PLL Freq Mult Ver 2 (Still prelim) text

Scott Gravenhorst, Synthaholic chordman at ix.netcom.com
Sat Aug 26 21:56:51 CEST 1995


                    
                      P R E L I M I N A R Y
                            VERSION 2

PLL FREQUENCY MULTIPLIER MODULE            Scott R. Gravenhorst
                                         chordman at ix.netcom.com
                                                      23-Aug-95

This circuit is a multi purpose phase locked loop designed for 
frequency multiplication.  Although it was designed around a PAiA 
FatMan, it should be adaptable to almost any analog system.  The 
circuit is based on a 4046 type micropower phase locked loop IC 
and uses a divide by N counter inserted in the loop to force the 
PLL's VCO to run at N times the input frequency.  It's output is 
provided as a TTL square wave.  You must design an appropriate 
attenuator or amplifier, whichever is necessary for your system.  
You may also wish to add a low pass filter at it's output to 
reduce the odd harmonics present in square waveforms.  

This is a sound generating module, intended to enhance a synth by 
tracking a VCO to produce it's own waveform at a given frequency.  
Because it is a phase locked loop, it has characteristics of it's 
own, such as frequency multiplication and a controllable lag when 
the input frequency changes.  

Note that only two multiplication factors are provided by this 
design: 2x and 3x.  Other factors could possibly work by changing 
the number of flipflops used by the walking ring counter (74164).  
If this is done, other circuit parameters will likely have to 
change.  In particular, the design of the low pass filter, the 
size of the PLL VCO MIN/MAX pots and the PLO timing capacitor.  
Only these two were provided for drawing simplicity.  

PAiA FatMan INPUT INTERFACE:

The 1458 op amp (which can really be almost any dual supply type 
op amp) is nothing more than a level converter.  The FatMan uses 
a 555 timer IC, powered below ground, for it's VCO threshold 
detector.  The 4046, which (here) runs on +5 Volts (Vcc), 
requires a positive signal with respect to ground.  The 1N914 
diode clamp prevents the signal fed to the 4046 from going below 
ground.  Pin 3 of the FatMan VCO 555 IC provides a pulse every 
time the oscillator cycles.  This pulse is level converted to the 
required range and presented to the 4046.  In this design, the 
4046 needs a rather healthy input signal to stay in solid lock.  
If you are using a synth other than the FatMan, make sure you 
provide the 4046 a square/pulse wave that toggles between Vcc and 
ground.  (Not just TTL, which can be .3v to 2.6v or thereabouts) 

CONTROL SWITCHES:
                                                         
2F / 3F
In my version, this switch allows 3 different multiplication 
factors: 2 times or 3 times.  2F provides an octave above the 
input and 3F provides a perfect fifth plus an octave above the 
input frequency (19 semitones above the input pitch).  My version 
also includes a switchable div by 2 counter at the frequency 
multiplied output to allow for a 1.5x multiplication, i.e. a 
simple perfect fifth above the input frequency.  Please feel free 
to add even more taps for other harmonic values, as well as 
adding frequency dividers to the output.  I think I will need to 
go buy a couple of big rotary switches!  

1:1 / MULT
This switch cuts in or out the div by N counter.  In the MULT 
position, the counter is inserted and the PLL will multiply the 
input frequency by whatever factor is set.  In the 1:1 (1 to 1) 
position, the loop will run with no forced multiplication.  1:1 
can be used several ways:  With phase comp 2 selected, no 
frequency multiplication will occur, but the lag effect will be 
present.  With phase comp 1, the loop will go crazy locking onto 
harmonics of the input frequency.  Many other effects are also 
possible, as this module has many adjustable parameters.  

