting discovery today while testing and refurbing an old Radeon 9700 Pro that I picked up in excellent condition.

I gave it a quick test in my Abit NF7-S 2.0 (Nforce 2 Ultra + Athlon XP) system. After a minute or so the PCB behind the GPU die was definitely hotter than the heatsink, so I knew the contact was pretty poor. I shut it down, but there weren't any artifacts up to that point (in Windows XP, but without drivers installed yet), so I knew it was at least worth improving the cooling. When I took the cooler offer, I saw that someone had already been in here, and that the stock thermal putty\crust (whetever it was...) on the heatsink had already been carved away in a square around the die and there was a tiny dot of what looked like Arctic Silver in the center... obviously making terrible contact with the cooler. There also still seemed to be a bit of a "rim" of old thermal crud around the die, so I carefully scraped that down to level it out.

To avoid having to remove the notorious big shim around the die (sounds like a nightmare), I sandwiched a thin 0.3mm copper shim between the die and the heatsink with thermal paste. I also installed some new clips since they were a bit less fiddly than the originals. During testing the heatsink was basically the same temp as the back of the PCB so the copper shim definitely did the job.

However, after I installed drivers the system would just lose display signal when loading Windows. BUMMER. I tried a few things and basically was ready to call it dead. Then, I read some posts online about this happening all the time back in the day... especially on higher-end Athlon XP systems, even some speficially mentioning the NF7-S 2.0! In some cases a different power supply fixed it. Rather than mess with all that, I just dropped it into another system I have, running an EpoX 9NDA3J (Nforce 3 Ultra) and an Athlon 64 X2 4200+. The card seems to be working perfectly fine on the second system! I haven't tested and heavy 3D applications yet, but it is working with whatever drivers the system found to install by default, and I ran the brief Directx 9.0C dxdiag 3D acceleration tests with no issues at all.

So, if you've got a 9700 Pro that seems dead, definitely try it in another system. I guess these things were quite finicky back in the day. It is either due to a high demand on the voltage rails, chipset\BIOS quirks or possibly even drivers. If I am able to narrow it down any further, I will post back here again. I will say, both systems are running the same Seasonic 550HT 550W 80Plus White with 30A on the +5v rail and multiple 12v rails, so I'm not sure if the power supply would be to blame. If it is, then maybe +5v rail is the culprit. The board is running a 1700+ Thoroughbred B core at 2Ghz, so it shouldn't be too power hungry, but the A64 is probably more balanced toward 12v.

I thought for sure the thing was dead, so I am quite pleased. 😀

Thanks for sharing this ! I'm building Socket-A system here and acquired Radeon 8500DV, which probably doesn't have this issue, but we will see 😀

CMB75 wrote on 2024-11-28, 13:56:

Tested my ASUS P5A-B today after the PS/2 to DIN adapter arrived. Worked out great, no issues at all, no leaking caps, nothing ... out of the box from the bay. Great find.

That's great news! It feels like finding gold when an eBay board just works, especially SS7 and earlier parts. Do you have any special plans for the board? I've wanted one for a while - I've played with a couple MVP3 boards but not yet Aladdin V.

SpoXi wrote on 2024-11-30, 17:04:

Adding a PS/2 mouse port on a Micronics JX30 (09-00183-xx) (Gateway 2000 4DX2-66V MB).

By default not supported but provisioned on the MB. Added a header, some "dremel time" to resolve clearance issue with the KBD connector, populate W7. And final but not least, the BIOS from this post:

Re: Larger Hard Drives on a Gateway 2000 4DX2-66V

Nice work. Did you have to do any other modifications other than getting the connector physically attached? PS/2 on a 486-class board is very nice yet uncommon. I have a KVM setup that is all PS/2 based, so my 486 is the outlier and requires a separate serial mouse.

Cosmic wrote on 2024-12-01, 19:59:

Nice work. Did you have to do any other modifications other than getting the connector physically attached? PS/2 on a 486-class board is very nice yet uncommon. I have a KVM setup that is all PS/2 based, so my 486 is the outlier and requires a separate serial mouse.

Changed W7 which according to the manual from "The Micro House Encyclopedia of Main Boards" states "Factory configured - do not alter" and goes to the 8242 on the MB. The hint was from another manual (same source) that concerns JX30G (or JX30GP which I was interested in, since it goes with the PS/2 port preinstalled from factory), there the same jumper marked as W3 is described as "On board mouse enabled (IRQ12)".

I feel your pain. In general you should look around, for a decent socket 2 or 3 that you can find confirmation for a working PS/2 port or header. Have two other boards with SIS that have the PS/2 headers working out of the box (Soyo sy-4saw and Zida 4dps).

Planing to do a mod of the HOT433. No luck with the provisioned pin header (after bios change), even though the chipset is the later "BF" revision:

Success: Adding a PS/2 mouse port to the Shuttle HOT433 (using a dedicated keyboard controller chip)

I was testing a board, socket 939.
I was having trouble going past some POST codes and BIOS flashing issues.

PSU gave in taking the board to the realm of dead tech. Now it insta powers on. Tried a few PSUs but it's the same.
I have no interest in troubleshooting it as it was a low end board. CPU isn't getting Vcore, I know where the issue is and I'd have to order stuff and wait a lot of days for nothing.

RIP Asus A8N5X

Edit: everything else survived.

It happens. I've literally had motherboards go up in smoke in front of me during testing through no fault of my own.

When you work with enough tech, especially old tech, it's bound to happen through odds alone.

Cosmic wrote on 2024-12-01, 19:56:

CMB75 wrote on 2024-11-28, 13:56:

Tested my ASUS P5A-B today after the PS/2 to DIN adapter arrived. Worked out great, no issues at all, no leaking caps, nothing ... out of the box from the bay. Great find.

That's great news! It feels like finding gold when an eBay board just works, especially SS7 and earlier parts. Do you have any special plans for the board? I've wanted one for a while - I've played with a couple MVP3 boards but not yet Aladdin V.

That's true, sometimes we do get lucky. I want to reduce my collection and limit myself to a couple of working systems. This board is one of the competitors for a pure DOS build.
The chipset has its flaws but is overall quite good. It is advisable though to look out for differences between the revisions especially if considering AMD K6-2+/III+ systems and a higher RAM capacity.

Scored a bunch of old CPU on Ebay, but the FC-PGA Adapter was what really could my attention. Maybe the Celeron 533A is also fun to play with. I read somewhere that Intel forget to make some of them PGA incompatible.

/MZ wrote on 2024-12-02, 08:44:

Maybe the Celeron 533A is also fun to play with. I read somewhere that Intel forget to make some of them PGA incompatible.

It might depend on the stepping.

I have a Coppermine Celeron 600 which is officially supported by my Abit ZM6 and works just fine. It's of the cB0 stepping, and I specifically bought it because of that. Some old forum posts on Anandtech indicated that only the Celeron 600 variants of that stepping will work in that board.

Kahenraz wrote on 2024-12-02, 00:55:

It happens. I've literally had motherboards go up in smoke in front of me during testing through no fault of my own.

When you work with enough tech, especially old tech, it's bound to happen through odds alone.

Yeah, it happens but it still sucks... That's our life as collectors. 🤣
I have far better boards but I wanted to pass it on as I didn't need it. PSU died and probably killed the Super I/o or some other component in the POWER GOOD line.
That's hours of work/troubleshooting for nothing. To the spares bin it goes unless anybody wants it for repairs (free to grab).