Theretroweb says...

Also known as: […]
Show full quote

Also known as:

• GVC FR500
• Packard Bell PB850

BCM is known as an OEM for Diamond, GVC and others. OEM boards often have less capabilities (f.e. no OC functions, castrated BIOS). So everything seems fine with your board.

Concerning your second question... Everything is configured correctly? Did you try a CMOS Clear?

kind regards
soggi

Hi Soggi, yup - battery out and jumper pins bridged. I can get into the bios for about 4 or 5 seconds before it freezes - likewise if I leave it, it gets to post and then stops responding. Nothing looks blown, no damaged traces I can see, memory tested and fine.. unfortunately I don't have a volt meter and wouldn't know how to use one if I did - or what to look for!! I'm a long time builder, and certainly not an electrician 😉

It looks like a GVC FR500 aka PB850, but with various bits not populated. The jumpers seem to match. It's pretty common for revisions of boards to change which features are used. If PB could save a few cents by not having a front panel header or speaker header in a system that doesn't need, them, they would leave them out.

PB tended to have very different configurations in different markets. Which country is this machine from?

As for the not booting - I don't think you have a keyboard problem, it's more likely the system is hanging and so not responding to anything. Could be a dead battery, could also be a CPU (or board...) clocked beyond its capabilities. If the CMOS battery is new and it still hangs, try setting the lowest clock speed available (2x 50MHz is generally a safe bet)

2x 50MHz gave it a blue screen with some white vertical lines! I also tried it as an MII-200 through jumpers and same story. I've previously tested with it in non-cyrix mode with a Pentium 90, and in Cyrix mode again with a 66mhzx3.5x M11-300. New (ish) battery in 🙁

Probably not the issue but pull some ram, try with just 2 matching simms, preferable 2x 4Mb @ 60nS.

RETROKOMODO wrote on 2024-10-13, 19:15:

I can get into the bios for about 4 or 5 seconds before it freezes - likewise if I leave it, it gets to post and then stops responding.

Sounds like a overheating issue or VRM (bad caps?)....do the VRM mosfets get hot ?
Just thinking out loud....

Sorry, I can't help you with all as the Wayback Machine is still down (see https://archive.org/) - f.e. there can be a newer BIOS, but no chance to check this against the archived BCM pages at this time.

Since you already tested another CPU, try just using one bank of RAM (two paired sticks) and maybe you also have the possibility to change the RAM? Setting everything to the lowest speed in most cases isn't the best idea, often it doesn't work at all (like your blue screen showed).

kind regards
soggi

I recently had a similar issue with a salvaged mainboard. It came with cpu & ram so I hooked it up to a psu and it started, yet like yours it froze after a moment.
sometimes it managed to get through the memory test, other times I even managed to get into the bios. It was odd, until I noticed one thing when comparing it to online photos.

Somebody had removed all but two of the electrolytic capacitors and the two that were still on failed to meet up to the specs.
So I removed those two, replaced all the missing ones, and voila the board now booted ok, tested the memory, got into the bios, was able to boot from a test drive.

So I'm also guessing, that the electrolytic caps near the cpu are likely to be dry and in need of replacements.

Good luck & enjoy !

Thanks all - it could be the capacitors? They all seem to be in decent shape though. Nothing bulging or split through the cap. What I might do is buy in a bunch of replacement caps and do some soldering - although I dread that because I haven't soldered since school.. and now i'm 46. I do have other things to replace caps on (980 Ti/A7N8X-X/NF7-S v2.0) so I guess i'll add this to the pile! <shudders>
I'll try to identify what I need to replace them with but i'll likely need help as never done it before and don't know what to look for!

Try all the other (non destructive) first...then it can still be the caps, even if they seem to be in decent shape. Remind that these caps are more than 25 years old, so it's very likely that they aren't in good condition. I just replaced a cap one or two times, more bad than good I guess - this was ~20 years ago. And i have dozens of pieces where caps have to be changed. 🤣

kind regards
soggi

Ok.. The problem - and i'm fairly confident in this - is the ram slots. Furthest bank gets silly stuff on screen as does next in. Second out gets repeated beeping, and closest to the cpu gets to the post screen. interestingly both closest to cpu also gets to post screen!

