I took a look at the original box of the ECS NFORCE3-A939, inside there is the manual, a CD with drivers and software, cables for HD and FD and SATA, and of course the motherboard, with (currently) a VGA AGP installed, it is an ATI Radeon 7000, I think it is 64 MB, on the red PCB board it says R7000A Rev 1.0, honestly I don't know the manufacturer of this card, however I think it's not famous, because it doesn't say it.

The motherboard, has a purple PCB color, there are five PCI slots and one AGP 8X, there is integrated 6ch audio, and 10/100 LAN, four USB ports expandable to eight, two SATA and two PATA, one FDD 34 PIN.

For memories, it is possible to install a maximum of two DDR DIMMs up to 2 GB maximum, even in dual channel, with DDR 200 333 and 400, I read like this in a strange table, and it seems that there are particular conditions, at least from what I read in the manual.

Today a batch of three motherboards arrived, a quite particular one, I already have another always socket 370, it comes from a micro PC, and does not have the traditional power connector, here there is a six PIN, plus a socket like that of HD CD etc..., as you can imagine 💭, you need an adapter or the original power supply, the brand of this MB should be "PC100", basically PCChips or ECS or other brands of the group.

The capacitors have to be changed, but otherwise it's pretty tidy, apart from a clean one, I don't see 🙈 nothing strange.

The other two are Socket A 462, one is a Gigabyte, which I find very interesting 🧐, it has the dual BIOS, and aesthetically it is very beautiful 😍 🤩, but it needs a change of electrolytic capacitors, because they are swollen.

The other is an ABIT, this one is also interesting 🧐, but it's the basic model, with only two RAM slots, and a minimal endowment compared 🫡 to the most equipped model, aesthetically it's fine, but I saw 👀 that they lack electrolytic capacitors, around the CPU and maybe others in nearby areas, I should investigate to understand which ones it had, and of course they need to be replaced, for the moment I focus on the other two, if and when I can figure out what's missing, I'll proceed and start working on it.

As for the ABIT KG7 Lite motherboard, I have already tried to investigate, to understand the capacity and voltage of the missing electrolytic capacitors, a couple seem to be special, not the usual capacity of 3300uF, but even 4700uF!, it will be very difficult to find the same spare parts.
The others seem more normal, but I have to understand what capacity it is, maybe some are 2200uF, but there are various sizes, so it is possible that others are of lower capacity.
Overall the board is the cheapest, of three of the same model, if I could add the two missing RAM banks, it should become a KG7, but even if it seems possible, it is not said that by welding them, then they can work, to see I don't think anything else is missing, but if that were the case, I think that all the owners of Lite, with a minimum of skill in welding, would do this transformation in KG7.
Unfortunately 😣 I think the missing electrolytic capacitors were removed, because they were swollen and gave problems, and the removal work, was not done by an expert person, maybe he was not able to free the holes, and he gave up.

Today I looked better at both the ABIT and the Gigabyte, as far as the ABIT is concerned, I saw which ones are missing, and I made a list of the missing ones, I was also able to understand the voltage and capacity for each individual capacitor, I think the big ones of 4700uF 6.3V are missing, I don't think I have any but you never know 👎!!!
From what I see, the fan is missing on the SB (chipset), I will put a fairly similar one, going back to the capacitors, those on the MB are all TEAPO brand, the quality is notoriously below average, and when they fail, they create problems, which are usually solved by changing them, with other quality ones.
Moving on to Gigabyte, in addition to some capacitors to be replaced, I noticed that the ATX connector needs to be changed, the plastic part is missing, I hope 🤞 that after the removal and replacement of capacitors and ATX connector, the board will work again, as I said earlier I like it 😻 a lot.

Since lately I'm also working on stuff like Socket 462, I point out that for a long time, I had no idea how they were, that is, I had seen them, and many preferred them to Intel correspondents, but until a dozen years ago, I had never had one between my paws 🐾, at that time I had the opportunity to be able to withdraw a lot of decommissioned PCs, that was the first and only time, however among those there was a PC with MB PCChips SIS chipsets (practically the same as ECS K7S5A), with Athlon XP 1700, and compared to my P4 NW 2000, it was perhaps even more, later I updated it to XP 2400 and DDR RAM, and occasionally I use it to verify HD or transfer files.
Among the various cards I have to try, I have this LuckyStar, complete with original 📦 box and instruction manual, I would like to try it, just to do some bench, being one of the first generations, I would probably use one of the first Athlon Socket-A 462, I should have both the FSB 200 and the FSB 266, or alternatively the Duron both below 1000 and above.

