Usually with those codes, excluding hardware problems, reprogramming the BIOS chip makes the PC work again, if you are interested on Tony359's YouTube channel there is the video on an ASUS Slot1, which after several repairs, he solved by reprogramming.
Here I would say that if you are able to do some investigations, I would go looking for something like interrupted track, bad welding, or defective SMD, maybe it could be somewhere the problem, then having found a way to make it work, it is a great clue, that there is something wrong, unfortunately I can't suggest a solution or where to focus the research.

impressive detective work
afaik ram detection happens on slowest safe chipset settings so nothing to be gained there, not to mention 0d means ram was detected and initialized successfully at this point

> code 41 0d

0D in this bios is Video bios execution

1seg000:F93D                 mov     al, 0Dh
2seg000:F93F                 mov     dx, 80h
3seg000:F942                 out     dx, al          ; manufacture's diagnostic checkpoint (POST code)
4seg000:F943                 mov     ax, 0C000h
5seg000:F946                 mov     ds, ax
6seg000:F948                 assume ds:nothing
7seg000:F948                 cmp     word ptr ds:0, 0AA55h
8seg000:F94E                 jz      short loc_FF971
9
10seg000:F971 loc_FF971:                              ; CODE XREF: seg000:F94E↑j
11seg000:F971                 push    bp
12seg000:F972                 call    far ptr 0C000h:3
13...
14seg000:F977                 pop     bp
15seg000:F978                 mov     ax, 3
16seg000:F97B                 int     10h             ; - VIDEO - SET VIDEO MODE
17seg000:F97B                                         ; AL = mode

41 is decompressing rest of the bios BIOS checksum error

1seg001:F97D                 lea     ax, aAwardBootblock ; "Award BootBlock Bios v1.0\r\nCopyright "...
2seg001:F981                 mov     cx, 45h ; 'E'
3seg001:F984                 xor     dx, dx
4seg001:F986                 call    print
5seg001:F989
6seg001:F989 checksum_error:                         ; CODE XREF: seg001:F96F↑j
7seg001:F989                 lea     ax, aDriveAErrorSys+8Fh ; "BIOS ROM checksum error\r\nDetecting fl"...
8seg001:F98D                 mov     cx, 17h
9seg001:F990                 nop
10seg001:F991                 mov     dx, 0FF01h
11seg001:F994                 call    print
12seg001:F997                 test    byte ptr [bp+0], 1
13seg001:F99B                 jz      short loc_FF9AB
14seg001:F99D                 lea     ax, aDriveAErrorSys+51h ; "Keyboard controller errorKeyboard error"...
15seg001:F9A1                 mov     cx, 19h
16seg001:F9A4                 nop
17seg001:F9A5                 mov     dx, 0FF01h
18seg001:F9A8                 call    print
19seg001:F9AB
20seg001:F9AB loc_FF9AB:                              ; CODE XREF: seg001:F99B↑j
21seg001:F9AB                 test    byte ptr [bp+0], 2
22seg001:F9AF                 jz      short loc_FF9BF
23seg001:F9B1                 lea     ax, aDriveAErrorSys+6Ah ; "Keyboard error or no keyboard presentBI"...
24seg001:F9B5                 mov     cx, 25h ; '%'
25seg001:F9B8                 nop
26seg001:F9B9                 mov     dx, 0FF01h
27seg001:F9BC                 call    print
28seg001:F9BF
29seg001:F9BF loc_FF9BF:                              ; CODE XREF: seg001:F9AF↑j
30seg001:F9BF                 mov     al, 41h ; 'A'
31seg001:F9C1                 mov     dx, 80h

PC@LIVE wrote on 2026-04-30, 19:27:

Usually with those codes, excluding hardware problems, reprogramming the BIOS chip makes the PC work again

sure looks like it! Either bios still corrupted (first 64KB), eprom A16 address line not connecting reliably, or bad ram that wasnt caught during ram detection? whatever it is its crashing during upper bios chip decompression

