Phasers and univibes: sow's ear and silk purse

 :

4/15/1997 8:30 AM

Mark Hammer

Phasers and univibes: sow's ear and silk purse In principle, a Univibe is a phase shifter with a   few cosmetic changes. Instead of having several   cascaded stages with 90 degrees phase shift at the   same frequency for each stage, different value   capacitors are used for each stage to produce small amounts of phase shift at multiple frequencies, instead of large amounts at a single frequency. Yah, yah, I know there is some sort of voodoo magic about the light-bulb based swep waveform that creates a more pleasing sweep than   an orthodox triangle-wave, and I accept that.     Assuming one is none too fussy about the sweep waveform, does changing the capacitors in a   standard 4-stage phase shifter produce a tone more characteristic of a Univibe? Is there some magic combination about how far the cap values (and concommitant phase-shift notches) are spaced, or how the stages are cascaded?     In part, I ask this because I re-stumbled onto a schematic of an older Ibanez phase-shifter that I found in a circuit anthology some years ago (I forget which one). The circuit had six stages of phase shift, and used some sort of custom 6-FETS-on-a-chip for the swept element, none of which is too remarkable. However, every cap value in the phase-shift stages was different (generally half the value of the stage preceding it). I've never heard the particular device, but it conforms more to a Univibe topology than a standard phase shifter. Anyone know this device, or have any input on what it takes to make a phase-shifter sound vib-ish? Are phase-shifter owners sitting on the guts of a Univibe without knowing it, or is there something else to consider?

 

4/16/1997 8:51 PM

R.G. Keen

Hi, Mark.     The voodoo in the LFO is just that it's a sine wave kind of almost predistorted for driving that incandescent lamp.     There is an article in an old Journal of the Audio Engineering Society you'll want to read. It's from the early '70's, and treats exactly on whether all the same is better than spread caps. The upshot the article claimed was that all the same is better. It makes a good read.     Is the phaser you found the one from the National Semiconductor Audio databook? Sounds like it. They used one of their jfet analog mux chips for a set of matched fets.     Since trying to decipher the univibe and making a printed circuit board for it, I've damn near beaten that thing to death trying to figure out why the cap values are what they are. They're clearly not octave related, nor any pregression I can figure out that makes musical sense. The best I have done is to just swap out values in one of the boards I've made. It's not clear that the vibe-ishness is in the caps, but swapping those around can make things sound different.     I think it is partially the huge range of resistance values that the LDR's can do, partly the residual distortion of the transistor implementation, and partly the LFO waveform. Perhaps there is something to the matching of resistance values at the output, as I notice that when they are not right, so that you get good cancellations at some frequency, the effect gets thin. This may account for the large variation in phasing sound in original univibes. The first one I heard was NOT impressive. Then I fixed one that sounded like a deity's voice. Same circuit.

 

4/17/1997 11:31 AM

Mark Amundson

The question of same value caps or a few   different values is all in what the designer thinks   the best sound is.     The same value camp wants a deep filter notch or two for the phase shifter and the mult cap size family wants a comb of notches across   the instrument's audio band.     Both approaches sound good and combinations   of both approaches together sound great as   well.     Mark Amundson, (mamundso@mr.net)

 

Page 1 of 1