Hi there.
Depending on the source, information about filters will refer to phase shift and/or group delay. When viewing filters from a theoretical standpoint, with periodic sine wave inputs, "phase shift" and "delay" basically amount to the same thing don't they? What about when the input isn't periodic or sinusoidal? In that case, it seems to me the distinction becomes important. Is it actually a time delay or actually a phase shift? Is this even a valid question?
The reason I ask is because I think it will make a difference on the crossover circuit I'm making. I've got filters with theoretical 180 degree phase shift at the 3dB point - 4-pole Butterworth. I want to take the unfiltered signal and align it with the filtered. I could do this with an all-pass filter, but I could also do it with a unity gain inverting amp, right? That's where my question comes from - if the "phase shift" a filter produces is actually a time delay, I would need an all-pass filter to delay the unfiltered signal by the same amount. If it actually is a phase shift with no time delay, it could be accomplished by inverting the signal.
Hopefully I haven't made any completely dumb errors in thinking. It's entirely possible.
Depending on the source, information about filters will refer to phase shift and/or group delay. When viewing filters from a theoretical standpoint, with periodic sine wave inputs, "phase shift" and "delay" basically amount to the same thing don't they? What about when the input isn't periodic or sinusoidal? In that case, it seems to me the distinction becomes important. Is it actually a time delay or actually a phase shift? Is this even a valid question?
The reason I ask is because I think it will make a difference on the crossover circuit I'm making. I've got filters with theoretical 180 degree phase shift at the 3dB point - 4-pole Butterworth. I want to take the unfiltered signal and align it with the filtered. I could do this with an all-pass filter, but I could also do it with a unity gain inverting amp, right? That's where my question comes from - if the "phase shift" a filter produces is actually a time delay, I would need an all-pass filter to delay the unfiltered signal by the same amount. If it actually is a phase shift with no time delay, it could be accomplished by inverting the signal.
Hopefully I haven't made any completely dumb errors in thinking. It's entirely possible.