>After I accidentally shorted an output pin of a 74F244 with an input from the SiS496, and when powering off and on again, ¡the motherboard boots!

another impressive discovery 😮
which pin to which pin exactly?

rasz_pl wrote on 2026-04-30, 21:19:

impressive detective work afaik ram detection happens on slowest safe chipset settings so nothing to be gained there, not to men […]
Show full quote

impressive detective work
afaik ram detection happens on slowest safe chipset settings so nothing to be gained there, not to mention 0d means ram was detected and initialized successfully at this point

> code 41 0d

0D in this bios is Video bios execution

Copy code to clipboard

1seg000:F93D                 mov     al, 0Dh
2seg000:F93F                 mov     dx, 80h
3seg000:F942                 out     dx, al          ; manufacture's diagnostic checkpoint (POST code)
4seg000:F943                 mov     ax, 0C000h
5seg000:F946                 mov     ds, ax
6seg000:F948                 assume ds:nothing
7seg000:F948                 cmp     word ptr ds:0, 0AA55h
8seg000:F94E                 jz      short loc_FF971
9
10seg000:F971 loc_FF971:                              ; CODE XREF: seg000:F94E↑j
11seg000:F971                 push    bp
12seg000:F972                 call    far ptr 0C000h:3

41 is decompressing rest of the bios

Copy code to clipboard

1seg000:F977                 pop     bp
2seg000:F978                 mov     ax, 3
3seg000:F97B                 int     10h             ; - VIDEO - SET VIDEO MODE
4seg000:F97B                                         ; AL = mode
5seg000:F97D                 lea     ax, ds:0E000h
6seg000:F981                 mov     cx, 45h ; 'E'
7seg000:F984                 xor     dx, dx
8seg000:F986                 call    sub_FFAED
9seg000:F989
10seg000:F989 loc_FF989:                              ; CODE XREF: seg000:F96F↑j
11seg000:F989                 lea     ax, ds:0FEF8h
12seg000:F98D                 mov     cx, 17h
13seg000:F990                 nop
14seg000:F991                 mov     dx, 0FF01h
15seg000:F994                 call    sub_FFAED
16seg000:F997                 test    byte ptr [bp+0], 1
17seg000:F99B                 jz      short loc_FF9AB
18seg000:F99D                 lea     ax, ds:0FEBAh
19seg000:F9A1                 mov     cx, 19h
20seg000:F9A4                 nop
21seg000:F9A5                 mov     dx, 0FF01h
22seg000:F9A8                 call    sub_FFAED
23seg000:F9AB
24seg000:F9AB loc_FF9AB:                              ; CODE XREF: seg000:F99B↑j
25seg000:F9AB                 test    byte ptr [bp+0], 2
26seg000:F9AF                 jz      short loc_FF9BF
27seg000:F9B1                 lea     ax, ds:0FED3h
28seg000:F9B5                 mov     cx, 25h ; '%'
29seg000:F9B8                 nop
30seg000:F9B9                 mov     dx, 0FF01h
31seg000:F9BC                 call    sub_FFAED
32seg000:F9BF
33seg000:F9BF loc_FF9BF:                              ; CODE XREF: seg000:F9AF↑j
34seg000:F9BF                 mov     al, 41h ; 'A'
35seg000:F9C1                 mov     dx, 80h

PC@LIVE wrote on 2026-04-30, 19:27:

Usually with those codes, excluding hardware problems, reprogramming the BIOS chip makes the PC work again

sure looks like it! Either bios still corrupted (first 64KB) or bad ram that wasnt caught during ram detection? whatever it is its crashing during upper bios chip decompression

>After I accidentally shorted an output pin of a 74F244 with an input from the SiS496, and when powering off and on again, ¡the motherboard boots!

another impressive discovery 😮
which pin to which pin exactly?

