Don Pearce wrote:

Eeyore wrote:
Don Pearce wrote:
Eeyore wrote:

If your drivers are up to the task, just use a first order crossover ( 6dB /
octave).

By its very nature it can't screw up. It will always give constant power and
pressure and eliminate any phase 'nasties'.

No, a singe order crossover can't be linear phase - you need at least
second order to achieve that.

Nonsense.

The vector sum of a first order crossover is flat in both amplitude and phase.

Graham

Show me the maths. And then explain, if it is flat in amplitude, how
does it work as a crossover? Then you can go on to explain what is
happening to the phase around crossover as opposed to the passband. Do
it in terms of group delay please - that is an easier visualization
for linear phase systems.

I will do the same - lets see who blinks first, shall we?

Here's something simple I can type here and now.

In the passband both filters have no phase shift and the 'gain' is 1.

At the crossover point the voltage gain of each section is 0.7071 at a phase angle
of +-45 degrees.

Take the vectors and sum and you're back to a voltage gain of 1 and no phase shift.

The power at crossover is -3dB ( half power ), add those and it's 0dB again.

If the drivers have the bandwidth it's a great scheme.

Graham