Originally Posted by
sebackman
....
Later this week I will remove the remaining two 2451Be’s from the old system replacing them with 2453SL’s and changing them from 3-way way to 2-way.
The measurements I did on D2’s with and without the passive JBL M2 filter show that the capacitor is mainly a protection cap and replacing it with a 6db filter in the BSS/DSP achieves nearly the same curve. The cap in the M2 filter is not a Butter slope, the D2 extends a little further down with the cap compared to 6db Butter I used in the DSP. Please see picture.
The XO used in my measurement corresponds to HP 6db filter at 957Hz. The frequency is the result of running a backwards calculation from a calculated 20,8 ohms load that the 8mf in series in the original JBL filter would be seeing. The 20,8 ohms load seen from the cap is a result of the average impedance between 800Hz and 20kHz (26,3 ohms) of the D2 driver on the M2 and 13 ohms in series and the 11ohms shunt in the filter. Lowering the XO can probably mitigate some of the difference, but the passive cap is probably not a Butter curve with those resistor values.
If you are building an exact M2 clone (this build are rather ClownsJ ) I would recommend keeping the passive filter to mimic the exact JBL curve. However, I do think it is possible to change the slope and/or the cut off frequency (lower) of the 6db filter in the DSP do achieve an identical curve without the cap’s, for the one willing to invest some time.
.....