[Coco] 8 slot multi pak

Zippster zippster278 at gmail.com
Tue Jan 17 14:45:52 EST 2017


There’s an issue with going beyond 4 slots on an MPI, software support.

The arbitration scheme using *CTS and *SCS traditionally only supports
for slots, so old software that uses these signals for selecting carts would mean
you’d only be able to use the first four slots and stay compatible.

Also, the *CART interrupt goes to the slot receiving *CTS, so autostart carts
would be limited to the first four slots as well unless the arbitration scheme is
extended.

Now some things don’t use those signals, like the DACs on the Orch90, and items
like that could go into extended slots without changing old software.

A thought that occurs to me is that with the unused bits in the MPI register at $FF7F,
the use of the *CTS/*SCS scheme can be easily extended to as many as 16 slots.
In fact, it looks like the bit locations might have been chosen with this in mind.

New software could easily support this, and perhaps old software can be patched,
though at some point additional bus buffering may be needed.

Thinking about it more, maybe this isn’t much of an issue.  You’d just have to choose
what slots you use which carts in carefully depending on the software you’re using.

Someone please tell me not to make a 16 slot monstrosity just to make use of those
register bits….   lol…  :)

- Ed



> On Jan 17, 2017, at 12:47 PM, Michael Brant <brant.michael.l at gmail.com> wrote:
> 
> Jim, it was my understanding is that you were building the mpi with the
> capability for 8 slots via 2 banks of 4.
> 
> On Jan 17, 2017 1:39 PM, "RETRO Innovations" <go4retro at go4retro.com> wrote:
> 
>> On 1/16/2017 10:40 PM, Brian Goers wrote:
>> 
>>> What is the status of the 8 slot Multi pak.
>>> How is the testing going?
>>> 
>>> 
>> I am finishing the final PCB design to the 4 port unit (with some relays
>> for power), but I am not aware of any 8 port testing.
>> 
>> 
>> Jim
>> 
>> 



More information about the Coco mailing list