What about RAM? Is it FPM or EDO? Did you try to change type of RAM?

I've tried the RAM that was in the board when I bought it. I haven't checked to see if it is EDO or FPM. I did try two sticks of what I know is FPM. These sticks I use for testing in general and haven't had an issue. Are there some socket 5 boards which only accept EDO? I can hunt down some EDO-only sticks to test.

I should mention that I've already tried a different floppy drive and cable. I've also confirmed that the diskette works to boot another computer.

Was the P100 just bare like that while troubleshooting? Adding an HSF doesn't change the result, does it? (I'd also peel off that label, for good measure.)

Still think your problem is related to your boot disk. Himem.sys is loading, memory should be OK. I would try a DOS 6.22 disk next.

Did you wait some minutes if something happens after it hangs?

Disable shadow RAM and all unneeded devices like parallell/seriell port and IDE controller in your BIOS.

I took the heatsink off just for the photo. Yes, I plan on removing the label.

Yes, I waited a few minutes. I waited an hour even. It just hangs up. The floppy disk light stays on for about 1 minute, then goes out.

I've used this boot disk with dozens of other systems without incident.

I haven't tried disabling the serial ports specifically, but I can try. I tried changing a lot of options in the BIOS, but don't specifically recall if the shadow RAM was one of them. I'll check that again.

feipoa wrote:

I've tried the RAM that was in the board when I bought it. I haven't checked to see if it is EDO or FPM. I did try two sticks of what I know is FPM. These sticks I use for testing in general and haven't had an issue. Are there some socket 5 boards which only accept EDO? I can hunt down some EDO-only sticks to test.

I've read the manual. The mainboard supports both type of RAM so in theory the type doesn't matter.
I asked just to clarify this.
I understand you tried to load MS_DOS from floppy. But what about HDD ? Have you tried to load MS-DOS from HDD ?

I was able to boot the system using the latest AWARD BIOS, however, there is a caveat, you cannot use this CMOS setting:

L2: Write-back
and
L2 bit length: 7-bit

This is the default setting though. WT + 7bit, WT + 8bit, and WB + 8bit all work though. I'm not sure what is up with WB + 7bit.

I then decided to put the original Phoenix BIOS back into the motherboard - this automatically resets the RTC memory when you swap BIOS types. And now the original Phoenix BIOS will boot. Strange. I'm now testing a much newer Phoenix BIOS from 10/2/1996. The original Phoenix was from 1/20/1995.

I guess the issue is sort of resolved. However, one odd bit was that the AWARD BIOS read the installed 32 MB as 8 MB, whereas Phoenix correctly reads 32 MB.

My objective of getting this board was to be able to run a Cyrix 6x86 at 80 MHz. I'm almost there, but not quite. I am stuck again. I think the issues are BIOS related. Does anyone have newer BIOSes for this board?

The issues I'm having with the BIOSes are as follows:

Original Phoenix BIOS from 1/25-1995: BIOS screen comes on, but system won't boot with cx6x86. P54C OK.

Phoenix BIOS from 2/14/1996: Cyrix 6x86 detected, but after booting from floppy, DOS says "divide overflow error". Same error when running cachechk. P54C OK.

Phoenix BIOS from 10/02/1996: Screen stays blank at power on with Cyrix 6x86 installed. P54C OK

AWARD BIOS from 8/17/1995: Detects M1-S, counts memory, then does not progress w/boot. Hangs at "AWARD Plug and Play BIOS Extension v1.0A". P54C OK

AWARD BIOS from 8/29/1995: Detects and boots cx6x86 just fine, however, this BIOS doesn't count the memory correctly. For 64 MB installed, it counts up to 16 MB. I've tried numerous memory sticks, but the issue remains. Using a P54C does not fix this problem.

No idea. Maybe you can compile a new one out of the existing ones with this 😉

https://sourceforge.net/projects/awdbedit/

I suppose if I had sufficient past experience or sufficient study time, which I don't. Alternately, someone might come along with a newer BIOS.

What is interesting is that the AMD K5 works with the middle two BIOSes shown above. I have a theory about why - the K5 ignores the fixed 1.5x multiplier and perhaps the BIOS knows that. The Cyrix 6x86 only has integer multipliers and for some reason the BIOS is conflicted between the 1.5 and 2X this CPU needs to run at.

I also suspect that the AWARD BIOS which doesn't work properly with memory counting is probably for the board with PCB revision 2.x. I am using 1.x.

I found a work around for getting the Cyrix 6x86 working on this motherboard. You must use the original Jan 1995 Phoenix BIOS and you must set the "system BIOS cacheable" to disabled. The BIOS will report the CPU as a Pentium 100, but cachechk shows the proper benchmark speeds for a Cyrix 6x86. So here's a board which will work with the original Cyrix 6x86-80.

The 40 MHz FSB is JP7 with jumpers on 1-2, 4-5, and 8-9.

Feipoa, any chance you could attach the original Jan 1995 Phoenix BIOS?

Is someone intending to use a Cyrix 6x86 on this board?

I've attached in a zip file all the BIOSes I had found for this board back when I was testing it. Also enclosed is a readme file to decipher the pros/cons of each BIOS version.

Cheers Feipoa. I just wanted to keep the option open 😀

Your file includes all known BIOS releases beside version AWARD v1.2 which was posted by Robin4 in Re: ECS SI54P AIO full manual

I'll get them to update http://www.win3x.org/uh19/motherboard/show/2174

OK, I have updated my archive.

I have added the bioses/manuals to the UH19 entry. Thank you for uploading/tracking down the info.