MOTM

You can also use phase-locked loops. The number of serial flip-flops you put
in the loop determines the number of octaves shifted.

 -----Original Message-----
From: 	Dave Bradley [mailto:daveb@...] 
Sent:	Friday, 28 July, 2000 4:21 PM
To:	MOTM List
Subject:	[motm] FW: [AH] Frequency doubling for VCOs

(Amazing - a nice programming trick amidst the dreck that is usual to AH -
Moe)

Frequency Doubling for VCOs

Here's an idea to create a sound one octave above the normal pitch. Maybe
it will help us focus on analog stuff...

We all know you can get sub-octaves by putting flip-flops after the VCO
output.
One pulse sets the FF output to high, the next pulse switches the output to
low, and so on.

You cannot use this technique to get higher than original sounds simply
because you don't know when the next pulse is due.
Or do you?

If you take an ascending (for now) sawtooth, you can use a CPR
(comparator) module to get a square wave. The second half of the wave will
yield a positive output.

If you invert that pulse-out, and add it to a the sawtooth output (after
the CPR), you get two rising ramps for every VCO cycle.

If you adjust the levels so that these two ramps are identical, you have
created a sawtooth wave one octave above the VCO's pitch.

Better still: If you still have modules to spare you can repeat this trick
for more octaves!

Sounds like a new PIC-based MOTM module...?  :) 

 -----Original Message-----
From: 	The Old Crow [mailto:oldcrow@...] 
Sent:	Friday, 28 July, 2000 4:45 PM
To:	'motm@egroups.com'
Subject:	RE: [motm] FW: [AH] Frequency doubling for VCOs


On Fri, 28 Jul 2000, Tkacs, Ken wrote:

> You can also use phase-locked loops. The number of serial flip-flops
> you put in the loop determines the number of octaves shifted.

  Not to mention you can add a divider after the PLL multiplier output and
end up with a frequency that is the ratio of the multiplier and divider:
M/N.  Make the 'multiplier' (loop divider) and 'divisor' (final divider)
programmable, attach a bit of microcontroller fun to acquire control
voltages to set the M and N integers, and you have something I call a
"Voltage-Controlled PLL Ratio Divider"

  Want a 5th above the tone? select M=3 and N=2.  And so on...

  (I have one of these on a protoype PCB).


Crow

/**/

On Fri, 28 Jul 2000, Tkacs, Ken wrote:

> You can also use phase-locked loops. The number of serial flip-flops
> you put in the loop determines the number of octaves shifted.

  Not to mention you can add a divider after the PLL multiplier output and
end up with a frequency that is the ratio of the multiplier and divider:
M/N.  Make the 'multiplier' (loop divider) and 'divisor' (final divider)
programmable, attach a bit of microcontroller fun to acquire control
voltages to set the M and N integers, and you have something I call a
"Voltage-Controlled PLL Ratio Divider"

  Want a 5th above the tone? select M=3 and N=2.  And so on...

  (I have one of these on a protoype PCB).


Crow

/**/

>Frequency Doubling for VCOs
>
>Here's an idea to create a sound one octave above the normal pitch. Maybe
>it will help us focus on analog stuff...
>.....

Or you can just use the trick Tony Clark showed me.  Just take
a triangle wave and run it through a full wave rectifier.  Presto!
2x Fx.  He has some very interesting VCOs.  Larry Hendry and anyone
else who attended AHMW2000 can attest to that.

-Cary

Cary Roberts wrote:
Or you can just use the trick Tony Clark showed me.  Just take
a triangle wave and run it through a full wave rectifier.  Presto!
2x Fx.  He has some very interesting VCOs.  Larry Hendry and anyone
else who attended AHMW2000 can attest to that.
---------
I have done that before too just messing' around.  You can get some nasty
distortion for other waveshapes though.  I will certainly attest as Cary
said that Tony's VCOs were very interesting.  They are not something for
sale, but ones that he made DIY for himself.  He must of provided
waveshaping for all waves AFTER the fullwave rectifier as his VCOs had all
waveshapes out plus all waveshapes at 2X frequency. However, Tony is one of
those banana jacks guys.  And, the crazy thing is his circuit boards were
all SMD (hand soldered).  Now, very few of us have that kind of patience.
Tony is a talented guy.  Next year at AHMW I just MUST spend some time
playing with his modular gear.  Aside from his DIY stuff, he has one of
those Fenix semi-modulars and some interesting Modcan stuff.  The Fenix is
kind of wild looking, but has stuff way too crammed into tiny spaces for my
personal taste.  I guess they only made about 75 of those.  The Modcan stuff
he had looked like nice quality, but the panels were all mixed with jacks
and controls.  I am used to the non-cluttered MOTM approach to jacks and
controls on separate parts of the control panel.  All-in-all, I thought it
was a very interesting set up.

Cary don't you still have some photos at your site?

Larry Hendry

>Cary don't you still have some photos at your site?

Yep.  Check out the following.  Click on the pic for a bigger shot.

http://www.retrosynth.com/slideshow/ahmw00/tony/03.html

Also visit:

http://www.germanisches.com/AHMWNoR2000/DCP_0138.JPG
http://www.germanisches.com/AHMWNoR2000/DCP_0141.JPG



-Cary

> Or you can just use the trick Tony Clark showed me.  Just take
> a triangle wave and run it through a full wave rectifier.  Presto!
> 2x Fx.  He has some very interesting VCOs.  Larry Hendry and anyone
> else who attended AHMW2000 can attest to that.

Take it one step further and build a "periodic" full wave rectifier.
(The Interpolating Scanner can do this.)
Now you can get various frequency ratios by changing the input level:

f_out = N * f_in  ,  N = function of input drive.

For certain input levels, a triangle input will result in a triangle
output, only at N times the frequency. Intermediate input levels
will result in a sound similar to oscillator sync (but it works
with *one* oscillator, of course).

And no built-in lag effect as with PLLs.

If you want a similar thing with pulse waves rather than triangle
waves, I have a very simple little circuit built from a few
quad comparators:
http://home.debitel.net/user/jhaible/jh_720_vco_scanner.gif

The word "scanner" is misleading here, because nothing is
actually scanned. The name is derived from the other
Scanner, which can do this frequency multiplication
as a "side effect". (http://www.synthfool.com/diy/jh_ipscan.html)

My suggestion is try the simple comparator based one
to do some first experiments, but let's have the second
one as a regular MOTM module some time to cover
the whole scope of its applications in one module.

JH.

In a message dated 00-07-28 16:48:20 EDT, you write:

<< Not to mention you can add a divider after the PLL multiplier output and
 end up with a frequency that is the ratio of the multiplier and divider:
 M/N.   >>



crow,
great to see this idea surface again !  i had kludged up a non-v.c. version 
from an article in "synapse" years ago called a "digital dronezilla". sadly, 
i shorted it out and killed the divider chip and couldn`t readily locate a 
replacement. the ratio ( m/n ) aspect made it great for  non-12  music.
best,
dave