Here I am again, maybe my attempts won't lead anywhere, because probably someone before me has tried to fix it, but I'll make a few attempts, just to see if I can at least figure out what's not working, clearly if it's a chip failure (UMC), I'm not able to repair it, the MB is the 4FLUD-1.0 ISA-OLB, and currently I can't display any post code, except the initial ones (— —), even the LED lights seem to be fine.

I performed the CMOS reset, and nothing has changed, I moved the jumper, for a while, at most a couple of minutes, if by chance they seem few to you, let me know, so I'll try again.

In the meantime I'm preparing the PINs to check, I want to see if I have voltage (right), in the slots in the memories and in the CPU, I guess you can find the +5V almost everywhere.

Before doing the voltage check, I notice that a trace near the battery, it is put quite badly, it could be interrupted, and it could prevent the PC from starting, I think that before doing other tests, it is to be repaired, and you have to check the close ones, the damage I guess is due to the battery, it must have lost some acid, and it corroded the copper tracks.

It's not an easy point to reach, but maybe it can be repaired by welding tin, after exposing the copper.

This type of work, is quite frequent on the 386 and the first 486, fortunately at some point, CR2032 batteries were used, which do not have this problem.

Before starting the repair of the track (or more), I check the voltages if they are present, first I see that voltages come, from the ATX-AT adapter, there are six wires of different color, black is the ground, the red is the +5V, the white is the -5V, the blue is the -12V, the yellow is the +12V, the gray should be the PG (Power Good), on these I measure +4.91V (Red), -4.91V (White), -11.39V (Blue), +11.99V (+12V), +4.87V (Grey).

Looking at these values, everything is ok, that means that it is powered correctly, I forgot to write, that I tried to hold slightly down some of the biggest chips, in case there is a non-perfect contact, unfortunately nothing changes.

I measured some points at random, and I find the +5V present, for example I measured F1 L1 L4 L5 L6, in all there are just over +4.90V, so far everything is ok

But measuring L2 and L3 I find 0V, this could be a problem, and it could be that damaged trace, to cause the absence of voltage, I have to do some checks, to understand if they are ground or not, if they are it is normal to find 0V.

On the L3 and L2 (0V) measurements, I verified, and the result is as follows:

L2 is directly grounded, so the measured value is correct, so this is fine .

L3 instead is connected to R6, R6 has a common connection with R7 and the minus of (-)C6, the track continues to reach something else, and passes on the back, and maybe even to the intermediate tracks, so it must be fixed, because otherwise the other components remain disconnected from ground.

Initially I thought there wasn't much connected, but looking better, I saw what's left of the track.

Looking at other similar MB, I notice that the IPC chip is SiS (instead of UMC), I don't think it makes any difference, for those who don't know, IPC stands for Integrated Peripherals Controller.

I believe that the problem of not starting this MB 486 (4FLUD-1.0), is caused by corrosion (battery acid leakage), visually there is no big damage, only a track to be redone, and probably still ️ one or two, this is what you see, however, there would be to remove the KB-BIOS chip, under there should be two other chips, and you have to see in what state are the tracks below, maybe if there is a lot of oxide, you have to unsold the socket of the KB-BIOS chip, there are 40 PINs, but if there are not too many difficulties, having free time, it must be done A try.

The result could be, that nothing changes, or that there is some sign of start, in this regard so far, I have tried the card without connecting a keyboard, but you have to connect one, to see if starting the three LEDs flash.

Another problem could be the KB-BIOS chip, that there is a problem of bad contact due to oxide, just clean up the pins well.

For today I am forced to suspend, I will be back shortly, maybe not so much, probably tomorrow.

Back to work a little while ago, on the 486 4FLUD-1.0, I already found some potential problems, I removed both the KB-BIOS and the BIOS (AMI) chip, as I imagined under that KB-BIOS there are two other chips, U1 is an HD74LSOOP, U2 is a DM7406N, both have 14 pins on two rows of 7.

Under the BIOS chip, surprisingly there is a U6, which is a 74F04PC, this one also has 14 pins on two rows.

Apart from the fact that I don't know anything about those chips, and what they are or what they do (?), on U1 I noticed that at least 3 pins, have poor tin solders, this could prevent the contact of at least one pin, and cause a boot problem (???).

