VOGONS


Reply 40 of 46, by rasz_pl

User metadata
Rank l33t
Rank
l33t

No fan monitoring chip on this board = cant wait for tach unless there is something discrete implemented, but this is low end OEM board so unlikely.

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 41 of 46, by Shponglefan

User metadata
Rank l33t
Rank
l33t

I was curious about the grounded tach, so I did some probing.

I first probed the SYS_FAN header (the one with the non-grounded tach). It connects to a resistor just below the Southbridge chip, and from there connects through a via to a resistor array just below the AS97127F chip. A pair of vias are going in to that resistor array, which in turn is connected to two pins on the AS97127F chip.

I don't know if this means the PR_FAN header is supposed to be connected to that array from one of those vias. I tried probing around a bit but couldn't find anything else that array is directly connected to.

I'm also not sure if that tach pin on the PR_FAN header is intentionally grounded. If I had a comparable motherboard for comparison, I could test to see if there is any difference between them. But unfortunately I do not.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 42 of 46, by rasz_pl

User metadata
Rank l33t
Rank
l33t

oh, so AS97127F does have fan monitoring after all? WTF is this ASUS weird world 😁 Which pins did you trace it to? Does it have Fan speed readout in bios? and/or in Windows diag programs?
Im looking at P3B-F diagram and its AS98127F does monitor Fans on pins 90 91 92. Diagram fragments in original size 😐
ASUS doesnt call those pins speed anything, resolution is terrible, but to me it looks more like FSTP1 FSTP2 FSTP3 as in some crude Fan stop detection?
EDIT: actually diagram with AS98127F matches AS99127F datasheet pinout! so it looks like AS97127F is the odd one out

https://github.com/raszpl/FIC-486-GAC-2-Cache-Module for AT&T Globalyst
https://github.com/raszpl/386RC-16 memory board
https://github.com/raszpl/440BX Reference Design adapted to Kicad
https://github.com/raszpl/Zenith_ZBIOS MFM-300 Monitor

Reply 43 of 46, by momaka

User metadata
Rank Oldbie
Rank
Oldbie

Hehe, nice to know you got the board working again (like it always was?) 😁
Perhaps you have the magic touch? 🤣
When I worked in a repair shop some years back, I remember this happening a few times where owner/boss gave me some boards to look at, telling me they didn't work at all... and when I did look at them, miraculously, they all worked just fine. So he started joking every time when we get something in for repair, that I should just go touch it and it would magically be fixed afterwards.

Anyways, back to the topic at hand... I suppose one easy way to know if the board really has tacho monitoring and protection on CPU fan RPM failure is to just disconnect the CPU fan prior to powering on the board and see if it halts again. If not, then the fluke must have been somewhere else.

Repo Man11 wrote on 2024-11-16, 20:26:

There was a time when I smelled a burning electronics smell coming from the PC I had built for my mother. It was working fine, and I couldn't isolate the source of the smell. Maybe six to nine months after I had visited her and first noticed it, her PC stopped powering on. It turned out that the mediocre power supply had a loose fitting power cord, and the resistance eventually cooked the power cord's contact. Fortunately, I had come prepared with a spare power supply just in case. I later used a dead power supply to scavenge a socket, and used it for repair and the power supply worked fine.

Interesting.
I suppose a cheap & nasty power cord with loose contacts could indeed do that.
I have a few on hand like there... and always wondered if they would end up doing that. Probably should get rid of them.
BTW, I've also seen the 1/0 (On-Off) power switches on cheap PSUs end up bad and not make good contact... so that's another component that could do this.

Reply 44 of 46, by Shponglefan

User metadata
Rank l33t
Rank
l33t
momaka wrote on 2024-11-19, 21:28:

Hehe, nice to know you got the board working again (like it always was?) 😁
Perhaps you have the magic touch? 🤣

I wish I had the magic touch. 😅

I'm currently 4 for 6 in motherboards I've tried to repair recently. Though I just started another one, so we'll see how that goes.

Anyways, back to the topic at hand... I suppose one easy way to know if the board really has tacho monitoring and protection on CPU fan RPM failure is to just disconnect the CPU fan prior to powering on the board and see if it halts again. If not, then the fluke must have been somewhere else.

I could try that, though from what the OP posted, it sounds like they weren't using the CPU fan header to begin with.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 45 of 46, by Shponglefan

User metadata
Rank l33t
Rank
l33t
rasz_pl wrote on 2024-11-17, 05:14:

oh, so AS97127F does have fan monitoring after all? WTF is this ASUS weird world 😁 Which pins did you trace it to? Does it have Fan speed readout in bios? and/or in Windows diag programs?
Im looking at P3B-F diagram and its AS98127F does monitor Fans on pins 90 91 92. Diagram fragments in original size 😐

I'll take another look at the board and try to recreate my testing and take some pics this time. From what I recall, I don't think it was those pins though.

The BIOS doesn't have any hardware monitoring. Haven't tried the board with Windows yet, just under DOS using my test bench install.

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards

Reply 46 of 46, by Shponglefan

User metadata
Rank l33t
Rank
l33t

Btw, there is one more minor defect on the board I'll probably repair. The Primary IDE socket has some broken plastic on the edge.

It's purely cosmetic, but I do like to fix those things. If nothing else, it gives me one thing to legitimately repair on this board. 😁

Pentium 4 Multi-OS Build
486 DX4-100 with 6 sound cards
486 DX-33 with 5 sound cards