I hope 🤞 you want to share your discovery, even if it doesn't necessarily work on other motherboards, but I'd be curious to see if it's possible to fix the motherboard, and I don't know why it restarts with the error post codes.
Of course I would also suspect that the BIOS chip could have a defect, but thinking about it, it could be a pin with welding to be redone, it goes with the trick, and it doesn't go without it, maybe the trick bypasses the defective pin (?)

rasz_pl wrote on 2026-04-30, 21:19:

afaik ram detection happens on slowest safe chipset settings so nothing to be gained there, not to mention 0d means ram was dete […]
Show full quote

afaik ram detection happens on slowest safe chipset settings so nothing to be gained there, not to mention 0d means ram was detected and initialized successfully at this point

sure looks like it! Either bios still corrupted (first 64KB), eprom A16 address line not connecting reliably, or bad ram that wasnt caught during ram detection? whatever it is its crashing during upper bios chip decompression

>After I accidentally shorted an output pin of a 74F244 with an input from the SiS496, and when powering off and on again, ¡the motherboard boots!

another impressive discovery 😮
which pin to which pin exactly?

Yes, memory seems to be detected, because without any SIMM installed, the beeps are definitely different (continuous, but much longer) and the POST codes are totally different.

I checked continuity between all BIOS pins and the southbridge (85C497), directly from pins to pin (to avoid missing a false contact in the DIP socket or something). But I might consider reflowing at this point... I just tried to avoid reflowing the chipset as that operation is a bit risky and things can get worse.

The pins shorted were A2 (DRAM address line #2, coming out of a 74F244 and going to pin #14 of SIMM modules) and one of the chipset (85C496) outputs (either address line #5 or #6 seem to have worked, which corresponds to 85C496 pins #143 and #144).

This "trick" usually just hangs the motherboard and nothing more. Just sometimes (maybe one out of ten tries), the PC boots perfectly after next reset (to be precise: power down / power up the power supply).

PC@LIVE wrote on 2026-04-30, 21:32:

I hope 🤞 you want to share your discovery, even if it doesn't necessarily work on other motherboards, but I'd be curious to see if it's possible to fix the motherboard, and I don't know why it restarts with the error post codes.
Of course I would also suspect that the BIOS chip could have a defect, but thinking about it, it could be a pin with welding to be redone, it goes with the trick, and it doesn't go without it, maybe the trick bypasses the defective pin (?)

Sure! On my last post I already shared the shorted pins which sometimes allows the machine to boot.
I don't know how to reply to several users on the same post, so I posted this new post to reply to yours (please, let me know if there's a better way to do it).

I already tried 2 different BIOS chips (AT27C010) with two different binaries (the one that originally came with board, and the one on retroweb, which is: 03/14/96-SiS-496-497/A/B-2A4IB000C-00).
The behavior is identical 😳
That is, post code 41 0D, short continuous beeps, and I managed to boot at least one time with either them.

Also checks continuity of all pins to the southbridge 85C497. But, as I said in my previous post, maybe is time to reflow both big SMD chips. Just wanted to do everything possible before reflowing, as doing it can damage the board and make things worse.

The really weird thing is that the shortcircuit "trick" happens in the northbridge (between SIMM module and 85C496) but the BIOS is connected to the southbridge (85C497).
How could that short in northbridge bypass a defective pin in the other chip?
Not saying that is impossible.
But right now I cannot think on any (solid) explanation.

Just for fun, here's the replacement of the 74F244. It looks a bit messy, but it works just as well as the original SMD chip.

(The yellow adhesive tape is just kapton tape to protect components when desoldering the original SMD chip)

RetroVein wrote on 2026-05-01, 10:15:

Sure! On my last post I already shared the shorted pins which sometimes allows the machine to boot. I don't know how to reply t […]
Show full quote

PC@LIVE wrote on 2026-04-30, 21:32:

I hope 🤞 you want to share your discovery, even if it doesn't necessarily work on other motherboards, but I'd be curious to see if it's possible to fix the motherboard, and I don't know why it restarts with the error post codes.
Of course I would also suspect that the BIOS chip could have a defect, but thinking about it, it could be a pin with welding to be redone, it goes with the trick, and it doesn't go without it, maybe the trick bypasses the defective pin (?)