The socket for the KB-BIOS chip, is 40 pins, removed the chip you notice a strong oxidation, in the pins on the battery side, maybe instead of wasting time ⏱ cleaning up and removing the oxide, I replace it, so I have to disassemble it to see the tracks below, and at that point we might as well change it with a new (or recovery ???).

The KB-BIOS chip, it needs a treatment, that brings the pins back to new, what I usually do, is to pass a very fine abrasive paper, and it works, I did it in the 386DX40, and it worked!!!

The BIOS chip on the other hand, has a couple of pretty crooked pins, which I have to straighten, and I think I'll clean them up, as I will do with the KB-BIOS chip.

For corroded tracks, at least one is quite damaged, probably broken, two others are in bad shape, but I don't know if there is an interruption somewhere.

Maybe before doing the repairs of the slopes, it is better to clean well, I don't know whether to use alcohol or an invitating for contacts ???

Also today I have to suspend the work, I will be back as soon as possible, with the removal of the 40-pin socket, and the repair of the tracks.

If I might ask, what are you doing with all these mainboards after repair?

Minutemanqvs wrote on 2024-01-01, 17:24:

If I might ask, what are you doing with all these mainboards after repair?

I have a collection of PCs, some complete, others are just the working MB, this one and some are missing pieces, I have another 486, but they are all VLB, this MB is ISA, and I have others to repair ISA and PCI.
After the repair, if it is successful, I do some tests and some benches, if all goes well, I try to complete the configuration, adding sound card, any other cards, or increase the RAM or put a faster CPU.

The repair work of the 486, will continue in the next few days, I have to remove the paint on the damaged (oxidized) tracks, to be able to cover them with tin, so that I can insert any bridges, and ensure continuity along the track, if it is damaged to the point, where the oxide has completely eaten the copper.

On cards 386-486, in the past I simply scratched with very fine paper, but in this 486, I would like to remove the paint using a different method, with a very small brush 🖌, soaked in acetone, I sprinkle the traces to be regenerated (or fixed), the other neighbors I can avoid unpainting, covering them with tape.

At work finished, I will see to do the welds, more than anything else to ensure the continuity of the tracks, this is important, and could cause the PC to boot, if there will be a step forward, I will be able to see it from the ISA-PCI post card.

Speaking of this, I have a problem that I thought I would solve, when I insert the card into the ISA (or PCI) slot, the display is not easy to read (idem the leds), unfortunately I do not have the cable with dual display, and so I thought I would use an ISA Riser, which I had in an old PC, and later I put it in the 5x86 133 P75, because I have the slim case, and the cards more than 7 cm high do not allow the lid to close.

I use the one with ISA slot on both sides, the other riser (green) is not ISA, I don't know what PC you jump out of, and I keep it for any strange motherboards, on which I would put my paws in the future

I'm doing an update on other motherboards, waiting for verification or repair, the first is an S.7 FIC PA2010+, VIA VPX chipset (similar to Intel VX), it has a capacitor to replace plus a (I think) transparent diode, this one is broken in two pieces, I think I can find one in a scrap MB.

The second is an HP AM2 BTX with Athlon 64X2 3800+, on this one everything is quite in order, the LEDs and the power switch are present.

The third is an ACER 8I945AE, the CPU is missing (775), I have another one to try, both are visually fine.
Then there are two ASROCKs, an ALive-NF7G-FullHD AM2 with Athlon 64X2 4400+, and a P4VT8 with P4 1.8A 478, the latter has some capacitors to replace, otherwise it is quite smoked, very dirty with black smoke, maybe you need to give him a nice bath with hot water and soap, or with a cleaning product (?).
For any suggestions on how to clean the MB, I listen to you and thank you.

Tonight I go back to 486 4FLUD-1.0, first of all looking better, I saw that the broken track, it is continuous, it does not connect in the middle of R6 and R7, the drawing I had made is therefore wrong!

I started with the removal of the enamel on the copper, scratching with very fine paper near L3, I saw that up to R7, the copper vanished.

So one track is interrupted, unfortunately it's not the only one, between R6 and R12 there could be another, but it doesn't end there!

