Lessee - as I recall it's 45 degrees per pole at the 3db point. But
that's plus 45 degrees for one filter, and minus 45 degrees for the
other (minor handwaving: without looking it up I'm not sure which
network exhibits the lead, which the lag)*. So you end up with a total
of 90 degrees per pole. Therefore with 2 pole networks ->180 degrees
(the bad one); 3 pole networks -> 270 degrees; 4 pole networks -> 360=0
degrees; 5 pole networks -> 450=90 degrees; 6 pole networks -> 540=180
degrees, etc, etc. But note that the angular reduction (angle >= 360)
can be fallacious since this really represents time delay. For a 500 Hz
crossover point, 500 Hz=2ms and 360 degrees is equal to a 2ms delay.
Audible? Maybe, maybe not...
But then, what do I know?
keith
* Let me try to figure this one out. For a woofer, a simple crossover
is a series inductor. I get noise out of the speaker by putting a
voltage out of the amplifier, causing a current to flow in the speaker
coil. With an inductive load (remember I put an inductor in series)
I'll get - ELI the ICE man - the voltage leading the current -> the
current lags the voltage.
For a tweeter a simple crossover is a series capacitor. Same reasoning
as above, I get the current leading the voltage. So It looks like the
phase is "plus" for the high-pass network and will be "minus" for the
low-pass.
But then, what do I know?