Sure! On my last post I already shared the shorted pins which sometimes allows the machine to boot.
I don't know how to reply to several users on the same post, so I posted this new post to reply to yours (please, let me know if there's a better way to do it).

I already tried 2 different BIOS chips (AT27C010) with two different binaries (the one that originally came with board, and the one on retroweb, which is: 03/14/96-SiS-496-497/A/B-2A4IB000C-00).
The behavior is identical 😳
That is, post code 41 0D, short continuous beeps, and I managed to boot at least one time with either them.

Also checks continuity of all pins to the southbridge 85C497. But, as I said in my previous post, maybe is time to reflow both big SMD chips. Just wanted to do everything possible before reflowing, as doing it can damage the board and make things worse.

The really weird thing is that the shortcircuit "trick" happens in the northbridge (between SIMM module and 85C496) but the BIOS is connected to the southbridge (85C497).
How could that short in northbridge bypass a defective pin in the other chip?
Not saying that is impossible.
But right now I cannot think on any (solid) explanation.

Just for fun, here's the replacement of the 74F244. It looks a bit messy, but it works just as well as the original SMD chip.

The attachment 74F244 DIP.JPG is no longer available

(The yellow adhesive tape is just kapton tape to protect components when desoldering the original SMD chip)

Thank you very much for sharing this trick, for the moment as I had already written to you, I have not had a problem similar to yours, that's why normally after reprogramming the PC works again.
And as you wrote I would rule out a problem with the RAM, because I usually use RAM working heads, although there may be a doubt about some bad contact, but since you have found a way to make the PC work, I would say that your idea of redoing the chip welds, it could be possible that there is a PIN (or some?) That could have non-optimal contacts, before trying to renew the welds, I would try to hold the chip down and see if it can go beyond post code 41.
For multiple response no problem 😉, I do "reply" I select everything and copy, then I go back and do "reply" on the second user, and paste the selection of the first user, finally I add the answers for both users, it works but I can't tell you if it's right to do it that way.
Finally, on what the circuit or the BIOS does, when you use the trick, I don't know, I think there is a point where you jump over the post code 41, or maybe even if you find an error at 41, you can still continue.

I was confidently wrong previously 🙁. C3 is BIOS decompression and is skipped because bios checksum fails at POST 0Ch.
You can confirm this by rewinding POST codes <- afaik thats what that one button on POST card does.

C1 is:
1 checks checksum of very small routine at FB020(very crude ram detection).
2a If checksum good calls FB020(very crude ram detection):

1seg001:B0DD                 mov     word ptr ds:400h, 5678h
2seg001:B0E3                 mov     word ptr ds:800h, 5678h
3seg001:B0E9                 mov     word ptr ds:100000h, 5678h
4seg001:B0F2                 mov     word ptr ds:200000h, 5678h
5seg001:B0FB                 mov     word ptr ds:400000h, 5678h
6seg001:B104                 mov     word ptr ds:800000h, 5678h
7seg001:B10D                 mov     word ptr ds:1000000h, 5678h
8seg001:B116                 mov     word ptr ds:2000000h, 5678h
9seg001:B11F                 mov     word ptr ds:0, 0AA55h
10seg001:B127                 mov     word ptr ds:4, 55AAh
11seg001:B12F                 xor     ebx, ebx
12seg001:B132                 cmp     word ptr ds:0, 0AA55h
13seg001:B13A                 jnz     loc_FB1DE
14seg001:B13E                 cmp     dh, 2
15seg001:B141                 jnz     short loc_FB151
16seg001:B143                 mov     bl, 5
17seg001:B145                 cmp     word ptr ds:400h, 0AA55h
18seg001:B14D                 jz      loc_FB1DE
19seg001:B151
20seg001:B151 loc_FB151:                              ; CODE XREF: seg001:B141↑j
21seg001:B151                 cmp     word ptr ds:200000h, 0AA55h

etc. tldr it writes 5678h marker to <1MB, 1 2 4 8 16 and 32MB, then 0AA55h to address 0 and checks when it stops being mirrored (wrap around). Flips Register "48h ~ 4Fh DRAM Boundary" and tries again.

