Magnus freq doubler.

[Magnus Danielson]

From: "Bjorn Julin" <bnillson at hotmail.com>
Subject: Re: Magnus freq doubler.
Date: Sun, 17 Sep 2000 20:56:29 GMT

> Hi.

Hi there,

> >Phew... I just thought I made a fool of myself just because I did not 
> >actually build it..
> 
> Hahaha, "who" are now transpiring salt water from his back? :-D
> (not to mention all the salt i allready have transpired.)

Well, since I don't take this that serious with myself there is no heavy
transpiration due to this little incident... ;-D

> >The whole trick is to build a 50 % PWM to be used to add another >reset 
> >jump. When one makes a sawtooth phase shifter, one takes the >same idea but 
> >instead moves the jump. This causes a DC error, but >since this is linearly 
> >dependent with the phase voltage, it is just >a matter of adding the phase 
> >voltage at the
> >output summing to be home free.
> 
> Exactly, this is allso done in the ARP chroma, but the
> chroma used to many components for what i was looking for.

Well, both the doubler and the phase shifter is a low-count job. I put some
effort into keeping the component count down for both and dipping the diode
will also allow the trowing of a resistor and the curcuit should preform
better ;)

> Actually, my reason for this frequency doubling and soft sync
> issues is that i have been laborating around a design for some
> time now, that uses "one" analoge oscillator to create additive synthesis in 
> a combination of 0-180 deg phase shifting and voltage controlled soft sync 
> to create new waveforms. So i had tried a
> load of different solutions  and become desperate for a simple
> sawtoth freq doubler.Ofcourse one can do all the stuff with a
> bucket load of componenets, but the trick is to do it all with a absolute 
> minimum of components. So far the oscillator sounds wery interesting and i 
> have managed to get over 40 different wave shapes, quite intresting is the 
> simple continous phase shifting (0-180 deg.)
> 
> Phase modulation of both square and sawtoth is possible, as you
> know it produces modest vibrato effects, if one was just able to
> shift 360 degrees!

Well, my sawtooth phase shifter is a 360 degree phase shifter, or at least
very close to 360 degree... just vary the comparator voltage over the full
sawtooth voltage and the comparator drops out the PWM wave and the summing
network will do the rest. Maybe it is time for me to make a schematic of that
and put up along the others?

> Some of the problems i have to solve is amplitudes,
> some sort of DC ALC seams to be a soulution.
> 
> >I guess I should have made a more comprehensive explanation of the 
> > >curcuit,others would have been even more confused about this little 
> > >trick.
> 
> Just some brief explenation would be enough.

Well, if I only spend a few more sentences things will be much clearer for
most people.

> > > >Maybe time to breadboard this thing...
> > >
> Ah, so you ment, "you" was going to bread board it,
> i tought you ment me!!

No... just saying that maybe I should have breadboarded it before putting the
schematic up... so I just though I was out in trouble just for being too lazy
to breadboard it myself.

> > > BJ (going blind in the desperate shearch of a freq doubler)
> 
> >Hopefully will this one do the trick, or at least act as an source of
> >inspiration. Maybe I should redo it without the diode...
> 
> What important is exact amplitude balancing of the saw and square.
> The diode is ok if one wants unipolar wave only .

Ah, but the thing is that the whole thing is balanced!
I get a DC error which I adjust by the resistor to negative feed. Also, this
is an ugly thing since I assume the saw to go from -5 V to +5 V as well as
assuming the powerlines to be at +/- 15V. That is the ugly side of it.
Removing the diode so that I get a (fairly) balanced output from the comparator
also removes the DC error from that signal, so then I can remove the DC
correction. I actually spent part of my efforts just to ensure that I still had
DC offset at 0V!

When Jörgen took the doubler schematic and turned it into a phase shifter based
on my hints he realized this and things became even simpler.

> Why not using a three way switch?

Uhm, I am not sure that I know how it would help... ah well..

Cheers,
Magnus