This almost never happens upon first boot-up, but it seems like things are working as they should before I add the HDD. I realize that I'll need to use a program to test the cache because the report of 256KB on boot isn't reliable.

I've never seen a GUI-like BIOS on a 486 board.

Did the fan work?

chrismeyer6 wrote on 2020-05-09, 15:48:

Did the fan work?

Yes, like a charm.

boxpressed wrote on 2020-05-09, 16:01:

chrismeyer6 wrote on 2020-05-09, 15:48:

Did the fan work?

Yes, like a charm.

Good job ! You got past the worst of 2 hurdles: No boot at all or boot with KB error and can't get into BIOS.

Excellent news! Glad all is working out so far. I wouldn't worry to much about fake cache I personally haven't seen any fake cache sticks just dummies soldered on

386 and 486 single 72pin dimm is one bank. FPM only. 80ns to 60ns

Pentium and later used 64 bit data path and requires two 72pin matched set is one bank. EDO or FPM 60ns or less.

Cheers,

I transferred the CF HDD from another 486 to this board. I used the "Auto IDE" function on the main BIOS page, and it detected it correctly. Then I ran CACHECHK and SPEEDSYS (sorry for the bad photo -- didn't want to mess with the PCX file), and everything seems okay. I don't really know what all the numbers mean or the graphs on SPEEDSYS, but I'd appreciate it if someone said whether anything looks unusual.

I'm using the FPM SIMMS shown in an earlier post. I also swapped out the CR2032, which looked like it was bulging a little.

Thanks to everyone for their help getting this board up and running! I like it quite a lot and love that everything is brand new.

All looks fine.

Cachechk... numbers tell you what is the access time. Lower is better. You can see when access time increases and that means it went from faster to slower cache/memory.

Your Intel DX4-100 has 16KB L1 cache and this is the fastest cache available - see the SpeedSYS numbers - it has around 66MB/s transfer speed (higher is better). After 16KB it goes to L2 cache that is the COAST module and it goes until 256KB - transfer rate of around 37MB/s. After the caches run out, the bare memory is accessed and you can see it is really slow - depends on the operation (read, write, move) - average of around 27MB/s.

kixs wrote on 2020-05-09, 20:19:

All looks fine.

Cachechk... numbers tell you what is the access time. Lower is better. You can see when access time increases and that means it went from faster to slower cache/memory.

Your Intel DX4-100 has 16KB L1 cache and this is the fastest cache available - see the SpeedSYS numbers - it has around 66MB/s transfer speed (higher is better). After 16KB it goes to L2 cache that is the COAST module and it goes until 256KB - transfer rate of around 37MB/s. After the caches run out, the bare memory is accessed and you can see it is really slow - depends on the operation (read, write, move) - average of around 27MB/s.

Thank you for the explanation. Maybe I'll try the EDO DRAM and my trusty Cyrix 5x86-120GP after a while. I assume the ability to enable the "Cyrix LSSER bit" is a feature of this board. I've never seen it on another board.

I took some photos of the chipset BIOS options in case anyone is interested in what they are:

I'm glad you were able to get this board up and running I'm looking forward to seeing your finished build!!

chrismeyer6 wrote on 2020-05-09, 22:07:

I'm glad you were able to get this board up and running I'm looking forward to seeing your finished build!!

Thanks for your help and encouragement! I bought this board as a backup to this great SiS-based 486 VLB/PCI board that is my DOS test bench, but now I might have to retire another VLB board and use the case for the M919.

Your very welcome my friend.