DavePlatt is correct.
I designed the mod that was the basis for this design, I'm the P in PZ-2.1 see post #75 in
this thread:
http://www.audiokarma.org/forums/in...akers-measurements.199790/page-4#post-2355102
I designed this based on looking at measurements previously made by others on
the Minimus 7 drivers. I always use a larger than text book woofing inductor in order
to provide some baffle step compensation.
I hooked up the woofer with a 1.6 mH inductor to my measurement system and tuned
the R and C for the lowest R value that did not cause peaking and a C value that
gave a -3dB point around 3 KHz. A pure 2nd order filter will not work on the woofer
due to the voice coil inductance which resonates with the C and causes peaking in
the response. It would be nothing like text book without the R in series.
I tuned the tweeter response in a similar way, wanting to reuse parts resulted in the
.4 mH for the shunt inductor, then I chose the R and C to provide roughly 3 dB of
attenuation and a -3 dB of about 3 KHz.
All of the values were ball park and would be tuned as necessary based on a the
measured system response and reverse null.
First thing I did was listen to it, and the initial impression was not bad, adjusted the
tweeter R for a reasonable balance to my ears.
Checked the reverse null with the drivers out of phase and it pointed up by about
10 degrees. It sounded so good that I made no further changes.
I commented that the woofer inductor could be reduced to 1 mH for less baffle step
and higher sensitivity which Zilch did and he made a few other minor tweaks.
R3 is deliberately in a somewhat odd position the value of 40 ohms is a bit lower than
the 47 that I suggested and that is why we put it "behind" R2. The only purpose for
that resistor is to flatten the input impedance. R3 also provides a bit more attenuation
on the tweeter but the value is so high that it is a very minor effect.
The electrical networks are both 2nd order that probably combine with the acoustical
response of the drivers to form very roughly 4th order rolloffs.
This was very much of a quicky design with as few measurements as possible.