[Coco] Color Computer cart addresses / CoCo PSG

[Zippster]

> On Dec 22, 2016, at 6:38 PM, RETRO Innovations <go4retro at go4retro.com> wrote:
> 
> On 12/22/2016 2:00 PM, Zippster wrote:
>> *SCS is active for $FF40-$FF5F if I understand correctly (R&W).  I think if I place
>> the 5 addresses I need at the upper end of this I should be ok staying out of the way
>> (and have the help of *SCS in doing so).
> 5 addresses?  I thought the YM and the AY needed 16 addresses?  Are you doing indirect addressing?

That should be 6 actually.  I have banking registers for the first 3, a control register on the 4th.

The YM chip does have 16 registers internally, but communications with the chip uses 2 ports (addresses).
You directly address the chips through these 2 ports.  One address to select the internal register, a subsequent
read or write to the second address loads or reads out the register via the data bus.

The AY chips are the same AFAIK.

>> 
>> I had it at the lower end of the *SCS region, but it looked like the SDC modifies a reg
>> on it’s initial search for other disk controllers (just a guess, I don’t know the procedure used).
>> Not a big deal as it can easily be corrected in software once running, but if I can stay
>> completely clear I’d rather.
>> 
>> So I’m thinking $FF5B-$FF5F in conjunction with *SCS will be ok for decoding register
>> access.  Easy to change if an issue crops up at least.  :)
> You're OK in an MPI, but with a Y cable, FDC underdecodes all of *SCS, so you'll conflict.

Good to know.

> 
> Not saying you should move, just noting it so you're aware.
> 
> Jim
> 
> 
> -- 
> Coco mailing list
> Coco at maltedmedia.com
> https://pairlist5.pair.net/mailman/listinfo/coco