So, I looked at the Gigabyte GA-7VAXP REV 1.1, everything seemed ok, except for the damaged ATX 20 socket, actually looking better, without using a magnifier, I noticed two missing SMDs, they should both be capacitors, one is clearly written, the other I didn't quite understand where it's written, but maybe the missing one is written next to it, so currently, I should get the following spare parts:
ATX 20 PIN socket
3300uF 6.3V electrolytic capacitors,
1200uF 6.3V electrolytic capacitors,
A pair of SMD capacitors from MB scrap.
Re-welding everything that is missing or broken, there could be a good chance that the motherboard will work again, for this MB I was thinking of an Athlon XP Barton 2800+ (333), maybe 🤔 not the best, but I think it could be more than enough, for a PC like that.

Finally, the last motherboard of the batch of 2 462 + 1 370, I have a particular motherboard between my paws, it comes from a Book PC, it should be an ECS BKi810 or equivalent Amptron, the Version of this is V3.3A, the chipset is as imaginable Intel i810, the card is integrated of everything, or almost everything, because the expansion slots are missing, typical of the i810 i PCI, this means that it is not possible to connect any PCI card, so if you do not have big pretensions, the PC has the essential to be used, at the limit, having the USB ports you can Connect something from there.

As written earlier, I have another same board (I don't know if the same REV), and in that I performed some repairs, but three or four SMDs are missing, in short, for lack of time, I put it aside, but this is in better condition, except for the capacitors around the CPU, there are seven out of eight swellings, and they are G-Luxon 1500uF 6.3V, near the 40 PIN IDE1 connector, there is another G-Luxon 1000uF 6.3V bulging, but in other areas there are others the same, which are visually are ok.

I circled and highlighted, two SMDs welded together, initially I thought the diode was detached from one side, but looking closer it is so factory.

I think to do the repair as soon as possible, it's interesting the support for Cyrix CPU by placing a jumper on 1-2 the FSB is 133, while on 2-3 it's normal (I guess 100 ), even if the MB is natively for the Mendocino, there could be support for the Coppermine, in this case also the P3, but with FSB 100, I don't think those from 133, let's say that a PIII of about 1 GHz, would be enough, but more realistically a 600-750 could be interesting, if it were possible to overclock raising the FSB to 133.

Trying to understand if the MB BKi810, was the same as the one I already had, I found out that it is the previous model, you can see it here:

Test and troubleshoot PC@LIVE motherboards

So in summary, I already had the first version of this MB, now I have the next version, so the first supports only compatible Mendocino or Coppermine CPUs (Modifited), the second exactly I don't know, maybe there could be the native support if Coppermine, and if so, it was, it should even accept the P3 FSB100, I don't think it's possible to go further, but I think that if the BIOS allows some overclocking, you can easily go beyond 100, but it also depends on the CPU used, some are able to work, only doing light overclocking, for example a Celeron 700 I don't think it can work With FSB over 100, while a 600 could work (?)

Anyway, it is not the ideal card for overclocking, much better ASUS ABIT type cards etc..., the CPU I would use in this MB, with the most recent revision, would be a 600-750 MHz P3, possibly with at least two 128 MB RAM each, in total 256 MB or possibly a single module always 256MB, in this case, I could if necessary add more RAM, ultimately it is not the ideal PC to play with it, the integrated i810 video cannot be replaced in any way, also because the PCI slots are missing, perhaps the best thing is to use it, for simple tasks or Connect it to a LAN, but if you really wanted to play it, it could be used with old DOS software or even Windows 9X, clearly nothing too recent.

Looking at the old Vogons pages, there is something really 😨 interesting 🧐!!!

Working on the Soyo SY-5BT, at a certain point it no longer started, and the Reset LED remained on, the important point is precisely this, I don't know why it got stuck (Reset), this prevents the correct start, usually the Led flashes, and then it stays off, in short if it stays on it is practically impossible for it to start.

But the interesting point 🧐 is that there were no HW failures in that MB, in reality there is a missing pin in the cache chip, but by disabling the L2 cache, there is no problem, so I can assume, that the problem is a BIOS block (random?), in short, having other MBs with the reset LED on, it is possible (or very likely) that they can be unlocked and work again(?)

Yes but what to do? You could probably try to rewrite the BIOS, this way it would work, unless the chip is damaged, in which case you need a new chip, or you can try directly with a new reprogrammed chip, and the blocked one you can try to restore the original content.

I will see today to start the work, on the Gigabyte GA-7VAXP, just to make a summary, you have to change four electrolytic capacitors from 3300uF 6.3V, plus one from 1200uF 6.3V, for this last job, I will also change the other one next to it, even if it is visually fine, so in total there would be six to replace, four from 3300 and two from 1200.

But before changing them, two other problems must be fixed, visible in the ATX 20 PIN area, the 20 PIN socket must be changed, and when it is removed, two SMD capacitors (missing) must be re-weld, they have probably been damaged, in an attempt to remove the plastic part, where the power supply is connected.

