phase 90 and feedback resistor...Mark?

[swt]

I have just finished my second phase 90. In both cases, if i use the 22k feedback resistor i get distortion. Take it out, and it sounds great. The fact is that i'd like to put a switch with a fixed resistor, and have it working with or without feedback. Is that ckt right?. I remember Mark saying something like aplying the feedback to the second stage, or maybe to the joint of the 10k and .05 cap, instead of the inverted input. Any ideas?. Thanks a lot for any replies!!

[idlefaction]

that's interesting, i always have been annoyed by that distortion...

[Exactopposite]

how about a pot for adjustable feedback?


just an idea

[idlefaction]

(reads your signature)

if only!!   :twisted:

[Paul Perry (Frostwave)]

(reads your signature)

if only!!   :twisted:

"if only" indeed....
how about, they are too expensive, keep breaking down, and are too noisy as wel!!! but as they say themselves, it's only because some guy is always stomping on them!
(moderators, you can kill this one if you like.. but i just HAD to say it)

[swt]

i have tried the pot idea as rg suggested, but it's got no sound or distortion, up to a certain point, and then at a high resistance, you beguin to hear the feedback. I think it's because it should go to a different place, as i already mention. I'll try different places and let you know. Keep in touch!

[Mark Hammer]

Try feeding the regenerated signal back to the second phase shift stage.  At the very least, it will sound different because of the inversion.

The other thing to keep in mind is that regeneration has the effect of increasing the signal amplitude.  One of the difficulties of FET-based phasers is that they are limited in terms of dynamic range.  Mismatched FETs, from what I understand, make this problem worse.  One of the ways of compensating for this is to use a cap and resistor between the gate and drain of the FET, as you see in Korg phasers and some others.  Apparently this detracts from their ability to track very fast LFO speeds.

If you look at RG's drawing of a generic 4-stage P90/DOD clone, you'll see a 10k fixed resistor in series with a 47k variable one.  Try sticking in a variable resistance.  It may be that reducing the amount of feedback (as opposed to cancelling it) may be helpful.  I'd also suggest sticking a DC-blocking cap in the feedback loop.  Something in the vicinity of .1 to .047uf might be suitable with a resistance value like you use.  That will also trim some of the bass so that the feedback signal won't be too huge.