Seems like you are testing with SIMM, on the board pic first posted RAM voltage jumper was set to 3.3V, on some boards this only affects the SDRAM slot, on others it needs to be set to 5V for SIMMs too. In view of mkarchers comments about the BIOS decompression, I have observed previously that sometimes memory seems to work at too low a voltage until you give it some hard work, then it messes up.

BitWrangler wrote on 2024-11-08, 14:13:

Seems like you are testing with SIMM, on the board pic first posted RAM voltage jumper was set to 3.3V, on some boards this only affects the SDRAM slot, on others it needs to be set to 5V for SIMMs too. In view of mkarchers comments about the BIOS decompression, I have observed previously that sometimes memory seems to work at too low a voltage until you give it some hard work, then it messes up.

The voltage jumper on this board only affects the DIMM slots. I've measured the SIMM voltage pins and they all read ~5V regardless of the DIMM voltage setting.

I did some thermal imaging of the board when powered up, and the SIMMs don't even get warm.

They're hard to make out since they don't have any temperature difference, but in the below picture I've circled them with a green box.

When I've tested other systems that successfully boot up, the SIMMs would heat up as they consume power. Not sure if this is normal or not when it gets stuck like this during POST.

I was looking up code C5 and came across this thread: Octek Hippo 15 stuck on C5 code

In that thread, it looks like using an incorrect BIOS was causing the system to halt on C5.

So I did some more systematic testing of this board using the chip's original BIOS and the V2 BIOS from Retro Web. For these tests, I used a pair of 8MB SIMMs:

With the original BIOS, first boot I got the usual C5:C3 code.

Upon powering off and powering back up, the system halted at C1:00. Performing a soft reboot, it would then halt on C3:0C. This remained consistent: powering off would halt on C1, hard reset would halt at C3.

I then switched back to the V2 BIOS from the Retro Web, and the system would consistently halt on C5:C3. Didn't matter whether it was a cold boot or hard reset.

With SIMMs removed, both BIOS versions would halt on C6 consistently.

I did another detailed inspection of the SIMM slots. They have some general wear on the pins, but nothing dramatic enough that I suspect there is an issue with the slot itself.

I don't have any regular 168-pin DIMMs to test with, so I ended up ordering a few to test those to see if there are any differences.

Shponglefan, mind that some motherboard BIOSes are extremely picky about the type of RAM modules installed. For example I have two s.7 motherboards (one with SiS 5598 chipset and another with VIA MVP3 chipset) with slots for both 72-pin SIMM EDO/FP-RAM and 168-pin DIMM SDRAM and they both hang on POST when "single sided" SDRAM module is installed. They "eat" only "double sided" SDRAM sticks. I never tested them with EDO or FP-RAM modules, but they might be picky about those kinds of SIMM modules too.

analog_programmer wrote on 2024-11-08, 18:01:

Shponglefan, mind that some motherboard BIOSes are extremely picky about the type of RAM modules installed. For example I have two s.7 motherboards (one with SiS 5598 chipset and another with VIA MVP3 chipset) with slots for both 72-pin SIMM EDO/FP-RAM and 168-pin DIMM SDRAM and they both hang on POST when "single sided" SDRAM module is installed. They "eat" only "double sided" SDRAM sticks. I never tested them with EDO or FP-RAM modules, but they might be picky about those kinds of SIMM modules too.

Oh, I'm aware of the pickiness of some motherboards 😉

I've tested about a dozen or more pairs of SIMMs with this particular board. Some don't get past the C6:C1 POST code (same as no RAM installed). Whereas others will get to C5:C3. Although even that hasn't been completely consistent in my testing.

This is also why I've ordered some DIMMs, to help rule out whether it's a problem with the SIMM slots.

I was reviewing my diagnostics notes and realized I forgot one of the symptoms of the MSI MS5145 board: that without a PCI video card installed, it shows POST codes of C5:C3. But if I install a PCI video card, it stops at C6:C1.

I re-tested this and sure enough, with a video card it went to C5:C3. But trying a PCI video card in each of the four PCI slots, every single time it stopped at C6:C1.

I'm not entirely sure why this would be the case? Evidently there is something about installing a video card that the system does not like.

Shponglefan wrote on 2024-11-09, 19:20:

I was reviewing my diagnostics notes and realized I forgot one of the symptoms of the MSI MS5145 board: that without a PCI video card installed, it shows POST codes of C5:C3. But if I install a PCI video card, it stops at C6:C1.

I re-tested this and sure enough, with a video card it went to C5:C3. But trying a PCI video card in each of the four PCI slots, every single time it stopped at C6:C1.

I'm not entirely sure why this would be the case? Evidently there is something about installing a video card that the system does not like.

I’ve seen in some cases where motherboards will not boot with the diagnostics card plugged in. Have you tried to boot the motherboard without it plugged in? Probably will not resolve the issue but anything is worth a try at this point.

zuldan wrote on 2024-11-09, 20:04:

I’ve seen in some cases where motherboards will not boot with the diagnostics card plugged in. Have you tried to boot the motherboard without it plugged in? Probably will not resolve the issue but anything is worth a try at this point.

Yup, I've tried that. I tried powering it up with no cards installed, just the speaker connected. It doesn't even beep.

I've also tried with just an ISA video card installed, but that doesn't work either.

One more thing, I have been examining the Northbridge chip. It does have some slight discoloration, but I'm not sure if this is a result of past overheating or just some general discoloration in the plastic.

It doesn't get overly hot to suggest an internal short or anything. I've also tried pressing on it. For the most part that doesn't seem to change anything, although pressing on the edges did occasionally result in no POST codes (--:--). But not pressing on it, it reverts back to the usual behavior (e.g. C5:C3 or C6:C1).

I've considered that maybe there is a bad solder joint and I should try reflowing it, but before doing that I'd like to rule out other possible issues.