I had a pentium pro board which wouldn't get past the post screen because of a dead battery it would have been a soldered ODIN or DALLAS, it displayed a message that the battery or cmos settings was wrong, that board got tossed in the bin and I never buy boards with those batterys.

SO yes it could just be the battery causing the system to hang but to be honest ive come across more than one 486 mainboard that hangs at that screen, sometimes it doesnt have the message displayed below the table of information and gets stuck or it will be himem testing extended memory, unfortunately this is one of the issues when 486 boards dont behave.

I've only experienced that once, when I had an IRQ conflict between the graphics card and the IDE controller

Yes it has that ODIN RTC. Now I got some DALLAS replacements ordered from eBay and will replace it when my soldering gear arrives. Really hoping it will fix it 🙁

It doesn't remember any settings, even while being powered on. For example you go into BIOS, change something, hit F10 and Yes and go back in and the setting hasn't been saved.

Why can't things just work...

Success! 😵

I did heaps of Google searches, tried them all until I ripped the cache chips out.

It boots now 😀

Might try inserting them again now that the ESCD got updated?

Now just got to wait for the RTC replacements. I might solder a socket in rather than the whole chip, what do you guys think?

socket is best so you can exchange the RTC when needed easily... but now that you have spares, why not hack the old ODIN one so it uses a normal CR2032 battery?
If successful it will be a better solution IMHO (a little ugly though) since you won't have to worry about special rtcs anymore
If you kill it, who cares? just use one of the DALLAS chips you got 😉

True, true...

Having issues getting the Cache to work. But maybe this has to do with the battery and it not storing CMOS data. I'll leave it at that for the time being.

I studied all the jumper settings and everything makes sense. But for some reason it either doesn't detect the cache at all, or gets stuck during POST.

Did you tried new cache chips? Maybe the old are just bad.

Robin4 wrote:

Did you tried new cache chips? Maybe the old are just bad.

Yes!

I tried with nice brand new 15ns chips (256K Cache). While it POSTs now, it gets stuck when booting from the floppy drive. And when fitted with 512K Cache the cache doesn't even get detected properly. It's quite disappointing as the missing L2 cache really slows things down 🙁

Mau1wurf1977 wrote:

It's quite disappointing as the missing L2 cache really slows things down 🙁

It is only 5% performance loss on a 486. Not that much of deal.

vetz wrote:

Mau1wurf1977 wrote:

It's quite disappointing as the missing L2 cache really slows things down 🙁

It is only 5% performance loss on a 486. Not that much of deal.

I swear it was a LOT more than that on my Acer 486. Especially in cache sensitive areas like games and with a fast video card.

But because currently the BIOS settings don't stick, it could be that the cache timings aren't compatible? So I will wait until I get my de-soldering gear and then go about replacing the ODIN.

Question: What socket do I need to order if I want to socket the replacement RTC?

I think it is a DIP-24. Try these,
http://www.ebay.ca/itm/110957986626

It is possible that your default BIOS cache settings aren't agreeing well after POST. I've seen some boards which set the dirty bit incorrectly by default, depending on L2 WB or WT mode. If it were me, I'd wait for the new RTC to reduce uncertainty with troubleshooting. Also, I like to use double-banked 256 KB cache for troubleshooting. I've noticed that with greater cache sizes, lower cache timings often needed. This depends on your FSB and PCB design though.

Yes hopefully it all works out 😀

Double banked = fill both banks? Yes I have all 9 sockets filled just like on the 486. They are 15ns.

256K or 512K double-banked cache?
Depending on how the new RTC works out, you might want to get an SRAM tester. I even have some good ole original cache chips go bad, not just the newer Chinese chips.

I have so far tried 3 sets of 9 cache chips and they all do the same thing. The machine starts booting from the floppy, then the floppy light goes dark for a small moment, continues to load and then stops with the floppy light going dark. Remove the cache chips and all is good.

Here are shots of the cache chips:

Board came with 9 of these:

I fitted two different sets of 9 of these:

And I tried 4 of these for 512KB but they don't even get detected:

Here all the BIOS options:

Make sure your cache configuration is correct. Some boards need very specific TAG chip for different total cache sizes. Check this in the motherboard manual. It could also be that some of the cache chips are f***ed.

I would check first for correct configuration first, to see if cache has correct TAG size for the total amount of cache.

Use Modbin 4.5 to unhide the TAG bit in the BIOS.

Use 7+1 for L2 WB
and 8+0 for L2 WT.

I suspect your BIOS has this flipped by default. Since you cannot save your CMOS settings, you cannot change you L2 cache mode.

Just wait for the new RTC...

Thank you!

So I would have to mod the BIOS? Oh dear.

Here an image of the cache chips.

Yes the configuration is correct. The only one I can't test is 128KB because I don't have the correct TAG.

Mau1wurf1977 wrote:

Why can't things just work...

Welcome to the world of 486's my friend! You've had it way too easy with your SS7 retro machine! 😀

from the BIOS screen this looks like an awesome 486 mobo, is it PCI Mau1wurf1977 ?
You need to get this going!