Re: CFFA and my Apple ][+ equals: timing touble

[Rich <rich@dreher.NOSPAM.net>]

In article <20040521032055.08010.00000946@mb-m04.aol.com>, 
mjmahon@aol.com says...
> The timing differences between the //e and the older ]['s are
> pretty well-known.  As you might expect, if a card was designed
> to work with the older machines, it will (usually) work in the //e,
> but the reverse is not the case.  The //e designers knew that
> the machine had to support almost all the ][/][+ cards.  (One
> card that was notorious because of its syncopated timing and
> DMA access was the Microsoft Softcard--it worked in some //e's
> and not in others, resulting in a redesign for the //e.)
> 
> I'm sure that you will find a timing that will provide ample setup
> and hold times for both classes of machine.
> 
> Ironically, the new timing may also correct some current cases
> of marginal operation (such as with fully-loaded, accelerated
> systems).
> 
> -michael


Michael,

Indeed, it would have been cool to pour over a set of design notes
of some old peripherial designer before I began this project.

But, I think I have finally fixed the timing, in such a way that
both ][+ and //e and //e enhanced machines should all work.
I still have to get out a GS and try it.
The timing *may* also help some brands of CF cards that didn't work 
perviously now work. But don't get too excited, because I tried my
Lexar card and it still did NOT work :(

Anyway, since I really only have less than 25 cards left, that is, most
are already in customers hands. I will come up with some very reasonable
(low cost) plan to get new CPLD's into the people's hand who want them.
I will figure out the details later...

Thanks, Michael, for getting me to work on the 6502 based firmware,
and as a result find these timing issues.

Rich