[sdiy] Pots vs Encoders, was Re: [sdiy] dave smith *instruments*

[Jason Tribbeck]

> Interesting thread for me because I'm in the thinking stage for a digital
> (FPGA) synth design that I want to be "knobby".

Oddly enough, I'm doing the same, although it'll be an ARM instead of PIC for the controller.

I did an early design with one rotary encoder (with switch), a 160x128 display, three switches and 3 LEDs (and two PICs [one for UI, the other for MIDI command dispatching], no FPGA).

I decided that it either:

a) only needed 1 switch (in addition to the rotary encoder switch), and no LEDs

Or:

b) needed lots of everything (one for everything)

The UI you could quite happily switch between the various controlees (the other switch was an 'escape' or 'abort' switch).

Having a number of controllers that is two or more, but less than what you need actually compromises the usability, and increases code complexity as you start having to multiplex the controllers. Although having, say, half of the number of controllers to controlees isn't so bad.

I really do like using optical encoders - they tend not to break down as much. But they are expensive.

The display I'll be using is a colour QVGA display, but it has no on-board controller (to speak of - it does have the voltage generators), so the FPGA will be doing some of that work (I should have enough spare logic).

I did use a nice 160x128 OLED display for a completely different project - unfortunately, it was a bit small, and I needed some undocumented features to make it work correctly.

> Since I would be using a PIC to handle the LCD anyway and it's over powered
> for doing
> only that job, I could also use the PIC to read the encoders and it's 10 bit
> ADC to
> read the pots.

I'm using the ARM for MIDI decoding and oscillator assignment as well (it communicates with the FPGA via SPI - or will be once I've debugged an issue with it).

How are you separating the functions?
-- 

Jason Tribbeck