oh2ftu wrote on 2023-08-31, 12:36:

Just repaired a broken Voodoo Banshee AGP. Lot's of funny text in bios which changed when pressing on a ram-chip. Also, said chip was very hot.
Replaced chip after waiting for two weeks for the chip. Seems to be working, ran both glquake and Q2 timedemos, now running 3dmark2000 😀

Ah, I wish my 2x broken banshee cards with RAM faults had such a detectable problem - the chips are all soldered down properly and none get hot so figuring out which of the 8 chips is bad has pretty bad odds. I'm planning to use an ATI Rage pro turbo AGP as a donor since it's got 8x of the same density / voltage / speed QFP SGRAM chips, and it's not as fast/good as a banshee card so the sacrifice is okay imo.

But that's wonderful your card is working again 😀 congratulations, the banshee is imo a pretty cool card being somewhere between a Voodoo 2 and Voodoo 3.

Today i built a AMD 486 DX 40 VLB system with 8mb Ram and 400MB Seagate HDD out of spare parts.

Then i realised the 5 1/4 bays need rails to mount drives into them.... wich i of course dont have.
I instantly lost the interest in completing it and will put it into another corner where it can stay until im not angry anymore...

DerBaum wrote on 2023-08-31, 18:48:

Then i realised the 5 1/4 bays need rails to mount drives into them....

I feel your pain. I have a couple of Antec cases with the same questionable "feature", and it of course also applies to the 3½" bays as well, both internal and external!

A lot of Antec cases use the same rails and may be compatible between models. Good luck if you lose the 2.5" bay cover though!

I have a habit now of putting spare parts in a plastic bag and taping them inside on the bottom of the case. That way I never lose them.

DerBaum wrote on 2023-08-31, 18:48:

Then i realised the 5 1/4 bays need rails to mount drives into them.... wich i of course dont have.

That's really irritating. I've had that happen before and ended up cannibalizing rails from another system.

The other option is to fabricate something out of wood or plastic. Wouldn't need to be too complicated judging by the pics of your case.

Wood is a good idea. You can match the mounting holes by using masking tape to transfer their locations onto the block for drilling.

amigopi wrote on 2023-08-31, 18:52:

aI feel your pain. I have a couple of Antec cases with the same questionable "feature", and it of course also applies to the 3½" bays as well, both internal and external!

I think i got lucky. The 3 1/2" and HDD "caddy" are still here. Only the stupid 8 large rails are missing.
It even has all factory super shiny slot covers except of 3 and all front covers.

Kahenraz wrote on 2023-08-31, 19:11:

I have a habit now of putting spare parts in a plastic bag and taping them inside on the bottom of the case. That way I never lose them.

You can see my spare parts bag in the second picture 😉

Shponglefan wrote on 2023-08-31, 19:14:

The other option is to fabricate something out of wood or plastic. Wouldn't need to be too complicated judging by the pics of your case.

Thats exactly what i will do.
I thought about 3d printing some, but my printer is broken. When i am back at my workshop (its in another city) i will get some material to make them by hand...

Finally got around to archiving the .HLP manual from my first own PC, the glorious black Fujitsu ICL Indiana.

https://archive.org/details/fujitsu-icl-india … eries-pc-manual
(A few screenshots: https://imgur.com/a/SKuJxTW)

Seen this beautiful machine mentioned and posted on here a few times over the years, even some requests for documentation.

PcBytes wrote on 2023-08-31, 06:47:

Well, I'm back to report I've been hit with both the most fampus and the most weirdest to fix bug on that P6VBX7. […]
Show full quote

Well, I'm back to report I've been hit with both the most fampus and the most weirdest to fix bug on that P6VBX7.

Remember how on Virtualbox you'd encounter a race condition due to a bug in the 2k IDE driver?