The difficulty will not be particularly high, for those who work on these cards, on the contrary all in all it could be a simple job, which can be done in a short time, I will take a little longer, but what matters is the final result, if there are no other problems, the card should be working again, and for this card I would have decided to use a Barton 2800 with FSB 333, which should be the maximum FSB speed supported by the MB, even if in overclocking maybe it is possible to go up to 400.

To replace the 20 PIN ATX connector on the Gigabyte, I found on a scrap board, a Biostar K8M800-M7A, which has a big scratch on the back, an ATX connector to be unsoldered, the board is in excellent condition (except for the scratch), and the connector is almost like new!

I'm starting to work on the Biostar, and the attempt to unsolder the 20 PINs succeeds almost halfway, some unfortunately I couldn't get rid of them, there was tin left in the hole 🕳, unfortunately I don't have a solderer that you suck up, but a simple solder with a tip and a manual pump (tin vacuum), so I would say that despite the attempt, at the moment I haven't been successful.

I'll see later to try other methods, with the tools I have, maybe I should heat the area well, before removing the pond, I'll see what I can do, the alternative if it should be impossible, the removal, is to buy a new ATX socket, but I hope it's not necessary.

I do not exclude the return, to some MB that I had tried to repair some time ago, lately I am selecting those that for me are not repairable, especially for lack of adequate tools, after several unsuccessful attempts, you have to move on to something else, but after several years, having the necessary time, you can try again trying something else, to note that accumulating experience, you already know from the start if there is any possibility of succeeding, to understand if you do not start, and after years you try again, it will continue not to start, but if you try to look for a failure or a different problem (from those tried previously tested), maybe Who knows if we can take steps forward, or even solve it!!!

I have a number of motherboards, which have stopped working, or do not give any sign of life, despite the pressure of the PWR-ON switch not starting.

A board that I had started to repair, and that seemed to have good possibilities of operation, it was detected instead still ️ broken, it is an MSI MS6199VA, Slot1 chipset VIA NB VT82C693A, I don't know if it comes from a brand PC (like HP Compaq etc...), after replacing some capacitors and trying to start, the VRM twins were fried, it seems that there is the +5V short, or at least something in that line does not work well.

Q4 and Q5 initialed CEB703AL, now fried must be replaced, but not only those other capacitors including a swollen one, must be changed, but perhaps it would be better to replace all those of the +5V line, the success or not can be visible by checking the +5V line, if trying with the tester, it stops playing, most likely, it could be tried, with some probability of success!!!

So, although I would like to but I can't, at least for now, I mean other cards to try or repair, years ago you could find quite a few at a low price or even as a gift, today it's no longer the case

But at least I have a decent supply, which allows me to continue, and do further experiences with dated PCs (some not too much), I have never cataloged the ones I have, maybe I have more than I imagine, however I have some cards ready, just to try, and others that have visible problems, usually swollen electrolytic capacitors, apart from some strange value, I should have enough (capacitors) to not run out, but there is a type that goes in almost all the boards of the 21st century, they are the 3300uF 6.3V, unless you buy large batches, arranging some cards, You end up having to tidy them up!

So, I should see how many boards those capacitors have to repair, and then decide which ones have priority, the choice will not be easy, because replacing them does not guarantee the operation, even if it is usually the case.

A summary of the next work, I have to try an S.423 with SDRAM MB 845 PRO VER1, with a Pentium 4 CPU installed, I imagine it's a 1500-1700, I hope it's an 1800, but we'll see this if at startup there is a video BIOS screen.

Three other different MB, they need the change of electrolytic capacitors, there is an S.370 QDI Advance 10T, very interesting for the presence of ISA and AMR, even if the VIA chipset is not among my favorites, then an ASROCK 775V88+ with P4 3.00 GHZ, and finally an ASUS P4B266, this is the first with 478 and DDR RAM, very interesting, even if the FSB 400 is not the best, the CPU can be any, I hope there is not a Celeron, but just in case I have a 2.6 or a 2.8 400, to put in.

PC@LIVE wrote on 2024-11-24, 09:42:

I will see today to start the work, on the Gigabyte GA-7VAXP, ...

what a nice looking motherboard!

Ujeen wrote on 2024-11-28, 23:16:

PC@LIVE wrote on 2024-11-24, 09:42:

I will see today to start the work, on the Gigabyte GA-7VAXP, ...

what a nice looking motherboard!

Thank you ☺️ very much my friend, I'll try to work on it this weekend, I hope 🤞 to be able to recover it, and that it works.

