From: Ethan Dicks (ethan.dicks_at_gmail.com)
Date: 2008-07-08 07:35:39
On Tue, Jul 8, 2008 at 4:44 AM, William Levak <wlevak@sdf.lonestar.org> wrote:
> The EPROM's are mapped to $9000-9FFF. Resistors pull the address lines from
> $Fnnn to $9nnn...
That makes sense. I was wondering how they overrode the usual reset
vector - just
force the processor to fetch from $9xxx when it thinks it's running at
$Fxxx (but of
course, code _at_ $9xxx would still run fine, as would any accesses to
zero page,
the stack page, and screen memory, since you only have to hold down A13 and
A14). Slick.
> At this point a message is displayed to remove the clip. When it
> detects that the clip has been removed, it continues with the rest of the
> tests in an endless cycle. The PET must be turned off to terminate the
> test.
Right. I knew about that part, but I was missing the very first step - how to
jump into the ROM in the first place.
> If you have the version with 2 4K EPROMS, there is a switch to select which
> EPROM is run. This switch was originally a momentary contact push button,
> but I had trouble using two push button switches and the clip all at the
> same time, so I replaced it with a toggle switch.
Makes sense.
> If you have anything at $9000-9FFF, it must be removed.
Also makes sense.
So if you don't have a real clip, it _sounds_like what might work is to burn
a copy of the image into a ROM and stuff it into the socket at $9000 (at
least for all PETs after the original static board), then if you had a clip with
just the resistors and a reset button, it would still work as advertised, but
with an internal ROM. Of course, it means that more of the PET board
would have to be functional than using an external ROM (bus buffers,
74154 address select...), but so much of the PET has to be working anyway
for the clip to be effective, and this would be an easier replica to build.
Thanks for the info, William. I look forward to seeing the images on
zimmers.net.
-ethan
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.