Re: MAX Machine PLA equations

On Sat, Aug 06, 2016 at 06:13:40PM +0200, Michał Pleban wrote:
> > Right, that makes sense...  Except SID is treated "properly"!
> 
> Yes, that's true. But maybe it's because they wanted to make SID write-only?

No, the paddle regs need reading, and the counters do (wave form,
envelope of chan 3).  The MAX has the paddles wired up (with a 4066
no less) so it is meant to be used :-)

> It's hard to explain it. When I hook up the GAL instead of 6703, we can
> try to see if "fixing" the equations makes any difference - maybe
> there's some reason to them, or maybe not.
> 
> >> Maybe that's not very relevant, since the VIC memory map seems to
> >> repeat itself every 16 kB.
> > But not the 6703 equations!
> 
> They do with regard to RAM and ROM. The IO chips are an anomaly, but
> they don't overlap with RAM and ROM in the VIC memory map. So maybe
> nobody bothered to fix the equations for them.

Yeah.

> It would be interesting to write some software which tries different
> settings of VIC registers to see if it really "sees" these chips. Plus I
> still don't know what drives the A15..A14 lines on VIC access - are they
> just left floating?

Probably.  There are two schematics of the MAX in circulation, both
different, and both wrong it seems.  Or they are for different generations
hardware.  But neither has this dealt with afaics.

If nothing drives an address line it will hold its value for a value,
and eventually drop to 0.  There is nothing that would pull the line
one way or the other as far as I can see.


Segher

       Message was sent through the cbm-hackers mailing list