[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
more on the Mill 6809
- Subject: more on the Mill 6809
- From: Alex Freed <alex_news@alexfreed.com>
- Date: Fri, 06 Jul 2012 21:46:58 -0700
- Newsgroups: comp.sys.apple2
- Organization: Aioe.org NNTP Server
- User-agent: Mozilla/5.0 (Windows NT 5.1; rv:13.0) Gecko/20120614 Thunderbird/13.0.1
As mentioned before the simple counting test from the Mill floppy won't
run on enhanced 2e because the 2 bytes at the end of ROM taken as a big
endian address point to a ROM location. I didn't figure out how to
switch to "language card" RAM and still get PRODOS to work so I have
modified the BASIC part of the DEMO1 a little.
Since I have a jumper to enable OS9 style memory translation I decided
to use it. With the jumper in the ON position $1000 in 6809 space
becomes $2000 in physical memory, so I changed line 140 to
BLOAD DEMO1.BIN, A$2000
Added a line POKE 8194, 246
This patches the 6809 binary to use $f628 as the screen address which is
$0628 physical. And finally wrote $1000 into $FFE because $FFFE becomes
physical $FFE.
With these changes the counting test runs perfectly well while 6502
still runs the BASIC program. I used a 74LS14, just as on the original
Mill to preserve the exact timing and things look good on the scope.
Data written by the 6809 overlaps the PH0 on both end, so there can be
no issues there. Data read by the 6809 also seems to be there when
needed. So I think I can declare the board operational.
Now I was NOT able to run OS9, but I'm pretty sure that my boot floppy
is bad. I didn't have a physical disk and only found an .SDK image.
I somehow expanded it to a floppy, but the disk is not a DOS3.3 or
PRODOS formatted - it has its own structure that CiderPress is not aware
of. When I try to boot from it, the boot sector finds the 6809 but the
code fails to load due to the wrong disk structure and the 6809 jumps to
uninitialized memory. The new card owners are welcome to figure out how
to make an operational OS9 floppy.
One mystery remains. The DEMO1.BAS seems to enable an interrupt to 6502
from the Mill card. Why? Where is the vector set? Currently I don't have
the IRQ line connected.
I have 3 names in line:
Sean F (2 cards), Steven H and Egan F. Any more?
You can have your cards now or wait till I get a good OS9 floppy and
test it. Send me e-mail.
-Alex.