For what 😨 I want to thank all those who participated in these discussions, related to motherboards, I'm not going to mention them all one by one, but I'm telling you:
Thank you 1000!
On the MSI MS-6741 VER:1, I just finished working on it, I replaced the damaged electrolytic capacitor, and I added the missing one, respectively a 1000uF 6.3V and a 1500uF 16V, for the replacement I used a Rubycon and a Panasonic, now the VRM area is all Rubycon except the Panasonic, which I put in place of the missing one, but ultimately at least this part should not have future problems, due to poor quality capacitors.

Previously I had written, of an identical motherboard but without integrated video, well it seems that despite having the same name, they are two different models, I don't know if the BIOS is incompatible, already because you have to be very careful to update it, here the chip is welded, so you can only reprogram from floppy, assuming that the Recovery BIOS is working.

The model that has the integrated video card is K8M800, and it is written on a sticker attached in a chip near that of the BIOS, the other card without integrated video, should be a K8T800, in practice two different chipsets for the same motherboard, and I would be quite sure that even the BIOS are different (incompatible!), I just hope that the previous owners did not make an incorrect BIOS update, and that therefore they found an unusable motherboard, this one in particular with K8M800, is perhaps more modern than the other, but this interests me relatively, the important thing for me is that it works, And that you can make a nice S.754 PC, which represents a period of passage between the 462 and the subsequent AM2, I don't forget the 939 but it's from the same period, while the 940 I think is very rare especially in a version with 4 CPUs, here's something like that you could admire it at the time, and for many it was a dream configuration.

Ultimately the card will have to be tested, using everything I used on the ASUS K8U-X, if I have the same problems, I will try something else, if it works well, I will have to try to understand if the K8U-X is repairable, we will see and let's hope well.

Well as I suspected, it's the ASUS K8U-X that doesn't work, it must have some problems, you can speculate towards the RAM, but whatever it is, it could be something else, currently it can't reach the BIOS screen, which it should do if there was no problem.

Why did I come to the conclusion that that ASUS K8U-X doesn't work?

For the simple reason, that I tried a lot of RAM, and I didn't have any change, then it can happen that some RAM in a PC doesn't work, but so many is impossible I would say, and the problem is not the CPU, because it comes from a working PC, and therefore the only way to verify that everything works, is to try it in another motherboard!

Said and done, finished fixing the MSI MS-6741 VER:1, there is a small problem, the hook for the CPU heatsink is missing, it is a type used only on the 754 and maybe other subsequent sockets, with only two screws, decided to disassemble it from the K8U-X, unfortunately the lower part is attached with a double-sided adhesive, it does not seem easy to detach it.

I had almost given up, but for a short test, I do without the hooks, and put the heatsink on the CPU, install a 128 MB DDR 400 RAM and connect PCI post card and video cable and ATX power supply.

I try to start and after a few beeps, there is the image, first the MSI logo and then also the first BIOS screen, usually at this point everything works, I should try to connect an IDE CF adapter with CF and DOS memory card, I will do it in the next few days, because I want to screw the part where the heat sink is hooked, without disassembling the lower one.

That's it for today, and I hope in the future to have the opportunity to recover other interesting motherboards, for the Intel 478 D845xxx board, I have the same problem of the occluded holes, but it is not possible to use the mini tips, because there are too many left, and most protrude and do not allow the centering of the hole, I think I will have to look for another system, to free those holes.

I will return to the MSI MS-6741 with Sempron 2600+, later, because first I have another ASUS 754 to try, the format is the normal ATX (which I prefer), the model is K8V-F, both the audio and the LAN are integrated, in addition it has a FireWire port accessible from the back of the PC in the shield area, and a socket next to the NB to add other FireWire ports, this technology was used for the connection of external camcorders or HD, the utility is relative, but the presence avoids the use of a special PCI card, leaving the place free for more.

Going back to this K8V-F REV. 1.03, it comes from an Olidata PC, and is dated 12/2004, I thought it was more recent, but anyway we are always facing a twenty-year-old technology, so if we want to use this card today, we have to take into account the limits, which are especially the use of single-core CPU, and first-generation DDR RAM.

The presence of expansion slots is abundant, we find an AGP 8X slot for video cards, five PCI that can be used to add a modem or other types of cards, the CNR is not present, even if it was expected, maybe there is in a possible Deluxe version, finally there are two DDR memory slots, and from what I remember, on the 754 there are usually two, while on the 939 we find up to four.