2b If FB020(very crude ram detection) checksum is bad bios skips it and ... calls exact copy of that routine located at FE0B0 (lol wtf). Not that it matters because it will just blindly assume presence of at least 256KB ram and keep executing no matter what.
Both FB020 and FE0B0 just initialize slowest safest memory controller settings possible and nothing more. Both also disable L2 cache.

3 Checks just first 64KB ram - if that test fails bios freezes at C1 with no beeper.

4 Clears 256KB ram

5 Bios copies itself to ram at address 64-192KB

6 Enables CPU L1 cache unless disabled in Cmos.

Then POST 0C takes over:
1 Calculates checksum of BIOS module copied earlier to ram address starting at 64KB, length E4AAh

2 Calculated checksum of Award Decompression Bios chunk.

If both are ok goes to POST C3 and decompresses full bios.

If any one is bad it skips C3 and goes straight to POST 1. I think this is boot recovery procedure:
1 clear_first_640KB ram

POST 5:
1 check if warm / cold boot
2a if warm go to POST 0C
2b if cold initialize keyboard controller/keyboard and go to POST 0C

POST 0C:
1 initialize IVT_Vectors

POST 0D:
1 VGA BIOS check & optional VGA BIOS init
2 prints "Award BootBlock Bios v1.0" ""BIOS ROM checksum error"

POST 41 tries to boot recovery floppy

You said "code 41 0d and continuous short beeps.". That only happens if it cant find a floppy drive, in that case it prints "Drive A error. System halt" and calls BEEP routine in a short loop.

>When there's a floppy drive attached, the BIOS tries to read from a floppy (I suppose that BIOS checksum fails, and it is trying to find a binary on the floppy to flash itself).

so it keeps executing with floppy, at least there is that

Award 4.51 is already a weird bios with decompressing modules and diverging execution paths. I still have a hard time tracking it all out and wrapping my head around it. I dont understand couple of things:
1 How computer arrives at 41 with no VGA graphics initialized 😐 maybe VGA bios checks for ram presence, fails and exits without initializing?
2 Where exactly Award decides between bootblock loader and successful BIOS decompression. My current theory is POST C3 is the diverging path - either its executed and thats where rest of proper BIOS initialization happens ending in OS load, or its skipped and anything after is Recovery BootBlock Loader.

RetroVein wrote on 2026-05-01, 09:57:

Yes, memory seems to be detected, because without any SIMM installed, the beeps are definitely different (continuous, but much longer) and the POST codes are totally different.

That detection is very crude and there is no real ram test beyond first 64KB this early.

RetroVein wrote on 2026-05-01, 09:57:

I checked continuity between all BIOS pins and the southbridge (85C497), directly from pins to pin (to avoid missing a false contact in the DIP socket or something). But I might consider reflowing at this point... I just tried to avoid reflowing the chipset as that operation is a bit risky and things can get worse.

At this point I believe BIOS itself is fine. What is failing is copying BIOS contents to 64-192KB ram window. This gets detected and execution turns into recovery path. But why no graphics?

RetroVein wrote on 2026-05-01, 09:57:

The pins shorted were A2 (DRAM address line #2, coming out of a 74F244 and going to pin #14 of SIMM modules) and one of the chipset (85C496) outputs (either address line #5 or #6 seem to have worked, which corresponds to 85C496 pins #143 and #144).

