Re: VIA shift register (was: Re: Did Commodore cheat with the quad density floppies?)

From: Francesco Messineo <francesco.messineo_at_gmail.com>
Date: Thu, 10 Jan 2019 08:58:22 +0100
Message-ID: <CAESs-_wYCKdC38BxUamUb=XX=4kKYUm+=w+fCG1pFzpaZmeJtw@mail.gmail.com>
On Thu, Jan 10, 2019 at 8:26 AM André Fachat <afachat@gmx.de> wrote:
>
>
>
> Am 10. Januar 2019 08:20:16 schrieb groepaz@gmx.net:
>
> > Am Donnerstag, 10. Januar 2019, 08:11:04 CET schrieb André Fachat:
> >> Am 9. Januar 2019 12:17:17 schrieb Francesco Messineo
> >>
> >> <francesco.messineo@gmail.com>:
> >>>> The major issue was actually the speed - not really of the bus but of the
> >>>> 1541 implementation with 6522 software workarounds making the bus look
> >>>> worse than it really was. Common thing was to add at least the parallel
> >>>> connection to 1541, bringing it somewhat closer to the original GPIB. And
> >>>> yes - it still allowed more than two devices and obviously not only the
> >>>> floppy drives.
> >>>
> >>> Yes, that was due to a bug in the 6522 chip.
> >>
> >> Never having fixed the shift register bug in the VIA 6522 e.g. in a new
> >> revision is my biggest gripe with MOS and Commodore. It would have made so
> >> many things so much easier.
> >>
> >> Well, that's history. Anyone know if there is a fixed CPLD/FPGA VIA?
> >
> > the CMOS versions (which are still available) have the external clocking bug
> > fixed, AFAIK
>
> Did they? Do you get them still?

65C22 can be still purchased new, I believe mouser has them. I don't
know if the 65C22 has that bug fixed, but all it takes to fix the bug
is a 74LS74 in front of the VIA (I remember checking this many years
ago, si I might be wrong), so it's not such a "big" problem if one
wants to use the old 6522.

>
> After sending I remembered the WDC but they are not plugin replacements as
> they have active IRQ drivers and need an extra diode.

well, little extra components aren't that bad anyway, if one really
wants to use the shift register.

Frank
Received on 2019-01-10 09:02:08

Archive generated by hypermail 2.2.0.