Active xover delay stages

[5th element]

I know that there are specific delay stages you can build to time align drivers on a flat baffle, but do normal filter stages delay the signal at all? I remember reading somewhere thay they do, which makes sense, but by how much?

Is there a time that all circuits ie with one opamp, do delay the signal by or is it circuit dependant?

Also if it is circuit dependant how do you calculate this time delay. Would a simulation program model this accurately? For example build the circuit and then send a sine wave in at one end and measure the output wave incomparison to the input and thus calculate the time delay.

Cheers Matt.

[5th element]

Well simulating does indeed seem to work. I have found out that the +-6db level on linkwitz site has 180 degree phase reversal at any frequency.

[sreten]

You multiply the phase angle by the wavelength to get the delay,
or lead (for low pass ?).

Bessel filters have the property that the phase angle changes
at such a rate as to be equivalent to a constant offset of the driver.
The higher order the filter, the more the offset.

So called linear phase which equates to a fixed time delay.

Other types of filters set assymetrically can compensate for the
offset of two drivers at the crossover point, but they are not
maximally flat delay like the Bessel.

some info here :

http://www.rane.com/note147.html

though it deals with group delay, not pure delay.





sreten.

[markp]

Check out a filter design handbook(not just a crossover book) to find the delay characteristics of each type of filter. Every type of filter has a different set of characteristics like Butterworth is max flat amplitude and Bessel is max flat delay but the trade off is the Bessel is not as steep per order as a Butterworth and rings slightly. There are many more factors involved but a book will outline them all for you.