On Wed, 7 Sep 1994, Charles Manning wrote:

> On Sep 6, 11:13, David Tait wrote:
> [stuff deleted]
> >
> > * I developed this project using a 20MHz 386DX machine and have
> >   subsequently found the C version to be unreliable when used with my
> >   33MHz 486DX.  Obviously a timing problem and I will try to fix this
> >   when I get some spare time.  The Qbasic version still worked though,
> >   and the C version worked with turbo mode switched off.
>
> I've had this kind of problem before with other cctry connected to PC
> parallel ports. One way to fix it is to edit the CMOS configuration of
> the PC and slow down the 8-bit I/O access by adding a few wait states.
> Maybe this can work in this case too.

In many cases the chips of the parallel card are too slow to alllow the
high speed switching of a fast card.  For instance, you might have a bt
set low, you then pulse the bit high, then a 486 clock cycle or two later
return it to low.  All fine in theory, but the card hasn't had time to
switch to the high state before it was instructed to return low.

Yeah, putting in wait delays will work.  Another one you could try is
feed the output of one or two of the bits back into parallel port. That
way, you could read when the card has made the switch.


Oh, by the way, I'm new to the PIC group.  Can anyone tell me something
about the PIC?  What it can do, features, nice stuff, etc?  Thanks.


Catcha,

Regards,

   Michael Kunstelj.