I make a brief jump to a Slot1 motherboard, bought some time ago, and never tried, the card was sold as working, in the images there was a video BIOS screen, so actually it should be working, but I usually don't try them without looking well, and I discovered that:

1) there is a U1 pin that needs to be re-selded, it should be a transport damage, luckily, that the PIN is quite reachable, and therefore welding it should be simple, I don't know what problems it can create, the chip is branded GD75232, and it seems somehow connected to the COM1, so it is not essential for the operation of the MB, but it could be important, for the operation of the COM1 port, which is usually connected to a serial mouse 🖱, in this card there is the possibility of connecting a PS/2 port for the mouse 🖱, but a special bracket provided is needed With the motherboard, which I don't have, I only have the one for the audio and the GAME port.

2) Known of the small 100uF 16V electrolytic capacitors, with a slight swelling, it is barely visible, but in case of problems, they would be the first candidates for replacement, after all the quality of these, is soon clear, the brand is LUXON, I let you imagine the origin and the quality.

3) On the NB chipset, there is no heatsink, I don't know if it was like that, or if it was removed, it may not be necessary the FSB is 66 MHz, but you can also select 100 MHz, at this frequency instead it could be useful, unfortunately it is not possible to install those of other motherboards, here there are no two holes (for the clips), the heatsink should be attached with a double-sided adhesive.

I forgot, the motherboard is (MB-686BXA1) equivalent to a PCCHIPS M725 REV:1.2, AT format for Intel Slot1 CPU, it has an ATX connector (in addition to the AT), there are three SDRAM sockets, and regarding the slots, there is an AGP three PCI and two ISA, you could complain about the lack of some other slot, but there is integrated audio, and beyond the PS/2 port there are a couple of USB ports, on a special card called ATX form card.

This motherboard, so by today will be ready to be tested with an Intel PII 300, initially with a VGA PCI, and later with an AGP, then if there is no operating problem, I should expand the RAM, and find a 2X AGP video card suitable for the period of this PC, which should be from 1998, plus a sufficient RAM supply to make it work at its best.

I finished the welding work of the PIN (U1 chip), of the motherboard the same as the PCCHIPS M725, a Slot1 in AT format with ALi chipset, as I wrote the chances that it works are almost certain, because among the seller's photos there was that of the BIOS screen, of course the doubt could have come if it had been taken from some other ad, but this seemed difficult to me.

Today is the time to try this card, since most of the necessary things, I find them in PC AT (with the ATX, instead I should use some adapters), I unpart the 486 VLB ZIDA Z4DVP from the bench, and connect everything to the M725, of course I will not be able to use the VGA VLB, instead of that I use a PCI SIS, which is fine in any PC, whether more modern or previous, in short it works on both PCI 5V and on the 3.3V ones.

Added a RAM, which I remembered was 16 MB, and a CPU PII 300, started the PC, everything goes without any problem, even if then at the first BIOS screen, it stops to wait for the press of a key, the RAM turned out to be 64 MB, maybe it is (?), however for the moment I'm interested in looking at the BIOS, the main menu of the AMI BIOS, seems quite complete, with many options available, then at a certain point, I decide to exit but instead of doing it by confirmation, choosing to save or not the settings, I decide to reset from the keyboard , then I see some codes, but the final ones are D4 D3, codes that I should have already seen in the past.

Here practically for the moment I'm stopped, because even if I turn it off and on again I always see those codes, and you can't see anything on the screen, I did some tests to change RAM and move the VGA, nothing has changed, I'll wait a bit before restarting, maybe these codes could be memorized, if so I'll do other tests, or one of the attempts, would be to do the CMOS reset, but in case I could remove the BIOS chip, to revive the contacts, or possibly evaluate whether to reprogram it, if I succeed with this card, I'll know what to do on other motherboards that show error codes Identical, I hope it's just an unfortunate case of temporary blocking, and at this point, I would advise in indecisive whether or not to save from BIOS, not to perform keyboard reset, rather I would say to turn off the power supply.

As soon as I finish performing some tests, I start by saying that it was useless to keep it off, the codes shown are always d4 and d3, even checking the paper manual, if you have any doubts, well I really don't think they vanish, sometimes the indications are quite generic, so the doubt remains, is it a HW or SW problem?

If the card had arrived, showing those codes from the beginning, I would say that it would be very likely that there is something wrong with the HW, but having seen the card work and especially after performing a keyboard reset, I would say that the HW is very likely that it has nothing wrong, this could be a lesson for the future, in case with identical codes, we want to try to solve.

