Phasers - All made with all pass filters?

[savethewhales]

Hello!

I've been searching on the internet around the schematics of the most known phasers, such as the MXR P90, the Boss PH-3 and the EH Small Stone and I couldn't find whether if only the P90 has all pass filters on the phase shifting stage or if the other two also have it.. I'm wondering because it seems like most phasers seem to have this kind of construction for the phase shifting, which makes sense..

Thanks in advance!

[Mark Hammer]

Yep.  That's how they all work.  Of course, as the difference between the P90 and Small Stone illustrates, there are different ways to achieve an allpass function, depending on what the "control element" is: FET vs OTA.

[savethewhales]

Yep.  That's how they all work.  Of course, as the difference between the P90 and Small Stone illustrates, there are different ways to achieve an allpass function, depending on what the "control element" is: FET vs OTA.

Nice! Thanks Mark!

[ElectricDruid]

Hello!

I've been searching on the internet around the schematics of the most known phasers, such as the MXR P90, the Boss PH-3 and the EH Small Stone and I couldn't find whether if only the P90 has all pass filters on the phase shifting stage or if the other two also have it.. I'm wondering because it seems like most phasers seem to have this kind of construction for the phase shifting, which makes sense..

Thanks in advance!

I recently tried making a phaser a different way and thereby discovered *why* they're all made that way. I should have realised if I'd thought about it, but I ploughed in and found out the hard way.

"Oh", I thought, "you just need a few notches, so let's string a few notch filters together and then sweep them." Simple, right?
Except if you try it, what you find is that the sloping sides of the notches interact with the sloping sides of the nearby notches and reduce the overall volume markedly. You finish up with a big drop in volume right in the middle of your signal where you'll hear it most. And it gets worse the closer together your notches are.

This doesn't happen with the phase-shift based notches, because as you add more phase stages, all you do is change the points at which there's 180 degrees of shift (and therefore a notch) and the points at which there's a 0 degree of shift (and therefore no cancellation). So the "Classic" phaser structure manages to produces multiple notches without the volume drop in the centre of the spectrum that you'd get with other methods. It really is the best way to do it. The only other way to get a similar effect would be a delay line, and we have a different word for that - notches produced that way are "flanging", but they do share the same property of not losing signal between the notches.

[savethewhales]

I recently tried making a phaser a different way and thereby discovered *why* they're all made that way. I should have realised if I'd thought about it, but I ploughed in and found out the hard way.

"Oh", I thought, "you just need a few notches, so let's string a few notch filters together and then sweep them." Simple, right?
Except if you try it, what you find is that the sloping sides of the notches interact with the sloping sides of the nearby notches and reduce the overall volume markedly. You finish up with a big drop in volume right in the middle of your signal where you'll hear it most. And it gets worse the closer together your notches are.

This doesn't happen with the phase-shift based notches, because as you add more phase stages, all you do is change the points at which there's 180 degrees of shift (and therefore a notch) and the points at which there's a 0 degree of shift (and therefore no cancellation). So the "Classic" phaser structure manages to produces multiple notches without the volume drop in the centre of the spectrum that you'd get with other methods. It really is the best way to do it. The only other way to get a similar effect would be a delay line, and we have a different word for that - notches produced that way are "flanging", but they do share the same property of not losing signal between the notches.

This is really interesting, as I did a phaser project for my college. It makes total sense what you're saying. But as I didn't research too much out of the all pass filter notching, I wouldn't know how to make a notch without the P90 and etc method.. I'm just thinking that maybe if the notches have a Q that's very tight, what you said shouldn't happen, or should it? Meaning that I assume your Q's weren't that tight (?)

[antonis]

Just the opposite..

https://sound-au.com/articles/active-filters.htm  7.5

[savethewhales]

Just the opposite..

https://sound-au.com/articles/active-filters.htm  7.5

So for notch filters:

Twin-Tee Notch Filter
Bridged-Tee Notch Filter
Fliege Notch Filter

Did you (Electrodruid) used any of these?

[iainpunk]

"Oh", I thought, "you just need a few notches, so let's string a few notch filters together and then sweep them." Simple, right?
Except if you try it, what you find is that the sloping sides of the notches interact with the sloping sides of the nearby notches and reduce the overall volume markedly. You finish up with a big drop in volume right in the middle of your signal where you'll hear it most. And it gets worse the closer together your notches are.

that's why you should use both moving notches and a really broad band pass, to equal out the mid scoop. and then a bunch of gain to bring it up to level again. i feel inspired to try this!

or you have a peak (or more) about 1/2nd of an octave above and/or below the notch(es), so you get a auto-wah-phaser- thingy

cheers, Iain

[ElectricDruid]

Just the opposite..

https://sound-au.com/articles/active-filters.htm  7.5

So for notch filters:

Twin-Tee Notch Filter
Bridged-Tee Notch Filter
Fliege Notch Filter

Did you (Electrodruid) used any of these?

I *looked* at those, but actually I used a wien-bridge notch design. But the specific notch filter implementation doesn't matter, they're all going to hit the same issue.

that's why you should use both moving notches and a really broad band pass, to equal out the mid scoop. and then a bunch of gain to bring it up to level again. i feel inspired to try this!

It seems like a lot of circuit to fix something you didn't need to build broken in the first place, but if you feel inspired by it, then sure, go for it. Maybe the result is a good sound, I don't know. But I doubt it's a classic phaser - maybe more wah-esque, like you suggested.