Today I received the cache for the Siemens D969, it would be a cache module, unfortunately I did not receive the module in the photo with three chips, but a similar module with four chips, the latter should come from Fujitsu PC, and could probably be usable, but this depends on the pinout, i.e. if it uses a standard one, it should work, otherwise if it uses a proprietary pinout, it won't work.
I think I have at least one or two different coast modules, but I don't know anything, if they work and for which MB they are usable, it would be interesting if somewhere there was a guide, or an archive of coast modules, and on the motherboards in which to install them .

I just noticed that the three-chip module has the central notch moved compared to the other, so it cannot be installed in the MB one, while the four-chip module has the notch in the right position, having already previously verified that it could install.

Tonight I spent some time ⏱, dusting off all the hidden corners, and not very accessible, the card had a layer of dust, quite high, now it looks nice, it is an ASROCK 775Dual-VSTA, a particular card because it has both an AGP slot and PCI-E 16X, in this way you can use both old video cards and recent cards, but I don't think it is possible to use both.

I don't know why it was decommissioned, but it should be working (so I was told), the CPU should be a single core, but it can be replaced with a dual core CPU, definitely better in performance.

The VIA chipset won't be the best, but it should be reliable, and the use of DDR2 RAM is definitely faster than regular DDR.

Finally, for tonight I conclude, installing the Slot1 CPU in the MB PCChips, I had this Celeron 333 Slot1, which I kept in the collection, and I installed it.

The PC is quite integrated, since the card is a Baby AT, there is no video, you can install a card in the AGP slot, or use a VGA PCI, or even an ISA, initially I will put an Intel 740 AGP card, and later I will decide whether to change it.

Having the socket for the ATX power supply, I will not use AT ATX adapter, while for the RAM I start with a 32 or 64 MB module, unfortunately I do not have the bracket with the audio ports, but I have the same in other motherboards, so I could temporarily use one, disconnecting it from another PC.

Tonight I tried the BXCEL motherboard, a PCChips M726, I put the i740 and a PC 100 RAM bank, I tried the boot and nothing, I check the battery and it was low, I change it with a new one, and nothing changes nothing.

I change RAM and VGA, I put a 64 MB Kingston PC133, and an ISA NT-200B with WD chip I think with 128 KB or 256?, anyway no results, the post card always displays — —, I tried to press on the chips, but nothing yet ️ no life signal.

So, I would say that maybe I should check the voltages of the mosfets, it's quite simple, because I have three, in a position not too difficult to reach, we'll see if I find some strange value, but they could be fine normal voltages, we'll see after the measurements what comes out.

I made a quick measurement, on the three mosfets, I didn't take note of the precise values, but I already have an indication that could explain, because it doesn't start!

On the two VRM mosfets I found, what I think is the voltage of the VCORE, which for a Celeron Slot1 should be about 2V, unfortunately I have a lower value, almost 1.7V on both mosfets, the low value could prevent the CPU and MB from starting, according only to this value, the problem could be nearby, there could be one or more capacitors with ESR beyond the tolerance, and the result could be the lowering of the VCORE voltage.

This possible problem with capacitors, however, is not visible, because they do not have bulges, but the G-Luxon capacitors are certainly not the best qualitatively, so I would first like to do some more tests with other CPUs, and I will check the jumpers, hoping that the problem is only a wrong setting.

I have not yet made ️ changes to the system, but the situation is the one visible in the photo, starting the Led Reset turns on and then turns off, even the keyboard LEDs turn on and off, all this is quite normal, as it should be normal that the +3.3V Led remains off, finally the IRDY Led sometimes flashes and others instead turn on, unfortunately the codes do not start.

For now that's all, later I'll do other tests with different HW, maybe if it's just a jumper or RAM problem or I'll try to use IPA in the Slot1 contacts, in case the contacts were superficially oxidized.

The capacitors can blow out at the bottom on some of those PC Chips boards, so look around the bases for the browned gunk, or if any seem to have the plastic plug pushed out at an angle.

Edit: also their slot 1 sockets need a heck of a lot of cleaning and jiggling to connect sometimes.
EditII: I have also known them to corrupt their BIOSes.

BitWrangler wrote on 2024-03-22, 00:47:

The capacitors can blow out at the bottom on some of those PC Chips boards, so look around the bases for the browned gunk, or if any seem to have the plastic plug pushed out at an angle.