Well, my board seems to be the rare case where that happens on a real hardware - setup runs well enough, but exactly a few seconds before you'd get to the components installation part (last two progeess bars - one is for that IIS component, the other is the classic component registering and saving settings) it unexpectedly closes, PC reboots and you're back to the first GUI setup screen for 2k.

Anyone has an idea how I'd tackle this 🤣? On Virtualbox I'd just adjust the delay for the IDE driver but not sure how I'd circumvent this issue on a real mobo, aside from using an Epox 3VCM to install 2k first on and then move it to the LT P6VBX7. (since they both use the same VIA drivers, more or less.)

I've been messing with Win2k on a Pentium MMX PC lately, and I ran into similar issues. I managed to complete the install when I disabled DMA on the IDE controller in the BIOS. After install, I disabled DMA on that channel in Device Manager, and then turned it back on in the BIOS.

I wonder if the early Intel IDE controllers that implemented DMA had some bugs.

Torzelan wrote on 2023-08-31, 20:12:

Finally got around to archiving the .HLP manual from my first own PC, the glorious black Fujitsu ICL Indiana. […]
Show full quote

Finally got around to archiving the .HLP manual from my first own PC, the glorious black Fujitsu ICL Indiana.

https://archive.org/details/fujitsu-icl-india … eries-pc-manual
(A few screenshots: https://imgur.com/a/SKuJxTW)

Seen this beautiful machine mentioned and posted on here a few times over the years, even some requests for documentation.

For those who don't know, winhlp32.exe has been removed from Windows since Vista, but it can be reenabled with a patch from Microsoft or with a special .bat file.

See the following KB articles (original and current):

https://web.archive.org/web/20120613220621/ht … t.com/kb/917607
https://support.microsoft.com/en-us/topic/err … e7-1a149da3973b

I still prefer to use some very old .HLP references for programming with older Win32 APIs. They contain plenty of useful information and also preserve documentation that is no longer available from modern Microsoft reference SDKs, such as the older Win16 APIs.

Kahenraz wrote on 2023-08-31, 22:52:

For those who don't know, winhlp32.exe has been removed from Windows since Vista, but it can be reenabled with a patch from Micr […]
Show full quote

Torzelan wrote on 2023-08-31, 20:12:

Finally got around to archiving the .HLP manual from my first own PC, the glorious black Fujitsu ICL Indiana. […]
Show full quote

Finally got around to archiving the .HLP manual from my first own PC, the glorious black Fujitsu ICL Indiana.

https://archive.org/details/fujitsu-icl-india … eries-pc-manual
(A few screenshots: https://imgur.com/a/SKuJxTW)

Seen this beautiful machine mentioned and posted on here a few times over the years, even some requests for documentation.

For those who don't know, winhlp32.exe has been removed from Windows since Vista, but it can be reenabled with a patch from Microsoft or with a special .bat file.

See the following KB articles (original and current):

https://web.archive.org/web/20120613220621/ht … t.com/kb/917607
https://support.microsoft.com/en-us/topic/err … e7-1a149da3973b

I still prefer to use some very old .HLP references for programming with older Win32 APIs. They contain plenty of useful information and also preserve documentation that is no longer available from modern Microsoft reference SDKs, such as the older Win16 APIs.

Does that work for most files for you? I never had luck with the patches in Windows 10. I can see the text, but have had trouble with images. I've actually had better luck using Winhelp in Linux with Wine.

Soldered some smd caps on a motherboard and a cpu.

Some sot23 on a m/b but forgot mosfets' order. 😀

fosterwj03 wrote on 2023-08-31, 22:44:

I've been messing with Win2k on a Pentium MMX PC lately, and I ran into similar issues. I managed to complete the install when I disabled DMA on the IDE controller in the BIOS. After install, I disabled DMA on that channel in Device Manager, and then turned it back on in the BIOS.

I wonder if the early Intel IDE controllers that implemented DMA had some bugs.