For this weekend, I hope to be able to complete the work on the Gigabyte GA-7VAXP, in particular I have to unsolder the pins from the donor card (ATX 20 PIN), and recover two SMD capacitors from another scrap board, then having the spare parts, I have to work on the Gigabyte, replacing the swollen electrolytic capacitors, and the damaged ATX connector, and finally solder the two missing SMD capacitors, that should be all.

When the work is finished, you have to do a boot test, I have not decided whether to initially use a Duron type CPU, or opt for a definitive solution, an Athlon XP 2800+ 333 (I think it's a Barton?), in this regard, there has been a lot of talk and written recently, about the correct installation of the AMD S.462, and how it is possible to damage them, many of these CPUs can rust, simply because the heat sink does not rest well, or for lack of thermal conductive paste, thinking about it in the past it is possible that it has broken some CPU, for the rush to try it, but recently no, I will pay the necessary attention, and I will use the largest possible heatsink, among those available.

A little while ago I started work on the Gigabyte GA-7VAXP, for the moment I removed the pins of the damaged ATX 20 PIN connecter, as you can see, some were burned, and you can see that they melted the plastic, however only after removing them, I thought it was better to understand if they were of a line, or of the mass, unfortunately the only way now is to look at the photos, and from there understand what pins they were.

Unfortunately, due to the rush, and the limited equipment, I could have damaged a pitch, but I can't say for sure, because I have to try to clean the rings from the excess pond, I can already say that it's not a big problem, because I'm on both sides, but in case I should make a bridge, to have continuity between above and below.

The next step will be the removal of a couple of SMD (capacitors) from a scrap MB, these must be soldered before the ATX socket, about this one, I removed with some difficulty, that of the Biostar S.754, and I should reuse it, even if I would prefer to put a new one, for two reasons, one to avoid other problems, and two that in case there is still ️ the same fusion problem, I would be calm that it is not due to the socket.

The work of replacing the bulging electrolytic capacitors, is the last to do, it is not usually too complicated, the result is usually quite decent, the welds may not be like the factory, but in my opinion it is enough that there is continuity, the aesthetics takes a back seat.

I continue the work, on the Gigabyte GA-7VAXP, I removed the capacitors to replace, one even if it seemed ok, I replace it, in total there are six, four from 3300uF and two from 1200uF, the removal went quite well, but a couple of the holes of the 3300 were left full of tin, I couldn't get rid of them, it's something that happens often, maybe the tracks have a large surface, and they cool down instantly? However, despite everything, I managed to create the sufficient passage, to install the new electrolytic capacitors, the quality of the factory ones, is poor, in short, it is what it is, the new ones are Nichicon and Panasonic, they will not be the best, but the quality should be there, unless it is imitations, and in this regard, I would not know how to distinguish them.

The last job I have left is to recover the SMD capacitors, from the scrap MB, but before welding all the replacement components, I noticed a couple of things, the area under the Mosfets around the inductors, in the back of the PCB is burnished, a sign that they must have heated up a lot, for this reason I should first check the power lines, in order to avoid problems, another job to do, it is a nice cleaning of the electrolytic capacitor areas, in one or two you notice stains, obvious sign of a leak, I will make them disappear using a degreasing product.

PC@LIVE wrote on 2024-11-30, 12:40:

Unfortunately, due to the rush, and the limited equipment, I could have damaged a pitch, but I can't say for sure, because I have to try to clean the rings from the excess pond, I can already say that it's not a big problem, because I'm on both sides, but in case I should make a bridge, to have continuity between above and below.

BTG re headers desoldering I recently came across this video. I haven't tried this yet, so can't say if this method works or not 😀
https://www.youtube.com/watch?v=Vou2xlJkuoU&l … RSIVosR&index=1

BTW you got cool ESR meter 😀

Ujeen wrote on 2024-11-30, 22:08:

BTG re headers desoldering I recently came across this video. I haven't tried this yet, so can't say if this method works or not […]
Show full quote

PC@LIVE wrote on 2024-11-30, 12:40:

Unfortunately, due to the rush, and the limited equipment, I could have damaged a pitch, but I can't say for sure, because I have to try to clean the rings from the excess pond, I can already say that it's not a big problem, because I'm on both sides, but in case I should make a bridge, to have continuity between above and below.

BTG re headers desoldering I recently came across this video. I haven't tried this yet, so can't say if this method works or not 😀
https://www.youtube.com/watch?v=Vou2xlJkuoU&l … RSIVosR&index=1

BTW you got cool ESR meter 😀

OK Thank you very much my friend.
I saw the video of your link, the trick of using a copper wire, is absolutely to be tried, then I don't know if it succeeds, because speaking for me, I have a different soldering iron (very cheap), and I have never used the flux, but I will soon order a tube, because I realised, that it is always used to make welds, and as I saw in the video, it could be the ideal solution to clear the holes with the braid.