New Design - The Glitchshifter.

[nelson]

Hi folks,

This is going to be my entry for the FX-X August competition for a noise making circuit.

An analog continuously variable pitch shifter.

Schematic here

I haven't breadboarded it yet, but will find the time this weekend and work out any bugs.


It is a bare bones phase locked loop based pitchshifting pedal. It uses a 555 set up as a variable frequency divider in the loop.

Walking through the circuit function from the input. We have a standard transistor buffer to provide a suitable input impedance to the guitar, this is then fed into a fuzz face esque two transistor square wave generator and low pass filter. This is fed into the signal input of a CD4046. The CD4046 is set up absolutely bare bones and is very similar to the "simple guitar synth". Changing the Glitch pot alters the low pass filter of the CD4046 allowing the VCO to oscillate and will cause lots of locking errors, hence the glitch title. The Oscillation pot controls the frequency of this oscillation - this is the noise maker portion of the circuit. The pitchshifter pot controls the frequency division of the VCO out frequency, variable from dividing (not sure on this) by 1 to dividing by 5 - so a range of 5 octaves *up* is possible because this is then fed to the phase comparator input. At this point the VCO out is tapped and fed into one side of a one pot mixing stage, of which the other side is fed by the straight square wave signal tapped off before the signal input of the CD4046.  This signal is then boosted by a simple one transistor gain stage which is then fed to the volume pot.


That is how I *think* it will work.

I guess I am looking for corrections and for the idea to be shot down, it can't be *this* simple to create a continuously variable analog pitch shifter - no matter how error prone it will be.

Also, given this, does anyone have any design ideas of how I could also get pitch shift down, any ideas for a resistively variable frequency multiplier and divider in the loop as opposed to just the frequency divider?

[John Lyons]

Sounds interesting.
Looking forward to seeing/hearing it.
Sorry, no advice...over my head. 

john

[ambulancevoice]

well, if it does work, could you add something to stop the variability of the pitch an turn it into a regular shifter? like a switch or something, it'd be cool to be able to switch between set pitch and variable pitch

[Pedal love]

I guess I just don't understand your use with this circuit. Are you using it to condition another circuit after this one? Will it be used with an envelope filter?

[markusw]

Cool! 

Looks interesting! Will have to check out how that 555 part works....

5 times btw is a bit more than two octaves up. 5 octaves would be 2*2*2*2*2 = 32
Anyway, I'm keen to hear some samples 

[nelson]

well, if it does work, could you add something to stop the variability of the pitch an turn it into a regular shifter? like a switch or something, it'd be cool to be able to switch between set pitch and variable pitch

That would be possible, you could just switch out the pot for various resistors.

I guess I just don't understand your use with this circuit. Are you using it to condition another circuit after this one? Will it be used with an envelope filter?

It is a fuzz/pitchshifter/noise maker/Octave pedal.

I don't think much else is needed....

I will think do some control mods when I get it running. I was thinking that it would be neat to have the pitchshift pot controlled by either an expression pedal or a probe like circuit.

Cool! 

Looks interesting! Will have to check out how that 555 part works....

5 times btw is a bit more than two octaves up. 5 octaves would be 2*2*2*2*2 = 32
Anyway, I'm keen to hear some samples 

Of course you're totally right about the frequency multiplication Markusw - it was late when I posted the thread.... What I wasn't sure about: what's the range of the frequency division of the 555. It's straight out of a Mimms book! I won't know until I finish bread boarding it and scope it.

If I can get it to work expect to see some youtube vids of it. 

I have a night of pizza, music and breadboarding ahead of me - having to remove the Cmos based Foxx Tone machine which was a tad disappointing.



 

[alex frias]

Good luck, my best wishes on your adventurous quest!

[foxfire]

I have a night of pizza, music and breadboarding ahead of me - having to remove the Cmos based Foxx Tone machine which was a tad disappointing.

sounds like a good night to me. sure the neighbors won't want to hear it when you get it working but oh well. rylan

[bioroids]

The pitchshifter pot controls the frequency division of the VCO out frequency, variable from dividing (not sure on this) by 1 to dividing by 5 - so a range of 5 octaves *up* is possible because this is then fed to the phase comparator input.

Sounds interesting, but I think when you divide the frequency you get lower pitch sounds, not higher. Divide by two is an octave down, not up. Am I right?

Regards

Miguel

[soggybag]

This reminds me of something from the Stompbox Cookbook. I posted a picture of it a while back I'll see if I can dig it up and post again. It was a HF oscillator controlled by the instrument that drove a phase lock loop. It was called FM Distortion. Here's a picture. The missing section is just an output buffer. I tried building this on a breadboard, but could not get it to work.

Here's a link to the thread. http://www.diystompboxes.com/smfforum/index.php?topic=65215.0

[nelson]

The pitchshifter pot controls the frequency division of the VCO out frequency, variable from dividing (not sure on this) by 1 to dividing by 5 - so a range of 5 octaves *up* is possible because this is then fed to the phase comparator input.

Sounds interesting, but I think when you divide the frequency you get lower pitch sounds, not higher. Divide by two is an octave down, not up. Am I right?

Regards

Miguel

Not when it is used in the loop of a phase locked loop.

I got to the breadboard last night, then realised I had no CD4046!

So, I set up the 555 as a frequency divider. But couldn't get it to work dividing down the frequency of a 555 set up as an oscillator.

After more research, it is not *actually* a frequency divider, it just extends pulse width. It sort of divides frequency, but the way it does it is it increases the initial pulse width by an amount set by the pot and 0.1úF cap to ground. Not allowing the input to trigger another pulse until the initial pulse is finished. Unlike more standard CMOS frequency dividers which essentially redistribute the signal on each pulse therefore dividing the frequency while maintaining pulse width. Unfortunately, I don't think this will work for the purposes here.



I want to do something with a PLL so, I am going to make an analog arpeggiator instead. I will try to make it as simple as possible, but inevitably it is going to be much more complex.
 

[Pedal love]

Can't you use a flip flop like a 4013 or something with the PLL. You might find something that works.

[nelson]

Can't you use a flip flop like a 4013 or something with the PLL. You might find something that works.

I'm probably going to use a CD4024 and some switching.

[pinkjimiphoton]

paul, just curious, did you ever get this project working?
looks really neat!

[nelson]

paul, just curious, did you ever get this project working?
looks really neat!

I never actually got past the breadboard stage. I abandoned it as I couldn't get the 555 to behave. 

I moved onto something else. Apparently I have the attention span of an ADHD diagnosed child on mescaline being entertained with shiny objects.

I might revisit the idea some day.

I'm sure something cool could come out of it.

[pinkjimiphoton]

that would be really....oooooh, shiny!!!!!