I'm starting to think that more than a failure, it's a block, I don't know what exactly, that is to say BIOS is too easy, I don't think it's the CMOS, because the card was tested without a battery, so after a certain period of inactivity the changes would be lost, and the card restarts from the initial factory values. I can't say what the keyboard reset works on, the reset via button connected to the jumper, should bring the BIOS back to the PC's boot codes, some say that pressing it by starting the PC, it can restart stuck cards, but with my cards it didn't work, here I didn't use this technique, but I only tried if it worked when the PC was blocked, the pressure restarted the codes, but then they always stopped on the d4 d3.

At this point I performed some tests by changing HW, after a few attempts, I solved it by changing RAM, maybe having read 64 MB instead of 16 MB (?), may have created some points of no return, however taking advantage of the fact that it works, I ran some diagnostics and Bench (including those of Phil's), to continue the start I simply pressed F2, and I was careful not to do any keyboard reset, the next step would be, expand the RAM and try a VGA AGP 2X, for this last modification, some benches will have to be redone, but the results can improve a little Or nothing, or much, depends on the type of bench, and the use of the GPU by the programs.

Since I solved it with the M725, which displayed only a few post codes, and then stopped on d4 d3, I remembered that I already saw those codes, not too long ago, and trying to remember if it was the Intel D850GB card, I had confirmation that it was just that!

At this point, I think I have excellent possibilities of resolution, with that motherboard, I will do more tests to see if I can unlock it, I would really like it, because I only have an S.423 with RDRAM, and I would like to have another one, even if slightly slower and with less total RAM.

Mainly, I should try various RAM, but I will do other tests trying to repeat the same things, which I did on the M725, undoubtedly it could be almost certain, that there is no HW failure, the problem would be as I believe SW, unfortunately here the BIOS chip is welded, and it is not possible to rewrite it, except that it is removed.

We just have to hope that there is no damage to the SW written in the chip, because otherwise the only solution is rewriting, at the moment it is not possible, because as I already wrote, the chip is welded.

Later, I have other cards to work on, including various S.462/A boards, including some ASUS, which I wrote about recently, but I have other boards from other brands, which are waiting to be tried, plus different AM2 motherboards, perhaps less interesting, but still quite dated (today), and with such performance, which can still be used today, if you do not make heavy use, with demanding programs.

I don't know if today I'll be able to try to exhume the Intel D850GB motherboard (Garibaldi), to remember what card it is, it's an S.423 for the first P4 with RDRAM (Rambus), this particular type of memory, it wasn't successful, let's say it was a lost bet (at the time the DDRs were produced), these DDRs despite costing less they were more or less the same, then following the abandonment of RDRAM, Intel adapted its chipsets to the use of traditional memories, first PC1XX with poor results (limited performance), and then with the DDR it went much better, in the meantime it was Baked the new socket 478 for the P4, as far as I know there are no S.423 with DDR RAM, in short I don't know what could have been wrong with the 423, moreover there are 423-478 adapters, and therefore many of those 55 extra pins, would probably be useless or simply doubles of others.

Perhaps, however, the S.478 was designed for future P4 Dual Core CPUs, which were later never produced, or for the support of additional instructions that were not added (?), in short, they could be predispositions to make the S.478 last a long time, even if they then gave up stopping at almost 4 GHz, and the next step was the S.T (LGA 775), which between Core FSB and MHz, created a lot of confusion (at least initially), on which CPU was actually faster, that is, for example a 2 Core 3 GHz could be slower than a 4 Core from 2.5GHz? Still ️ today it's not easy to answer, it could be in some cases yes and in others no.

On the M725 I noticed something, not having the battery, all the settings saved in the BIOS would be lost, in reality it is not like that, the PC is disconnected from the power grid, it is never in ST-BY, after saving through F2, now it turns on and arrives directly to the OS loading, without asking to press F1 or F2, which instead I would have expected it to do, so in practice the battery is only used for the RTC (clock and date), the rest seems to be stored in the BIOS chip, for the rest I mean the PC configuration, RAM Drive etc...

If that's the case, then just a CMOS reset, could reset this information, and restart the BIOS, like the first boot after leaving the factory.

Maybe it's possible that the CMOS reset is not necessary, especially if the motherboard is kept stationary and unused for a long time, I don't know exactly how long it takes, but I've often found cards that started after several years, restart as if it were the first time, that turn on.

