Active crossover for JBL L300 will split frequencies between the woofer and the mid-highs. Then a passive crossover will split the frequencies between the mids and the highs. This hibrid approach is because today I have only 2 amplifiers that I would like to use in my system.
For the Active Crossover I am using Nelson Pass 2-24 Active Crossover, you can read the article about it here: https://www.firstwatt.com/pdf/art_diy%20biamp_6-24_%20crossover.pdf
As a starting point, I will configure the 2-24 crossover as an LR4 24db/oct at 800hz. We will have to determine the RC values of the circuit. Let's start simulating the circuit to make sure our calculations make sense. This is the circuit and the response for the example provided by Nelson for an LR4 1Khz. Simulations using Microcap 12.
Things look right, the crossover dips 6db at 1Khz and phase shifts 360 degrees. Let's calculate now our LR values to move the frequency to 800hz.
Take a look at 1976 Linkwitz's paper for L-R filters https://www.linkwitzlab.com/JAES/jaes_papers76.htm, If you compare the proposed LR4 filter to Pass'example they are almost identical but Pass has swapped some of the values of the resistors and the stated formula for cut frequency Fc = 1 / 2*PI*sqrt(2)RC doesn't apply anymore. After playing with the formula a little bit it turns out you can calculate Fc for Pass' examples as Fc = 1/ Pi*sqrt(2)RC with C in Farads and R in Ohm and here we go. With C=10nF and R=28100ohms we get Fc=800.1hz (some resistors and capacitors are halved)
This is the 6-24 Active Crossover during assembly, it should be a good starting point to start listening and measuring the L300s
Measurements
A measurment of the AXO responses for right channel (red) and left channel (blue) is quite spectacular and the matching between channels is also superb
Looking outside of the box and exploring the effect of moving the HI-level Pot from maximum to minimum.
I am quite surprised the filters don't continue falling and they plateau.I am getting a rejection band of around 30db at maximum level but it decreases as I lower the Hi-level potentiometer. Also the filter response suffers, it seems like a small amount of level adjustment is OK but if too much is done things could go wrong. If serious level adjustment is needed I will try to do it at the amplifiers instead of the XO.
I was able to increase the rejection to almost 50db putting the soundcard at max volume and decreasing the gain of the measuring input. That is much better, but I am starting to think that that plateau down there is real
Comments