Edit: also their slot 1 sockets need a heck of a lot of cleaning and jiggling to connect sometimes.
EditII: I have also known them to corrupt their BIOSes.

Thanks a lot friend, looking at them they seem fine, both above and below, I'll try to look at some of them under the microscope, maybe we could see if there is a leak, but I thought about the capacitors, because the VCORE is that of the Coppermine almost 1.7V, while it should be around 2.0V, could that 0.3V less be the reason for the failure to boot?
For Slot1, I'll do a cleaning of the slot, and then maybe a splash of IPA, who knows maybe the contacts will liven up.
For the BIOS I thought there might be a problem, but checking the voltages I didn't find them normal, in that case I would have excluded the problem with the capacitors (at least those of the VRM), so not having a suitable programmer, I would have to get a chip already programmed, or an adapter for the USB programmer I usually use.

I was thinking coppermines went from 1.65 to 1.85 so 1.7 sounds okay, wasn't it dechutes that wanted 2.0? ... which come to think of it might be better for testing the board and flashing newest BIOS because I'm not sure they supported later cores with original BIOS.

BitWrangler wrote on 2024-03-22, 20:16:

I was thinking coppermines went from 1.65 to 1.85 so 1.7 sounds okay, wasn't it dechutes that wanted 2.0? ... which come to think of it might be better for testing the board and flashing newest BIOS because I'm not sure they supported later cores with original BIOS.

Yes, it's actually an idea, put a Coppermine with Slot1 adapter, and try if it starts, if it's just a voltage problem, don't let the PC start.
As for the BIOS, updating it would be possible, but I don't have the adapter for the BIOS chip, except that I take an adapter and put a 32 pin chip in it, without reprogramming the chip that is on the MB (DIL 32), so it would be useful a DIL-PLCC adapter.

I thought that if the problem of not starting, on the PCChips BXCEL, is due to the low VCORE, that is, instead of about 2.0V there are less than 1.7V, things are two, I could use an adapter with VCORE regulator on board (via jumper), it remains to be seen though, if the jumpers for the multi X are ignored, or you have to change the setting on the MB, but it is not necessarily that there are multi higher than 8X, although in other MB, I have seen that some multi bass are being remapped.

Alternative to this use a Coppermine Slot1, or via adapter, as long as the BIOS does not create problems, but in the past I have even used a Celeron Tualatin with adapter, and the MB mistakenly identified it as Xeon, so I will try to gather some CPU, and start the tests, hoping to find a combination that works.

Returning to the MB Slot1 BXCEL PCChips M726, I have interesting news, I started by checking the contacts of the CPU, which I had previously cleaned, then I change CPU and switch from Celeron 333 to a PII 400, with the other CPU the VCORE goes up a little about 1.9V, a little low but still better than almost 1.7V, after other useless boot attempts, I try to tilt the CPU a little, to see if something changes, in fact a couple of codes appear, I think 43 and 42, I didn't take a picture, but it made me understand that the problem is of contact with The slots of the Slot1, so I pull out the CPU and start cleaning the pins of the Slot1, I removed a surface layer of blackish oxide, then I try to spray the IPA, I put everything back on and try again, it doesn't go!

I check the voltage of the VCORE, it has risen beyond 2.20V, a little high but acceptable, and then as we will see, maybe the tester is slightly above the real value .

I put the Celeron back, and after a couple of attempts, I see the post codes starting to scroll, a sign that the PC starts up, in fact shortly after there is a double beep and a screen on the screen !!!

I check the VCORE on the Mosfet, it is always above 2.2V, but from the BIOS the indicated value is just above 2.0V, a sign that maybe there is some component that brings the voltage to standard values.

However, for the moment I am more than satisfied with the result, now I just have to connect the usual IDE CF adapter with memory card, and from DOS do some benches and diagnostics, I do not expect results of a certain level, indeed the opposite, but who knows, from what I know is that the ALi chipset, does not shine for speed, but it is quite good in reliability.

Here we are, I just finished running some diagnostics and bench DOS, the results do not show anything exceptional, at least for what I see, then maybe they could also be good for such a PC, undoubtedly the PC apart from a normal slowness, it doesn't look so bad, considering that it is a 333 MHz, in short, it can go very well with the first Windows, maybe with more RAM even 2000 or XP would be usable, but given the age and speed, I would say Windows 95 or 98SE, and if you want something more modern Windows ME, but since it has three RAM banks I would add another 64 or 128 MB, maybe 256 would be more than enough.