In short, when you think you know enough, you realize that in some cases, despite having put so much effort to find a HW problem, you don't have to do anything but try again after a while, and if it starts, it can't be just luck, because now you know that there are some problems not due to HW failures, so try and try another HW, can in the right combination, make the PC unlock, and give you the satisfaction of making it work!!!

On the M725 I did some tests, to improve the current configuration.

The biggest problem I solved, is that of the RAM, from the initial 32 MB, I went to 64 MB, adding another SDRAM of 32 MB, and so far so good.

When I added an Ada 64 MB RAM, the PC did not start, and trying and trying again by removing and putting the RAM, I always saw the same code — — on the post card, for some strange reason, I thought that a RAM a could be defective, or there was a power problem, after several attempts they are always at the same point, so I think that maybe I could try later, after a few minutes but I try again and nothing!!!

I think that maybe there may be a contact problem, and so I try to push on some chips, then after several attempts, I try to tilt the CPU a little, and this has an effect, you see some code, at this point after some removals and reinserts, the PC works again, so I start adding RAM again, until I reach 128 MB, using two SDRAMs of 32 and one of 64 (32+32+64), this is enough for Windows 98SE, if necessary you can remove one or both of those from 32, and change them with others from 64, in total We will be at 160 or 192 MB, then beyond this it is possible to go up, but for a PC like this, I don't know how useful it can be to have more than 128 MB, or have more than 256 MB.

A test instead that from what I've seen, it went wrong, it's the change of video card, try a couple of VGA AGP, a Matrox G200 and a TNT2 M64, but nothing on the screen, it crashes on the codes visible in the photo, exactly I don't know if those cards are working, I think I've tried them in the past, but after a long time that I keep them aside, it might be possible that they don't work anymore, I don't think it's the case to try other cards, rather I should simply remove a card from a similar PC, and see if the Matrox G200 and the TNT2 M64 work (?), at that point I would be quite calm to try the other VGA on the M725, if this one doesn't display anything either, ai ai ai the AGP has something wrong!

Of course Slot1, although it is interesting as a solution, because through adapters it can be used with different CPUs, it is sometimes problematic with contacts, it is not uncommon that it is not optimal, and that the card is unusable, some then strangely only work with some CPUs, there is probably a minimum of contact surface, and if insufficient it does not allow the BIOS codes to start, but something like this can happen even on normal sockets, even there it takes little to work or not, a recent example is the PINE PT2068.1 that only works with the 486SX, the other DX CPUs Try, they don't show codes, or they show some but not enough to start.

On the M725 I made some adjustments, to improve the configuration, first of all from a Phil's bench, I noticed that both SDRAM PC66 and PC100 were present, and this can be fine if you only use CPU with FSB 66, but it could create various problems or even the failure to start, if changing CPU, we use one with FSB 100 .

I identified the slowest module, because in the chips it was indicated 10, instead of 8, I removed it and ran the SpeedSYS 4.70 program (it is located in Phil's Bench), the memory was now all PC100, in total there are 96 MB, for the moment it's fine.

Checking between the 32 MB RAM, I see that most are all with indicated 10, I assume that therefore a good part of these, if not all are PC66, then there are also others indicated as 9 or 7.5, it is possible that these can be PC100, the 7.5 definitely, the 9 maybe .

I searched among the other PC100 RAMs, even among the more spacious ones, these are all quite fast, and often there is the word PC100, however after evaluating various modules, I decide to use a 32 MB one, in the future if I need to increase the total beyond 128 MB, I can remove the 32 MB ones, and replace them with others of greater capacity, I don't think it will happen, but I could change my mind, if after loading Windows 98SE, I noticed that more memory was needed.

So, I brought the total RAM back to 128 MB, with three PC100 SDRAMs, occupying all three available desks, this should not lead to a higher memory speed, it did not yet exist ️ single channel or dual channel, then anyway considering the CPU speed of only 300 MHz, I don't know how much a faster RAM can do, even if you could switch to FSB 100, lowering the multi to 3X, and this could perhaps speed up the PC, in all processing.

Chapter VGA AGP, here I previously tried a couple of AGP video cards, with disappointing results, they did not display anything on the screen, to avoid further evidence that can confuse ideas, I removed a working VGA AGP from another PCCHIPS Slot1 motherboard, now I don't remember the model, it's an Intel i740 with 8 MB of RAM, the card later I remembered that it was present in the first PII-300 I had, and this M725 looks a lot like it, apart from this everything is fine

The card works very well, and also the AGP of the motherboard, I definitely removed the VGA PCI SiS, and so all that remains is to add a drive, where you can load Windows 98SE, a typical disk of that period was for example a 4.3 GB, or something similar, today having old disks in good condition, they are the ideal solution, but there is always the alternative IDE CF adapter with 4 or 8 GB CF memory card.

