[Coco] Introduction, cartridge slot proto boards, J&M controllers, and EDTASM pak.

[Matthew Stock]

On Thu, Aug 22, 2013 at 9:37 AM, Mark J. Blair <nf6x at nf6x.net> wrote:

>
> On Aug 22, 2013, at 04:19 , Matthew D Stock <stock at bexkat.com> wrote:
> > Hi Mark, welcome!  I'm still new to the party myself, but I have a few
> suggestions.  Regarding the prototyping board you're looking for - I was
> never able to find one from any of the usual sources that would fit the
> Coco.  That's one of the reasons I created my bus buffer (
> http://www.bexkat.com/products.html).  It breaks all of the bus pins out
> to a header which you can connect to a breadboard or other project.
>
> Thanks! That looks good. I watched your video of it, and I can see that we
> have some similar ideas for CoCo hardware projects. I've been thinking
> about making a CoCo cartridge with a Xilinx Spartan 6 FPGA on it to
> experiment with. My ideas about what to do with it are mostly the usual
> stuff; emulate ROM paks, emulate floppy drives, implement a real floppy
> controller, VGA video output, interface a CF card or IDE drive, and so on.
> Nothing original, I assure you!
>
>
Cool.  All my design stuff is open, so feel free to take a look and use
what you can.  If you want to buy one of my boards as a way to bootstrap
your own work, I won't complain about that either.  :-)  I didn't go with
an FPGA direct on a board since I didn't want to deal with BGA.  And I
wanted to see how flexible I could make things but still stay at a
reasonable cost.


> Very nice. I've also thought about hanging an ethernet port off the CoCo,
> but I haven't convinced myself yet that there's a good enough use case for
> it. Does a CoCo have the horsepower to run a TCP/IP stack and have any room
> left over to do anything with it? I considered using something like a
> Raspberry Pi to run the network stack. That seems silly, since the
> processor running the TCP/IP stack could probably emulate the whole CoCo in
> its spare cycles. One of my more entertaining co-workers likened it to
> having a team of PhD's spoon-feeding a vegetable in a hospital bed. :)
>

Well, the WizNet chips that are out there offload a lot of the very
low-level stack, and so it's more feasible than you might think.  I'm
probably going to leave the interface to that chip fairly "raw", and rely
on someone to code a driver on the Coco side.  In essence, the interface
can be as simple as a handful of registers - two bytes for the WizNet
address, 4 for data send/receive, and one for command and status info.  It
can be simplified even more, but a wider data path will make the data
transfer a little more efficient.  The Coco will not be doing huge data
transfers, but giving it a way to interact natively with the network seemed
fun.

Let me know where you go with your projects, I'd be happy to help where I
can.
  -Matt