After some help from Kanecvr,i managed to start the board

Saddly the Amd dx4 120 mhz is not working,i managed to get a 5x86 133mhz,which works fine,but the actual bios sees it as an 120mhz.

Atm i have 2 problems with the board:

-i can't get it to run l1 cache in write back mode,the cpu acording to what i read about it online should have we l1 cache,but if i set the jumper to wb the board does not post.If i set it to wt or take the jumper of completly the board works.There is an option in bios about activating l1 cache in wb mode but doesn't seem to have any effect.

-second problem is with hard drive size,altough the bios sees the right size of the hard drive,when i try to partition the drive,fdisk(i used fdisk from windows 98se boot disk) sees it only a maximum space of 504mb,i tried with a 800mb and 1,2gb harddrive.

I ran speedsys,it's the only testing program for dos i have and i got this results,are they ok for a 5x86?

You need to use a different mode. Old Int13h is limited to 1024 cylinders, and 1024*16*63* 0.5MB = 504MB. Try setting it to "LARGE" instead of "NORMAL".

Thx setting it to large worked,altough now the bios shows a lower number of cylinders then the number written on the hard drive label,it also works with it set to lba,but not sure this old drive takes advantage of lba.

LARGE works by setting up a translation scheme between the Int13h and the HDD. Basically it reduces the number of cylinders to a value that can be used by Int 13h (under 1024) and increase the number of heads to compensate, and does some math whenever DOS calls int13h when accessing the HDD to convert the CHS values to the ranges you see in NORMAL mode.

Oh yeah about your Cache settings jumper issues - try the following:

1. Short (close) J7 (reset BIOS) and turn on computer. This will reset your bios - it might be required to get the board to work properly with a new cpu
2. Turn off computer and remove jumper from J7 - or you could remove the jumper after you turned the machine on - just remember you need to short the jumper and turn the computer on to clear CMOS (unlike on newer machines).
3. Turn computer back on and set L2 (external) cache to Write Trough in bios.
4. Jumper JP7 (L1 cache mode) to 1-2 and see if your CPU runs in WB mode. As far as I know the 486-VIP-IO does not like to run external cache in WB mode - that will cause failure to POST or hanging during boot.

Nahkri wrote:

Thx setting it to large worked,altough now the bios shows a lower number of cylinders then the number written on the hard drive label,it also works with it set to lba,but not sure this old drive takes advantage of lba.

When it was working, my board was easily accepting drives up to 8GB with LBA mode. The numbers of course will be different since they are the results of some translation (i.e. not physical but logical/faked numbers), so it is ok. You can also use an overlay utility like the ones from Ontrack, if you want.

kanecvr wrote:

Oh yeah about your Cache settings jumper issues - try the following: […]
Show full quote

Oh yeah about your Cache settings jumper issues - try the following:

1. Short (close) J7 (reset BIOS) and turn on computer. This will reset your bios - it might be required to get the board to work properly with a new cpu
2. Turn off computer and remove jumper from J7 - or you could remove the jumper after you turned the machine on - just remember you need to short the jumper and turn the computer on to clear CMOS (unlike on newer machines).
3. Turn computer back on and set L2 (external) cache to Write Trough in bios.
4. Jumper JP7 (L1 cache mode) to 1-2 and see if your CPU runs in WB mode. As far as I know the 486-VIP-IO does not like to run external cache in WB mode - that will cause failure to POST or hanging during boot.

Ok i tried,but no results,tbh shorting j7 doesn't reset the bios,according the manual it's only for password clearing.
I also tried taking out the coin baterry,starting without batery,leaving the batery off for an hour,loaded bios defaults,setting l2 cache to wt,pretty much tried everything i could,i can't get the board to run with the jumper in wb mode.
It's weird couse i can hear the hard drive spinning up,but i don't get video on the monitor,who knows maybe with a vlb or isa videocard it would work,but i have only pci videocards,so until i can get 1 of those i can't try it out.
Either that or maybe my cpu doesn't support wb l1 cache,or simply the motherboard sucks in this regard,or it needs a bios update.