The board looks pretty good, the chipset heatsink is missing, and the hook for the CPU heatsink, on the back there are a couple of deep scratches, but they are on areas with wide tracks, and do not interrupt the track, so I would say that after a dusting, the board can be prepared for the boot test, since the electrolytic capacitors do not have bulges.

News about the ASUS K8V-F REV. 1.03, it seems that it is a particular card produced by ASUS for PC manufacturers, so it was not sold in PC stores (from what I understand), in this case Olidata was supplied by ASUS for their PC lines, and from what I understand, the same card is used in other PCs such as Fujitsu, but this could make any update difficult, because you can get confused with other similar models always K8V, honestly I don't know how the normal K8V and the K8V-F differ, however in case of doubt I would avoid any update Of the BIOS (in particular).

I have concluded the boot test, and the motherboard seems to work, I see the BIOS screens, but at a certain point it crashes, due to a Fatal Error, I tried to enter the BIOS, before you see the last screen, but it does not seem to be possible to enter it.

The CPU I used is the Sempron 3000+, unfortunately it is not recognized, and maybe this is the problem that causes the fatal error, in the end I will have to change the CPU and try again, I don't know if the Sempron came out years after the Athlon 64, and maybe this causes the lack of recognition, that is, there is no CPUs stored in the BIOS, or it could be that the 3000+, is one of the last to be supported, and therefore requires a BIOS subsequent to version 1007 present in the BIOS chip of the motherboard.

Speaking of this ASUS K8V-F, I tried to interpret the error messages displayed, one is very clear, the CPU is not recognized, this could depend on the BIOS version, which lacks support for the fastest CPUs, in theory it should definitely recognize the Athlon 64 (like the various Hammers), but there could be slightly revised versions, which came out later, in short, maybe first it's better to look for info on the CPUs we have, and in any case if we have one of the first releases, it's better to use that one.

A second problem could be, caused by the low battery, I did not check if the voltage is normal (just over 3V), but if by chance it was very low, this would prevent the saving of custom settings, and after a short time you return to the initial ones.

Even if this is actually not a problem that usually causes the PC to crash, in fact often resetting the CMOS means that it can be reused, in short, trying does not cost anything, but zeroing should not be causing other problems, if anything solve them.

To begin with, I performed the voltage check on the CR2032 battery, it is just above 0V, decidedly insufficient to store something, so I would rule out a problem related to the low battery.

In the meantime, I changed VGA, I currently use an ATI AGP, Radeon 7000 32 MB during the next tests, I do not rule out replacing it with another one, but for the moment I will use this one.

So I start trying a different CPU, not by much though, among the 754 available I have a Sempron 2800+, and I start trying this, hoping to get progress, unfortunately it doesn't change anything, exactly I don't know what the difference is between the two Sempron (3000 and 2800), but maybe they are speeds apart, very similar if not even the same, so I'm almost resigned, but I do a third test.

I get from the MSI MS-6741, the Sempron 2600+, it should be old enough, and since I don't know how it will go, I hope to have more luck.

I start the PC and I don't notice anything different, then at the end the codes 85 84 appear, but the beep is only one, and finally the first BIOS screen appears, without that annoying fatal error, from then on everything is normal, with F1 I access the BIOS, and I can choose what to view, or what to change.

Ultimately, it was a success, at least for now, the next step would be to connect the usual IDE CF adapter with CF and DOS memory card, start the various bench and diagnostics, but first I change RAM, remove the test one, and put two DDR 400 of 256 MB each, for the moment the total of 512 MB is fine, later we'll see if we need more.

Of the PC with the ASUS K8V-F and Sempron 2600+, looking better at the images of the BIOS screen I understood, what was the difference, with the faster Sempron, you can read in North Bridge RevUnknown, while in the other there is RevCG, normally the chip is external to the CPU, and therefore changing CPU does not change, here instead on AMD K8 CPUs, it is integrated into the CPU, obviously in the 2800+ and 3000+ it was a subsequent revision, and it was not recognized, this consequently was the cause of the error.

But there is more, the Fatal Error is usually related to memories, in particular ECC memories, so here maybe using a DDR ECC RAM maybe, you could be able to continue, without the PC crashing during startup, but it depends on whether DDR ECC memories are supported on 754, this info is usually found in the manual.

