Tried my hand a slight overclock with my ABIT NF7 v2.0 + Barton 2500+.

Can report a happy success with 210x11 setting, for a OC of 100MHz over the 3200+ speeds 🤣

They have been known to do 2.4-2.5Ghz without needing more voltage, but Nforce boards can start to top out around 220, some went a lot further but there's a lot of messing around with swapping RAM to get them stable.

Nexxen wrote on 2024-03-31, 21:04:

1st: I had that with a dead bios chip; but also with a dead RTC, and a dead crystal. If it helps at all. 3rd: don't you need to […]
Show full quote

PC@LIVE wrote on 2024-03-31, 19:16:

Today I tried to recover three MB, an ASUS P2B-F, a Compaq Camaro (Mitac), and an ASUS K8N-E Deluxe, with the first I have to do […]
Show full quote

Today I tried to recover three MB, an ASUS P2B-F, a Compaq Camaro (Mitac), and an ASUS K8N-E Deluxe, with the first I have to do a test, to rewrite the BIOS via floppy (recovery BIOS), but if not the only solution is to change the BIOS chip with another (probably new one), if that doesn't work, I don't know what else to check or try.
For the second one, the Mitac I have to try some tracks on the back, maybe one or more are interrupted or damaged, if I find one interrupted, it could be fixed and the card maybe works again.
To finish the K8N-E Deluxe, I tried to install Windows on a 40 GB WD SATA disk, via CD reader, no version installs, tried from 98 to XP and even 7, when writing to the disk the installation stops block.
The only thing 😨 I can do is change the location of the SATA cable, but I don't think it will make any difference.

1st: I had that with a dead bios chip; but also with a dead RTC, and a dead crystal. If it helps at all.
3rd: don't you need to slipstream SATA drivers into the installation ISO image? IFRC sometimes loading from floppy fails. IDE works?

Edit: isn't 2E 2F video card related?

codes also point to

2E Detect and initialise hard drive
2F Detect and initialise math coprocessor

Bios chip rotten too?

Thanks a lot friend, so for the P2B-F, I'll try tomorrow if by chance, we can make some progress via floppy, but I also thought that the BIOS chip might be dead, I hope the RTC and the crystal.
For the Mitac (Camaro), you gave me a very useful suggestion, I'll try to connect an HD ATA, maybe it's looking because it's present in the BIOS, maybe I'll try to reset the BIOS, but I don't think the chip is dead, because it shows some codes, rather it may have been rewritten badly, and it was not possible to reprogram it, because it is soldered!
Finally, the ASUS K8N-E Deluxe, after various hardware and Windows changes, I managed to install Windows XP SP2, and for me it's fine, the PC seems to work well, I've already done some benching, the images are on my Test page and troubleshooting…

BitWrangler wrote on 2024-03-31, 22:47:

They have been known to do 2.4-2.5Ghz without needing more voltage, but Nforce boards can start to top out around 220, some went a lot further but there's a lot of messing around with swapping RAM to get them stable.

Not with the cooler and RAM sticks I use.
I'm using standard Corsair ValueRAM + random Sycron module that can do DDR400, and the cooler is a Spire FalconRock II.
Fairly sure they're not suited for 220FSB - 210 or 215 might likely be their limit.

Kinda same goes cooler-wise with my Soltek SL-75FRN2-RL as well, except that one might have an slight advantage due to using 2x ADATA Vitesta DDR500 and a single PQI Turbo stick.
IIRC the ADATA is Hynix based and the PQI might be Samsung, so there's some chances of a higher OC although I use a mobile 2600+ on that one.

I finally figured out what was wrong with the shitty HP laptop. I have a modern SSD in an IDE converter, and the converter uses a through-hole header for the IDE connector instead of a surface mount header like every 2.5" hard drive ever made. The through-hole pins scraped through the insulator under the drive, resulting in intermittent shorts to the metal under it. I added some more insulation, and while the drive is an extremely tight fit, it doesn't lock up randomly anymore.

