Variable phase op-amp circuit

Hello,

I would like to design a phase shifter (0 to 180 degrees) based on an op-amp.
Which configuration is the best and which resistor and capacitor values to choose (the phase must be adjusted with a potentiometer).

Any ideas ?
Thanx !

YannLu said:

Hello,

I would like to design a phase shifter (0 to 180 degrees) based on an op-amp.
Which configuration is the best and which resistor and capacitor values to choose (the phase must be adjusted with a potentiometer).

Any ideas ?
Thanx !

Click to expand...

Better depends on what you are trying to accomplish?  There is a guitar phase shift effect (called "phasor") that can generate a notch when summed with a dry signal. There should be lots of examples out there in the webverse.

JR

Not an effect. It is just to align phases from 2 microphones.

You could find the schematic of the BSS FDS360 ,its an analog crossover from years back ,it incorporates a phase alignment stage giving a 0-180 degrees shift,adjusted with a preset pot . These units were really well engineered and last for years ,so you can be sure their circuit works very well .

YannLu said:

Not an effect. It is just to align phases from 2 microphones.

Click to expand...

If two microphones are picking up the same source over different length paths there will be arrival time, not phase, differences.

A phase shift correction might look better for a single frequency sine wave, but complex waveforms need time correction not phase shift..

There are all-pass filter networks that deliver a modest amount of time delay over a useful bandwidth (at least wide enough frequency range for loudspeaker crossover driver delay alignment).

There have been simple phase shift networks offered for delay correction like you suggest, but IMO they are the wrong tool for the task.  Of course opinions vary.

JR

PS: The phase shift part of a phasor will give you what you are looking for, often as simple as a unity gain inverting op amp, with an RC feeding the + input...The tuning of the RC will determine what frequency it changes from inverting to noninverting.  I do not recommend this for your application, but since this is DIY try it and find out what you think.  :

Thank you John for your interesting opinion.
I guess it is just a question of vocabulary.
Commercial device like  the Radial Phazer (http://www.radialeng.com/phazer.php ) is based on "phase variation".
Probably very small time delays equals phase shifting for people.

Discussion here : http://www.gearslutz.com/board/so-much-gear-so-little-time/381506-phase-align-vs-time-align.html

Variable all-pass filter

I have an schem somewhere
That uses a quad opamp and ganged pot. Have to look around

Oh thanks ! A double stage allpass...

YannLu said:

Thank you John for your interesting opinion.
I guess it is just a question of vocabulary.
Commercial device like  the Radial Phazer (http://www.radialeng.com/phazer.php ) is based on "phase variation".
Probably very small time delays equals phase shifting for people.

Click to expand...

Not vocabulary a question of physics...  A phase shift circuit can only mimic time delay at a single frequency, if that is good enough to make you happy I already told you how to make one... (inverting op amp, with RC feeding + input)

I could say in an advertisement that this circuit performs miracles, but it doesn't.

JR

Yes it is a standard all-pass filter. I am going to try it with different values of Rx and C. The R is the variable resistor.

I owned a Little Labs IBP for many years, the result as John hints, is usually different but not necessarily better.  I sold mine.

The real trick is to avoid sound arriving to two different mics at the same time, using polar patterns, positioning and less microphones.

YannLu said:

Not an effect. It is just to align phases from 2 microphones.

Click to expand...

You need a digital delay, or you could just slide the regions around in your DAW to align the two mics.

As JR and others note, you're not aligning phase, you are compensating for path length differences, and the right way to do that is with time delay.

Tubetec said:

You could find the schematic of the BSS FDS360 ,its an analog crossover from years back ,it incorporates a phase alignment stage giving a 0-180 degrees shift,adjusted with a preset pot . These units were really well engineered and last for years ,so you can be sure their circuit works very well .

Click to expand...

The phase align in the FDS360 works because it needs to work only in a very narrow BW, namely in the area where two transducers overlap. That's a fraction of an octave. That's why a 1st-order all-pass filter is adequate in this role.

abbey road d enfer said:

The phase align in the FDS360 works because it needs to work only in a very narrow BW, namely in the area where two transducers overlap. That's a fraction of an octave. That's why a 1st-order all-pass filter is adequate in this role.

Click to expand...

Yes, I mentioned this earlier,,,  (There are all-pass filter networks that deliver a modest amount of time delay over a useful bandwidth (at least wide enough frequency range for loudspeaker crossover driver delay alignment)).

you can stack up multiple stage all pass analog filters and get a usable short delay over more than one octave range. These were the SOTA for active analog loudspeaker crossovers several decades ago when digital delay to deal with single digit inches of driver offset was prohibitively expensive .

These days digital delay is cheap, or free inside some digital recording/mixing environments so why not use the right tool for the job?

JR