MOTM

The concept behind the CHAIN switch is that you have a need to expand 
small sequencers into large ones. There are two types of expansion: 
creating longer rows, and creating more rows.

To create more rows, you don't need a second full fledged sequencer 
with its own clock counter, position setting, etc. In fact, that's 
wasteful. You just need a row of pots or switches and an output or 
two.

To create longer rows, you do need those things.

So that leads to two types of modules in a system like this: dumb 
expander modules that blindly take their address from some other 
module, and multiple 'master' sequencer modules with some type of 
smart chaining ability.

For this system, where compactness and cost-effectiveness is an 
important issue, I feel that the use of normalled connections behind 
the panel between masters and expanders is justified. A small ribbon 
cable of 8 or less connections would be sufficient. Expanders would 
be connected to their master on a more or less permanent basis. 
Masters would also physically connect together using this ribbon 
cable method, but from a different bus. They would be configured by 
the user via the front panel CHAIN switch. I don't have all the 
details worked out right now - the switch graphic is kind of a 
placeholder for the panel design. Maybe you wire a bunch of master 
modules together behind the panel, and then set up one or two 
independent chain groups via the switches.

I'll put up a few expander module designs, a few more explanations, 
and let it sit for awhile. I've got to teach myself PIC programming, 
the hardware design is otherwise pretty straightforward.

Moe