[sdiy] divider ratios

[Scott Gravenhorst]

I think what you're describing is what a 'top octave generator'
does, plus the use of binary dividers as well.  The 1/8, 2/8
etc. are the binary dividers and as you say, they produce all
of the Cs below the input C, which is the highest.

So instead of creating the octaves first and then the notes in
between, a TOG creates the top 12 notes of the western scale
(assuming that is what you need), hence the name "Top Octave
Generator", as it generates the 12 notes of the highest or top
octave.  Once you have these 12 frequencies, it's easy to use 
a single chip of several binary dividers to generate the 
lower octaves for each of these 12 signals.  One chain of
binary dividers is required for each of the 12 top octave notes.

Unfortunately, AFAIK, no commercial IC is available to do
this with flipflops.  MK50240 (MKS50240?) was made by MosTek,
but is no longer manufactured.

A few people on this list have produced alternatives, such
as using a PIC to do this.  One important thing, though is 
that regardless of whether a PIC is used or a discrete chip
or a bunch of flipflops, your 12 notes will be phase locked
to each other.  Although this is not bad in and of itself,
it does lack the character of an organ that uses 12 separate
oscillators to make the top octave, dividing this down for
the octaves needed using that same binary divider chain as
for a flipflop based TOG.

Someone else here is considering building an organ with the
12 top tones generated by 12 separate oscillators, namely
the quad version of the 555 timer.  Issues here are parts
count, but moreover tuning stability.  It would be best if
you could tune it once and essentially forget it.  (I am
not suggesting it can't be done with this chip).

IMHO, if you can get 12 separate oscillators to be tuning 
stable, this would produce the warmest sound.

And now a message from Mr. PLL:  (c:
Another possibility might be to use a flipflop based TOG and
then put those 12 signals into 12 PLLs, one for each top octave
note.  Each PLL would also have an LFO influencing the VCO in 
it to cause small frequency variations.  The binary dividers
the produce the octaves would then be driven by the PLLs output
square wave.  This method would be easier to tune than 12
oscillators, and warmer than a straight flipflop system.  Perhaps
not as warm as 12 separate oscillators, but I don't know that
for sure.  It will increase the part count since each PLL is 
a chip, and requires some resistors and capacitors for each.

Finally, one thing is certain:  The flipflop method's accuracy
of the scale produced is directly proportional to the number
of flipflops used.  Too few and it will sound bad.  More flip
flops means more parts, or in the case of a PIC, more memory and
a higher input clock frequency which is problematic because of 
the PIC's finite maximum CPU clock (not to be confused with the 
clock required for producing the notes).

<soundgood at web.de> wrote:
>I have been thinking about my problem and if I am right I don't 
>need the division ratios I gave. Lets consider that the organ 
>oscillator creates the highest note (say C) and the decision chip 
>divides this into the Cs below this than the division ratios 
>would be: 1/8 2/8 3/8 4/8 5/8 6/8 7/8 8/8 the input frquency.
>1/8,1/2, 1/4 are easy with a four bit counter but the ones in the 
>middle would need multiplying the bottom octave(1/8)by 7,6,5 and 
>3. Well I have to measure the working chips and see if my 
>assumptions are right. Any tips on multiplying a square wave 
>frequency than? Thanks, T 
>____________________________________________________ 
>Berufsunfähigskeitversicherung von Mamax bei WEB.DE. Jetzt 

=========================================================  
- Government: The other religion.
- The media's credibility should always be questioned.
- Lambs who lie down with lions are lunch.

-- Scott Gravenhorst | LegoManiac / Lego Trains / RIS 1.5
-- Linux Rex         | RedWebMail by RedStarWare
-- FatMan: home1.GTE.NET/res0658s/TekLab_FatMan_WebSite/index.html