alex.freed....@gmail.com wrote:
On Apr 10, 12:06 am, "Michael J. Mahon" <mjma...@aol.com> wrote:
alex.freed....@gmail.com wrote:
There is no need to use a microcontroller to read/write SD cards. I
have
one on my Pseudo Disk only because it emulates a floppy. In principle
you can use bit-banging by the 6502 itself to access an SD card via
SPI protocol. A single $1 CPLD that will serialize/deserialize data
while
converting the levels 3.3 <-> 5V is all it takes for a practical
solution.
Just a matter of software drivers.
And speed. A byte-parallel interface to the 6502 driver code
makes for reasonably fast data transfers. A software bit serial
interface, particularly with handshaking, will run much slower
(and take more RAM).
Maybe my English teachers didn't do a good job four decades ago or
I'm saving keystrokes too much.
Michael, you seem to argue and yet rephrase what I was writing.
Yes... Your writing was clear, I just didn't read clearly what you
were suggesting. You're right, we're on the same page.
So again to make it clear my points are:
1. Bit banging is possible in PRINCIPLE. Not practical.
2. My proposed serializer built around a CPLD in case of a //c
solution
can take care of low level SPI.
Our Carte Blanche has a connector for an MMC/SD card. Somebody needs
to write a simple driver to use it as a ProDOS volume. Ideally there
should
be a way to select from several logical volumes as the cards now days
are
considerably bigger than the 32 MB max for ProDOS.
And the FPGA should provide plenty of room for a state machine ...
The FPGA has enough room for a whole new computer (or two or three).
Right. I was being ironic--should have used the ;-)
I don't think a ProDOS driver for the card will be a problem.
It would be pretty close to those that already exist.