B&C 15CXN88 (Active Crossover)
15" Coaxial Two Way Design
15" Coaxial Two Way Design
Crossover Design:
The active version of the B&C 15CXN88 crossover is based largely on the passive version that was developed first since I felt it I had it well dialed in. Crossover point and acoustic slopes are essentially the same as is the overall response shape. There are a few addisional fine tune corrections on the active implementation like more precise notching of the HF breakup peak but the overall voicing is much the same. The biggest difference is going to be bass extension, the passive design has a fullspace f3 of roughly 100Hz in the sealed cabinet it was designed in while the default active tuning extends this to ~55Hz giving a much more full and solid low end.
Even though this is an active crossover there are passive components used as well. Essentially a 1st order filter on both drivers. For the high frequency driver the capacitor acts as a form of low frequency protection as well as dropping the level in combination with a fixed resistive L-Pad which helps lower hiss on noisier amps/DSPs to acceptable levels. The damping from the parallel resistance also prevents the HF diaphragm from resonating when the woofer is being driven, an issue I only realized was happing during well into the passive crossover development.
On the woofer side I am using a 1mH air core in series with the driver, I used a 15 gauge coil to keep DCR low but it could be an 18 gauge with only a very slight loss in output. The increasing series impedance provided by the inductor as you move upward in frequency provides an obstruction to the back EMF voltages that are generated by the driver while being driven. In some way it can be thought of as an external shorting ring. The benefit is a reduction of odd order harmonics in the midrange of ~5dB when compared to direct wiring the woofer to the amplifier.
Suitable Crossover Parts that fit the PCB:
C1(8.2uF) - L1(1.0mH) - R1 (1.8 Ohm) - R2 (8.2 Ohm)
Note: I realized after I had designed this that 1.8 Ohms is not a commonly used value for 10w resistors used in crossovers parts. If you have trouble sourcing one in its place you can use a 2 ohm which will drop down the top end ~0.3dB or combine a 3.3 and 4.0 Ohm resistor in parallel which will give you 1.8 ohms.
Crossover sim with individual driver responses and reverse polarity notch:
Crossover Filter Slopes:
B&C 15CXN88 Crossover Schematic (Passive Portion):
Active Crossover Filters:
Note: The active crossover was developed on a Crown DCi 4|300N, these filter settings may not be compatible with all DSPs due to capability of filter types and numbers of usable filters.
High Frequency:
Highpass Filter - Butterworth 36dB/Octave - 860Hz
Gain = -2.0dB
PEQ1 - Frequency = 950Hz - Gain = 2.2dB - Q = 2.0
PEQ2 - Frequency = 1440Hz - Gain = -2.3dB - Q = 2.2
PEQ3 - Frequency = 2000Hz - Gain = -1.8dB - Q = 1.0
PEQ4 - Frequency = 3000Hz - Gain = 1.0dB - Q = 6.0
PEQ5 - Frequency = 4000Hz - Gain = -3.0dB - Q = 2.5
PEQ6 - Frequency = 13700Hz - Gain = 3.5dB - Q = 10
PEQ7 - Frequency = 16700Hz - Gain = -7.0dB - Q = 10
Low Frequency:
Lowpass Filter - Butterworth 24dB/Octave - 850Hz
Delay = 0.65ms
Low Shelf (1st order) - Frequency = 35Hz - Gain = 10dB
PEQ2 - Frequency = 65Hz - Gain = 6.5dB - Q = 1.4
PEQ3 - Frequency = 175Hz - Gain = -2.4dB - Q = 0.8
PEQ4 - Frequency = 350Hz - Gain = -3.8dB - Q = 1.0
PEQ5 - Frequency = 500Hz - Gain = -0.5dB - Q = 4.0
PEQ6 - Frequency = 650Hz - Gain = 1.2dB - Q = 6.0
PEQ7 - Frequency = 1000Hz - Gain = 2.0dB - Q = 2.0
PEQ8 - Frequency = 1200Hz - Gain = -4.4dB - Q = 7.0
B&C 15CXN88 Design Measurements - Active Crossover
Measurements gated at 14ms and blended to diffraction adjusted nearfield woofer response below 250hz.
1/24th Octave Smoothing Applied. Output level adjusted to equal the passive tests.
On Axis SPL & 10 Degrees Off Axis
CTA-2034 Style Spin Data
Harmonic Distortion @ 85, 95, 100, 105, 110 & 115dB/1m - (Measured at 50cm)
110 and 115dB sweeps started at 60Hz to prevent amp clipping due to bass EQ in the active design though there is some clipping still present at the beginning of the 115dB sweep.
Compression at 85/95/100/105/110/115dB normalized against 75dB:
110 and 115dB sweeps started at 60Hz to prevent amp clipping due to bass EQ in the active design.
Note: some amp clipping was still present at the start of the 115dB sweep which is likely why you see much higher compression there below 90Hz. A Crown DCi 4|300N was used which gives ~300w per channel @ 8 ohms.