BIOS Updates forum

Hello.

I got a Socket 7 Motherboard which is a Gigabyte GA-586ATV Rev. 4B, a later revision which uses a Raytheon RC5041 regulator (https://pdf.dzsc.com/RC5/RC5041.pdf) and can supply the core voltage from 2.1v to 3.5v. Out of curiosity I decided to try the patched BIOS that includes support for K6-2, K6-III and K6plus, and for large HDD up to 128GB found at: http://web.inter.nl.net/hcc/J.Steunebrink/k6plus.htm

I used a Pentium CPU to be able to flash the BIOS successfully, then upgrade the CPU with an K6-2+ 450/ACZ running at 6 x 66 Mhz (using the x2 multiplier since it is interpreted as x6). The POST recognizes the CPU properly as a K6-2+ 400mhz and started the memory test, however after the memory test passed and before the floppy and IDE drives are being detected the computer starts rebooting with a black screen and the POST screen appeared once again to perform the memory test, once it has passed it reboots once again, rinse and repeat. I can't even enter the BIOS setup, since it also reboots the computer.

I have cleared the CMOS stored on the RTC battery and made sure that the Video and System BIOS Cacheable was disabled prior replacing the CPU, also I unplugged the floppy and IDE drives from the motherboard but to no avail. Only after switching back to the Pentium CPU, everything worked fine.

I wonder it it has to do anything with EDO RAM (using 2 x 32MB Nanya 60ns), the inability to set the voltage manually on the motherboard (there is a voltage-related DIP switch which forces 3.2v for the original AMD K6-233. Does the patched BIOS sets the voltage to 2.1v when using a K6plus?) or something else.

The support for large HDD worked fine though.

I have decided to try again with the K6-2+ and replacing the EDO ram with 2 x 8MB to see if the restart loop has something to do with this and to my surprise I was able to past the POST memory count without issues. However, the good news ends here because when I restart by doing a soft reboot or Ctrl+Alt+Del the screen stays black and when I try to install more than 16MB of RAM, the reboot loop reappears.

I finally managed to get a voltage regulator adapter to try again. This adapter supports 2.0v, 2.2v and 2.4v for the core voltage, and the CPU multiplier up to 6x using jumpers. Installed the K6-2+ CPU into the adapter and configured the jumpers accordingly. However the reboot issue still persists despite installing the adapter into the motherboard socket.

Also got a K6-III 450 AHX CPU which requires 2.4v for the core voltage to test with and the same issue still occurs.

It seems that the issue is between the amount of RAM memory and the K6 write allocation feature, as the BIOS was able to POST successfully with 16 MB or less (warm reboot still doesn’t work with those CPUs though), or using an officially supported CPU with 64 MB ram.

For what is worth, the voltage regulator adapter in question says P55C-K6-S2 Rev. C.

Finally got the CPU to work on this specific GA-586ATV motherboard with more than 16 MB of RAM. Had to use the Award BIOS 1.22 since 1.28 (regardless of whether using the patched one by Jan or the official one) has the reboot issue during POST and with Award BIOS 1.23 doesn't even POST at all when using the K6-2+ and K6-III CPUs.

I patched the older BIOS by using the http://www.rom.by BIOS patcher, and used MODBIN to add "System BIOS Cacheable" and "Video BIOS Cacheable" into the BIOS menu and disabled them by default.

The older BIOSes were found at: https://www.ultimateretro.net/en/motherboards/2732, since the official Gigabyte site only has the Award BIOS 1.28.

Once again I ran SpeedSys with the K6-2+ CPU and write allocation was enabled according to the log, and no issues with warm reboot.