On the side of the - BT1, from the socket of the KB-BIOS a track comes out and passes in the back, of this there is only one piece left between the boundaries of BT1 and U3, then there are three others under the plinth, these are not reachable, the pline must be removed, it is to be removed and if possible it should be replaced, some BT1 side pins are very oxidized, I don't think a splash of deoxidizer is enough.

It also facilitates the welding of the bridges, for the restoration of the interrupted connections, and you can make some welds on U1, because they are scarce of tin, when the work is finished, I should get an awakening of the MB, I hope you work again.

I forgot, the two BIOS chips, they were fixed (pin straightening), and I scratched the oxide, passing the pins on the very fine paper, now they are much better, they are not like new, but almost.

I work in parallel with the 486 and the FIC PA2010+, I examined the S.7, I think of replacing the broken diode (soon), and replacing the capacitor (swollen), there are other little problems to solve, the first of the 72 PIN RAM slots , it has a deformed contact, perhaps it doesn't touch the neighboring one, but it needs to be fixed, to avoid "shorts".
I noticed that the BIOS chip had a pin bent on itself, I removed the chip, and fixed all the deformed (crooked and bent) pins, the socket needs cleaning, to improve contact with the chip pins , it is not as beautiful as it should be, the alternative would be to replace it.
I noticed that the motherboard version is 4.1, unfortunately it doesn't have the jumper for 2.1V (for the K6), the minimum there would be 2.8V, perhaps it could go down to 2.5V by removing the jumper, but that's not certain , it could work differently, maybe there are secret settings, to select lower voltages.

I continue with the work on the 486 4FLUD-1.0, I worked on the two biggest tracks, I scratched the surface with a fine paper, until it becomes copper color, well even if apparently they seem interrupted, tried the continuity with the tester on diode test, those tracks of his turn out to be still ️ ok, so a part of the track has remained, since there is continuity.

Different speech instead for the traces under the socket of the KB-BIOS, being very close to the - of the battery, it is possible that the corrosion is greater, also the socket must be deoxidized or replaced, so this is where you have to concentrate the work, of course if I change it, I avoid cleaning it.

Also by removing the plinth, I can redo the welds on U1, there are at least three few of tin, also I could solder two wires, on the + and - of the battery (BT1), so that I can place it away from the circuits, and avoid future corrosion problems.

I started the repair work of the FIC PA2010+, first I found a scrap MB, from which I take a few pieces every now and then, luckily there is the same transparent diode, with the two black bars, I removed it in both MB, and I put the intact one in the FIC, that is, I welded it instead of the broken one, unfortunately it is only the first of the welding jobs, there is to replace the capacitor in the VRM area, but there are other work to be done, for example in the RAM sockets, on SIMM1 there are two very crooked pins, currently not It is also possible to use the second bench (SIMM2) accordingly, maybe I can recover a socket from a MB 486 scrap, it is a complicated job, but at the moment it is not necessary, because the card can also use DIMM, maybe the broken lever of the DIM2, could create problems?

If you also have to change the DIM2, but for the moment I will not even make this change, we will see later if there will be a need to change it, and always on the subject of changes there would be to change the USB ports near the PS/2, one of the two is broken, the last jobs to be done are the cleaning of the contacts, of BAT and BIOS socket, they are far from clean and shiny.

As soon as I finish all these jobs, I would be ready for the start-up test, but maybe there are other points to clean, some points I have already done using alcohol, but there are some points that need to be cleaned up well.

Unfortunately I noticed that two SIMM banks are to be replaced, in practice both SIM1 and SIM4 have broken hooks, in the SIM1 there are a couple of crooked PINs, and I don't think it is worth trying to straighten them, in short, both 72 PIN banks are not usable, in the SIM4 there is a crooked PIN, maybe it would be possible to straighten it, but there is a broken hook, so I would say that for the moment just give it a clean, and leave them like this, in any case I will use the DIMMs, or at least a DIMM.

Looking at the setting of the VRM, it is currently selected for 3.5V, but you can change it to 3.3V, or select lower voltages, there are three positions, for 3.2V (K6-233), for 2.9V (K6 166-200), and 2.8V for P.MMX and 6X86L.

The multi instead is 1.5X, but there are only (alternatively) 2X 2.5X and 3X, in short, at most you can get there with some CPUs at 3.5X (1.5X) or with the K6 CXT at 6X (2X).

