Date   : Tue, 13 Jul 1993 10:28:36 +0100 (BST)
From   : "I Stephenson" <ian@...>
Subject: Re: Tube connections (was RE: colour Graphics)

>Output _from_ the computer?
Yeah - Everything is an output FROM the host except the data lines.

>I assume it suffices to isolate the Ntube on a second processor to disable
>it.... (Hmmm, and NIRQ I guess).

More or less - NIRQ doesn't go through the tube (otherwise everything would get  
VERY confused). IRQ's are generated independanly on the 2P when data is written  
to the tube hardware(ish). The circuit (in addition to zapping NTube) also  
holds the reset line low on the disabled processor. All the lines are buffered,  
as they now need to drive a greater load.
(though bear in mind that I can't get this circuit to work! - it seems to be  
doing sensible things when I test it here at work, but when I take it home and  
plug it in it just dies :-().

>So, this seems to indicate only the low 8 locations are required by TUBE
>hardware. Does the splitter ensure splitter switching writes don't make it
>to the second processor?
Yes - NTube is only passed on when A3 and A4 are low.

The next 8 locations are used by the tube splitter itself to switch processors,  
but provided A4 is high your in the clear.

You probably want to check out the Master circuit diagrams if you can, as it  
can switch between internal and external processors. It probably works almost  
identically (this circuit is described as the ABM tube splitter, so it predates  
the Master).

>How many second processors have you got!? :-)
2 - a 6502 and a 32016.

Ian