feipoa wrote on 2024-12-17, 21:47:

Thanks for checking. The Wiki article on Intel chipsets claims the Intel Aries chipset can handle up to 50 MHz, but there aren't […]
Show full quote

Thanks for checking. The Wiki article on Intel chipsets claims the Intel Aries chipset can handle up to 50 MHz, but there aren't any motherboards with a 40 or 50 MHz option.

These CPUs certainly can handle 40 MHz:

AMD Am5x86 at 3x40 and usually at 4x40
Cyrix Cx5x86 at 3x40

but it sounds like the motherboard doesn't output at 40/50 MHz. What clock generator is on that motherboard?

I’m at work now, but I’ll have a look tomorrow morning when I get home.

feipoa wrote on 2024-12-17, 21:47:

What clock generator is on that motherboard?

IC Designs 2028SC-A13 (also marked 9417FKV, but I'm guessing that's a date code)?? I can't find a datasheet, but if I had to guess that would be the IC I would pick. Right next to the crystal on this motherboard. Looks vaguely clock generator-esque in number of pins.

I don't see any other candidate chips on this board.

Maybe found a datasheet. Assuming it works the way I think it does (one jumper pulls up S0 and S1 on the clock chip, the other jumper selects the clock between the clock gen and a crystal socket (that this board doesn't have)):

Jumper 1, 1-2 pulls up S0, Jumper 1 2-3 pulls up S1.

Jumper 1, 1-2-3 might be 40MHz (pulling up s0 and s1). I need to do some testing to make sure it's actually working as I think it is.

These jumpers don't go to the clock chip, at least not to S0, S1, S2. They must be for something else (PCI/ISA dividers?)
There is another 3pin jumper near the CPU labeled clock that seems to go to the clock chip. 1-2 is supposed to be 33MHz, 2-3 is supposed to be 25MHz. The datasheet I found seems to indicate that S0-S2 are active high (no prime bar above the pins on the diagram). This is surprising because S2 is pulled high with a resistor right next to the generator IC. So I'm working on the assumption that they're in fact, active low.

According to the datasheet, 0 0 1 should give me 24MHz and 0 1 0 should give me 32MHz. The three pin header seems to go:

1. Either 5V or ground (5V to ground on this board is only 5Ohms, it's hard to tell which is which (!?))
2. this goes to S1
3. this goes to S0

So, 1-2, the 33MHz setting seems to just be a place to park the jumper, doesn't seem to do anything. Confirmed by leaving the jumper off, still get 33MHz. 2-3 seems to tie S1 and S0 together, meaning the datasheet I have isn't correct. Which isn't that surprising given the active low/high issue above.

TL;DR:
It's almost certainly possible to get 20MHz, 24MHz, 32MHz, 40MHz, 50MHz, 66MHz, 80MHz, and 100MHz from this clock generator by lifting s0-s2 and breaking them out to jumpers. There probably aren't adequate dividers for PCI and ISA bus speeds, but there are probably undocumented settings available given the two 3 pin jumpers that seem to control this. More work than I'm willing to put in, the pins are very very small 😀

Nice work.

I'm not sure about the 100 MHz option, but the others may be possible. More likely that 100 MHz option is 60 Mhz.

Without a proper datasheet, I also wouldn't be very motived to experiment further. I'd mostly want to find 40 MHz to run Am5x86-160. Does L1:WB work on this motherboard with i486 and Am5x86?

I've seen FSB jumper headers which also had that placeholder position.

feipoa wrote on 2024-12-19, 21:49:

Nice work. […]
Show full quote

Nice work.

I'm not sure about the 100 MHz option, but the others may be possible. More likely that 100 MHz option is 60 Mhz.

Without a proper datasheet, I also wouldn't be very motived to experiment further. I'd mostly want to find 40 MHz to run Am5x86-160. Does L1:WB work on this motherboard with i486 and Am5x86?

I've seen FSB jumper headers which also had that placeholder position.

It would be the perfect motherboard if I could get it to boot from any PCI IDE controller card. Every single one I have hangs once post is complete. Sadly this board has the CMD0640 IDE Controller, so it's pretty much worthless. Promise Ultra33 and Ultra66 don't post. Highpoint 366&302NLF don't post. None of the VIA cards I have post. The only thing I haven't tried is getting a Silicon Image one (that's an updated and hopefully less buggy version of the CMD0640), but I doubt that would have drivers for Windows 95. Not having DMA support is kinda a buzz-kill. It's really slow.

I'm having to use XTIDE anyway, so it's stuck in PIO-1.

I'd try SCSI but if it doesn't boot from a PCI IDE controller I doubt it would boot from a SCSI controller either.

My experience has been that the Adaptec 2940W, 2940UW, or 2940U2W series of PCI cards have been the most agreeable on these older systems. Most of my testing was done with the 2940U2W.