MXR Phase 45 with JC Maillet's mods + my Resonance mod

[frequencycentral]

I just built this circuit. It is indeed a very subtle phaser. A little too subtle for my liking, so I've been tweaking about with it and come up with an extra mod of my own. I wanted to add resonance, but I can't afford the space of an extra knob, I have got space for another switch though. I've added a switch and a 68K resistor between the input and the output. Very simple. Throwing the switch of course adds 'feedback' or 'resonance'. The 68K value was chosen to prevent self oscillation. I measured 55K as the absolute lowest limit at which it's possible to prevent self oscillation. The amount of resonance can then be controlled with the 'Mix' pot. The result is what I would describe as very 'chewy' and filtery. I'm finding this pedal much more interesting with this mod, just thought I should share it. I hope no one minds if I post up the schematic with my extra little mod on it.



EDIT: Here's a link to JC's site with details on his mods: http://www.lynx.bc.ca/~jc/pedalsPhase45.html

[Cliff Schecht]

nvm

[frequencycentral]

nvm

no visible meaning?

[gigimarga]

NeVerMind

[frequencycentral]

NeVerMind

....and that means what exactly?

[mdh]

I find that I really like the stock Phase 45 with the rate nearly maxed for sort of Leslie-ish sounds for blues.  But if you want an over-the-top swooshy phase, it's definitely not the circuit.

I am curious about your resonance mod, though... maybe R.G., Mark, or any of the other modulation gurus could comment on it.  As far as I know, the typical resonance arrangement for phasers is to feed back over an odd number of stages, so, e.g., in a 4-stager, the feedback would come from the output of stage 4 to the input of stage 2, not the input of stage 1 or the circuit input.  I think there's a reason for this.  Of course I would also think that you would want to feed back to a point after the input buffer, not before.

[frequencycentral]

I'd have to agree. Stock, it's a better vibe than a swooshy phase. I'm planning a few different phasers and I wanted this one as a point of reference. I'll try feeding the output back in after the input cap as you suggest. I think the way I've got it set up I'm mixing the out of phase signal back into the input, I don't think there's another way to do this in a 2 stage design. And yes I was kind of hoping Mr Hammer would chip in at some point. I think tomorrow I'll also try the equvalent of Mark's Ropez phaser/filter idea on this circuit on either the first or second stage.

[moosapotamus]

If you're tinkering with it, maybe you could also try taking the feedback directly from the output of the last stage, instead of after the mix pot, and feeding it back into the inverting input of the input buffer. If that sounds any good, maybe a three position toggle switch (center off) to select connecting the feedback to either the non-inverting input of the input buffer (positive feedback), or the inverting input (negative feedback), or off (no feedback)... just thinking out loud.

~ Charlie

[frequencycentral]

If you're tinkering with it, maybe you could also try taking the feedback directly from the output of the last stage, instead of after the mix pot, and feeding it back into the inverting input of the input buffer. If that sounds any good, maybe a three position toggle switch (center off) to select connecting the feedback to either the non-inverting input of the input buffer (positive feedback), or the inverting input (negative feedback), or off (no feedback)... just thinking out loud.

~ Charlie

Thanks Charlie, I'll try that idea too! there might be more mileage than I originally thought.

[Eb7+9]

I just built this circuit. It is indeed a very subtle phaser. A little too subtle for my liking

to give the 45 more goose you may want to replace the 10k linearizing resistors by 22k or 33k values - together they also limit the scope of resistance change in the jFET's ... I've posted this idea before but maybe wasn't clear about it - try it and see if it does anything for you

[frequencycentral]

I just built this circuit. It is indeed a very subtle phaser. A little too subtle for my liking

to give the 45 more goose you may want to replace the 10k linearizing resistors by 22k or 33k values - together they also limit the scope of resistance change in the jFET's ... I've posted this idea before but maybe wasn't clear about it - try it and see if it does anything for you

Thanks JC, but whick 10K's? There are dozens of them! Do you mean the two from the vref?

[Eb7+9]

I'm talking about the ones (four in all) that are connected across the Drain and Source of the two jFET devices - they're connected in series with the mid-point leading to a 0.01u cap that's also tied to the Grid of the jFET devices ... together (with the cap) they form a linearizing network that parallels the non-linear channel resistance of the jFETs ...

In effect, a total shunting resistance of 20k is applied to the channels in a stock 45, which is quite low for phasor circuits (the lowest I know of anyway) ... sometimes you might see equivalent (single or split) values go up to 100k in other phasors, so in theory 47k's should produce a dramatic change from the 10k's ... the only downside is you might perceive an increase in signal distortion

Obviosuly there's a tradeoff between phasor range and linearity ... the 45 was designed as a "clean" studio quality phasor - and this is the gist of how it was done

[Mark Hammer]

No fabulous insights to offer here, but it is worth noting that feedback can occur from many different places going to many different places.  The usual route that phaser feedback takes (from last stage to 2nd stage) is predicated on the idea that phase-shifted and original signal are summed and that the number of stages results in no wholesale phase inversion (although obviously there is "shift").  So, if you wanted to produce a more emphatic effect, added feedback through an odd number of stages yields that phase inversion and an enhanced differencing.

The Small Stone, on the other hand, takes the feedback all the way back to the input stage before the wet/dry split.  Different use of feedback.  I remember some discussions with Mike Irwin regarding the sonic difference between feedback loops in flangers that were outside the compander vs inside.

So, what you have with the P45 is essentially equivalent to the Small Stone in its functioning.

[frequencycentral]

@ JC: Thank's I'll try that. I built this on a 37x9 perf with all the pots attached to the board. It was a bit of a feat to get everything on, but I'm determined to get every last drop of goodness out of it and generally mod the hell out of it before I box it up, so I'll try those resistor changes. I'm not finding the Bias pot so useful, so I'll probably replace it with a trimpot, and maybe try adding a Depth pot in the space I have spare. I'm thinking that the 3M9 resistor could be varied in some way with a 1M pot to good effect.

@ Mark and Charlie: I've tried various different feedback arangements, some don't work at all, some are not variable with the Mix pot like the initial one I posted about. So I'll stick to the original plan. Is there any reason why I shouldn't do it this particular way?

[Mark Hammer]

@ Mark and Charlie: I've tried various different feedback arangements, some don't work at all, some are not variable with the Mix pot like the initial one I posted about. So I'll stick to the original plan. Is there any reason why I shouldn't do it this particular way?

My answer is "No".  Your feedback path is essentially feeding a signal with a notch in it to the input, and is tantamount to feeding the output of one P45 to the input of another, except their tracking each other.  It's obviously not gonna be as awesome as feeding back a signal with 2 or 3 notches, but if it works, it works.  My only real concern might be how it interacts with other pedals before or after, given that the feedback path is a fairly low resistance one.