i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

zuldan wrote on 2026-07-05, 22:47:

Ouch! What make/model PSU is that?

It's a Codegen "520W" model - rated with 42 amps on 5V (the real limit would be 30 amps, as that's the rating of the 5V rectifier diode). A not so great PSU. However I don't want to go through the efforts of checking ancient CPUs, when going to 12V for the CPU is possible. I halt my overclocking until the 12V mod is done. No need to risk anything.

Living wrote on 2026-07-06, 02:03:

i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

Well.. fast Socket A is a wild range. If you overclock in extreme ranges, the CPU consumes just a hecking lot of power. Here's 2V VCore + SuperPi 1M on an A7N8X - this is the 5V rail:

We're having 4.86V at the ATX connector with a current draw of 34A - which results in 165,24W of power going into that board

Ozzuneoj wrote on 2026-07-05, 22:11:

I doubt I would ever do anything like that, but I find the idea totally fascinating. I'd love to see pictures and read about how you do it.

How thoroughly have setups like this been tested? Are there any long-term negatives to doing this aside from the usual stress of overclocking on the CPU?

I wouldn't see any serious issues with that. They've been proven during overclock to run with 12V for the VRM just fine.

STM specifically states that L6917B (used both on the A7N8X and K7S8X) is capable to run with 12V just fine. You'll get a drop in efficiency though.

Living wrote on 2026-07-06, 02:03:

i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

Hi,

You might not need to abandon that idea just based on that, while most Socket A motherboards use 5 Volts to derive the CPU and other voltages there are some, especially later boards that use the 12 volts rail for that. Unfortunately most motherboards doesn't have: "I use 12V" printed on it (or even in the manual), but there are various online lists that you can check for example:
https://www.vogonswiki.com/index.php/List_of_ … _A_motherboards
Note: That unlike what the name suggests, that site is not officially affiliated or run by vogons.org. Also as it is a Wiki, while it requires registration which suggest that the data is more reliable than for a Wiki not needing it, it might still be inaccurate.

MagefromAntares wrote on 2026-07-06, 05:41:

Hi, […]
Show full quote

Living wrote on 2026-07-06, 02:03:

i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

Hi,

You might not need to abandon that idea just based on that, while most Socket A motherboards use 5 Volts to derive the CPU and other voltages there are some, especially later boards that use the 12 volts rail for that. Unfortunately most motherboards doesn't have: "I use 12V" printed on it (or even in the manual), but there are various online lists that you can check for example:
https://www.vogonswiki.com/index.php/List_of_ … _A_motherboards
Note: That unlike what the name suggests, that site is not officially affiliated or run by vogons.org. Also as it is a Wiki, while it requires registration which suggest that the data is more reliable than for a Wiki not needing it, it might still be inaccurate.

Oh, nice! I guess I'd forgotten that my Abit NF7-S had a 12v VRM... it's been a while since I've heard much about that.

Just so I'm clear on this... Does that mean that a high powered (or overclocked) CPU in a board like that doesn't increase the +5v rail requirement at all and instead puts the entire CPU load on the +12v rail of the PSU?

Along those same lines, does a Socket A board using a +12v VRM change anything with regard to powering AGP or PCI cards that don't have auxiliary power connectors? Or is this only relating to CPU load?

Ozzuneoj wrote on 2026-07-06, 06:04:

Oh, nice! I guess I'd forgotten that my Abit NF7-S had a 12v VRM... it's been a while since I've heard much about that. […]
Show full quote

MagefromAntares wrote on 2026-07-06, 05:41:

Hi, […]
Show full quote

Living wrote on 2026-07-06, 02:03:

i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

Hi,

You might not need to abandon that idea just based on that, while most Socket A motherboards use 5 Volts to derive the CPU and other voltages there are some, especially later boards that use the 12 volts rail for that. Unfortunately most motherboards doesn't have: "I use 12V" printed on it (or even in the manual), but there are various online lists that you can check for example:
https://www.vogonswiki.com/index.php/List_of_ … _A_motherboards
Note: That unlike what the name suggests, that site is not officially affiliated or run by vogons.org. Also as it is a Wiki, while it requires registration which suggest that the data is more reliable than for a Wiki not needing it, it might still be inaccurate.

Oh, nice! I guess I'd forgotten that my Abit NF7-S had a 12v VRM... it's been a while since I've heard much about that.

Just so I'm clear on this... Does that mean that a high powered (or overclocked) CPU in a board like that doesn't increase the +5v rail requirement at all and instead puts the entire CPU load on the +12v rail of the PSU?

Along those same lines, does a Socket A board using a +12v VRM change anything with regard to powering AGP or PCI cards that don't have auxiliary power connectors? Or is this only relating to CPU load?

I think these lists are about the primary VRM of the motherboard, so it suggest that the whole CPU power comes from the 12V regulator, although this doesn't preclude the possibility of having a secondary VRM powering the cards or various chips from a different rail. According to the front page of vogonswiki.com, even if that page is not officially affiliated with vogons.org many of the contributors of that Wiki are also members of this forum, so someone who helped put together that list might read this and can confirm if this page were put together with the possibility of a secondary VRM present or just considered the power requirements of the CPU.

