Just use Bank 0 and just, the already installed, 256K.
The RTC may be the culprit but one can't be 100% sure, also you may invest a few bucks in a diagnose card.

I've come across boards that do absolutely nothing with dead RTC, but that's pretty abnormal behaviour. More common is that everything (except saving settings) works with dead RTC, but if you remove the RTC, the board completely refuses to boot. You haven't removed it, so that doesn't sound likely.

got a POST card in, as expected it just displays -- -- and speaker is silent as well, +5V/+12V/-12V/CLK light up.

seems to me like a good chance the BIOS is wiped/corrupted, i had already suspected this since the EPROM sits a bit slantedly in the socket, which suggests that it has been tampered with at some point. anyway i suppose i could get a TL866 and try that, but how likely is it to find a suitable BIOS for this board?

In the photo, I see 16 pieces of 32kx8 cache modules. Does that make it 512K of cache rather than 256K?

If you suspect the bios and you have a tl866 or you are planning to get one, it will be useful for other stuff down the road anyway you can save the bios image then view the contents in tl866 app, usually there are clear English strings inside the bios, for example copyright 1998
A corrupt image probably won't have any clearly readable lines.

Also another test you can do if one of the chipsets is getting hot and might be too hot check the continuity on the pins.

Use one probe on ground, like the screw holder or the ground pin on isa then using the other probe measure readings on each pin on the chipset. The pins that will have full continuity will be ground pins the pins that have no visible traces connected to them. most of your pins on the chipset shouldn't have full continuity only some level of resistance to ground. If you find that alot of your pins especially the ones from the chipset that are connected to visual traces are showing full continuity to ground something is shorted, possibly inside the chip. Bit time consuming for every pin but nothing to disassemble or desolder

btw u must of missed it feipoa but its not a picture of his board

treeman wrote:

btw u must of missed it feipoa but its not a picture of his board

I didn't miss it. He said his board looked identical, so I assume that is cache as well. So I'm wondering if the cache is set properly.

treeman wrote:

Also another test you can do if one of the chipsets is getting hot and might be too hot check the continuity on the pins.

Use one probe on ground, like the screw holder or the ground pin on isa then using the other probe measure readings on each pin on the chipset. The pins that will have full continuity will be ground pins the pins that have no visible traces connected to them. most of your pins on the chipset shouldn't have full continuity only some level of resistance to ground. If you find that alot of your pins especially the ones from the chipset that are connected to visual traces are showing full continuity to ground something is shorted, possibly inside the chip. Bit time consuming for every pin but nothing to disassemble or desolder

surprisingly i found a datasheet for the 82433LX chip (https://html.alldatasheet.com/html-pdf/66096/ … 5/S82433LX.html), i did a bit of probing and the VSS pins have continuity to ground while VDD shows something like 760 on the display, which i assume is the resistance you were talking about. this goes for both of these chips, i've no idea if this is an expected readout or not.

feipoa wrote:

I didn't miss it. He said his board looked identical, so I assume that is cache as well. So I'm wondering if the cache is set properly.

what i meant to say is the layout is identical, but as i noted in the OP my board only comes with bank0 loaded and it's also jumpered for that, so no issues there. the clockspeed jumper was set at an undocumented position though and i wonder if that could have caused any damage.

That sounds like expected behaviour vss is ground so all vss pins in a perfect scenario should have direct continuity to ground, vdd is positive and that resistance sounds about right as it is going through the chip circuit but not shorted to the ground. I would say if both chips are similar and they giving close readings going by common logic they are ok, if one chip was showing alot of continuity to ground through non vss chips that would be a sign of a alarm.

But if 1 chip stays cold and 1 is warm then I would look at the path to the chip that is cold, I know I know there is alot of traces and paths to it but first a close visual inspection, anything going to the second cold chip that looks cracked or burnt. If not you can do the same check for any resistor leading to the second cold chip.

Could also check any resistors on the path to the second chipset for being open, since the second chip is cold check all the traces for vdd pins, no power no heat.
So locate each trace from each vdd pin, trace it as far back as possible and check continuity, since they will pass other components it probably won't be full continuity but some resistance. If you find 1 with infinite resistance meaning it is open focus on that line and go closer and closer to the chip until you find a component that is open.

It is a start, bit of a shot in the dark but at least your not changing random chips and hoping for post.

So from here, if you have power, the bios, cmos, chipset, cIockgen I think are subsystems that could prevent error checking and beep codes.

ok thanks for the advice, i checked all the resistors around those two chips using the codes printed on them and everything looked in line with spec, also VDD is connected to 5V on both chips so all seems right there. nothing looks burned on the board either. by the way i didn't mean that one of the chips was cold, they all get warm quickly, just the topmost chip seemed a bit hotter. maybe this could also depend on which RAM bank has sticks plugged in?

either way the most promising shot right now seems to be checking the BIOS so i'll report back once i got a programmer.

i have obtained a tl866 ii plus since and have dumped the bios. not sure about the chip manufacturer but as 27c010 is silkscreened on the board i just went with an intel one. had to turn off check ID as well for some reason. anyway there is data and some readable text as well but obviously i can't tell whether this is a bad rom or not. i've attached the image anyway.

probably a very long shot but does anyone have a bios that has a reasonable chance of working on this board?

1  Vendor: Award
2 Version: v4.50G
3  String: 06/06/94-MERCURY-ZB-2A59BS21-00
4 Sign-on: REV ZB-B
5 Add-ons: NCR3
6PCI ROMs: [1000:0003] Broadcom/LSI 53c825

modbin can open the dump so it looks legit.

Added this board to the UR database:
https://www.ultimateretro.net/en/motherboards/10085

If we ever find jumper settings or a manual they will be added there.