😮
That scope screen of A2 doesnt really look bad, still over 3V and no obvious glitches.
BIOS ram detection only ever really checks address _0_ of 1 2 4 etc megabyte. If A2 was bad if would still happily detect full size of installed ram, not to mention bios keeps executing even if ram detection fails completely.

I would check every pin in SIMM sockets, maybe they also need deoxit or you got one bend.

PC@LIVE wrote on 2026-05-01, 10:42:

I would say that your idea of redoing the chip welds, it could be possible that there is a PIN (or some?) That could have non-optimal contacts, before trying to renew the welds, I would try to hold the chip down and see if it can go beyond post code 41.

I tried, not only on the chip, but on other places. But no luck.

Also, after many tries, the motherboard was able to go through POST correctly (and I managed to go into BIOS, etc.) without any previous short-circuit.

So, once in a while, the motherboard boots even without the short-circuit trick. But that happens 1 out of 100 tries, or less. It just seems that the short-circuit makes it more probable. But I cannot confirm this with real data.

rasz_pl wrote on 2026-05-02, 08:51:

At this point I believe BIOS itself is fine. What is failing is copying BIOS contents to 64-192KB ram window. This gets detected […]
Show full quote

At this point I believe BIOS itself is fine. What is failing is copying BIOS contents to 64-192KB ram window. This gets detected and execution turns into recovery path. But why no graphics?

😮
That scope screen of A2 doesnt really look bad, still over 3V and no obvious glitches.
BIOS ram detection only ever really checks address _0_ of 1 2 4 etc megabyte. If A2 was bad if would still happily detect full size of installed ram, not to mention bios keeps executing even if ram detection fails completely.

I would check every pin in SIMM sockets, maybe they also need deoxit or you got one bend.

Thank you very much for all that detailed information, which I read carefully.

Even though A2 signal doesn't look too bad, it was by the time the only suspect.

I checked contuinity of all SIMM pins (directly on the contacts) to the 74F244 and chipset ICs. On 4 sockets. They all test ok.

During the long initial testing sesions, there'were indeed 2 pins bent. But I spot them, and I straightened them.

About why there is no video output... Could it be because I'm using a PCI video card? Anyway, I'll try to use an ISA video card tomorrow, and let you know if there's video output.

UPDATE.
It seems fixed.
Strange problem it was. Strange fix too.

It took me a couple of sessions to repeat all tests before posting, just to make sure.

Long story short, the motherboard now boots, and I'm able to boot into DOS and run speedsys. The L1 and L2 caches are enabled, and I get reasonable benchmark results.

Now... what fixed it?
I'm not 100% sure. But every thing seemed to change after trying an ISA video card.

• When I plugged the video card into one ISA slot, the motherboard still didn't boot neither I had video output. But the continuos beeps were gone, and no more the system stalls with the dreadful POST codes 41 0d.
• I tried another ISA slot, and after a little bit of fiddling, I got video output and a good POST.
• I swapped the video card to the previous PCI I used before... and it also worked !

In fact, I didn't get a single POST failure, except sometimes (when using the ISA card, which sometimes seems to make some false contact, and the card is not detected).

After many years working with vintage hardware, this is one of the weirdest repair sessions.

Any idea of what could happened?
My best guess is that the ISA slot where I inserted the VGA card had some dirt that conducted some leakage current and corrupted the BIOS data transfer to the southbridge... After inserting the ISA card, that "short" opened.
I tried the post card on that ISA slot before, and still got the problem described in the OP. But the ISA card is 16 bit. Maybe the problem was on the 16 bit portion of the connector.
Does it sound too far-fetched?

Weird. No video while the error codes had nothing to do with video were also weird.

rasz_pl wrote on 2026-05-16, 14:58:

Weird. No video while the error codes had nothing to do with video were also weird.

Yes. And even though the system now works, this kind of magical things in troubleshooting (be it hardware orsoftware related) just makes me feel uneasy. And curious.
Any idea of what could have happened?

Hi, @RetroVein,
Do you have any idea how long the motherboard has not been energized, before you switch on the PSU for the first time now?