Arrived today, with a certain delay, due to the strike of the shippers, the batch of three motherboards, would be these:

ASUS K8U-X REV. 1.02 S.754

ASUS M2V-TVM/V S.AM2

INTEL D845GEBV2 S.478

Apart from a certain layer of dust, which I brushed away, the boards seem to be in excellent condition, in two you can see the swelling of the capacitors near the CPU, so I think that after the replacement, both the S.754 and the S.478, they should work again, or if they work like that way, they should not give errors or blue screens, while running programs or when starting the PC.

As for the S.AM2, instead nothing strange is seen, here are the things, it works but has a relatively slow CPU, and has been replaced for this, or it has a failure somewhere, and I would not exclude the electrolytic capacitors, even if visually fine, there could be some or even one, which are out of tolerance, translated they should be replaced with others of better quality.

This is what you see, then we know very well, that it will be the tools to tell us if the fault is not visible, from my experience, the two boards with bulging electrolytic capacitors, are easily repairable, in both cases I will replace the swollen ones, but I do not rule out being able to replace others, of the same capacity and voltage, which are nearby, this only to prevent future failures, but sometimes it has happened that even some of the good ones (visually), were out of tolerance, even if little.

All three cards have CPUs installed, in the S.754 you can see which one it is, AMD Sempron 3000+, from memory I remember that it goes to about 1800 MHz, all in all not very absolute, we are about half of the fastest CPUs (MHz), but it seems quite efficient, arriving at performance similar to a 3000, in short for those who want a PC with a not very common Socket, the 754 compared to the 939 is less widespread, but there is a certain disadvantage, the CPUs that can be used as upgrades are all single core, and for certain uses it could be a little slow, or a lot Slower, it depends on what we would like him to do.

On the S.754, I saw that the PS/2 keyboard and mouse ports are missing🖱, luckily the work was done quite well, so I should recover the ports from a scrap MB, and re-solder them on the K8U-X, but this initially will not be necessary, I will use the USB ports, only later, if the card works again, I would put the PS/2 ports back.

The Intel motherboard, has a second generation S.478, so FSB 533, it could be possible to go up further, but usually Intel boards do not have this possibility, the BIOS should be of simplified type, and to see there are no jumpers to change FSB.

On the AM2 card, in addition to the audio and the integrated LAN, there is also the integrated video, but if it is not fast enough, you can add a PCI-E 16X video card, or alternatively also a less common PCI-E 1X, apart from this there is the shield included, useful for installing the MB in a case.

I'm finalizing the configuration of the PCCHIPS M725, I haven't updated anything, but I've added a couple of PCI cards, which I won't currently use, but in the future they can be very useful, either for transferring files or folders, or for direct connection to a server.

Let's see what I added:

A PCI LAN card Realtek RTL8029 10Mb chip, in this card dated 1997, so just before the MB, there is a socket for DIP28 chip, usually I saw used in that socket, a BIOS XT-IDE chip, in reality it was used on ISA cards, but I think there is no problem, if inserted in this PCI card.

Another peculiarity is the presence of two inputs or outputs, one is the traditional RJ45, the other is the one for the coaxial cable, widely used in the networks of the 90s, and later this port for coaxial cable, was eliminated on the cards of the 2000s, where the RJ45 port is still used ️ today.

The second one I added is a modem card, PCI Conexant HSF 56K HSFi (Churchill), I think it can also be used as a FAX, as well as a modem, obviously internal, the card in fact is equipped with a buzzer, and two ports to connect LINE and PHONE.

For this I don't know if it's worth adding it, of course if you want to connect your PC to a telephone line, finding a server to call, you could connect to the internet, or rather you would connect to the internet, today I'm not saying that it's impossible, but it's quite difficult, but it could very well be used to send and receive faxes, between distant offices, if the PCs are not connected to the Internet.

Today's news, I finished repairing the ASUS K8U-X motherboard, the work to be done was quite simple, but not always everything goes smoothly, in short, at least this time the problems have been solved.

