Phase 90/180 - Extra stages noise problem

[Slade]

Well, I've done a phase 90 with the R.G.'s mod to have a 8 stages phaser. The thing is that I made it selectable by a toggle switch to change between 4 and 8 stages. I made it this way:



The red part are the stages I inserted (I inserted four, not two).

I'm using NE5532 Opamps for the LFO and 8 matched 2N5952 transistors.

Well, it works perfect for 4 stages, but when I switch to the 8 stages a high "electrical hiss" appears that only changes when I move the "feedback" pot, no matter the rate, though there is also a hig phase oscillation noise that obviously changes with the rate when I'm not playing.. So I have two noises there... But I think they are related..

Any help is much appreciated.
Regards.

[Barcode80]

double check your switch wiring. If you accidentally wired the center lug swapped with one of the outer lugs in the diagram, you will get oscillation because the output of the extra stages is going right back to the input of them.

[Slade]

No problems with the switch, it's working ok and the wiring is correct, thanks

[Mark Hammer]

It is common practice when phasers use a large number of stages to stick a small-value cap in the feedback path every 4-6 stages.

In view of your wiring arrangement, I would suggest the following:
1) Stick a 1200pf cap in parallel with the 10k feedback resistor in the second phase shift stage, where the feedback path returns to.
2) Stick a second 1200pf cap in the feedback path of the 4th stage of your added 4 stages.  In 8-stage mode, this will give you some treble rolloff at stage 2 and at stage 6.

The two caps will provide a 6db/octave treble rolloff starting around 13.3khz in 4 stage mode, and a 12db/oct rolloff at the same frequency in 8-stage mode.  How well that balances off hiss and bandwidth will depend on what you use the phaser with.  If you need more hiss control, then up the cap-values to 1500pf to drop the rolloff down to 10.6khz.

[Slade]

2) Stick a second 1200pf cap in the feedback path of the 4th stage of your added 4 stages.  In 8-stage mode, this will give you some treble rolloff at stage 2 and at stage 6.

Thank you, Mark, I really want to try that.
This second capacitor should also be in parallel with the 10k resistor, right?

[Slade]

It works great! The tone loss is not such high.. Soon I'll be posting a demo of my new pedal, I'm really glad with it.
Thank you very much, Mark, your tips completely solved the problem! I won't forget your help when I'm posting it in my site.

Regards,

Fernando.-

[Mark Hammer]

Trust me, it is no brilliant insight on my part.  If you look at as many phaser schematics as I have, you will see it used in a number of them.  It's just that so many of us here tend to build 4-stage pahse shifters that you never see it needing to be used.

Take a look at the earlier Ross Phaser here.  One of the more interesting things about it is that U3a is a fixed allpass stage (i.e., unswept).  because that means the feedback path goes through an uneven number of allpass stages, that allows the feedback path to go back to the first stage, instead of the second.  But note, as well, the presence of C11 in that additional feedback stage.  It rolls off the treble above 16khz.  Not as strong a filtering effect as what I suggested to you, but it aims for the same goal: keeping noise from being recirculated and accumulating.

[Slade]

That's right, Mark.
Well, for the feedback I put a trimmer in the feedback resistor and set it to the maximum feedback before it get distorted, and changed the value of the potentiometer, and now I have an awesome feedback control!

Regards.

[Slade]

Well, here is what I promissed

Clean:

http://www.youtube.com/watch?v=6FZprd9kQk4&feature=related

Post Dist (Using my Beast Face II  )
http://www.youtube.com/watch?v=5C3AqFLbts4



Thanks!