In any case, the solution to make the card work without errors, with Sempron 2800+ and 3000+ CPUs, would be a BIOS update, or a second alternative, would be to modify the current BIOS, updating the CPU recognition part, or simply replacing it with a socket from a subsequent BIOS, in short, a little complex to do, but without adequate experience in this regard, you can only do damage.

For my part, I don't need to use faster CPUs, the 2800+ or 3000+ ones, I will use them on the MSI MS-6741, already because I have two cards with the same name, but they differ in the chipset.

So currently I don't need to update the BIOS, except that in the future, I can put my paws on faster CPUs, and they wouldn't be recognized, unfortunately I don't think it's possible in the short term, compared to the S.939 754 CPUs are usually more expensive, and so if I need another 754 CPU, I would go look for a cheaper one, and this usually corresponds to models lower than 2800+.

I will do another update, another change of VGA AGP, I reasoned with a colleague that on this PC, it would be better to use an FX5200, instead of the Radeon 7000, or alternatively also an ATI 9000 or 9200, but I would say that if the FX5200 works, this would be the choice I would make.

I noticed in the BIOS of the ASUS K8V-F, a line in red, on the CPU fan speed line, it seems to want to report an abnormal value, maybe it could actually be like that, but the heat sink I used comes from a Socket subsequent to the 754, I think it's from an AM2, because there is a 4 pin socket, instead of the usual 3 pin.

I don't want to write random numbers, but I imagine that more or less around 3000, it can be considered normal, the current one of just above 1500 rpm, is about half, then if it is enough for cooling, or it is actually too low, I can't say, there would then be another possibility, that is that the revolutions do not correspond to the actual ones, and therefore the sensor is not precise in reading.

For the moment not having a disk with Windows, I can't check if there is a match or not, and I have no way to check the CPU temperatures, so as not to touch strange values.

I have selected the VGA I will use, it is an AGP 8X from Sparkle, with 128 MB DDR, the chips are 4ns, which should allow fairly safe overclocking, and theoretically you could reach 500 MHz (250 actually), compared to the current 400 MHz (200 now), but I don't think that increasing the speed of the RAM makes a big difference, even if it could earn something.

We'll see if tomorrow I can have the necessary time to install this board, and make some benches, with the other VGA I've already done several, including those of Phil's, we'll see the differences which and how many will be.

Comparing my initial idea, for a VGA suitable for the K8V-F, they confirmed that the ATI 7000 was a bit too cheap, and using a GF FX5200 the PC would have been suitable for most of the programs, which use VGA heavily, it is definitely not the best but it should be quite good.

After all, even the PC is currently not at its best, the CPU is a Sempron 2600+, so a 1600 MHz with only 128 KB of L2 cache, unfortunately the problem of non-recognition, prevented me from using the 3000+, which was a little faster, but if we had the availability of a top-of-the-range Athlon 64 754, we could get quite beyond the 3500+, with more than 2000 MHz, absolutely not a big difference, but quite relevant, and I think also visually it could be seen.

This VGA, I recovered among the ones I keep for future updates, it seemed good and not excessive, for this PC, it is Sparkle production, the exact model is SP8834, it has 128 MB DDR with 4 ns chips, these if necessary could allow the overclocking of the VGA, honestly it will not be necessary, but anyway the chips are slightly faster, than those commonly used (5 ns), despite all this card was among the low-level GF FX, more common were the 5500, while if I'm not mistaken the high level was the 5800-5900.

Here, it would have been important if the S.754 had the possibility to use multicore CPUs, unfortunately from what I know, you can only use single core CPU and single channel RAM, so the idea would be that of a dual boot, with DOS and Windows 98SE, you could even install a Windows XP or higher, but the RAM should be increased, and in my opinion with the current 512 MB DDR 400 (2X 256), it is not worth going further using Windows 98SE.

All that remains is to try again ️ the PC, and redo the bench including those of Phil's, when instead I have connected the disk with Windows 98SE installed, I could run other bench 3D programs, which can tell us if something is not at its best, I don't know but usually, it is very likely that it is some drivers, and it can be solved using others, but this is to be seen later.

