Sorry for the necro-posing, but I am a Necromancer. 😀 Sad there are no more post how it went with this particular board. I can confirm the problem might have been in loose pins on the QFP chips. I had dead board of the same type and one chip (SIS 503) had almost left the board. The other one had some bad pins as well.
However I ma still struggling a bit. The board runs only if the L2 cache is set to WT mode. If I set WB in bios, it always get stuck on post screen right before booting.
Pretty weird issue...

Solved - I have flashed older bios from 1995 (the last one is 1998) and the cache and all the other functions are working OK now and the board behaves as it should.
The bioses all available here;
https://theretroweb.com/

Bump.
Found this thing in my 'need to fix one day' box and found it to still be behaving as described before. Still don't have an oscilloscope, but I have a few other things - ISA and PCI clock measuring cards and Necroware's guide to de-lidding an in-place DS128(8)7 RTC to replace batteries.

That's the good news. Bad news is: didn't help:
- PCI clock is 25MHz (correct for jumper settings running CPU at 75MHz)
- ISA clock is 6.25MHz
- new battery is happily over 3V.
- still zero life in the board :'(

So I ruled out clock generation and RTC being the issue.

H.W.Necromancer wrote on 2024-02-03, 13:45:

Sorry for the necro-posing, but I am a Necromancer. 😀 Sad there are no more post how it went with this particular board. I can confirm the problem might have been in loose pins on the QFP chips. I had dead board of the same type and one chip (SIS 503) had almost left the board. The other one had some bad pins as well.

At first glance I couldn't see any issues, but one side of the 85C501 looked like it was a tiny bit short. Unfortunately I decided to put some flux on and quickly run over it with my soldering iron to be sure. Bad idea. I created two bridges while doing that and while trying to fix the second one I managed to break the leg entirely. It was pin 115, REQ 3 - pretty necessary for the PCI bus to work.

Given this I've decided to give up this board.

dionb wrote on 2024-12-28, 21:26:

At first glance I couldn't see any issues, but one side of the 85C501 looked like it was a tiny bit short. Unfortunately I decided to put some flux on and quickly run over it with my soldering iron to be sure. Bad idea. I created two bridges while doing that and while trying to fix the second one I managed to break the leg entirely. It was pin 115, REQ 3 - pretty necessary for the PCI bus to work.

Given this I've decided to give up this board.

You can dremel pretty far in to solder some enameled wire.
Kinda of good news.
Corners are easier.

Nexxen wrote on 2024-12-28, 21:28:

[...] […]
Show full quote

[...]

You can dremel pretty far in to solder some enameled wire.
Kinda of good news.
Corners are easier.

Theoretically possible, but beyond my current skills+equipment, I'm afraid.

dionb wrote on 2024-12-28, 22:01:

Nexxen wrote on 2024-12-28, 21:28:

[...] […]
Show full quote

[...]

You can dremel pretty far in to solder some enameled wire.
Kinda of good news.
Corners are easier.

Theoretically possible, but beyond my current skills+equipment, I'm afraid.

Take this back and make us proud!
You were born a Vogoner!

dionb wrote on 2024-12-28, 21:26:

Bump. Found this thing in my 'need to fix one day' box and found it to still be behaving as described before. Still don't have a […]
Show full quote

Bump.
Found this thing in my 'need to fix one day' box and found it to still be behaving as described before. Still don't have an oscilloscope, but I have a few other things - ISA and PCI clock measuring cards and Necroware's guide to de-lidding an in-place DS128(8)7 RTC to replace batteries.

That's the good news. Bad news is: didn't help:
- PCI clock is 25MHz (correct for jumper settings running CPU at 75MHz)
- ISA clock is 6.25MHz
- new battery is happily over 3V.
- still zero life in the board :'(

So I ruled out clock generation and RTC being the issue.

