XOVER State Variable Filter

[steph_tsf]

This is a typical application of the 5-channel FFT-based Audio Analyzer. See the attached .fsm

ch1 : reference channel (no need to graph it)
ch2 : lowpass signal
ch3 : bandpass signal
ch4 : highpass signal
ch5 : reconstruction (lowpass + bandpass + highpass)

reference : ch1
impulse response graph : ch2 (we can also graph ch3, ch4, ch5 impulse response)

For implementing a 3-way crossover operating at 1 kHz, we need to configure the State Variable Filter with Fc = 1.0 kHz. The Q factor can be set anywhere between 0.3 and 1.0.

The resulting graph illustrates the State Variable Xover behaviour and properties.
The Lowpass slope is - 12 dB/octave.
The Highpass slope is + 12 dB/octave.
The Bandpass exhibits two 1st-order asymptots.
The reconstructed audio is perfect. The global phase is intact.
The speakers operate with considerable relative phase shifts : 90 degree for the bandpass, and 180 degree between the Lowpass and the Highpass. At the Fc frequency, the Lowpass and Highpass output the same sound amplitude, however due to their 180 degree relative phase shifs, they annihilate each other. At Fc, the Bandpass is supposed to deliver the sound. The Lowpass + Highpass annililation is easy to get on-axis, and difficult to maintain off-axis. The polar radiation pattern will thus exhibit variations. The solution for minimizing those variations is to rely on concentric or coincident drivers.
A practical solution is to rely on a d'Appolito setup for the Lowpass / Bandpass section.
Say two 20 cm drivers for the Lowpass, surrounding one 8 cm driver for the bandpass.
What about the tweeter? The tweeter needs to be a miniaturized tweeter, Visaton CP13 or Dayton ND16, put in the centre of the 8 cm bandpass driver, kind of car audio solution.

Cheers,
Steph