I finished the comparison between the two video cards, using the ASUS K8V-F motherboard with Sempron 2600+, but I don't remember if with the tests done with the RADEON 7000 I had 128 MB of RAM, while currently I have 512 MB, with two DDR 400 of 256 MB each.

I found some differences, in Phil's benches, with other DOS programs running the benches, the gift values are the same or do not change much, because the results are obtained by using the CPU, and in this case it is always the same!

Let's see what's different:

In parentheses the values of the Radeon 7000

2) 418.1 (347.3)
3) 465.0 FPS (391.4)
4) 83.0 FPS (69.1)
5) 366.0 (320.0)
6) 40.6 (33.8 )

A) 504.66 (472.72)
B) 129.22 (108.72)
(C) 273.2 (257.2)
D) 110.2 (99.8 )
E) 27.6 (60.0)

You can see that only the last one with the letter e) is lower, but it happened with different video cards that the result is sometimes strange, so I wouldn't know which of the two is correct.

As you can see, there is a certain improvement, not as obvious as you would expect, but in any case you gain various FPS, maybe it won't make a big difference, surely the FX5200 is better, at least seeing these results, to see some benches in Windows 98SE, it will take a long time, now with this motherboard I would have for the moment finished, in conclusion I would say that I had some doubts that it could work, instead it seems that it also works quite well.

Of course until now, I've used it with a fairly normal configuration, nothing special, but at least so far, I haven't had any unpleasant surprises, of course I should compare this motherboard with other similar ones, to be able to say if it's better or worse, but it doesn't seem bad at all.

Today we change Socket, and move on to Intel, this time a Socket T (LGA 775), ASROCK 775Dual-VSTA REV. G/A 1.02, this acronym can be interpreted in various ways, but I don't think the Dual refers to Dual Core CPU, but it would instead (I imagine) refer to video card ports, in fact there are both the AGP and the PCI-E 16X, the previous owner told me he tried to use two video cards, but he couldn't see anything in one of the two cards, I think they can't be used at the same time, and so he upgraded with another PC, and gave me this motherboard plus a 256 MB VGA AGP 8X ATI 9550 (PCB Red), so after a long sleep, today it's time to get going again 🏍, this motherboard a little peculiar, from the silkscreens on the board, it says that it supports CPUs: Conroe Presler up to 1066 FSB and Dual Core.

I don't know what CPU is installed there, but I don't think it's a Dual Core, in the past when it was working, I had seen a Pentium 5XX or 6XX, I think about 3 GHz, he asked me ️ if it was possible to switch to a Pentium D, but I don't know if he then upgraded, in short it could be a surprise, but I don't think that after doing such an upgrade, he would have disused it.

I had given a good dust in the past, but today I will do some points, which were a little dirty, using a small brush 🖌, then you will need the usual equipment used in test, VGA PCI card PCI Post Card, DDR2 RAM, a power supply, keyboard and video, regarding the ATX power supply I usually use a 24 PIN, here you need a 20 PIN, the problem could be a very close capacitor, which would prevent the insertion of the 24 PIN one, I should do a test, however if it is not good, I should try to recover another ATX power supply with 20+4 or 20 PIN, nothing Complicated, but it would make me waste a lot of time, and I might not be able to do the start test today.

More on the board, there is no integrated video, it is in ATX format, there is audio and integrated LAN, for various expansions, there are four PCI slots, four RAM banks but only two DDR2 or alternatively two DDR, the BIOS chip is not welded (PLCC32), BIOS AMI I think 1.50 (?), VIA chipset, and RAM with the possibility of Dual Channel.

With the ASROCK 775Dual unfortunately I have to disassemble from a PC an ATX power supply that goes well, maybe a relatively simple solution would be a 24-20 PIN ATX adapter, in this way I could always use the 24 PIN ones, and adapt them to the motherboards where you can't insert, in this case, despite the space around it, there is that electrolytic capacitor that is at the 4 additional PINs, and this made it impossible to insert.!!!

It came between my paws, the 486 PCI SMT Terminator motherboard, I had put on an AMD 486 DX4-100, and then I had modified the jumpers to configure this CPU, there are quite a few, some are very complex, you have to be careful and the manual is not really easy to understand, in short, after wasting a lot of time, I do it and it's ready to start.

Or rather you can try after adding the essential, a 4 MB 72 PIN RAM (which is 100% working), a PCI VGA and the usual ISA PCI post card, as soon as everything else is connected, I start and move the switch from 0 to 1.