The most obvious one was, the group of 4 swollen electrolytic capacitors, all 1500uF 6.3V, and these were replaced by Nichicons, I don't know if they are original, however not having any others I used those, always 1500uF and 6.3V, same dimensions (diameter and height), so I start with the removal.

Once removed, I checked the ESR values, the OHm 🕉 are between 2 and 3, quite high I would say, the capacity is not measurable, the tester warns that they are "In Circuit/Leaky", so in short they are KO.

I proceed with the replacement, being careful, the white part printed on the motherboard, indicates the positive (usually the negative), this can sometimes create confusion, especially if we remove them all, but I usually take photos before and after the replacement, so as to notice if there is an error.

A fifth electrolytic capacitor of 1500uF 6.3V, is in a nearby area, along with others of 1000uF 16V, but this one is visually fine, I have not replaced it and I think (and I hope) it is not necessary, I should have the answer when I try to start it.

The last job to do is to free a couple of holes from the pond, PS/2 keyboard and mouse sockets 🖱, I had to make some attempts using different techniques, but then I released them, even if for the moment I have not removed the sockets from a scrap MB, I will do it if the K8U-X works, even if you wanted you could use the USB instead of the PS/2.

I seemed very close to getting a video BIOS screen, with the ASUS K8U-X and Sempron 3000+, the BIOS of the MB is AMI, the Version maybe 0703 (it says on the chip), I made some attempts, and with each next step it seemed to go better and better, but at a certain point, all the progress made vanished, it came back to freeze almost immediately on the post codes d4 b7, but let's see what happened.

Initially no code on the post card, I decide to make some attempts including the CMOS reset, then after several attempts I start without RAM, and you can hear three repeated beeps, sign that there is a functionality and it signals a RAM problem, and at the restart, it marks b7 d3 (preceded by C5 C4 d1 13 d0), with a DDR 400 ram it marks E0 d6, I add the PCI VGA and the codes become d5 d4, later also dA d9, after tests with other RAM I arrive at C7 FF, on two or three times beep and you see the VGA BIOS screen, the main one In DOS, on the other hand, you can see some fragments of the characters, but you can't read anything.

So I was so close to resurrecting it, unfortunately then every other startup attempt, it didn't get that close, at most I can see d4 b7 (preceded by d3 C5 C4 d1 13 d0), few codes compared to before.

For the curiosity of those who want to see the sequence of the post codes, I add the previous ones to C7 FF, and these are:

C7 FF C4 C2 C1 C0 dA d9 d8 d7 d6 d5 d4 b7 d3 C5 C4 d1 13 d0.

I don't think it's important for the operation, but I performed these tests without the PS/2 keyboard and mouse ports, and without connecting USB keyboard and mouse, at the limit I can try by connecting a USB keyboard, but I don't think the absence prevents the PC from starting, regarding the battery, it was empty and I changed it with one that should be ok, but I saw that if the codes are not present it displays them, and nothing should change.

Let's say that for now, I didn't get the result I hoped for, so I don't know whether to insist, trying something else, or change the motherboard, switching to another 754 board, here I think that for the moment I don't know if I can make progress, I was quite confident about recovering this motherboard.

However, I don't think all the RAMs tested are defective, I don't rule out that some may not work, but having tried different modules both in capacity and number of chips, I think that among these there could be the right one, I haven't tried modules lower than DDR 400, and maybe they could also work, but at this point I have temporarily exhausted the available time, and I have to postpone to the next few days, further tests.

After trying in vain to change RAM on the ASUS K8U-X, I decide to try DDR 333 RAM and later DDR 266, even here there is no significant change, the codes always remain the same (d4 b7), here I start to think that the problem is not the RAM, but it could be another.

At the moment I exclude a BIOS problem, and therefore the only thing to try, is an S.754 CPU, which has an almost direct connection to the RAM, the RAM controller is integrated, and communicates at the frequency of 800 MHz, I look among those in the collection, and find a Sempron 2600+, I replace it with 3000+, and try the boot again.

Initially I don't see the same post codes, but after a few attempts, a video BIOS screen appears, you can clearly read that there is an error, and that it starts the Recovery BIOS from floppy, but this is not possible because there is no floppy connected!

At this point, I could really be close to the solution, I have to remove the BIOS chip, insert it into the programmer, and after downloading a later version, rewrite the chip and reinstall it, and try again if it works!?!

