Re: [stella] Parker Brother's E0 Bankswitching

Subject: Re: [stella] Parker Brother's E0 Bankswitching
From: Joe Grand <joe@xxxxxxxxxxxxxx>
Date: Fri, 25 Apr 2003 12:44:43 -0400
At 12:13 PM 4/25/2003, Adam wrote:
On Fri, 25 Apr 2003, Joe Grand wrote:
> The cartridge would need to know which slices are mapped to which segment
> at any given point in time, since it will send back different code to the
> system depending on which slice is actually mapped into the segment, right?
> This sounds complicated to me from a cartridge point-of-view, especially if
> you think in 1982 terms, since there will need to be some state machine
> circuitry or something to keep track of the state of each segment (so
> probably 9 bits of storage, 3 bits to specify the current slice 0-7 for
> each segment).

9 bits is a lot if you're using 74xx series chips, but not if you're
manufacturing custom RAM chips in high volume.  Nowadays it'd be easy to do
this in a CPLD.

Good point. The only resource I current have is a 20V8 PLD, and I don't think I can do this up with CUPL (yuck).

> Maybe another way to ask this is: When programming, once a slice is mapped
> into a segment (with LDA to one of the hot-spots, LDA $1FE0 for example),
> that memory map will stay the same until another slice is later loaded into
> the segment (LDA $1FE1, for example).
> Am I understanding this properly?


OK, thanks for clarifying. This makes sense now. :)


---------------------------------------------------------------------------------------------- Archives (includes files) at Unsub & more at

Current Thread