VCO FREQ CAP SWITCHES: CAP 1, CAP 2, CAP 4
These 3 switches allow for 7 different cap values (you can add 
more).  It is a simple binary system that allows capacitance 
values from .0169 down to .0022 in steps of about .0022 uF as 
follows: 

CAP 4     CAP 2     CAP 1     total capacitance (uF)
off       off       off      PLL DISABLED
off       off       on       .0022
off       on        off      .0047
off       on        on       .0069
on        off       off      .01
on        off       on       .0122
on        on        off      .0147
on        on        on       .0169 

Should you decide to expand the design by adding more switched 
capacitors, each cap switch added will (almost) double the number 
of combinations available.  You can add cap values at either end 
i.e. larger ones or smaller ones.  Smaller values will extended 
the high range of the PLL while larger ones will allow it to 
oscillate lower.  My spec sheet on the 4046 does not specify a 
maximum or minimum cap value.  It does, however, specify 1.3 
megahertz as the maximum VCO frequency, well above audio 
harmonics.  

Note 1: The labels on the switches follow binary powers.
Note 2: With all switches off, the PLL may oscillate at an 
extremely high frequency due to stray circuit capacitance.

DAMPING CAP
This switch cuts in or out an extra capacitor in the low pass 
filter circuit.  Again, it's use is not strictly defined and has 
an effect on how quickly (or if) the loop settles down after a 
pitch change.  It has a dramatic effect on the playing range and  
the sound produced.  Here too, more caps with switches may be 
necessary to allow proper and desired performanbce if you expand 
the system by adding more taps to the divide by N counter.  

RESET
This switch forces the divide by N counter to reset to zero.  
This counter may have random bit settings when powered up and can 
give wierd (you might like them) results if not reset.  Random 
values may also occur when the 2F/3F switch (sets the divide 
ratio) is changed.  Also if the 1:1 / MULT switch is changed.  
Press and release to reset the counter.  Note that if this switch 
is pressed while a note is playing, the counter (as well as the 
output) stops while it is held depressed.  Since while depressed, 
the phase comparator input is receiving 0 hertz, when the RESET 
switch is released, the loop must re-lock.  When this happens, 
you can get an interesting 'sproing' sound, depending on the 
settings of the other controls.  Kind of metallic.  Use caution 
here, you might get a loud 'pop'.  

COMPARATOR
This switch allows using either comparator built into the 4046 
IC.  Phase comparator 1 is a standard XOR type that has a 
narrower lock range and will lock on harmonics of the input 
frequency.  Phase comparator 2 is a leading edge detector type 
and has a much wider lock range and is resistant to locking onto 
harmonics of the input signal.  

--------------------------------------------------------------

POTENTIOMETERS:

PLO F MAX
Sets the maximum frequency the PLL's VCO can produce.  To raise 
the maximum frequency, turn this pot to a lower resistance value.  

PLO F MIN
Sets the minimum frequency the PLL's VCO can produce.  The MIN 
and MAX pots will dramatically effect the operation of the loop.  
The MAX control should be set to a lower resistance value than 
the MIN pot to allow proper PLL operation.  To lower the minimum 
frequency, turn this pot to a higher resistance value.  

DAMP
This pot controls the damping factor of the PLL.  It can be set 
so that the tracking lag is short, long or even to allow 
significant over/undershoot.  Over/undershoot causes a 'sproing' 
sound when the input pitch changes.  This will be more pronounced 
as the setting of this pot approaches zero ohms.  When set to 
higher resistance values, various damping effects occur.  You can 
get an interesting 'squink' sound at the attack of any note that 
causes relatively large input pitch changes.  

HARMONIC PAN
This control selects how much signal is supplied by the PLL 
multiplied output and how much is supplied by a divide by N 
counter (frequency divider) at the PLL VCO output.  It is one of 
the timbre controls of the sound produced.  

I have had alot of fun designing and using this circuit.  It 
allows creation of yet more color and texture for electronic 
sound and music synthesis.  

I would like to thank Don Tillman for his input.

This is a preliminary document.  Please send comments by email to 
me (chordman at ix.netcom.com).  The final text and GIF documents 
will be posted when this project jumps onto a piece of perfboard.  













-- 
There is no 12 step program for synthaholics.  Thank your Superior Being.




More information about the Synth-diy mailing list