Moog phaser, frequency shifter schematics

[ExpAnonColin]

http://home.clara.net/dragonser/circuit/circ.html

Gigantic schematics... doubt anyone will build them, but hey.

-Colin

[Mark Hammer]

The Stage Phaser is yet another super-deluxe Small Stone (12 stages of CA3094 allpass filtering).  One of the more intriguing things about it is the use of several compression and limiting stages, all optically-based.

Before anybody goes nuts and starts thinking they've found an analog harmonizer, note that frequency shifters work by moving ALL frequency content by a fixed amount (e.g., 150hz).  This completely screws with the relationships between harmonics since proportions are NOT preserved.  This is entirely different than harmonizers where harmonic relationships ARE preserved.  They are both musically valid effects, but are nothing at all like each other.  You can hear sound samples of one Mike Irwin designed for MODCAN here: http://www.modcan.com/sound.html  Scroll down a bit and you'll see links.

[ExpAnonColin]

The Stage Phaser is yet another super-deluxe Small Stone (12 stages of CA3094 allpass filtering).  One of the more intriguing things about it is the use of several compression and limiting stages, all optically-based.

Before anybody goes nuts and starts thinking they've found an analog harmonizer, note that frequency shifters work by moving ALL frequency content by a fixed amount (e.g., 150hz).  This completely screws with the relationships between harmonics since proportions are NOT preserved.  This is entirely different than harmonizers where harmonic relationships ARE preserved.  They are both musically valid effects, but are nothing at all like each other.  You can hear sound samples of one Mike Irwin designed for MODCAN here: http://www.modcan.com/sound.html  Scroll down a bit and you'll see links.

Thanks for the insight  Frequency shifters ARE still very interesting devices though, even with guitar.  Not for 12 string simulation but for a more interesting effect.

-Colin

[Tim Escobedo]

Frequency shifters are usually pretty pricey. A couple analog synth companies make them, the modules sometimes ranging above $1k!

Alesis makes a small digital effect that supposed to do frequency shifting, among other things, called the Bitrman. Last time I checked, they cost only $50, though no CV inputs or other goodies. At that price, it's probably more attractive than trying to build.

[ExpAnonColin]

Frequency shifters are usually pretty pricey. A couple analog synth companies make them, the modules sometimes ranging above $1k!

Alesis makes a small digital effect that supposed to do frequency shifting, among other things, called the Bitrman. Last time I checked, they cost only $50, though no CV inputs or other goodies. At that price, it's probably more attractive than trying to build.

The bitrman really does that?

I think it would be interesting to compare the sound of a 4 quadrant multiplier and a frequency shifter.

-Colin

[computerjones]

a 4 quadrant multiplier(balanced modulator) sounds exactly like both the up and down shift together on a frequency shifter.  in fact they are doing the same thing.

[ExpAnonColin]

a 4 quadrant multiplier(balanced modulator) sounds exactly like both the up and down shift together on a frequency shifter.  in fact they are doing the same thing.

Well, thank you for clearing that up

-Colin

[Tim Escobedo]

I thought frequency shifters used some kind of rejection of one sideband, a effort that complicates ring modulator design several times over with a result that still sounds pretty much like ring modulation.  :wink:

[ExpAnonColin]

I thought frequency shifters used some kind of rejection of one sideband, a effort that complicates ring modulator design several times over with a result that still sounds pretty much like ring modulation.  :wink:

:lol:

[R.G.]

I thought frequency shifters used some kind of rejection of one sideband,

That's one way to do frequency shifting - called the heterodyne technique in the radio biz. You nonlinearly bang signal against a carrier, get sum and difference, then filter out the sideband you want. Ugh! Nasty, hard to compensate, drifts, all the things you can imagine.

The frequency shifter in the schemos works a different way. There is a quirk in the math of trig identities that says that if you multiply the sine and cosine of a signal by the sine and cosine of another signal and add/subtract the results just right, you get out just the sum or difference, not both. This is the hot ticket right now in the RF biz for Frequency modulation. It's called I/Q modulation, the Q standing for quadrature, not sure what the I's for.

The stuff in the shifter that looks like phasers but with no modulation is generating two signals that vary by 90 degrees over a wide frequency band. The two signals are then multiplied separately by the true/inverted version of a modulation signal then summed, and you get out the original signal shifted up or down by the modulation frequency.  It's much simpler than other ways of doing FM (if you can afford to stuff all that junk inside a chip that is) and so most RF ICs that do FM in cell phones, wireless phones, etc. use I/Q modulation.

The frequency shifter is just a setup to do the same thing with audio carrier and LFO modulation, down around 0 frequency on the modulation signal.

[ExpAnonColin]

Interesting insight as always, RG.

-Colin

[Paul Perry (Frostwave)]

?? can't seem to load down Phaser pages 4 & 5 (tried 2 days)??

[puretube]

the "I" stands for the "in-phase"-component (0 deg.), the "Q" (as stated by R.G.) for the "quadrature-component" (90 deg. shifted out-of-phase compared to the "I").

That mentioned 0 Hertz region (0.01 to 0.00 to -0.01) is the fascinating part....

(expecting lots of   about "negative freq.")...

[puretube]

?? can't seem to load down Phaser pages 4 & 5 (tried 2 days)??

go here:
http://freespace.virgin.net/dragon.servicing/circuit/circ.html

[Mark Hammer]

Although there is a categorically similar loss of pitched quality in both the frequency shifter and ring-modulator, they do not sound exactly the same.   In the case of frequency shifting the relationship between harmonics and between fundementals in the case of a polyphonic input changes.  I suspect frequency shifters are also more tolerant of polyphonic inputs.  You get some glimmer of this from Mike's soundclips.

One of the cool things about Mike's design is that you simultaneously get upshifted and downshifted versions of the input signal and rotation of the shift-amount control affects both at once.  Let me tell you it is a remarkable effect to have a music input pull apart in that way.  The two versions of the input diverge more, but they each become less pitched as the fixed amount of shifting starts to screw with the relationships between harmonics more and more.

[Jason Stout]

(expecting lots of about "negative freq.")...

Ok, you got me...tell me how it is so.