Top octave dividers: A few ideas

[The Old Crow]

On Fri, 5 Nov 1999, Rene Schmitz wrote:

> Anyone considered using the CMOS 4059? Its division can be set using a
> 4-digit BCD number (1-1000). Interfaceing should be easy to a uC with
> two 8-bit ports, I don't know if they contain a latch. It can run up to
> 12Mhz. Should make a fine DCO. But its out of the question for a
> TO-divider as you'd need 12 of them, and they cost about 2.5 EUR here. 

  The 4059 is a 3 to 15999 divider using BCD-coded jam inputs.  It is a
rather bulky device for making TOGs, but is wonderful for PLL frequency
synthesizers.  My function generator uses one for this.

> Another idea would be to use twelve fixed frequency oscillators, like
> two 40106s with caps + resistors. As was pointed out here, these CMOS
> schmitt triggers are quite temperature stable. Would allow a precise
> equally tempered scale. Tunes like a piano. Output is not a 50%
> squarewave, but this doesn't matter as you'll use dividers anyway. 

  Much easier to program a PIC chip for single-note or multiple notes.
Crystal-controlled, too.  Using a Scenix 100MHz part, you get 100MIPS of
processing power in an 8-bit uC.  A 50MHz SX18 chip will attain the
top-octave using a "down-counter chains" version of the program.

  I forgot I posted the PIC version on a website a few years ago.  See:
http://chip.aeug.org/oct54all.asm

> Last but not least: I saw an octave divider done with a PIC 16C5x but it
> was not the top octave. One could use a PLL to shift it a few octaves
> up. The PLLs will always be locked, so loop response is not an issue.
> (The source is at the site of "The old Crow", see the link at the
> synth-diy homepage.)

  Er, this is me. :)  I have since made TOGs from 50MHz scenix parts.

  --Crow

/**/