Here I don't see what I was hoping for, the codes are always — —, and despite a minimal sign of initial activity, making other attempts, moving the cards and the RAM, it always remains the same.

To be on the safe side I check the voltage on the linear regulator, I measure 3.27V which is very close to the selection of 3.3V, so nothing strange here, later I'll try again with a new CR2032 battery, but nothing changes, unfortunately this board has several very oxidized points, despite the cleaning, there could still be ️ too much oxide, I'll try something else in the next few days, but I think it's difficult to find a way to make it work.

Ciao Elio,

As you probably know, there are 2 different AMD 486DX4-100 CPU types. The Am486DX4-100 NV8T and the Am486DX4-100 SV8B.
These AMD types have different functions on several pins and need different jumper settings. Which CPU type do you have on the SMT486 Terminator board?

Cheers, Jan

Chkcpu wrote on 2026-01-09, 17:36:

Ciao Elio, […]
Show full quote

Ciao Elio,

As you probably know, there are 2 different AMD 486DX4-100 CPU types. The Am486DX4-100 NV8T and the Am486DX4-100 SV8B.
These AMD types have different functions on several pins and need different jumper settings. Which CPU type do you have on the SMT486 Terminator board?

Cheers, Jan

Ciao Jan
Thank you very much for the information, so on the SMT I have an AMD Am486 DX4-100, A80486DX4-100NV8T, 3V that would be slightly lower than the minimum selectable voltage (3.3V), I followed the instructions in the manual on TRW, and I hope I have not made any mistakes.
Aesthetically there are no obvious problems, except for some oxidations, present in both ALi chips and Cache L2 chips, but I think they were caused by washing in water, without drying correctly.
I would really like to make it work, if I want I have other CPUs, that I can try, maybe you could try it with some 486 5V, and maybe I would also try other RAM, currently the keyboard LEDs are flashing, and the LED reset too, it would seem that the ok is missing to start the post codes, maybe you just need to do other tests, and use some deoxidizing spray, at the limit I can also try an ISA video card, if maybe you can see some difference.
Greetings

PC@LIVE wrote on 2026-01-09, 17:53:

Ciao Jan Thank you very much for the information, so on the SMT I have an AMD Am486 DX4-100, A80486DX4-100NV8T, 3V that would be […]
Show full quote

Chkcpu wrote on 2026-01-09, 17:36:

Ciao Elio, […]
Show full quote

Ciao Elio,

As you probably know, there are 2 different AMD 486DX4-100 CPU types. The Am486DX4-100 NV8T and the Am486DX4-100 SV8B.
These AMD types have different functions on several pins and need different jumper settings. Which CPU type do you have on the SMT486 Terminator board?

Cheers, Jan

Ciao Jan
Thank you very much for the information, so on the SMT I have an AMD Am486 DX4-100, A80486DX4-100NV8T, 3V that would be slightly lower than the minimum selectable voltage (3.3V), I followed the instructions in the manual on TRW, and I hope I have not made any mistakes.
Aesthetically there are no obvious problems, except for some oxidations, present in both ALi chips and Cache L2 chips, but I think they were caused by washing in water, without drying correctly.
I would really like to make it work, if I want I have other CPUs, that I can try, maybe you could try it with some 486 5V, and maybe I would also try other RAM, currently the keyboard LEDs are flashing, and the LED reset too, it would seem that the ok is missing to start the post codes, maybe you just need to do other tests, and use some deoxidizing spray, at the limit I can also try an ISA video card, if maybe you can see some difference.
Greetings

Ciao Elio,

Okay, you have the Standard (Non-Enhanced) version of the Am486DX4-100.
This is actually a 3.3V CPU, so setting the JP20 jumper at 9-10 is correct.

