Honestly, I was not able to find a suitable forum for posting this question. Please direct me if this is inappropriate.
I am planning on building an active 4th order LR filter for my 2.1 subwoofer system. LR filters have a -6 dB point at the crossover frequency.
Adding the voltage at this point exactly produces 0 dB. -6dB corresponds with half the voltage at each output.
But the the power at each amplifier output is proportional to the square of the voltage, so a quarter of the power. Sum that and I get half the power.
So to me it seems that the voltage is constant. But the power has a -3dB dip at the crossover frequency.
Since LR-4 is quite popular and applied by many audio systems I think I make a mistake in thinking. Where?
I am planning on building an active 4th order LR filter for my 2.1 subwoofer system. LR filters have a -6 dB point at the crossover frequency.
Adding the voltage at this point exactly produces 0 dB. -6dB corresponds with half the voltage at each output.
But the the power at each amplifier output is proportional to the square of the voltage, so a quarter of the power. Sum that and I get half the power.
So to me it seems that the voltage is constant. But the power has a -3dB dip at the crossover frequency.
Since LR-4 is quite popular and applied by many audio systems I think I make a mistake in thinking. Where?
You haven't, it's all correct.
When the loudspeakers have equal sensitivity and equal distances from their acoustic centres to the ears of the listener, the free-field axial response is determined by the sum of the sound pressures from each loudspeaker, and the sound pressure is proportional to the voltage rather than the power. That is, in the crossover region where both loudspeakers are active, you indeed need 3 dB less power to get a certain free-field on-axis SPL.
When the loudspeakers have equal sensitivity and equal distances from their acoustic centres to the ears of the listener, the free-field axial response is determined by the sum of the sound pressures from each loudspeaker, and the sound pressure is proportional to the voltage rather than the power. That is, in the crossover region where both loudspeakers are active, you indeed need 3 dB less power to get a certain free-field on-axis SPL.
Last edited:
Hello Marcel,
That is a very clear explanation. I did not know that. I am an electrical engineer built a lot of speaker systems, but I am not familiar with this kind of loudspeaker physics.
Thank you.
That is a very clear explanation. I did not know that. I am an electrical engineer built a lot of speaker systems, but I am not familiar with this kind of loudspeaker physics.
Thank you.
Please keep in mind that for a satellite-subwoofer system, the equal distance condition is not necessarily met. If it isn't, Butterworth or something in between Butterworth and Linkwitz-Riley may work better than Linkwitz-Riley.
Darn... I did not think of that. The subwoofer and the satellites are in the same vertial plane (front/back) but there is a horizontal distance between the woofer and the satellites.
As a matter of fact the subwoofer is a TL of 1.6m wide and 35cm high. And the satellites stand on top of that box.
Currently the woofer is located in the middle of the box, but it has to be shifted to one side to supress harmonics. So the woofer will be below the right satellite.
The TL port is then just at the other side of the box right below the other satellite.
Now according to Hornresp the TL will have a hump between 80-120Hz and then drop off sharply. So I planned to set the crossover frequency to 100Hz. That might just compensate for that hump.
But all in all, forgetting about the horizontal offset does not make it more straightforward.
As a matter of fact the subwoofer is a TL of 1.6m wide and 35cm high. And the satellites stand on top of that box.
Currently the woofer is located in the middle of the box, but it has to be shifted to one side to supress harmonics. So the woofer will be below the right satellite.
The TL port is then just at the other side of the box right below the other satellite.
Now according to Hornresp the TL will have a hump between 80-120Hz and then drop off sharply. So I planned to set the crossover frequency to 100Hz. That might just compensate for that hump.
But all in all, forgetting about the horizontal offset does not make it more straightforward.
Last edited:
At 100 Hz, when you can keep the path length differences well below 43 cm, Linkwitz-Riley should work well (better than Butterworth anyway). I think that should be feasible with one satellite just above the woofer and the other just above the other side of the transmission line box.
The wavelength at 100 Hz is about 3.43 m. A quarter wavelength path difference would cause a -3 dB error that you could compensate for by using Butterworth instead of Linkwitz-Riley. Interpolating, somewhere around an eighth wavelength, 43 cm, they are both equally far off.
The wavelength at 100 Hz is about 3.43 m. A quarter wavelength path difference would cause a -3 dB error that you could compensate for by using Butterworth instead of Linkwitz-Riley. Interpolating, somewhere around an eighth wavelength, 43 cm, they are both equally far off.
Since the listening position is 4 meters in front of the loudspeakers the difference in path length is much less than the distance from bass driver to satellite.
It is the difference between the longer side and the hypotenuse of a triangle. It might be less serious after all.
Still not so smart to overlook it.
It is the difference between the longer side and the hypotenuse of a triangle. It might be less serious after all.
Still not so smart to overlook it.