Well i've blasted the memory sockets with contact cleaner and no change. Still get garbage on screen using J18 & J16. J19 beeps at me and J21 gets so far and then the system freezes. Back of the board around under the memory slots is factory fresh looking, so must be an issue somewhere else? Not sure from that if caps are the issue..
The way i'm looking at it now is for £37 I got a CPU and some working memory! 😀 Unless anyone can think of something else though, it looks like this board might be for the bin.. 🙁

Your posts aren't clear enough to me. Did you populate the RAM slots in pairs (as necessary for PS/2 RAM on socket 5/7)?

kind regards
soggi

Hi Soggi,

Rambling probably born of frustration!
Tried single working ram and pairs. I have tried a working pair in the two furthest sockets from the CPU - j16&j18, and also in j19&j21 closest to the CPU. I get the partial post in 19&j21, and nothing but a garbled corrupted mess in 16&j18.
With single sticks I get partial post from the closest socket of j21, j19 just causes system beeping with a blank screen, and either of the two furthest sockets I get the same corrupt image as in my last post.
No caps are leaking or bulging at either end. Nothing on the back of the board indicates damage - no traces seem damaged - it looks new from the back! No burnt out resistors can be seen on the top.
All connection points for memory have been hit with electrical contact cleaner.
Dropping Mhz ranges, voltages or multipliers has made no difference, and has been tested in combinations now on 3 different CPU's.
PSU has been swapped twice to see if that made a difference, and no difference detected!

soggi wrote on 2024-10-14, 14:25:

Your posts aren't clear enough to me. Did you populate the RAM slots in pairs (as necessary for PS/2 RAM on socket 5/7)?

Not all socket 5 /7. Some chipsets/memory controllers support half-populated memory banks, eg a single SIMM. The SiS 5596 is one of them:

https://theretroweb.com/chip/documentation/55 … 58988250320.pdf

Each bank can be half populated bank, but the DRAM module must be plugged in the
even SIMM of bank.

Of course, performance will be utterly abysmal when combining the UMA of 5596 with a 32b wide memory bus, but for testing purposes it's a pretty useful feature.

@RETROKOMODO:

Oh man, this took some time...these OEM boards are *$%#!

Your board has a BIOS version 00.04.03a, the original BCM has latest version 1.01 (see here), PB tells us the latest BIOS for the PB850 (FR500) is version 2.06 (see here, DL dead) and GVC itself seems to have no BIOS at all.

Do you have the possibility to hotflash or to flash the BIOS with a flasher? Then I would try the 4.08 BIOS listed at TRW, maybe it's the latest of the line which your board has!? You can also try the 2.06 and the 1.01 BIOS, there's even a patched one of the latter.

If the above and other memory sticks together, other CPUs and lower FSB frequencies like 66 MHz don't help ... then the caps could probably the reason.

----------------------

@dionb:

Ah OK, thx - I actually didn't know that there were chipsets with support for half-populated PS/2 banks.

kind regards
soggi

Hi Guys,

It could well be caps. If not - no idea! I'll look into re-capping! It's off the bench now as something new has arrived for me to tinker with, but i'll undoubtedly circle back again for another go.

Thanks again for all the assistance!

Best regards,

Fred

The GVC FR500 BIOS is here!
https://www.mediafire.com/file/ewuydrdqgi50n6 … n_4.08.img/file

Have you tried it with a pci VGA instead of the onboard? The graphic artifacts make me think it's either that, or possibly the memory controller part of the northbridge. Try to cool it with some dust-off, freeze spray or an ice cube in a bag to see if it makes any difference in being able to make the bios not freeze for a while.

Cleaning the entire board is also an option to try if all else fails

dionb wrote on 2024-10-14, 18:52:

Not all socket 5 /7. Some chipsets/memory controllers support half-populated memory banks, eg a single SIMM. The SiS 5596 is one […]
Show full quote

soggi wrote on 2024-10-14, 14:25:

Your posts aren't clear enough to me. Did you populate the RAM slots in pairs (as necessary for PS/2 RAM on socket 5/7)?

Not all socket 5 /7. Some chipsets/memory controllers support half-populated memory banks, eg a single SIMM. The SiS 5596 is one of them:

https://theretroweb.com/chip/documentation/55 … 58988250320.pdf

Each bank can be half populated bank, but the DRAM module must be plugged in the
even SIMM of bank.

Of course, performance will be utterly abysmal when combining the UMA of 5596 with a 32b wide memory bus, but for testing purposes it's a pretty useful feature.

I never realized it supported the use of only one SIMM, yikes. Imagine the performance penalty if you did that AND used the onboard video.... I suppose in some applications (office/low cost/industrial?) a minimalistic setup like that could make sense