On 12/25/2010 8:23 AM, Steven Hirsch wrote:
On Fri, 24 Dec 2010, Jim Brain wrote:
* I so want to start learning Verilog. I
thought, if I got a
basic environment going, I could then take a look at virtualizing
the pieces to an FPGA connected to the C128/C64. But, since I
know little of Z80 and CP/M (having used it long ago), I thought
going to FPGAs from nothing was too large a leap.
Learning is always a valid motivation for a project.
Yep, that's why I am
approaching it as I am. I thought bringing up a
standalone Z80 CP/M environment would show me the required portions of
the CP/M system, and it would give me a baseline for an FPGA-based
solution (or at least prove to me that an FPGA-based solution is not vaible)
* The next step after that is an accelerated
CP/M. This plays in
with my hope to obtain the rights to the SuperCPU for the
C64/C128. It lacked accelerated Z80/CP/M compatibility, which I
thought I could add, though maybe not.
Now you're talking! I suspect you'd have considerably more commercial
success with a reissue of the SuperCPU. If you use an actual WDC
65816 with verilog-based FPGA glue logic it should be a relatively
straightforward job. With a large enough FPGA, you might well be able
to synthesize both the 65816 and a Z80. I suspect the latter approach
might be cheaper in the long run.
I have already purchased a Digilent board with an
XC3S1200, the next to
largest Spartan 3E in production. If the design can fit on 1 or 2 of
those units (or the 1600s), I can make the system viable at a reasonable
cost. I am in the process of obtaining the necessary rights to the
SuperCPU design from the initial developer, but the actual design is no
longer cost effective. It would need to be virtualized into FPGA to be
viable.
So, there is a method to my madness, but it's a long road.
Steve
--
Jim Brain, Brain Innovations
brain at
jbrain.com
www.jbrain.com
www.jbrain.net (eStore)