In the past it happened with other motherboards the same thing, maybe the content of the PC configuration, it is recorded in the BIOS, and maybe the numerous attempts with different HW, may have caused the Boot Block error, all in all it is not an unsolvable problem, indeed since the BIOS chip is not welded, it is relatively simple to solve, of course, however we will be able to see if the problem is solved, only after reprogramming, but it is almost certain that by updating the BIOS (or rewriting it), everything should be back in place.

Update on the ASUS K8U-X, I removed the BIOS chip from the motherboard, and I reprogrammed it with version 1001, previously there was supposed to be 0703, actually I don't know, because I never saw the BIOS screen on video, where at the bottom is the information line of the AMI BIOS.

The reprogramming was carried out successfully, and it was verified by the program, in short, everything is fine

At this point all that remains is to put the chip back on the MB, and try the boot again, this usually solves all the problems and the card should start, as when it was new!

In reality, things went differently, almost nothing has changed, the motherboard continues to stop on the usual codes, or passes ️ a couple more, but it doesn't continue and doesn't get to display anything on the screen.

At this point after retrying various RAM and even an AGP video card, I decided to dedicate my free time on other 754 motherboards, reusing CPU heatsink and as much as possible for a boot test.

The only thing that might still need to be tried is another ATX power supply, this one has the +12V that measures about 11.64V, I don't think it can be a problem, but you can't know, a test would remove all possible doubts.

Today I could try another 754 motherboard, because with the ASUS K8U-X I can't reach the entire sequence of post codes, which shows the BIOS screen on video, and allows you to enter and change the BIOS settings, or more importantly, booting from disk and loading the OS, which currently would be to figure out which Windows (?).

The problem seems to be caused by the RAM, although I have tried a lot, sometimes it seems close to startup, the post codes stop just before they can display something, I still believe that almost all the RAMs are working, and the CPU is, because it comes from another PC updated to Athlon64, I performed a visual check, of the RAM pins, and I didn't see anything suspicious, so either there is a problem in the tabs of the desks, those that connect the individual PINs to the DDR RAM, or if these are also fine, I would say that excluding a problem with the Socket 754, the only thing I could suspect, are those 820uF 6.3V capacitors visually in good condition, in reality I don't know what conditions they are in, and it wouldn't be a surprise if one or more were out of tolerance, if these are also not problematic, then for me it becomes a bit too complex to solve, I would need a professional equipment that I don't have, that RAM area is connected directly to the PINs of the Socket 754, and it is also relatively easy to map each PIN of the RAM bench, but maybe some tracks are directed to the CPU without any SMD (?), and these clearly go Check only if they show strange or absent voltage values

Stil some controls on the ASUS K8U-X, the electrolytic capacitors around the RAM sockets, are quite decent like ESR, the capacity instead, it is not possible to measure it on board, that's if there is a problem, I would say that it is that, that is, capacity out of tolerance, but it could be quite normal, and so for now I'm suspending the work, I don't want to waste time replacing them, if I'm not sure there's a problem.

Usually, as years go by, electrolytic capacitors can swell, and this is the sign that they needed to be replaced, even if initially there was no swelling, then having many spare parts, you could make a replacement of all the capacitors, but often it is enough to change those in the VRM or chipset area, they are probably the most stressed, in short, it is almost automatic that they need to be changed, if the quality of the capacitors is not the best.

The next 754 board, to try is an MSI with red PCB, quite lively and quite interesting if the ultimate goal is to create a PC with a single-color theme, that is, with a lot of red, it should not be difficult to find ATi and RAM video cards with red PCB.

So I move on to fix, an ASUS motherboard (obviously!!!), uATX model P5KPL-AM EPU REV. 1.03 G (LGA 775), there are two crooked pins, which I fix with tweezers, the result is quite close to how they were originally, maybe there is a slight displacement, but I think it could be fine.

This card, accepts almost all 775 CPUs, and from what I have seen, also the modified 771, in fact I find some Xeons among the CPUs, but there are also Dual Core and even Quad Core, the complete list can be found on TRW, or on the ASUS website.

I think of using a CPU Core 2 Duo, or a Pentium Dual Core, while the RAM will initially be 1 or 2 GB, then if necessary I will go up to 3 or 4, more is not possible, because it has only two DDR2 banks instead of the usual four.

Before preparing the ASUS P5KPL-AM EPU, I have time to do a boot test with a Gigabyte GA-K8NF9 Ultra, an S.939 card with PCI-E instead of the AGP, dual BIOS chip, even if both welded, but still useful in case of problems with the BIOS, in short, overall a nice card, which deserves an Athlon 64 X2 CPU or even an Opteron X2, in short, something better than normal Sempron and Athlon 64 CPUs.

