Phase stage why not?

[Kipper4]

Why do more people not use this type of low pass (if I got that right) in phaser stages as opposed to the type seen in crontoxs little worker phaser (I'll find a link)

Here's what I'm talking about David Rolo's Liquid Mercury

https://www.davidrolo.com/2014/04/21/liquid-mercury-tap-tempo-phaser/

Cheers
Rich

[Kipper4]

Link to little worker phaser for comparison.

https://www.diystompboxes.com/smfforum/index.php?topic=107188.0


I note there's some discussion on the type of op amp used for some of the phasers.
The op amps bandwidth is important in some models.
So would it be possible to adapt/scale  the "low pass" type filter to work with say an
NE5532 or OPA2604

Link to a Ray Ring "low pass" type one again.
https://circuitsaladdotcom.files.wordpress.com/2014/11/ldr-phaser1.jpg


https://circuitsaladdotcom.files.wordpress.com/2014/11/pwm-phaser.jpg

[Mark Hammer]

Why do more people not use this type of low pass (if I got that right) in phaser stages as opposed to the type seen in crontoxs little worker phaser (I'll find a link)

Actually, no lowpass involved at all.  Using 8 stages results in the range of notches and peaks being nudged downwards.  If you make a phaser that can select between 4 and 8 stages, you'll notice that immediately.  Through whatever magic, David has goosed the feedback/resonance capabilities, which can make it sound like a a bandpass filter, especially when using the random-voltage setting on the TapLFO.  The high contrast between successive steps creates the sense that there has been a change in the resonant frequency

[Rob Strand]

Here's what I'm talking about David Rolo's Liquid Mercury
Link to little worker phaser for comparison.

For these two I'm seeing the "All-pass filter design which is based on a high-pass filter.
The first link has a low-pass filter on the feedback/mixer, which is a separate idea from the type of all-pass filter.

Link to a Ray Ring "low pass" type one again.

From that example I now understand where you are coming from. So in this case the all-pass filter is built-up from a low-pass filter.

The two types of all-pass filters both a achieve a similar goal. The difference between the two is the "high-pass" version inverts the signal at low-frequencies whereas the "low-pass" version does not.  When you use even numbers of stages this difference disappears and in fact both types of circuits would then be identical.   The main difference is more practical in that the "high-pass" version has the controlled resistors connect to ground (or Vref).     That lets you use dual Vactrols with a common ground point.   Conceptually JFET circuits are easier to understand with the sources connected to ground/Vref  however it is actually possible to wire the JFETs with the sources to the previous opamp output and use a low-pass version of the circuit.

A caveat to the two all-pass circuits being the same for even numbers of stage is when you use use of the all-pass stages as an economy mixing stage like the later MXR Phase 90's IIRC the two circuit start to behave a little different; as the all-pass filter gets stuffed up in different ways.

https://circuitsaladdotcom.files.wordpress.com/2014/11/pwm-phaser.jpg

As for the switched conductance filters conceptually these need the caps to hold the average voltage.     That's the general structure used for this type of filter.   However, I suspect you could get these to work in a a high-pass from as well.


   

[Mark Hammer]

I think what you're describing is the difference between "lead" and "lag" all-pass filters, differentiated by whether the amount of phase shift increases at higher frequencies or lower frequencies.  I always forget which one is lead and which one lag.