That´s a quite odd population of the memory slots. In most cases the sticks need to be plugged in quadruple groups side to side with no gaps between them. Try populating the 4 slots on the right in the picture.
You could also try with the 9 cache chips removed to rule out any defective cache chip.

Mel0n wrote on 2022-08-17, 23:54:

First post here but active user on LTT forums (not quite the best place for hardware of this age...) Recently picked up two 486 […]
Show full quote

First post here but active user on LTT forums (not quite the best place for hardware of this age...)
Recently picked up two 486 machines and two CRTs retired from my school, about to be scrapped.
One was a Packard Bell Packmate 27 PB420, of which the PRAM battery had not yet leaked which was fantastic. The other was this one.

Caught my attention because it seems to have sockets for the installed 486 as well as 386 and 387 chips.
Does anyone have any idea what this board is? The chipset in the picture is a little unclear so I will say it is a Uni U4800. 5x 16 bit ISA, 1x 8 bit ISA, 1x eISA.
In any ISA slot a POST test card displays only 3.3v and -12v, no +12 or +5. I swapped in the PSU from the Packard Bell system, it did not work still (and the PB system works fine). This system's PSU worked in the Packard Bell so that is not the issue.
I assume it's a capacitor, but with no electrolytics I'd jump to the blue ones between the ISA slots being dead. Either a hard short, or an open circuit.
Final thing is, I have reason to believe this PC (Micro Flex case badge) was assembled by a small company or a single person. It has a very unprofessional feel. Some of the ISA cards were not screwed in, there was hot glue on every connector (floppy cables, HDD Molex, etc - including ISA slots as you can see in the picture). Had PSU extension cables to plug into the board (as the PSU wires, which btw were from an IBM PC style PSU, were too short to reach), they looked hand soldered and were ziptied together. But on the back of the case it lists the processor options and speeds. "386DX mod: 33-40", same type of thing for 486S and DX.
Can anyone point me in a general direction?
Edit: clipped the Varta batt off, it does not seem to have exploded. It was hot glued down so that's all you see on the board up there. Didn't want to remove glue for fear of ripping traces. It did come with a HDD but I saw no reference to hardware of the board or manufacturer on it.

Sidus 386/486C64L-SID - try to find that board on ultimateretro, looks the same to me!

butjer1010 wrote on 2022-08-18, 06:30:

Sidus 386/486C64L-SID - try to find that board on ultimateretro, looks the same to me!

Indeed!
And they really made Bank0= 1,3,5,7 and Bank1=2,4,6,8.

majestyk wrote on 2022-08-18, 06:17:

That´s a quite odd population of the memory slots. In most cases the sticks need to be plugged in quadruple groups side to side with no gaps between them. Try populating the 4 slots on the right in the picture.
You could also try with the 9 cache chips removed to rule out any defective cache chip.

Yeah, noticed that thing with the memory too... the machine appears never to have been opened from factory, but I cannot be sure.

Sidus 386/486C64L-SID - try to find that board on ultimateretro, looks the same to me!

Thanks, that's definitely it.

Mel0n wrote on 2022-08-17, 23:54:

Didn't want to remove glue for fear of ripping traces.

isopropyl alcohol works great at penetrating between hot glue and whatever its sticking to

rasz_pl wrote on 2022-08-18, 12:10:

Mel0n wrote on 2022-08-17, 23:54:

Didn't want to remove glue for fear of ripping traces.

isopropyl alcohol works great at penetrating between hot glue and whatever its sticking to

Good to know, thanks. I'll try pouring a little on the board.

Have you replaced the battery or connected an external one?
And where´s the RTC on this board? Inside the all in one square chip with the sticker?

But first make sure +12V and +5V are distributed again. There could be some damage under the glue where the battery was or under the power connector...
The picture is extremely blurry when it´s zoomed.

Mel0n wrote on 2022-08-18, 11:59:

Sidus 386/486C64L-SID - try to find that board on ultimateretro, looks the same to me!

Thanks, that's definitely it.

You're welcome. Please, make another photo of this motherboard, and this time, don't make photos with Sony Ericsson K750 😀 It will be more helpful, just to see what could be the problem.

butjer1010 wrote on 2022-08-18, 21:09:

Mel0n wrote on 2022-08-18, 11:59:

Sidus 386/486C64L-SID - try to find that board on ultimateretro, looks the same to me!

Thanks, that's definitely it.

You're welcome. Please, make another photo of this motherboard, and this time, don't make photos with Sony Ericsson K750 😀 It will be more helpful, just to see what could be the problem.

Hah!
I do believe I've found the issue, just testing in circuit a bunch of those little blue caps have really high ESR, or are hard shorted. So I will try and replace those

Mel0n wrote on 2022-08-19, 12:52:

Hah!
I do believe I've found the issue, just testing in circuit a bunch of those little blue caps have really high ESR, or are hard shorted. So I will try and replace those

Don't jump to conclusions too quick: The ESR meters are usually impedance meters that measure (expressed in a simplified way) the sum of the real ESR and the capacitive impedance at 60-100kHz. For big electrolytics, the capacitive impedance is so low, that the displayed value is just the ESR.

If you are talking about the very small blue caps that are scattered all over the board: Those have a very low capacity compared to electrolytic, because they are 100nF ceramic capacitors, and readings around 10 Ohms do not indicate a bad cap.

Also, bad caps do not prevent power to be delivered to the ISA slot. If voltages are missing, it is definitely a bad contact or a broken trace.

On the other hand, usually those 100nF caps are between +5V and GND, and the ESR between those two power rails is supposed to be quite low. Measuring high ESR at a blue cap might indicate that the power delivery to that cap is interrupted and point to a broken trace on the board.

mkarcher wrote on 2022-08-19, 17:34:

Don't jump to conclusions too quick: The ESR meters are usually impedance meters that measure (expressed in a simplified way) th […]
Show full quote

Mel0n wrote on 2022-08-19, 12:52:

Hah!
I do believe I've found the issue, just testing in circuit a bunch of those little blue caps have really high ESR, or are hard shorted. So I will try and replace those

Don't jump to conclusions too quick: The ESR meters are usually impedance meters that measure (expressed in a simplified way) the sum of the real ESR and the capacitive impedance at 60-100kHz. For big electrolytics, the capacitive impedance is so low, that the displayed value is just the ESR.

If you are talking about the very small blue caps that are scattered all over the board: Those have a very low capacity compared to electrolytic, because they are 100nF ceramic capacitors, and readings around 10 Ohms do not indicate a bad cap.

Also, bad caps do not prevent power to be delivered to the ISA slot. If voltages are missing, it is definitely a bad contact or a broken trace.

On the other hand, usually those 100nF caps are between +5V and GND, and the ESR between those two power rails is supposed to be quite low. Measuring high ESR at a blue cap might indicate that the power delivery to that cap is interrupted and point to a broken trace on the board.

Oh good to know. Most of the ISA slots share common traces so I checked those slot by slot, and they seem to all be intact between the traces. I've been able to trace the 5V and 12v signals from the PSU connector down to the first voltage pin of the first ISA slot but it stops there (near a bank of capacitors/resistors? not sure which). I'll keep poking around but thanks for the assistance.