🚀 go-pugleaf

In an article, Al Cargill wrote:

> I am trying to put together a system based around an old 386DX40 
> motherboard. When it was pulled out a long time ago it was working 
> perfectly. However now it gets through the initial RAM check and then 
> as it is looking at the floppy drive it gives the following error 
> message

> 	ON BOARD PARITY ERROR
> 	ADDR (HEX) = (0000:0002)
> 	SYSTEM HALTED

> 	The motherboard has 64K cache and an AMI BIOS. Now I presume 
> the error message means that one (or more?) of the cache chips is 
> fritzed. I am right on this? I would like to know if I am on the 
> right track before I start pulling out the chips as I am sure to make 
> a bodge job of that and so I would like to know that I have to do it 
> before I do!! 
> 	If I set the jumpers such that it is set for 0K cache 
> even though the chips are in place will this cause any damage? I 
> have tried switching off the external cache in the BIOS but this 
> doesn't change what happens when it is switched on.

Al, are you sure that it is the cache memory that's at fault ?

I had a similar problem when I first installed a beta version of Windows
95, which, after a couple of crashes, bombed my CMOS entirely. After I had
set all the details back in, (diskdrives, etc. etc.) I suffered terrible
problems with memory parity errors.

To cut a long story short, I found that I had neglected to set the "Wait
State" part of the CMOS to 0 (zero) wait states (I run a 33MHz 486DX).
After I did that all was OK.

Just a thought .......

Regards,
Paul Calter, Lancashire, UK