Aside from that, I'm having the usual VIA audio "fun." The VxD driver out of 68MU220b.zip (the one that gives legacy DOS compatibility) sounds better than the WDM driver or the newer Vinyl drivers, but it also has compatibility problems with lots of Windows games. There's a DirectSound compatibility setting that needs to be checked for some games and unchecked for others, and if it's in the wrong state, you will have delayed audio, sound effects only, music only, or totally broken noise. Sometimes you'll have sound only with the setting enabled and music only with the setting disabled. There's no setting that works with everything. The WDM drivers work reliably with everything, but every version I've tried sounds wrong. You can get a sample in a Bits und Bolts video (on a completely different motherboard) here (where the Windows startup sound fades out in the middle - mine also sounds scratchy):
https://youtu.be/KfHhJQ1EfJg?t=694

In Windows XP, the driver slipstreamed into the ISO that I used had the same problem with game audio being totally broken, and it made a loud pop sound when a regular Windows sound started playing. I swapped it with the original HP driver, but that driver doesn't support Windows 98.

My Behringer UCA202 works perfectly in Windows 98, though. I don't get DOS compatibility, but I don't need it here.

Archived my Reveal sound card driver collection to Archive org. Nothing impressive, just a collection from mirror sites, ftp and what not collected over the last two years...if interested see my sig, is the latest post.
Will try to upload to to our Vogons Library if all goes as planned...

I'm waiting on an upgraded Gotek and an XTIDE with an SD Card adapter on it, but I've otherwise gotten my XT in to the condition I want it to be in.

Keyboard has been thoroughly cleaned, and the monitor and case have been magic erasered. Only thing I really need to do is figure out how to get all the cards running, and to replace the fan inside the PSU. It's a scratchy squeaky mess.

BetaC wrote on 2024-04-01, 17:56:

I'm waiting on an upgraded Gotek and an XTIDE with an SD Card adapter on it, but I've otherwise gotten my XT in to the condition I want it to be in.

Keyboard has been thoroughly cleaned, and the monitor and case have been magic erasered. Only thing I really need to do is figure out how to get all the cards running, and to replace the fan inside the PSU. It's a scratchy squeaky mess.

Holy smokes that's a nice looking system. Did you have do any work with the mainboard or internals?

I'm looking into OpenGL performance on modern systems. Apparently, Nvidia is the only choice for consistently good/high performance in gaming, due to deficient drivers from AMD and Intel.

Cosmic wrote on 2024-04-01, 18:23:

BetaC wrote on 2024-04-01, 17:56:

I'm waiting on an upgraded Gotek and an XTIDE with an SD Card adapter on it, but I've otherwise gotten my XT in to the condition I want it to be in.

Keyboard has been thoroughly cleaned, and the monitor and case have been magic erasered. Only thing I really need to do is figure out how to get all the cards running, and to replace the fan inside the PSU. It's a scratchy squeaky mess.

Holy smokes that's a nice looking system. Did you have do any work with the mainboard or internals?

Only thing I had to do was some percussive maintenance on the hard drive since it was completely stalled. Otherwise, swapping out the faulty MDA card with an EGA card that can do monochrome/hercules and throwing in my Sound Blaster 2.0 for the sake of it.

Building a HIDman converter. First time attempting SMD soldering, fairly happy with the results so far. Only thing that gave me real grief was the LED. Still a bit crooked, but I tested it and it'll do. Microcontroller is still in the mail (was supposed to arrive today 😡), so I can't finish it yet.

BetaC wrote on 2024-04-01, 17:56:

I'm waiting on an upgraded Gotek and an XTIDE with an SD Card adapter on it, but I've otherwise gotten my XT in to the condition I want it to be in.

Although I believe I got to play with Apple IIs as a child, I've only ever seen these IBM systems in photos. I've always wanted to experience just how slow they really are in person.

As an average human child of average human child fortunes, I did not obtain my own computer until I was gifted a hand-me-down 386 or a 486 (I can't remember which), which is well beyond the power envelope of these early 8088 machines.