I currently have the Intel 740 VGA AGP 8 MB, I think I'll change it with another, maybe I have an ATi Rage and something (I don't remember exactly what ), for the rest the audio is integrated, then there would be the possibility to connect USB and PS/2 ports, with a special card, but if you want it would be simple enough to self build, or you can use a PCI card.

In short, I thought I had to work hard, and that it was not easy to bring her back to life, but now that it works, I must say that I like it enough, and my collection of motherboards and PCs, has been enriched with an additional and particular PCChips card, they will not be the maximum performance, but at the time they were quite cheap.

Already the first update in the program, is to increase the RAM, but it will not be the only one, I do not have many AGP 2X VGAs, even if the 4X could work, I would prefer to use them on a PC with AGP 4X slots, unfortunately being the first version of this Slot, they do not have much RAM on board, and the GPU chips, are often the same as the VGA PCI, except for a few special cases.

For the CPU, being quite overclokable, I will try to change the jumpers of the FSB, so that it goes from 66 to 100 MHz, the result would be a jump from 333 to 500 MHz, but this I can do it after doing more tests, the end result depends on how stable the CPU is at standard VCORE.

Yeah a celeron might go to 500. Cheapest thing worth putting in the AGP slot would be a Rage XL, better would be a TnT card or Rage128, or whatever alternatives you've already got.

BitWrangler wrote on 2024-03-24, 13:56:

Yeah a celeron might go to 500. Cheapest thing worth putting in the AGP slot would be a Rage XL, better would be a TnT card or Rage128, or whatever alternatives you've already got.

Thanks a lot friend, as soon as I finish the RAM update, I think I'll get up to 256 MB, beyond that I don't think it's much use, I'll install a Windows, maybe I'd go to ME, then I'll try the FSB 100, if it starts and Windows doesn't show blue screens , everything should be fine, for the VGAs, I'm waiting for a response on some video cards on sale, there are a couple of AGP 2X ones, plus others that are obviously better, if you don't accept the proposal, I'll look for one among those to be repaired, there should be be an 8 or 16 MB ATi Rage, which needs a change of capacitors, or I recover the VGA of the ASUS MEZ-M, because the MB is still faulty, and therefore the VGA is unused.

Today I don't know if I can upgrade the RAM, I would have decided to add one or two banks, for a maximum total of 256 MB, beyond that amount I don't think it's necessary, except in special cases, like an HD with various Windows, from 95 to XP, but even so if they are enough for XP, for the others they are still enough, I would say.

For Windows it is necessary to connect a mouse 🖱, but here I miss the card with USB and PS/2 ports, I have that particular card in other MB PCChips, but with a certain free time, and with a certain skill in welding, you can self-build, the scheme is quite simple, but it is not taken for granted that the visual result is equal to the original, indeed it would be a handicrafted work, but aesthetics matters little, as long as it works.

I performed a first RAM upgrade, added a second 64 MB PC133 module, certainly the PC100 modules would also be fine, but we will see later, currently for me it is convenient to use these of relatively low capacity, and having used a PC133 module at the beginning, so as not to unbalance the system, I added a second one, so that they do not necessarily go to 100 MHz anyway, but in any case I can stay at 100 MHz with the CL2, while using PC100 modules I would hardly have found CL2 modules, they are usually most of the CL3.

Currently, I have reached the 128MB of total RAM, missing for the 256 MB, a third 128 MB module, which I will choose from the available ones, or I could remove the two 64 MB, and put one of 256 MB, I honestly do not know if it is better one or three modules, I usually use if possible three identical modules but I do not have them, I think maybe I can find two of 128 MB identical, maybe it could be the best solution.

I ran the Speedsys program, just to see what changes, it doesn't seem to me there is a significant difference, but if there is not very visible.

After reaching 256 MB of RAM, I added a third 128 MB module, so now I have 64+64+128 MB, which I think are enough for any old Windows, indeed for some it is even a high amount, but with a multiple installation of various Windows, you can use what we need if necessary.

Initially, I tried another module, but the PC wouldn't boot, however after several attempts, I changed the module and I managed, to boot with all the RAM banks full.

Now you have to try to make the audio work on board, it is currently turned off, but first you have to connect the audio ports, and install an HD, load a Windows and its drivers.

I did again ️ the Speedsys benches, I would say the same results, evidently they do not change with different amount of RAM.