Only thing that comes to mind is mechanical damage. Plugging ISA stressed mobo in the "right way" temporarily fixing whateevr broken connecting it has.

I'm asking about how long the motherboard has not been energized, because I had two similar cases with 30+ years old motherboards, that refused to POST normally at first start after they have been in storage for more than a decade. And after a couple of faulty tries they "magically" started to POST without any interventions and giving no further problems. I think this could be due to electrolytic capacitors reforming i.e. healthy electrolytic capacitors, but long time not energized, behave like bad capacitors and need some power-on time to regain their normal characteristics.

aVd wrote on 2026-05-17, 06:00:

I'm asking about how long the motherboard has not been energized, because I had two similar cases with 30+ years old motherboards, that refused to POST normally at first start after they have been in storage for more than a decade. And after a couple of faulty tries they "magically" started to POST without any interventions and giving no further problems. I think this could be due to electrolytic capacitors reforming i.e. healthy electrolytic capacitors, but long time not energized, behave like bad capacitors and need some power-on time to regain their normal characteristics.

That definitely is a very interesting observation.
It is not the first time I hear something like that, but honestly, I didn't considered it until now that you mentioned it.
As for the answer to you question... I don't know. This motherboard didn't belong to me until recently. But probably was not energized for a long time (it was very dirty. Came to me with no CPU neither SIMM modules, and there was dust all over the place.

rasz_pl wrote on 2026-05-17, 05:16:

Only thing that comes to mind is mechanical damage. Plugging ISA stressed mobo in the "right way" temporarily fixing whateevr broken connecting it has.

If that is so, then I'll have to handle this board with extreme care!
And I'll also have to resolder back the SMD 74F244! 😀

RetroVein wrote on 2026-05-19, 09:20:

That definitely is a very interesting observation.
It is not the first time I hear something like that, but honestly, I didn't considered it until now that you mentioned it.
As for the answer to you question... I don't know. This motherboard didn't belong to me until recently. But probably was not energized for a long time (it was very dirty. Came to me with no CPU neither SIMM modules, and there was dust all over the place.

Thanks for the answer! As soon as I read how your motherboard started working without any interventions after the initial failed attempts with hangs at POSTing, I recalled my own very similar and strange cases involving two very old motherboards, that has been sitting "in storage" for a very long time. Then I recalled the peculiarities of electrolytic capacitors, that have been stored out of service (or on the shelf) for a long time - their reversible loss of characteristics. I really have no other explanation besides my theory, that the electrolytic capacitors have to undergo reforming before the motherboard's "awakening".

aVd wrote on 2026-05-19, 13:15:

Thanks for the answer! As soon as I read how your motherboard started working without any interventions after the initial failed attempts with hangs at POSTing, I recalled my own very similar and strange cases involving two very old motherboards, that has been sitting "in storage" for a very long time. Then I recalled the peculiarities of electrolytic capacitors, that have been stored out of service (or on the shelf) for a long time - their reversible loss of characteristics. I really have no other explanation besides my theory, that the electrolytic capacitors have to undergo reforming before the motherboard's "awakening".

Did you ever replaced the capacitors with new ones?

RetroVein wrote on 2026-05-19, 19:16:

Did you ever replaced the capacitors with new ones?

Nope. Those two motherboards are socket 5/7 from the mid '90s and made far earlier, than the ones from the early 2000s "capacitor plague" era boards. They are still in their original state as they came out of the factory.

So far I only recapped some s. 370, s. A and newer like s. 939 motherboards with visibly bulged and/or leaked electrolytic capacitors.

aVd wrote on 2026-05-19, 20:00:

Nope. Those two motherboards are socket 5/7 from the mid '90s and made far earlier, than the ones from the early 2000s "capacitor plague" era boards.

Well, the motherboard I had the issue with is a late socket 3, so probably around the same age as yours socket 5, and very much free of any capacitor plague 😀