[sdiy] Thru-Zero, was as3340 PW/tuning issue

[David G Dixon]

Adam,

 

To answer your question, let's talk about triangle-core VCOs for a moment,
since this is what I have the most direct experience with.

 

The typical tricore VCO is a "tri-square" core - a comparator compares the
sum of the triangle wave and a voltage-limited (i.e., with back-to-back
zeners) square wave (generated from the comparator) with ground.  When this
signal passes through ground, the comparator and the square wave flip
states.  This changes the direction of current to the triangle integrator,
and defines the voltage of the triangle apices.  That's how a basic
tri-square oscillator works, such as you would find in a Dixie VCO.

 

To get Linear FM on this beast, a linear VCA is inserted in the loop between
the square-wave generator and the triangle integrator.  If the gain of this
VCA is taken to zero, then the oscillator stops.  If the gain of this VCA is
taken negative, what happens?  The triangle and the square wave are now not
in the correct phase relationship, so that their sum will move away from
ground, and will get stuck at one of the voltage rails.  Again, the VCO will
stop because the comparator will never flip.  In order for the VCO to
continue oscillating under this condition, the triangle signal being summed
at the comparator input must be inverted.  This requires a bit of logic to
read the polarity of the Linear FM VCA's gain (or some other convenient
indicator of the triangle current's direction) to drive an optional
inverter.

 

When this is all done correctly, the VCO will continue to oscillate normally
even though the Linear FM gain has gone negative (i.e., has passed "through
zero").  If the TZFM circuit has been designed well, then there will be no
perceptible "glitch" when the FM passes through zero, and the resulting
waveform will pass smoothly and symmetrically through a slope of zero.  This
is what is achieved with, for example, a Rubicon VCO.

 

If Linear FM gain is broad and passes through zero this gives a much deeper
and richer-sounding FM.  It is the basis of bells, mallots, and all sorts of
other strange and wonderful effects.  You generally do not find TZFM on
Roland, Korg or Sequential synths.  There was (is) nothing fundamental
preventing these makers from incorporating analog TZFM - they just chose not
to.  When I finally get around to building my own dream monosynth, it will
definitely have TZFM built in (it's actually pretty easy, and getting easier
- I've got it down to almost no additional circuitry beyond the conventional
VCO core - maybe two or three additional opamps and a JFET - thanks to my
beloved 2164).  Hopefully, that will happen sometime later this year.

 

Dave

 

  _____  

From: Synth-diy [mailto:synth-diy-bounces at synth-diy.org] On Behalf Of Adam
Inglis
Sent: Saturday, January 04, 2020 6:25 PM
To: ColinMuirDorward; Synth DIY
Subject: [sdiy] Thru-Zero, was as3340 PW/tuning issue

 

 

Thanks Colin

Someone in that thread says they would love a "through-zero version".

 

Seems to be a lot of talk about TZ all over these types of forums - why?
What's the big deal? How are these different from the VCOs in my Roland,
Korg and Sequential synths?

 

Sure I could google but I know I'd get better education here ;-)

 

Adam

 

 





On 5 Jan 2020, at 6:26 am, ColinMuirDorward <colindorward at gmail.com> wrote:

 

Just a note to say thanks to everyone contributing to this conversation.
I've applied lots of what I learned here to my design, and I've got the
PWM/pitch interaction down to what I think is an acceptable minimum.
Anyways, it's nothing special, but I've started a build
<https://www.muffwiggler.com/forum/viewtopic.php?f=17&t=225670>  thread over
at muffs, for anyone who needs yet another low cost 3340 vco. I need
several, so thought I may as well go public in case anyone else does, too.
I'll be using JLCPCB assembly for all the resistors and caps, and the rest
will be through hole at home. I have lots of as3340s, so can include with
the PCB if anyone needs. 3340 is running on +10/-5, so the circuit should be
at home in 12v or 15v setups. I'm in Yukon, Canada.

Thanks,

Colin

 

On Fri, Jan 3, 2020 at 12:39 PM David G Dixon <dixon at mail.ubc.ca> wrote:

The Intellijel Rubicon (and Rubicon-2) gives the option of generating pulse
waves from Triangle, Saw, or Double-Saw (twice the frequency).  It's not a
whole synth, just a TZFM VCO.

 


  _____  


 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20200105/d7401c1e/attachment.htm>