I'll be looking into this. In my case it isn't a Intel chipset that has this issue, but rather a VIA 693 chipset w/ 596A southbridge (which to be honest I've never seen in my life - all I knew were 596B and the 686 iterations).

Has the 596A ever been documented in regards to its issues? I know the 686 had the SB Live bug but otherwise not much stuff about the VIA southbridges (and I'm not specifically looking for documentation on those, but general knowledge and bugs).

PcBytes wrote on 2023-08-31, 23:55:

fosterwj03 wrote on 2023-08-31, 22:44:

I've been messing with Win2k on a Pentium MMX PC lately, and I ran into similar issues. I managed to complete the install when I disabled DMA on the IDE controller in the BIOS. After install, I disabled DMA on that channel in Device Manager, and then turned it back on in the BIOS.

I wonder if the early Intel IDE controllers that implemented DMA had some bugs.

I'll be looking into this. In my case it isn't a Intel chipset that has this issue, but rather a VIA 693 chipset w/ 596A southbridge (which to be honest I've never seen in my life - all I knew were 596B and the 686 iterations).

Has the 596A ever been documented in regards to its issues? I know the 686 had the SB Live bug but otherwise not much stuff about the VIA southbridges (and I'm not specifically looking for documentation on those, but general knowledge and bugs).

I think Win2k activates DMA for hard drives by default with the included ATAPI driver. I wonder if Via had their own driver you need to install during the text phase of setup to avoid this problem.

Oh I was busy yesterday. I also swapped a PS/2-port -tower. Low melt alloy ftw 😀

While gutting out some ATX cases from my shed for a colleague, who needed some beige cases, it occured to me, that I never checked my 386sx33 for the risk of Varta Plague.

So, yet another just-in-time save for a nice 386sx board 😀

Tested an Apple multippe scan 720 (17") with my Compaq prolinea with an adapter (the monitor has one of those apple connectors). it worked really well. It can do 85 Hz in 1024x768, I need no more. No noticeable burn in, but the screen is maybe not very bright. My first monitor crt in 20 years. Maybe it is normal. Now to cover up this god awful apple logo...

Installed Windows 7 on iMac 2011 27". Tried to get to windows 10. No cigar... Windows 10 is way better than windows 7 and protected... It wanted a product key. This iMac has a gorgeous lg 2k display with Inky Black's and nice internal speakers, really bright too. Beats my benq ex3210r in picture quality and it's great for media consumption. Best £100 spent on tech. Tried YouTube in win 7, don't know what magic Apple do but the sound is less full and the colours aren't saturated with iMac profile.

Took apart the FMA3500C battery to take a look at the insides. Might be installing a BMC and a Lithium ION Battery (I found out I can get one with the exact same specs as the NiCAD, and it's smalelr in size, so I can probably fit the BMC in there as well).

The next thing I want to do is take apart my FMAK9200D and take gut shots of that laptop and video for #DOSCember video. Trying to make updated profiles for all the machines I have, but I might just stop sticking to #blahblahblah videos after next year because it's kinda' keeping me from just enjoying the darned things and recording some truly candid footage of me gaming on them.

Tried to repair this Asus A7N8X-X motherboard.

Initially the board wouldn't stay powered on. It would power off automatically after a few seconds. Some of the 3300uF caps were bulging, so I removed and tested them. They were all bad.

Replacing them with new caps has now led to the board staying on, but still won't POST. Further testing revealed there is a short. I keep getting a 30 ohm resistance on all the caps I replaced, along with another of other components in and around that area (MOSFETs, inductors, SMD caps, etc.). So I proceeded to undo my recapping to see I'd maybe caused the short.

But even after removing all the replaced caps, the short remains.

Further testing and tracing led me to the underside of the nForce2 400 chip. One 'edge' of those caps shows the 30 ohm short, some of them 15 ohms, and some not shorted.

Now I'm wondering if the issue is the main nForce chip itself, unless I can find something else in the mean time.