H.W.Necromancer wrote on 2024-02-03, 13:45:

Sorry for the necro-posing, but I am a Necromancer. 😀 Sad there are no more post how it went with this particular board. I can confirm the problem might have been in loose pins on the QFP chips. I had dead board of the same type and one chip (SIS 503) had almost left the board. The other one had some bad pins as well.

At first glance I couldn't see any issues, but one side of the 85C501 looked like it was a tiny bit short. Unfortunately I decided to put some flux on and quickly run over it with my soldering iron to be sure. Bad idea. I created two bridges while doing that and while trying to fix the second one I managed to break the leg entirely. It was pin 115, REQ 3 - pretty necessary for the PCI bus to work.

Given this I've decided to give up this board.

--
Been there done that when I tried to fix such a thing some years ago. 1) always test all the pins with a thin medical needle (or something like that) if they wobble at the pad. 2) do not add solder - it is already there. (or only a super super tiny bit - just make your solder tip silver as it should be 3) yes - use flux but touch only the bad pins - I am not using "drag method" 4) if there is a short - bridge, add flux, DO NOT panic and use solder wick (copper braids) and suck all the solder in to it.
--
I am happy to help another Vogoneer - I might have a donnor board - usless OEM from some compaq and I think the chipset is there.
--
But do not panic. Train your drilling / grinding capabilities on some usless junk and than try to fix it. You can even do it manualy, it just takes time. The actual chip die is in the middle - you can expose the bonding metal and solder on to it.
--
where are you from? PM me if you wish.
Good luck!

Hi all .
Looks like the story will continue.
Motherboard ended up in ‘asus freak’ hands (me 😎) as a ‘display item’.
I’ll try to diagnose/repair it.

85C501 doesn’t look that bad - keep y fingers crossed 🙃

Great news...
At close inspection, 'missing' pin turned out to be bent and pressed under the IC !
It took while (a lot of fun BTW...) and I managed to straighten it without breaking.

The game is still on 😉

Locutus wrote on 2025-10-01, 23:23:

Great news... At close inspection, 'missing' pin turned out to be bent and pressed under the IC ! It took while (a lot of fun BT […]
Show full quote

Great news...
At close inspection, 'missing' pin turned out to be bent and pressed under the IC !
It took while (a lot of fun BTW...) and I managed to straighten it without breaking.

The attachment IMG_2273.jpg is no longer available

The game is still on 😉

I am still following this story! The board is in good hands. Amazing job. I did such repairs too but this looks almost as good as before - like if nothing bad happened!! CUDOz!
--
You have ecouraged me to get back ti one SiS based Abit where I failed and burned our some pads. Same issue - loose pins all over the board. But due to the layout it is harder to fix it

Hi.

H.W.Necromancer wrote on 2025-10-02, 15:38:

I am still following this story! The board is in good hands. Amazing job. I did such repairs too but this looks almost as good as before - like if nothing bad happened!! CUDOz!

Thanks!
Me & my 'chinascope' did our best 😉

H.W.Necromancer wrote on 2025-10-02, 15:38:

You have ecouraged me to get back ti one SiS based Abit where I failed and burned our some pads. Same issue - loose pins all over the board. But due to the layout it is harder to fix it

Good luck - fingers crossed !

I decided to power up the motherboard to check if 85C501 doesn't explode 😉
85C501 seems to be OK, CPU gets warm and the board is absolutely dead silent.
(Post card shows all voltages to be present but no post-codes on ISA/PCI).

I've checked all the pins around 'chipset' ICs and found out quite a lot loose legs in 85C503.
I think I'll resolder all of then in this chip.

Looks like the DALLAS was 'hacked' quite brutally from the top without desoldering.

I've got a bunch of DALLAS chips, so I didn't bother desolering it, just cut off all the legs and removed them one by one.
There is some damage under it -broken trace and cracked SMD resistors.

The fun continues...

Damaged SMD resistors have been replaced, broken trace rebuild and covered with UV soldermask.

I also found the socket that nicely masks all the repairs.