Ozzuneoj wrote on 2026-07-06, 06:04:

Oh, nice! I guess I'd forgotten that my Abit NF7-S had a 12v VRM... it's been a while since I've heard much about that. […]
Show full quote

MagefromAntares wrote on 2026-07-06, 05:41:

Hi, […]
Show full quote

Living wrote on 2026-07-06, 02:03:

i had to abandon the idea of a fast Socket A computer because of this. The pull from the 5v rail is insane and there is no modern power supply below u$s 150 that can sustain that (i refuse to use 20years+ PSU's)

Hi,

You might not need to abandon that idea just based on that, while most Socket A motherboards use 5 Volts to derive the CPU and other voltages there are some, especially later boards that use the 12 volts rail for that. Unfortunately most motherboards doesn't have: "I use 12V" printed on it (or even in the manual), but there are various online lists that you can check for example:
https://www.vogonswiki.com/index.php/List_of_ … _A_motherboards
Note: That unlike what the name suggests, that site is not officially affiliated or run by vogons.org. Also as it is a Wiki, while it requires registration which suggest that the data is more reliable than for a Wiki not needing it, it might still be inaccurate.

Oh, nice! I guess I'd forgotten that my Abit NF7-S had a 12v VRM... it's been a while since I've heard much about that.

Just so I'm clear on this... Does that mean that a high powered (or overclocked) CPU in a board like that doesn't increase the +5v rail requirement at all and instead puts the entire CPU load on the +12v rail of the PSU?

Along those same lines, does a Socket A board using a +12v VRM change anything with regard to powering AGP or PCI cards that don't have auxiliary power connectors? Or is this only relating to CPU load?

Yes. The CPU (only) is powered by the 12V rail - other components might be, but usually they're powered from 3.3V and 5V. This lifts a heavy load off the 5V rail. It only affects the power of the CPU. Naturally though, the 5V rail of all slots and the CPU power are shared, meaning the voltage drop to and in the board will be significantly lower if the CPU is not using 5V anymore. That would leave more headroom (less voltage drop on the 5V rail) for the expansion slots. I'm not sure if that really makes a difference in practice though.

In normal operation (not extreme overclocking) I'd say it's not a big deal.

VanillaFairy wrote on 2026-07-04, 23:33:

Haven't tried booting it as a floppy since I'm getting tired and I'd rather wait just for the IDE DVD-RW drive I have on the way to arrive, but I tried setting it to 500MiB and it still won't boot...
To be fair I wouldn't be surprised if part of the issue was me using a USB keyboard, but I don't have any PS/2 keyboards to try with anyway.

Hooo boy this disk drive might also be dead.

it at least detects there is a disk in it, and spins the disk too even, but it can't seem to read any of the disks I've burnt; it waits a minute or so and then repeats the same "DISK BOOT FAILURE, INSERT SYSTEM DISK AND PRESS ENTER" error.
I guess the laser could be dirty, or maybe it hates the Phillips DVD-Rs I got, or maybe something else is broken...
(For reference, it is a LG GH22LP20 that I got)

Do you have a way to extract the Vista CD onto the hard drive? Then you could make the drive bootable with bootsect.exe and run the installer from there, without the need to boot from CD or USB.

Could you explain your setup a bit more? I'm not sure if I understood it right, you have a retro PC with IDE only, and it has a SATA SSD connected over an adapter? And the only other PC is a modern one that has no SATA/IDE at all? If so, how did you get anything at all onto the SSD?

asdf53 wrote on 2026-07-06, 15:52:

Do you have a way to extract the Vista CD onto the hard drive? Then you could make the drive bootable with bootsect.exe and run the installer from there, without the need to boot from CD or USB.

I do have the ISO already, and even got it booting off a CF card but then it complained about not detecting any system volume information. when the installer was the thing that formatted the SSD.
Also can't get the Plop boot manager booting from the DVD-Rs either. currently the only thing I have been able to boot is when I put an already-installed Alpine Linux onto the SSD. (and I guess also Ventoy but I couldn't get anywhere from there)

Could you explain your setup a bit more? I'm not sure if I understood it right, you have a retro PC with IDE only, and it has a SATA SSD connected over an adapter? And the only other PC is a modern one that has no SATA/IDE at all? If so, how did you get anything at all onto the SSD?

I have a retro PC with IDE only, it does have USB ports but cannot boot from them, and the SATA SSD is connected with an adaptor.
The other PC is my daily driver with no IDE, it does have SATA ports (one of them is connected to a big hdd I impulse-bought, the other's empty) but I just used a USB-to-SATA adaptor to put files onto it. (and previously a USB to CFcard adaptor when I was trying to use CF cards.)

Then you could connect the SSD to your main PC and follow this guide to copy the Vista installation files to it and make it bootable. When done, put it back into your retro PC and boot from it to start the installation.