The card had four swollen electrolytic capacitors, and I had replaced them some time ago, and it was waiting to be tested, today it's her turn.

I complete the configuration with CPU RAM and VGA PCI, later I will add more RAM, and I will change the video card with a PCI-E 16X GF.

The CPU is AMD Athlon 64 3200+, a 2000 MHz single core, which I initially had on the Internet PC 939, later it was replaced by an Athlon 64 X2 3800+, so this one is definitely working.

For the RAM I have four 512 MB DDR 400 from Corsair, to try the PC I use two (so 1 GB), I will add the other two later, reaching a total of 2 GB, the maximum would be 4, but there is not at least for the moment, the need to have more RAM, it would be a possible upgrade for the future, together with a change of CPU, from single to dual core.

I connect everything I need, and get ready to start, I would be quite confident, and I expect to see several post codes scroll, and finally a BIOS screen to video.

Unfortunately it didn't go that way, I saw the LEDs of the PCI post card turn on and off instantly, and I saw a hint of fan rotation, it's obvious that there's something short, and the power supply protection intervenes to turn off!

On this problem, I don't have time to investigate, but later I'll see to take some measurements, I'm not very optimistic, because it could be some SMD, and it's not easy to replace them, however I keep it at hand, in case I have time to check it.

The last MB tested, did not give the desired results, here is perhaps in the case of the ASUS K8U-X, maybe there is still ️ some tests to be done, but if the problem is the RAM, for some reason incompatible or not well recognized, I could not do much, at the limit you could disassemble the one from another PC and try it, here if even this is not good, then there is definitely a hw problem, however I would not have mind being able to make it work.

So, now that I have the workstation ready for a motherboard to try, I take the ASUS M2V-TVM/V that was in the lot with the K8U-X, the card is complete, to try it you just need to add a DDR2 RAM, but also a 24 PIN ATX power supply, and the keyboard.

Note that you don't need the video card, because it's integrated (like audio and LAN), but if you want you can install a VGA PCI-E 16X, if we need something more powerful.

Being a uATX format motherboard, there are few expansion slots, two PCI plus a 1X and 16X PCI-E, but since both audio and LAN are integrated, PCI slots could be used for other types of cards, it depends on what we do with this PC.

So let's see, what happens when I try to start, initially I see codes scroll, and then stop on d4 07, they are not strange codes, in the past I have already seen them, but they correspond to something (AMI BIOS), I decide after a few tests to remove the RAM, and the codes displayed are 07 05, in addition to three continuous beeps, in short I decide to change RAM, and I put one without a label 🏷, and I try the start again.

This time I have the BIOS screen on video, and I can read the BIOS version, the CPU what it is, and the RAM how much there is.

The CPU is perhaps not what you could imagine, it is a normal 1800 MHz Sempron 3400+, an excellent starting point for a future upgrade, the RAM is 1 GB 533 MHz, of which 128 MB used by the integrated video card, currently not the time to try to expand it, nor to complete the configuration with the addition of a disk, and Windows loading.

But it's still good that it works, although I don't understand why it was sold in a batch of motherboards to be repaired, maybe there's some problem somewhere? I really hope not!!!

Here is the umpteenth change of motherboard, this time no ASUS (a brand to which I have been a customer since the beginning), this is still an MSI, and here I will try everything I can, CPU and RAM that on the K8U-F I could not see work.

Let's start with the format, uATX is not my favorite, personally I prefer the normal ATX, however here everything is integrated, there are audio video and LAN, there are only two DDR RAM banks, the Socket is 754, for the additional cards, we have an AGP 8X slot (I imagine), three normal PCI and a CNR, the latter was not very successful, and the most common CNR cards, are 56K internal modems, I think that shortly after it was replaced by the PCI-E 1X, but it is also true that USB Wi-Fi has made internal modems obsolete.

In addition to the normal IDE ports (1 and 2), and the one to connect the floppy drive, there are two SATA ports for connecting SATA drives, I think it is useful if the hardware has been recovered from a PC later than this, or in case you do not have an IDE-SATA adapter.

The MSI MS-6741 VER:1 card is in excellent visual condition, I only found a couple of crooked PINs, to be straightened, on the VIA VT6307 chip, I thought there was nothing else to do, in reality there is an electrolytic capacitor to replace between the RAM and the 34 PIN port of the floppy, it is not swollen but there are a couple of holes, and liquid has leaked, the brand is TEAPO (what quality???), from 1000 uF 6.3V, there are about 10 of the same, the others instead especially those of the VRM are Rubycon, and are in excellent visual conditions.

