assumptions with solutions are better than simply telling people to use a
video card when it was stated that was not the desire...
as for things being different between manufacturers, of course it is. As I
always told engineers when I was giving a technical presentation for
in-circuit emulators that I designed, "not everyone when to the same
school, therefore every design is different..."
A single CPLD from Xilinx can be put on a card and be the entire
"programmable" solution. The current consumption would be minor compared to
all the chips necessary for a real video card.
best regards, Steve Thatcher
At 11:35 AM 02/21/2005, you wrote:
>On Mon, Feb 21, 2005 at 11:26:33AM -0500, Steve Thatcher wrote:
> > in order for you to fool the bios you will need some hardware that will
> > "respond" to key areas that the bios is interested in.
> >
> > Picking a mono card as a target, you will need hardware that allows for
> > writes and reads of memory from b000:0 to b000:7fff (the end address
> may be
> > a lower location). I don't think the bios does a real memory test, so you
> > may be able to use one register that lets the cpu write a value to it that
> > can be read back (the register should respond to all addresses in the
> range.
>
>This is an assumption; I know for a fact that different BIOS
>manufacturers check in different ways; for example, Award vs.
>Phoenix vs. AMI. Some write to registers; some write to video RAM;
>some do both; some do neither (read $3d8 for verticle retrace
>status or other methods).
>
>The only universal/compatible way to fool the BIOS into thinking a
>card is in the machine is to *put a card in the machine*.
>--
>Jim Leonard http://www.oldskool.org/ Email:
>trixter_at_oldskool.org
>Like PC games? Help support the MobyGames
>database: http://www.mobygames.com/
>Or taste a slice of the demoscene at http://www.mindcandydvd.com/
Received on Mon Feb 21 2005 - 11:04:13 GMT