Re: companion product for GSBarnDoor

[Eric Rucker <bhtooefr@gmail.com>]

On May 19, 12:27 am, "Christopher G. Mason" <c...@my-deja.com> wrote:
> The accelerator has to slow down on certain I/O functions to 1Mhz due to
> timing sensitive operations, this is why there is so much logic on the
> TWGS and ZipGS. The TWGS actually has an LED that only lights when the
> accelerator is running. It flickers quite a bit during normal use. The
> only time its fully lit is when my IIgs decides to crash. Not too much
> I/O going on then. :)
>
> The best thing to do would be to mirror all the system RAM in RAM
> running at the speed of the accelerator (cheap to do these days), but
> that only breaks the 2.6Mhz barrier between the RAM and CPU, I/O is
> still stuck at 1Mhz.

I was simplifying a bit, the 1 MHz slowdown occurs even on a stock
system, too. ;) (And, I thought some I/O worked at 2.8 MHz, too? I
recall KEGS speeding UP to 2.8 to service 3.5" floppies, when held at
1 MHz, for example.)

On May 19, 2:23 am, "Michael J. Mahon" <mjma...@aol.com> wrote:
<snip informative explanation of how 65xx memory access works that
shows that I lack clue>

Fair enough, I was just guessing as to how the 65xx worked. Guess I
was wrong. I'll go read David Empson's post.