Continuously Variable Number of Phase Stages?

[varialbender]

I was thinking about how to fit a rotary switch to choose the number of phase stages, when I thought, I wonder if it could be replaced with a pot?
The idea is a bit strange, and isn't exactly correct.
The thinking behind this is that the number of notches is proportional to the number of phase stages, so getting rid of notches is sorta like decreasing the number of stages.
To do that, I thought it might be possible to do something like the phasefilter mod.
The last 2 stages of the phaser would be turned into a 4-pole low-pass filter, oscillating in sync with the rest of the phase stages.
The low-pass filter would have a variable center, which, at high frequencies, would allow normal phasing, but at lower frequencies, would begin cutting out some of the phasing.
At least, that's how I think it would work.
The higher the frequencies, the closer to 0 amplitude, and when added with the original, nothing would change.
In that sense, the low frequencies would be double amplitude, then there would be notching and peaking, and at some point, it would slope down to unity gain.
If it were desired, the same filters could be subtracted from the original signal and the whole thing doubled so that it would behave as normal:
Double lows, notching/peaking, double highs.
That might be overkill, especially when adding a center control to the phasefilter mod seems very easy, and would probably sound pretty good.

Anyone follow me, or am I crazy/stupid?
I tried searching for something similar, but nothing came up.
Thanks for reading through my crap.

[RaceDriver205]

am I crazy/stupid?
Thanks for reading through my crap.

Geez mate, sell yourself short why dont you! 

I recon thats a pretty good idea. The MXR Phase 180 at GGG (I think), shows how you add extra phase shift stages. Having a rotary switch that removed stages would be very cool, and not too complex to implement.
I think your onto something there.

Speaking of which, I wonder how a phaser with 100 stages would sound. Damn! now Ive just added something else to my build list!

EDIT: MXR Phase 180 with added stages is at 'geofex.com' under 'schematics and more'

[StephenGiles]

am I crazy/stupid?
Thanks for reading through my crap.

Geez mate, sell yourself short why dont you! 

I recon thats a pretty good idea. The MXR Phase 180 at GGG (I think), shows how you add extra phase shift stages. Having a rotary switch that removed stages would be very cool, and not too complex to implement.
I think your onto something there.

Speaking of which, I wonder how a phaser with 100 stages would sound. Damn! now Ive just added something else to my build list!

EDIT: MXR Phase 180 with added stages is at 'geofex.com' under 'schematics and more'

100 stage phaser..................I should think noisy, and will not fit into a 1590 box!!!!!!!!!

I should go for a 23 stage,

[RaceDriver205]

100 stage phaser..................I should think noisy, and will not fit into a 1590 box!!!!!!!!!

Lol, it'd need more like a "crate". I was exagerating... a little.

[varialbender]

Thanks for the support.

Anyone know if this'll work?
I know the slope of the filter is really important, and I'm not sure if 4-pole is steep enough.
Any feedback appreciated, thanks.

[R.G.]

I don't know if this counts as continuously variable, but you can use an LM3914 and ten analog switches, either JFET or CMOS, to select the output of two to twenty phase stages by turning a single pot. Using JFETs with ...really... slow... gate delays would slow the fades between them a lot.

It looks like: Buffer to phase line and output mixer; twenty phase stages; ten taps, at every second phase stage; output mixer fed by the taps and the buffer. The control pot runs the LM3914, and its ten outputs are rigged to run the taps. Is that anything like what you had in mind.

[varialbender]

Yeah, my first idea was similar, but I wasn't sure about it.
I like the idea, but filtering the phased signal has peaked my interest.
Do you have any idea whether or not my suggestion would work? Sound good?
It just seems like such an easy thing to do.
If someone already has a phasefilter mod on, say, a vactrol phaser, I think all you'd need to do is add a pot to limit the current going to the lowpass filter, right?
Anyone have a phaser they can try this on? (who knows what they're doing, so I'm not blamed if that causes exploding... )
Thanks R.G., I'll keep your method in mind.

[TELEFUNKON]

take care of what you`re feeding back!
esp. when you send the FB signal back
through a phase-shifting lowpass filter... (Barkhausen);


C`mon, R.G.: that`s not continuously variable, but "digital";
(or at least Vexish "seek-phasing").

check/google:  "Haible" & "scanner"...

[R.G.]

C`mon, R.G.: that`s not continuously variable, but "digital";

Yeah, it is in that it switches phase notches in and out.

The lowpass filter shutting off higher notches is also going to be a bit odd, in that notches will emerge from non-existance as they travel down into the frequency region that is passed in the phase line. The lowpass method is less abrupt in that the notches at least emerge from the high frequency region. They also emerge from the lowpass filter's phase shift region, as you note, so the notching right at the low pass region might get odd.

I guess like everything else, the devil is in the details. How fast the phase notches move in the frequency domain versus how fast the lowpass region moves and the direction that the phase notches move compared to the lowpass filter would be critical. If the phase notches move up in frequency faster than the lowpass cutoff, they'll all move into non-audibility at full sweep and only be audible as they move out of the lowpass region before the lowpass cuts them off again by sweeping lower.

[Transmogrifox]

I thought it would be interesting to do a 12-stage phaser & use a couple analog switch (like a CD4066), or multiplexor to have an 8-bit switched input to the final mixer: each input coming from each of the final 8 phase stages.   One could program a sequence, and as the different phase stages were allowed in the mix, there would be an overall signal power change (like sequenced tremolo) resulting, along with a different phase response.  I think it could get really strange if you started switching it at audio frequencies.