Btw i put some screens from how i set up the bios,as i said before my experience with 486 motheboards it's non existant,so if u spot something wrong with the settings let me know.

Does anyone have cachechk results for this motherboard using an Am5x86-133 or Am5x86-160 with L1 properly enabled in write-back mode? I'd like to see how slow these VIA boards really are; that is the roomer anyway.

feipoa wrote:

Does anyone have cachechk results for this motherboard using an Am5x86-133 or Am5x86-160 with L1 properly enabled in write-back mode? I'd like to see how slow these VIA boards really are; that is the roomer anyway.

Here you go: https://www.youtube.com/watch?v=hJ_9PT4a76k&feature=youtu.be

Cachecheck results can be seen at about 2:03 - same 133MHz am5x86-p75 CPU used.

L1 - 135.4 mb /sec 7.7ns /byte 3.9 clks
L2 - 42.9 mb /sec 24.5ns /byte 12.4 clks
Main Memory speed 15.5 mb / sec 67.5ns /byte 34.3 clks

CPU-wise they're pretty speedy. Mine also scored like 60 in speedsys @ 160MHz

@Nahkri - my boards work fine in WB mode with stock bios... maybe it's the CPU? But as far as I know the am586 only came with WB cache... something is not properly setup on the board and it fails to post with cache in WB mode... happened to me one but I can't remember what I changed to get it to work..

Ouch. I didn't think it would be that slow. There may be an incorrect setting somewhere. I know from experience with SiS496- and SiS471-based motherboards, that if the L2 dirty bit is not set correctly, the main memory speeds drops 3-fold. This looks like what is happening on your motherboard. For example, on the Asus VL/I-486SV2GX4, if you set the L2 cache to "write-back" in the BIOS (using the latest BIOS version 402.1), the TAG+Dirty is incorrectly left at 8+0 instead of 7+1, and the RAM speeds drops 3-fold. Luckily, setting L2 mode to "Auto" corrects the problem. The same thing happens with SiS 496-based boards, however these BIOSes often allow for manual control of the alter bit. Sometimes, the Alter Bit option is hidden and you must use Modbin to unhide it. Unless your motherboard has a specific DIP socket for the Dirty Bit, you should use 7+1 for L2 in write-back mode and 8+0 for L2 in write-through mode.

kanecvr wrote:

@Nahkri - my boards work fine in WB mode with stock bios... maybe it's the CPU? But as far as I know the am586 only came with WB cache... something is not properly setup on the board and it fails to post with cache in WB mode... happened to me one but I can't remember what I changed to get it to work..

I'll guess i'll have to wait till u get to your board,take a picture of how jumpers are set,set them same way u did and try again.

Pardon me necro'ing this thread, but can somebody please tell me what exactly is to be inserted to the RN19 socket for operating a DX4 / 5x86?

Looks like an 8-pin resistor network would fit. Watching kanecvr's youtube videos and pictures, it sure looks like such a SIP. If anyone could tell the numbers and letters stamped on it, I could try using single resistors of the right resistance in the right manner to replicate the part. Thanks!

DeafPK wrote:

Pardon me necro'ing this thread, but can somebody please tell me what exactly is to be inserted to the RN19 socket for operating a DX4 / 5x86?

Looks like an 8-pin resistor network would fit. Watching kanecvr's youtube videos and pictures, it sure looks like such a SIP. If anyone could tell the numbers and letters stamped on it, I could try using single resistors of the right resistance in the right manner to replicate the part. Thanks!

The manual says "RP 0Ω 8P4R"
I dunno what this is off the top of my head but someone else might.

Someone found a solution here.
Re: Question regarding 8-pin jumper

I decoded the part number. - Posted in the thread in the link above this.
.