While waiting for parts for the 12V conversion, I started to write a smbus kernel driver for windows XP so I can directly access anything it that I want. For nforce2 and 3.

asdf53 wrote on 2026-07-06, 16:13:

Then you could connect the SSD to your main PC and follow this guide to copy the Vista installation files to it and make it bootable. When done, put it back into your retro PC and boot from it to start the installation.

this worked - thank you so much!
(unfortunately, it seems my GPU that I got was a different one, it's showing up as an FX 5700LE in Vista instead of the GTS 7600 that I thought I was getting, even though it was in the 7600's box...... I think I got scammed probably. 😒 )

tehsiggi wrote on 2026-07-06, 04:52:

STM specifically states that L6917B (used both on the A7N8X and K7S8X) is capable to run with 12V just fine. You'll get a drop in efficiency though.

The attachment Screenshot from 2026-07-06 06-48-56.png is no longer available

How's the VRM cooling on the board? 5% decrease in efficiency would equal something like 7.5W extra power dissipation in that. That could increase the temperature by quite a bit.

[Edit] no idea why my quote includes that image and not the screenshot? I don't really get how that works

H3nrik V! wrote on Yesterday, 07:17:

tehsiggi wrote on 2026-07-06, 04:52:

STM specifically states that L6917B (used both on the A7N8X and K7S8X) is capable to run with 12V just fine. You'll get a drop in efficiency though.

The attachment Screenshot from 2026-07-06 06-48-56.png is no longer available

How's the VRM cooling on the board? 5% decrease in efficiency would equal something like 7.5W extra power dissipation in that. That could increase the temperature by quite a bit.

[Edit] no idea why my quote includes that image and not the screenshot? I don't really get how that works

The only heatsink they have is the PCB itself. Granted - the power consumption in the picture is for the whole board, that at least includes the memory. Still though, the point about the efficiency is completely right. The FETs are having a rough time:

This is with 2V VCore, which is pretty extreme in many regards. However it has to be said that a 120mm fan was constantly blowing air over them.

In normal operation though, I assume it would be a bit better. I'll still arrange a cooling solution for those FETs (and the PCB backside). What a joy 😁

a7n8x_2v_vrm_tmp.jpeg

This is with 2V VCore, which is pretty extreme in many regards. However it has to be said that a 120mm fan was constantly blowing air over them.

In normal operation though, I assume it would be a bit better. I'll still arrange a cooling solution for those FETs (and the PCB backside). What a joy 😁

That is pretty hot... I have to put some passive heatsinks at least....

SlowA wrote on Yesterday, 09:59:

a7n8x_2v_vrm_tmp.jpeg

This is with 2V VCore, which is pretty extreme in many regards. However it has to be said that a 120mm fan was constantly blowing air over them.

In normal operation though, I assume it would be a bit better. I'll still arrange a cooling solution for those FETs (and the PCB backside). What a joy 😁

That is pretty hot... I have to put some passive heatsinks at least....

Well, to be fair to the board: Under normal use with an Athlon XP2800 without any overclocking (and no additional airflow over the FETs even) things are way more calm:

tehsiggi wrote on Yesterday, 07:32:

In normal operation though, I assume it would be a bit better. I'll still arrange a cooling solution for those FETs (and the PCB backside). What a joy 😁

Sounds like the sensible thing to do 🤣

Part two of "abusing the smbus connector on the ASUS A7N8X".
I added read and write functions to my driver to allow for block and word reads and writes, as some parts have 16 bit data per register.

I then connected an INA219 breakout board I've got laying around to the smbus connector on the board. Fed 5V into it from USB as the "rail" voltage, so I actually got something to measure.

The INA219 uses the default address 0x40 on this breakout board and I can read it. Added some init routines and a custom command to read the data from the INA219 in my little commandline program. Works like a charm.

It'll get some use eventually, albeit in a different form.

Got Win95 setup and running nicely under PCem.

Some months ago I wanted to run older Win9x built on top of DOS for some reason... so I dug out my original Win95 and Win98 install CDs

98 installed without much trouble, but I couldn't get 95 CD to boot to the instasller.
As 98 was "good enough" for whatever I needed - I just put 95 away again and didn't work at it much more.

This morning while doing something else I came across my Win95 CD and my addled brain fired up with an old memory - I recalled than Win95 CD didn't boot and that setup had to be run under DOS.

So I fired up PCem, made a minimal install of MS-DOS 6.22, then mounted the Win95 CD and ran "SETUP" from it.
- An lo and behold - Win95 installed OK!

Managed to get audio working on my retro system; TheRetroWeb didn't list any drivers, but did mention that the motherboard's sound chip was the Yamaha YMF753. from there, a quick search lead me to the Internet Archive where that very chip had drivers for XP and, lo and behold, they somehow still worked on Vista too!

I found the drivers at https://archive.org/details/yamaha-semiconduc … -website-backup, I feel it might be worth putting these on TheRetroWeb although I'm not sure how to.

I guess now all that's left is to track down a real 7600 GS or GT, since unfortunately I'm stuck with an FX 5700 LE instead. still enough to mess around and have some fun with, but it's below min-spec for Spore (it does run somehow though, albeit incredibly slowly and it lags even on the main menu) and I was hoping to be able to mess around with Spore on more period-accurate hardware.
...maybe once my current savings accounts mature I'll seek out one-