All loose legs have been resoldered, checked sram ICs (one was dead) and... there is almost complete success !
Motherboard 'POSTs' and BOOTS from IDE.

BUT, there is still a problem with ISA BUS.
It is partially working - I get readings on the Diagnostic Post code Card, but when I plug VGA (even 8bit) it is not working.

Locutus wrote on 2025-10-02, 18:04:

[...]

Looks like the DALLAS was 'hacked' quite brutally from the top without desoldering.

The attachment View recent photos.jpeg is no longer available

Guilty as charged 😉

I was following necroware's method of using a hot-air gun to make the epoxy soft then pull it off.

I've checked the waveforms on each pin of ISA 8 with an oscilloscope.
The waveform on pin A9 (Data bit 0) didn't look 'normal' to me, as if something was pulling the signal high.

I've checked the resistance between pin A9 and +5V power... and it is only 9 ohms 🙁

Locutus wrote on 2025-10-04, 21:52:

I've checked the waveforms on each pin of ISA 8 with an oscilloscope. The waveform on pin A9 (Data bit 0) didn't look 'normal' […]
Show full quote

I've checked the waveforms on each pin of ISA 8 with an oscilloscope.
The waveform on pin A9 (Data bit 0) didn't look 'normal' to me, as if something was pulling the signal high.

The attachment RigolDS8.png is no longer available

I've checked the resistance between pin A9 and +5V power... and it is only 9 ohms 🙁

A9 is DATA 0

Where does it go to? It's tracing time, maybe it's just a bad res network. U4 maybe?

Nexxen wrote on 2025-10-04, 23:44:

A9 is DATA 0
Where does it go to? It's tracing time, maybe it's just a bad res network. U4 maybe?

Hi.
It connects to a transceiver 74245.
I optimistically assumed that the damage would concern this IC, but after desoldering it, resistance to +5V rail was still very low.

I have determined that FDC37C665GT controller is also directly connected to this line.

I've lifted pin 48 of SMC IC and resistance on ISA pin went to infinity.
Resistance between pin 48 of FDC37C665GT and +5V rail is 9ohm - this IC is definitely damaged.
I've resoldered 74245 and ISA is working !

Locutus wrote on 2025-10-05, 00:38:

Hi. It connects to a transceiver 74245. I optimistically assumed that the damage would concern this IC, but after desoldering […]
Show full quote

Nexxen wrote on 2025-10-04, 23:44:

A9 is DATA 0
Where does it go to? It's tracing time, maybe it's just a bad res network. U4 maybe?

Hi.
It connects to a transceiver 74245.
I optimistically assumed that the damage would concern this IC, but after desoldering it, resistance to +5V rail was still very low.

The attachment TTL_DES.jpg is no longer available

I have determined that FDC37C665GT controller is also directly connected to this line.

The attachment SUPER_IO.jpg is no longer available

I've lifted pin 48 of SMC IC and resistance on ISA pin went to infinity.
Resistance between pin 48 of FDC37C665GT and +5V rail is 9ohm - this IC is definitely damaged.
I've resoldered 74245 and ISA is working !

The attachment ISA_WORKS.jpg is no longer available

You desoldered the 74xxx, and after resoldering it ISA started working 100%?
Wow, this is wild!

Floppy drive controller FDC37C665GT definitely looks bad... There's a lot of them new on ebay, but prices are high

Nexxen wrote on 2025-10-05, 00:54:

You desoldered the 74xxx, and after resoldering it ISA started working 100%?
Wow, this is wild!

But "problematic" pin 48 of SMC FDC controller stayed disconnected !

MikeSG wrote on 2025-10-05, 12:19:

Floppy drive controller FDC37C665GT definitely looks bad... There's a lot of them new on ebay, but prices are high

Yup...
I've found in my pile of "junk" not so good looking Gigabyte motherboard that can be a donor of this this chip.
Not sure if it is operational, but since I have it, it is worth trying.