I have a second same motherboard, except in some details, and in the integrated video (absent), but it requires the replacement of the entire row of electrolyte capacitors of the VRM, they are 1500uF 6.3V of the KZG, in total there are eight, even if there are two swollen ones, they all need to be changed, to avoid having to redo the work later.

I hope tomorrow to replace the 1000uF capacitor on the motherboard with integrated video, I noticed that it has a capacitor maybe missing from 1500uF 16V, the other motherboard requires more work, and for the moment I don't have much time available, but later I can try

On the MSI MS-6741 VER:1 board I removed the damaged electrolytic capacitor, a 1000uF 6.3V, from a subsequent check it was not so bad, but since it has damage, also to prevent future problems, I will replace it with another equal in capacity and voltage.

I have both in this and another motherboard, the problem of the holes occluded by hard tin, even if on the one hand it melts, on the other it remains semi-solid, it is not a rare thing and often happens with the holes connected to large common areas, in the end I solved it in one way or another, but here it seems quite complex to free them, because the holes seem a bit delicate to me, I would not want to damage the seats, in short I would consider changing the system, on some videos I have seen the use of the braid and the Flux, while I use only the suction pump, And the tip of the welder on the other.

If you had a professional tool, that is, a welder with an extractor, it is likely that I can be able to free the holes, without having to struggle (in the most complicated cases), but there is a second possibility, to free the holes, the use of the drill and a very small bit, personally I have a set of mini bits, but I have never used this system, it could be a great alternative if we drill gradually, with bits of increasing diameter, but it can be risky, and exaggerating with the diameter of the bit, you can irremediably damage the track.

I have the same problem with the holes, on the Intel D845GEBV2 motherboard, an ATX socket 478 for CPU up to FSB 533, the card only has integrated audio, for the LAN and video you need PCI cards, or AGP for VGA.

In this motherboard, I had three swollen electrolytic capacitors, they are all three in the VRM zone, and they are the largest from 3300uF 6.3V, after removal I checked them, the ESR was around 2 - 3 Ohm 🕉, the capacity instead it was not possible to measure it, a sign that they are unusable, and then it is very likely that in these conditions, the board does not start, or if it does it shows stability problems.

On the MSI MS-6741 VER:1 board after removing the damaged electrolytic capacitor, I had to face the problem of the occluded holes (from the tin) on the +, I tried to use a new system (for me), first of all, I measured the diameter of the pins of the new electrolytic capacitors, the diameter is 0.4 mm, for those of 3300uF, so for this from 1000uF the diameter could be lower, but let's consider it equal to 0.4 mm, the diameter of the hole released from the tin, is about 0.8 mm, this is much more than necessary, however not all have the same diameter, some are Larger if the electrolytic capacitors are larger, others can be smaller in diameter.

So what to do to clear the holes?

The technique that is usually used is to try again other times to dissolve the tin and suck it, in most cases it works, but it can happen that there is some particular case, and you have to heat the surface a little, but if even so you can't?

Well undoubtedly, the use of professional equipment allows us to solve, even if it may not be easy, those like me who do not have expensive and professional tools, can get the same result with alternative systems, one of these is drilling with a mini drill bit for mini, I bought a set for dremmel type drills, the diameter goes from 0.3 mm to 1.2 mm, some of these like the larger ones are not needed, for the holes of the electrolytic capacitors present on the motherboard, I have never tried to free the holes from the tin by drilling.

So I decide to see if using the 0.3 mm one, I can remove the pond from the center of the hole, but I do it manually, because the drill could turn too fast, and I wouldn't want to do damage, I proceed and it doesn't take long, in a short time I can move from one side to the other, then it went to the diameter of 0.4 mm, and later even larger, everything is very simple but being very careful that the hole is perpendicular, in short, everything is fine and also quite simple to do, now that the holes are free I can move on to insert the new electrolytic capacitors, And weld them.

When the work is finished I will show the final result, which is then facilitated, by the relative softness of the pond, and I would say that for the cost of the tips, the result is faster and simpler than traditional methods, I would say that it largely repays the time lost normally, and avoids a lot of anger, in short if done with a certain caution, it is a fairly simple and fast system, but I strongly advise against it, for those who want to use it directly without testing on scrap motherboards.