[Coco] The BIGGEST PROBLEM with (So-Called) NEW CoCo Hardware...

Dave Philipsen dave at davebiz.com
Sun Mar 5 23:20:04 EST 2017


I like the idea of a $4 USB host device and there is some potential 
there but it still has to be interfaced to the CoCo.  The device for 
joysticks would have to drive four 0-5 vdc analog signals to the CoCo.  
The keyboard device would have to interface to the connector in the CoCo 
that connects to the keyboard matrix.  The MIDI device would have to 
interface to a serial port.  So the $4 device now becomes at least a 
$10-20 device because you have to add extra circuitry and connectors.

One of the things I like about the CoCo3FPGA is the ability to make new 
hardware "look like" the original hardware so that the firmware in the 
ROM doesn't have to be changed.  Gary Becker created a joystick 
interface which allows you to plug in standard RS joysticks.  The ADC 
chips which read the joysticks are newer and much different than the 
hardware used on the original CoCo (they actually transmit as serial 
data).  But since he knows how the CoCo expects to see the data he can 
read the data from the chips and present it to the CoCo hardware so that 
the CoCo doesn't know the difference.  The audio chip on the DE-1 is a 
CODEC with much more modern technology than what the CoCo had.  But the 
CoCo3FPGA uses stock CoCo ROMs.  The firmware "thinks" it's still 
talking to the old hardware. Same with the synthesis of the Orch90 
cartridge and the floppy controller.

That is one reason I like the CoCo3FPGA.  Essentially, it uses more 
modern components to synthesize the CoCo.  It's not a software 
emulation.  It's a hardware synthesis.  In addition, it doesn't require 
a PC to be running in order for it to work.  If you're really into the 
6809 (as I am) then you can also explore some things that were never 
really possible before with a CoCo like higher resolution graphics, 
larger arrays of memory, SD cards, etc. while still retaining a high 
degree of compatibility with the original CoCo 3.  All of the early 
"CoCo 4" designs failed because they could not offer backward 
compatibility with the CoCo 3.  Now the CoCo3FPGA does that.  The thrill 
for me is to be able to program the 6809 and actually do some things 
that are beyond the capability of the original CoCo 3.  I wrote a little 
demo for Gary last year that drew some graphics and used an VGA-style 
font to display characters on a 640 x 450 graphics screen with 256 
colors.  The graphics page required 288,000 bytes of memory.  Just 
clearing the page with an old 2 MHz 6809 would have been a little slow 
but the 25 MHz core of the CoCo3FPGA handled it with ease.

For the CoCo3FPGA your $4 USB interface could do some cool stuff because 
you don't necessarily have to create the extra hardware to interface to 
the CoCo.  For the most part you could just use one or two I/O lines 
from the CoCo3FPGA to talk to the USB device and then the HDL 
programming could be done to "fake" it into looking like the old hardware.

Dave


On 3/5/2017 8:47 PM, Allen Huffman wrote:
>> On Mar 5, 2017, at 8:30 PM, RETRO Innovations <go4retro at go4retro.com> wrote:
>>
>> 31250bps RS232 is easy, USB is hard
>
> This used to be true, yes.
>
> With the $4 USB host chip, it should now be as easy, and cheap, do hook up various dedicated USB devices. It won’t be as full featured as a PC, of course, but a pack with a USB port is now easy and cheap - but for one class of device. Firmware makes it read a mouse (like using an RS232 serial mouse today), or change firmware and it’s a USB MIDI pack to hook to modern music equipment, or another firmware load and it’s a USB keyboard drive JUST for OS-9, etc.
>
> Lots of potential there.
>
> And likely, there is a SATA solution like this too, but I’m not sure a SATA drive (SSD or otherwise) makes sense on the CoCo anymore with the SD storage being so cheap and fast.
>
> 		— A
>
>



More information about the Coco mailing list