Looking at the SMT486 manual from TRW, I see that this AMD CPU type needs the first row of jumper settings on page 7.
So set:
JP5: Open
JP12: Open
JP13: Open
JP14: 2-3 closed
JP15: 1-2 closed
JP16: Open
JP17: 1-2 closed
JP18: 2-3 closed
JP19: Open
JP21: Open
JP27: Closed

In addition, set the FSB jumpers for 33MHz:
JP8: Closed
JP9: Closed
JP10: 1-2 closed
JP11: Open

Hope this helps.
Greetings, Jan

Chkcpu wrote on 2026-01-09, 20:04:

Ciao Elio, […]
Show full quote

PC@LIVE wrote on 2026-01-09, 17:53:

Ciao Jan Thank you very much for the information, so on the SMT I have an AMD Am486 DX4-100, A80486DX4-100NV8T, 3V that would be […]
Show full quote

Chkcpu wrote on 2026-01-09, 17:36:

Ciao Elio, […]
Show full quote

Ciao Elio,

As you probably know, there are 2 different AMD 486DX4-100 CPU types. The Am486DX4-100 NV8T and the Am486DX4-100 SV8B.
These AMD types have different functions on several pins and need different jumper settings. Which CPU type do you have on the SMT486 Terminator board?

Cheers, Jan

Ciao Jan
Thank you very much for the information, so on the SMT I have an AMD Am486 DX4-100, A80486DX4-100NV8T, 3V that would be slightly lower than the minimum selectable voltage (3.3V), I followed the instructions in the manual on TRW, and I hope I have not made any mistakes.
Aesthetically there are no obvious problems, except for some oxidations, present in both ALi chips and Cache L2 chips, but I think they were caused by washing in water, without drying correctly.
I would really like to make it work, if I want I have other CPUs, that I can try, maybe you could try it with some 486 5V, and maybe I would also try other RAM, currently the keyboard LEDs are flashing, and the LED reset too, it would seem that the ok is missing to start the post codes, maybe you just need to do other tests, and use some deoxidizing spray, at the limit I can also try an ISA video card, if maybe you can see some difference.
Greetings

Ciao Elio,

Okay, you have the Standard (Non-Enhanced) version of the Am486DX4-100.
This is actually a 3.3V CPU, so setting the JP20 jumper at 9-10 is correct.

Looking at the SMT486 manual from TRW, I see that this AMD CPU type needs the first row of jumper settings on page 7.
So set:
JP5: Open
JP12: Open
JP13: Open
JP14: 2-3 closed
JP15: 1-2 closed
JP16: Open
JP17: 1-2 closed
JP18: 2-3 closed
JP19: Open
JP21: Open
JP27: Closed

In addition, set the FSB jumpers for 33MHz:
JP8: Closed
JP9: Closed
JP10: 1-2 closed
JP11: Open

Hope this helps.
Greetings, Jan

Ciao Jan,
Thank you very much for the jumper settings, following your instructions, I am quite optimistic, because I found four CPU jumpers to correct, those of the FSB instead are ok.
I won't make it in time for tonight, but tomorrow morning I can work on it to fix the four jumpers, which would be JP5 JP12 JP16 and JP17, maybe I got confused, and I set those jumpers for the other type of DX4-100, however it is very likely that those errors made it impossible to start the motherboard, we'll see but there are good chances that it will work.
If I manage to make it work, it will be very similar to my first DX4-100, although the motherboard is slightly different, in that I had a BEK P405, which had an extra PCI and RAM slot. Unfortunately that card together with the PC I don't know what happened to it, it was probably thrown away, at least that's what I was told.
In case it works, I already have a 1 MB Miro S3 PCI video card, some 16 MB RAM and various PCI sound cards, then if I want I should try to do some bench, including those of Phil's.
Greetings

I corrected four jumpers on the SMT 486 Terminator, some were closed instead of open, and now I would say that they are all in the right position, so if there are no problems due to anything else, it should work!?

I'll try again with the AMD 486DX4-100 CPU, it's a normal DX4, I'd say the same as the ones that came out first, in short, it's not a later version with additions or corrections (possible), I guess it's working, because I haven't had the chance to try it, but eventually I could if necessary, usually anyway these CPUs are working, of course an unfortunate example can happen, but I think it can work, later if I don't get progress, I'll do some tests with other CPUs, used in other PCs, definitely working.

For RAM I use a 4 MB (70ns) SIMM 72 PIN, it works great on the PT-2068.1 PINE, I removed it because I changed SIMM to a 16 MB one.

Super tested video card, the usual PCI that I use on all PCs, old and new works very well, I don't remember which PC it was installed in, however here it is only passing through, for this 486 PCI, as I already said I have a 1 MB PCI S3 video card from Miro.