Just to check, on the FIC PA2010+, I looked for other possible problems, and looking for them you can find them!

Inside the PCI slots, I found something, two are okay, one has the A1 and B1 pins crooked, the other has one of the pins (I don't know if A1 or B1) also crooked, in short, I don't know how they folded them, but maybe several extractions and card changes had been made, I guess for some problem, probably due to those crooked pins, I don't know what the A1 and B1 pins correspond to, however I imagine that if there is no contact with the card (video for example), it may even not have any screen on screen.

Also these pins, I will keep them so initially, later when I have finished the work, and the card works, I should try to straighten them, or in the worst case, I should also change the two PCI slots, add the two SIMM 72 PIN and that of the DIMM, plus maybe the BIOS socket, there will be a lot of welding work (eye to the eyes !!!).

Now, I'll try to get some indispensable things, like 40-pin and 28-pin chip sockets, I hope to find new ones, I'll try to go to an electronics store, I already know it's likely that they don't have them, and that they need to be ordered.

These sockets, I will need to remove the too oxidized ones, which I have in the 486, here here at home I have other 486 boards, two almost identical (one scrap), and another bought towards the new 96, which is no longer working, actually it only worked a couple of days.

The pair of 486 cards, is VLB that is Vesa Local BUS, one as I said is to pick up components to be reused in the other, the one to be fixed instead, there is damage from acid corrosion battery, various traces are broken (I guess), we will see what can be done, but in any case I have to get the BIOS chip and another missing chip, exactly I don't know what chip it is.

I ordered the 28-pin and 40-pin hookes, for the BIOS chips, unfortunately I forgot to order the 32-pin ones, always for BIOS chips, but I can add them (Monday).

At this point with the 486, I'm waiting for the spare parts, I think they can arrive on the weekend, and so for the moment I'll take care of the other boards, I'll do some repairs on the FIC PA2010+, I'll change the VRM condenser, and I'll try to fix some problems, which I found later.

I observed in detail a couple of points, and I found a broken track and a suspicious one (I don't understand if still ️ integrates?), then there would be a pin of the chipset to be straightened, it's slightly bent, it's a not too complicated job, which in the past I did on other motherboards, it won't come perfect , but it will still be fine.

There would be finally, to straighten the pins of the PCI slot, we will see what can be done, but the important thing is that it does not touch , the adjacent pins.

I finished the work on the FIC PA2010+, the swollen capacitor has been replaced, the original one was of the I.Q. brand, I don't think it means imagine quality, however for the moment I have replaced only the one (swollen), always in that area there are five others identical, plus some bigger and others smaller.

After disassembling it, I checked the values, through an Atlas ESR+, no surprise, the values were beyond the tolerable, in practice it is knockout.

For the replacement, I used a Rubycon YXG 1000uF 6.3V, both the values and the dimensions are identical to those of the I.Q., if I wanted I had a different brand, but I chose the Rubycon, maybe if all goes well, it will be the only one I replace, possibly there are ten others equal, plus five bigger, and others smaller.

Unintentionally I saw that the other chipset also, has some bent pins, fortunately they don't seem to touch, so I'll straighten them, trying to space them apart, so that they don't stay too close.

I hope that this “telling” in pieces, the works performed, does not bore you, and consider that some jobs, I do them by eye 👁, without any magnifier, or particular equipment, this makes it clear why the result can be visually discreet.

After finishing the work of replacing the capacitor, on the FIC PA2010+, they pointed out to me (Thank you very much), the generous dimensions of the heatsinks, on the VRM mosfets, this means that the regulators are of the linear type, and significantly limits the update with faster CPUs, it is usually better not to exceed 200 MHz, then it might not be the case, given that the K6 300 MHz are supported in the booklet, but perhaps it would not be a problem given that they have lower VCORE voltage (2.2V), the REV. the card I have is 4.1, I think one of the first, and the jumper for 2.2V is missing.
In any case, I can use other CPUs, I have some interesting single voltage ones, like the K5 166 or various Cyrix IBM 6X86, IDT C6 200 MHz, then there would also be the Pentiums (not MMX), at least these would be worth trying initially, perhaps the most suitable would be the Cyrix or IBM, certainly the VIA VPX does not shine in terms of performance, in fact it could slow down the bench results a lot, here it will be difficult to obtain scores above the average.