Date   : Thu, 06 Mar 2008 00:04:48 +1100
From   : msmcdoug@... (Mark McDougall)
Subject: FPGA BBC - yah!

Hi Guys,

I've been working on the sound chip whilst I await the 6502 core fixes and 
have basically finished it. The tone generators should be spot-on - the 
noise generator should be close, if not perfect, but I'm yet to do any 
significant testing on it. Bottom line - my BCC beeps! ;)

Tonight I bit the bullet and started studying the 6502 core in detail. Long 
story short - I've fixed the BRK problem! So now I can type syntax errors 
and press <ESC> 'til the cows come home!

This leaves me with a more-or-less fully functional BBC, except:
* no cursor as of yet (6845 doesn't have one)
* no MODE 7 as of yet (6845 doesn't support interlace)
   - a "special build" gives me 'hybrid' MODE 7 (ONLY) text mode
* no USER VIA, serial, comms, tube, 1MHz bus etc.

Not sure how much further I'll take the design at this point. It's a good 
basis for anyone wanting to pick up the ball and run with it - the source is 
freely available online. As I think I mentioned, my primary aim was to 
familiarise myself with the 6845 core - but the BBC turned out to be much 
more complex - and much more interesting - than I bargained for! :)

One particular area that most of my FPGA microcomputer designs are lacking 
is the implementation of storage devices - which makes them little more than 
'proof of concept' in terms of usefulness. This is something I'm looking to 
rectify - other designs are starting to use internal copro cores to emulate 
storage devices by reading FAT disk image files off SD cards, for example. I 
guess the BBC disk controller would be a prime target...

All this recent talk of copro's has also sparked some interest - the fpga 
implementation would be significantly quicker, faster and cheaper to 
prototype any copro design. The caveat is that the tube interface would have 
to be implemented and verified as correct to validate any prototype "copro 
board" design. But I think it's a very interesting option...

I'd be tempted to look at the Z80 copro schematics in an attempt to 
implement the tube and copro within the FPGA - but the aforementioned lack 
of storage emulation makes it difficult to transfer and execute the required 
software to exercise it. :(

To end my rant, thanks again for all those that have assisted in the 
process. I might go a bit quiet for a bit but rest assured I'll be perusing 
this mailing list with interest - I've grown quite fond of the "Beeb" in my 
short relationship with it! I've even got a *real* one on the way! And I'd 
be willing to entertain proposals for further development on the FPGA if 
someone has a pressing need or interest in it...

Regards,

-- 
|              Mark McDougall                | "Electrical Engineers do it
|  <http://members.iinet.net.au/~msmcdoug>   |   with less resistance!"