Signal phase slightly off (opamp)

Thread Starter

abu1985

Joined Oct 18, 2015
74
Playing with the circuit below, I noticed that this inverting output is off by about 90 degrees. It should be 180 degrees. (I may have the angle confused here) The circuit is built on solderless breadboard.

As I'm thinking about it, say I want to split the signals from the source, and then output in sync. Like, two channels to an amplifier. If I filter one channel, and not the other, would not the signal be out of sync? Maybe just buffer the other channel so they sync up? Are they ever truly in sync?


View attachment 194116
 

Thread Starter

abu1985

Joined Oct 18, 2015
74
Your attachment does not show.
Yes, a filter changes the phase which delays the signal.
The attachment is literally just a low pass filter circuit.

But my scope didn’t show phases directly 180 out. It was like, 90 out. So, question: what is typically used to bring the phases back together? Is there a chip that can sync 2 phases together?
 

Thread Starter

abu1985

Joined Oct 18, 2015
74
Is your signal audio or video?
Is your filter a highpass or is it a lowpass?
Without a schematic of the total circuit, and details about the signal frequency, we don't have a clue.
Yes, I should have been more descriptive. This is just a general applications question; I'm just researching. So in general, an audio signal (20KHz and down), would there be some type of syncing of the signal after filtering on a 2 channel amplifier? High pass, low pass, notch and stop filters included; or something like a graphic equalizer?

Attached is a low pass filter I was playing around with on my solderless breadboard. This was the attachment in my first post that did not work.

Thanks for your time.
 

Attachments

Marley

Joined Apr 4, 2016
502
Correct, then. The circuit is an integrator. A low pass filter. The phase shift will change with frequency.
To put it another way: the system has a delay caused by the capacitor in the feedback loop.
This will cause a phase shift.
 

MrAl

Joined Jun 17, 2014
11,396
Yes, I should have been more descriptive. This is just a general applications question; I'm just researching. So in general, an audio signal (20KHz and down), would there be some type of syncing of the signal after filtering on a 2 channel amplifier? High pass, low pass, notch and stop filters included; or something like a graphic equalizer?

Attached is a low pass filter I was playing around with on my solderless breadboard. This was the attachment in my first post that did not work.

Thanks for your time.
Here are the phase shifts and amplitudes for the frequencies in order:
f = 0, 20, 200, 2000, 20000 Hz
ph = 0.0, -63.56, -87.15, -89.72, -89.97
Amp = 10.0, 4.45, 0.5, 0.05, 0.0

So you can see the amplitude varies from 10 to 0 and the phase varies from 0 to 90 degrees.
 

OBW0549

Joined Mar 2, 2015
3,566
Attached is a low pass filter I was playing around with on my solderless breadboard. This was the attachment in my first post that did not work.
Your circuit is working just fine; you need to understand that ALL filters, no matter how they''re implemented, will show a phase shift between input and output that varies with frequency.

Playing with the circuit below, I noticed that this inverting output is off by about 90 degrees. It should be 180 degrees.
Wrong. A 90 degree phase shift is exactly what you should expect from a single-pole lowpass filter with a signal well above its cutoff frequency.
 

Thread Starter

abu1985

Joined Oct 18, 2015
74
Correct, then. The circuit is an integrator. A low pass filter. The phase shift will change with frequency.
To put it another way: the system has a delay caused by the capacitor in the feedback loop.
This will cause a phase shift.
Okay. Is that a problem If I have two filters for audio, and you want to amplify the signals individually after the filter? The output of each filter could be at a different phase angle to each other. I’m guessing it makes no difference in a audio application, and the output sound is not effected by the phase shift.
 

Audioguru again

Joined Oct 21, 2019
6,674
Two filters for audio are used in an active crossover circuit for a two-way speaker. The filter for the woofer is lowpass and the filter for the tweeter is highpass.
If second-order filters are used then their output phases at the crossover frequency is 180 degrees causing total cancellation when added together. Therefore the phase of one of the filters or one of the speakers must be inverted for a good combined output.
 

Thread Starter

abu1985

Joined Oct 18, 2015
74
Two filters for audio are used in an active crossover circuit for a two-way speaker. The filter for the woofer is lowpass and the filter for the tweeter is highpass.
If second-order filters are used then their output phases at the crossover frequency is 180 degrees causing total cancellation when added together. Therefore the phase of one of the filters or one of the speakers must be inverted for a good combined output.
Crossover is just another word for having one LPF and one HPF, correct?
 

crutschow

Joined Mar 14, 2008
34,285
Crossover is just another word for having one LPF and one HPF, correct?
Yes.

Below is the LTspice simulation of a Linkwitz-Riley subwoofer crossover filter, which consists of 2-pole Sallen-Key low-pass and high pass-filters with damping factors (Qo) of 0.5.
(The filter's resistors and capacitors are conveniently all the same size.)

Note the inverter for one stage (U3) which is required as AG mentioned.
This phase inversion can also be achieved by just inverting the connection to one of the speakers.

This filer has the advantage that their sum (blue trace) is perfectly flat (which includes the effects of the filter phase shifts) through the crossover frequency (80Hz here).


1575918706505.png
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,674
A crossover circuit cuts high frequencies for the woofer and cuts low frequencies for the tweeter.
One LPF and one HPF instead could make a bandpass filter.
 
Top