Beating my head against the wall...
I decided to put my Soyo 7VMA-B into my Lian Li PC50 case. I mostly completed the move, tested it to POST beep once, then just about had it finished and ready to install Win98. I have a Tualatin 1.26 that was working fine in this motherboard, but suddenly it wouldn't POST. I tried different power supplies, POST card etc., then I finally swapped in a Coppermine 1000 then it posted just fine? When I swapped CPUs I could smell a very faint burning odor. I had noticed that the modded CPU didn't seem to fit quite as deeply into the socket as it ought to, and the board was still dead when I swapped the Tualatin back in. So I grabbed my Tyan S1854 and tried the Tualatin in that, and it POSTed! I also noticed that the CPU seems to fit better in that motherboard's socket for some odd reason. It seems the poor fit in the Soyo board's socket caused poor contact on the power pin(s), and burned them in the socket, but it isn't a problem with the Tyan board (and the Soyo board seems fine with a Coppermine CPU). I'm installing win98 on the Tyan board with the Tualatin while typing this. The very small bummer about this is that the Tyan board doesn't have a front USB port while the Soyo one does.

Maybe it was arcing on the power rail. I also noticed that these interposer Tualatins are not a very deep fit.

It is still possible to replace the socket, if that's all it is.

Kahenraz wrote on 2024-04-02, 02:01:

BetaC wrote on 2024-04-01, 17:56:

I'm waiting on an upgraded Gotek and an XTIDE with an SD Card adapter on it, but I've otherwise gotten my XT in to the condition I want it to be in.

Although I believe I got to play with Apple IIs as a child, I've only ever seen these IBM systems in photos. I've always wanted to experience just how slow they really are in person.

As an average human child of average human child fortunes, I did not obtain my own computer until I was gifted a hand-me-down 386 or a 486 (I can't remember which), which is well beyond the power envelope of these early 8088 machines.

They aren't ZX81 slow, or even VIC-20 unusable. Just, like, "it really is taking a minute to count 640k" slow.

I am allergic to motherboard BIOSes which do not allow me to skip the startup memory check.

Kahenraz wrote on 2024-04-02, 04:21:

I am allergic to motherboard BIOSes which do not allow me to skip the startup memory check.

Don't ever buy a PS/2, then. They're just as bad. Mine also needs to let me know that the only RAM it likes either has been reserved or has lost some capacity. Never actually which one though.

Minutemanqvs wrote on 2024-03-28, 16:33:

I tested the Radeon 9700 I got by surprise the other day, and well...oh you Radeon 9700/9800, always killing yourself!
I tried pressing on the memory modules, the patterns don't change...so it's back to the seller.

It's almost always the GPU chip that's the problem on these. Pressing on the mem chips will do nothing. In fact, these cards use leaded solder, so neither the memory nor the GPU BGA will ever go bad. Rather, it's the shitty undersized coolers that ATI used that cause this issue. The cards run too hot and eventually either the GPU chip dies due to prolonged use at high temperatures -or- the GPU chip silicon develops problems with the BGA to the substrate. The latter is fixable... well, sort of / half the time.

That said, I've had fairly decent "luck" reviving these video cards by giving them a reflow. Only thing is, once fixed, you have to keep them much cooler than with their original stock cooler if you want them to last. For that, the stock cooler has to be replaced with one that can handle 60 Watts of TDP without going over 55C - i.e. something the size of a large Barton-sized CPU heatsink... or just one of the bigger passive coolers made for ATI/AMD cards fitted with a fan on top. I have an abundance of Xbox 360 CPU heatsinks (both v1 and v2 with the copper heatpipe), so I use those. They top out at about 60W TDP.

Let me say that reflow method has stopped having success on those... I have a 9800SE AIW that no matter how much I'd keep it cool would artefact again.

Heck, they're even more fragile than DIA-001 PS3 mainboards.

Ironically, a 6600 that came to me with no detect on POST managed to come back to life... although despite being at the beginning of the bumpgate era, its issue was likely BGA shock - there was a fairly ugly dented cap not far from the GPU itself.

Kahenraz wrote on 2024-04-02, 02:47:

Maybe it was arcing on the power rail. I also noticed that these interposer Tualatins are not a very deep fit.

It is still possible to replace the socket, if that's all it is.

I realized my mistake: with the less than perfect fit of this modified CPU I should have removed the heatsink and made sure that the CPU was fully inserted after handling the motherboard. Without the support of the case, the PCB could flex and the CPU could lose contact. It had been working perfectly for some time, then I moved it from one case to another and then it malfunctioned. I ought to have anticipated this, but it didn't cross my mind.