Re: PET 2001/2001N can "write" to ROMs

From: Francesco Messineo <francesco.messineo_at_gmail.com>
Date: Mon, 22 Apr 2019 22:25:51 +0200
Message-ID: <CAESs-_z1jjWaH7EFePuRXA4=dUyzBc6uQGw-=tguveZ_GORSOQ_at_mail.gmail.com>
On Mon, Apr 22, 2019 at 10:21 PM André Fachat <afachat_at_gmx.de> wrote:
>
>
>
> Am 22. April 2019 22:16:57 schrieb Francesco Messineo
> <francesco.messineo_at_gmail.com>:
>
> > Hi all,
> > Ok, the subject is a pun, but write is in double quotes anyway. I was
> > wondering why the PET (at least the 2001 and 2001N, the ones I have),
> > leave the ROMs selected during a write cycle on ROM space. This of
> > course makes a data bus contention because the selected ROM will drive
> > the data bus and the 6502 will drive the data bus too since it's on a
> > write cycle.
> > I wonder if this can be a factor in the high failure rate of the old
> > 6540s and the newer 2332/2316.
> > It was just lazyness/cheapness or do I miss something?
>
> My guess is that this was a cost measure. Under the assumption that the
> code would not write to ROM there would be no need to decode R/-W.
>
> André

thanks for confirming the problem André, I'll be modifying my RAM/ROM
replacement daughterboard to disable the ROM in a write access, I was
wondering why I made such a mistake when I realized that CBM did the
same :/

Frank
Received on 2020-05-29 21:28:03

Archive generated by hypermail 2.3.0.