From: Wolfgang Moser (womo_at_news.trikaliotis.net)
Date: 2008-12-12 15:30:32
Hello Craig,
Craig Taylor schrieb:
> I'm just wondering how predictable / reliable / reproducible driving the VIC
> chip from the REU would be? I can think of all sorts of fun, interesting
> tricks to play by disabling the VIC (mono-color) and then having the REU
> dump various values into the background register.
>
> Has anyone tried this?
I did systematic tests of this scenario for my REU
technical reference writeup. Version 1.0 of this
document is available at the former Funet.Fi archive
hosted on zimmers.net:
http://zimmers.net/anonftp/pub/cbm/documents/chipdata/CSG8726TechRefDoc-1.0.zip
I found out that you cannot reliably transfer data
between the C64's IO area and the REU. The grade of
unrealibility depends on the transfer mode choosed.
While normal transfers (REU->C64 or C64->REU) are
fairly stable (but not 100% rock solid), the verify
operation seems to be the most criticial one.
Further invetigations showed to me that it is
not the data itself that gets corrupted while a
transfer is done, instead the REU sometimes fails
to address the correct byte within the I/O area
and therefore another byte gets selected, which is
the correctly transferred.
So, in the end, the transfer itself is 100%
reliable, but the C64 I/O area adressing is not
and wrong bytes get selected.
Read about this in section 3.3 of the document
referenced above.
The best you can do regarding VIC programming with
the help of the REU, is doing it indirectly. You
could program the VIC so that it is showing the
ghost byte from 0x3FFF all the time and the program
the REU to constantly fill this 0x3FFF byte all the
time (using fixed C64 adressing mode).
You could also program the REU to constantly write
and overwrite the sprite pointers or something like
that that is located in the C64s main memory.
For my tests I checked three different C64
mainboard revisions which were all equipped with
the 65xx chipsets.
I don't know yet, if the REU addressing is more
stable with the later 85xx chipset mainboard
revisions or the C128(D) in C64 mode.
Womo
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.