After connecting everything you need, there is the audio cable of the post card broken, unfortunately if it beeps it will not be possible to listen to them, boot and nothing — —, I do some tests also changing the position to the RAM, and you can see some code, there is a minimum of vital activity, one I photographed, but other codes appeared in other startup attempts, even if after viewing codes, at the next startup attempt, it showed nothing.

From what I've seen, the problem could be contact, probably the pins of the RAM banks, they are very dirty and oxidized, I don't know what material they use to build them, maybe it's the same as the pins of the chips, I haven't done anything for now, but I would like to try to revive them, I think there may be one or more problematic pins, a look at the microscope should be taken, I don't know if it's a good idea to scratch them (?).

For the moment I have to stop here, but it seems that there is a minimal possibility of functioning, I will see to make some further attempts, and maybe with the right combination, I could be able to complete the start, for the moment I would say that I am very far from the finish line.

It seemed to be going better, and there had been a decisive step forward, on more than one occasion I saw post codes, which however stopped in a short time, analyzing what they correspond to, well there is little about it in the manual, and if you find any indication, you have to consider that it may not be reliable, in short, the ISA PCI post card helps a lot, the manual I would say little.

I am aware that the conditions of the motherboard are not terrible, but it makes many doubts about the contacts, in short there may be a hw failure, but the very dark welds and the pins of the RAM and the sockets, also alternate blackened parts, makes me think that without an adequate treatment, which renews the contacts, I don't think we can take any progress.

Then maybe it can be anything, and those suspicious contacts can be fine, in short there is not much else to do, since I have already done several tests, let's see the main ones:

Try other RAMs, including a 60ns one, in case the 70ns one is too slow, and creates further problems, I end up trying an EDO RAM, in case the normal ones for some reason are not well accepted.

CPU change, always AMD but 486DX2-66 5V, this one is definitely working, it was on my first DX2-66, so working works, first I have to change one of the jumpers, and then I have to change the voltage jumpers for +5V, despite this, nothing on the screen and post card always stopped on — —.

I try to start without RAM, to see if you hear the beeps but nothing, no beep, the audio cable has been fixed, so even this attempt was useless.

I try the ISA video card instead of the PCI one, and nothing! I tried several times to move and press on the chips, and nothing!

So, if we were in the early 2000s, I could easily find another motherboard, unfortunately they have been difficult to find for several years and above all they are not very cheap.

I would rule out, a BIOS problem in any case, because if it has displayed some codes, it is a sign that it works, definitely the initial part, and the same I could say for the DX4-100 CPU, but that one to be on the safe side, I can try it on another PC.

Often facing oxidation problems, it happens very often that the problem can only be solved on one side, I'll explain better:

If I have an oxidized RAM bench, I can hardly clean the contacts of the bench, while it is quite simple to clean those of the RAM, so clean the RAM well, most of the time it may not be enough, because the slot remains dirty, and the contact even if better, will not be enough to run the RAM, and consequently the PC.

Yes, you could use the IPA, but you can spray a little and hope that it is enough, sometimes I have seen the RAM inserted several times, and it could work, because maybe it has softened and in the contact points, it manages to thin and improve the conduction.

I looked for alternative systems and products, some recommend cleaning with ultrasound, this system is probably used in repair centers, also not to waste too much time cleaning the boards, for a hobbyist with sporadic use, I imagine it is too expensive.

Others recommend the use of professional deoxidizing products for electrical contacts, I think this is a possible solution, the cost of such a spray is not too high, that is, it costs but working it amply repays what was spent, already because many motherboards, despite not having any faulty components, do not start for some non-optimal contact, I believe that a single contact is enough, and it is not necessarily found in the RAM PINs, it could very well be in the CPU socket.

In short, I think not soon, to try this technique, and I would also spray it in the sockets of the BIOS chip, and in those of the L2 and TAG cache chip 🏷, or any removable chip, of course it might work, but I don't know if it works for a long or short time.

Anyway I don't recommend it, even if it would work (?) An unblocking lubricant spray, because it could leave an oily deposit, and over time it could create other problems, which we will gladly avoid, for now therefore, I invite those who follow me not to imitate me and not try these techniques, if I see in the future that it is effective, maybe it should be well evaluated whether to use it, undoubtedly if we use the spray on one thing at a time, we will be able to understand, where there was a contact problem, we can do several tests but, in the end we will know where the problem was, later maybe if equipped enough, the problematic parts could be replaced.