Hi guys, this is my first time building a 3 way speaker. I was wondering if my crossover design is any good. Id like the crossover to be 250hz/2500hz. Any tips on how to improve it, or is it better to measure the speaker after i purchase them and then make the crossover
Measuring would be good. Have you considered the baffle effects?
How about turning off system phase and turning on the individual traces?
How about turning off system phase and turning on the individual traces?
Your mid will need attenuating due to overlapping response gain from LF and HF, but also from bandpass gain and filter slope Q. It will end up sounding shouty the way its currently simmed.
For improving such a crossover, you need to build the enclosure and make tests = measurements and listening...
This seems stupid as simulators are very performant today, but all the loudspeaker builders firms still do this today, because there's always surprises between simulation and reality ! I mean : the simulation is at least a starting point.
But over the simulation, the final result can be quite different : attenuators or transistion frequencies may need ear tweaking in the end - that's often what I've seen in my DIY speaker builds...
But it's me OK ? 😉
T
This seems stupid as simulators are very performant today, but all the loudspeaker builders firms still do this today, because there's always surprises between simulation and reality ! I mean : the simulation is at least a starting point.
But over the simulation, the final result can be quite different : attenuators or transistion frequencies may need ear tweaking in the end - that's often what I've seen in my DIY speaker builds...
But it's me OK ? 😉
T
The only correct way to design a speaker system crossover is to measure the actual driver' responses, multi axis at least 0-90 degrees (0-180 if possible).
The actual measured FR will vary from the factory provided IEC-baffle FR, due to the baffle step, diffraction etc.
The measurements should be 2-channel, providing the correct time impulse.
Those are just the very basic steps, many more details are involved in the process.
Suggested software for a simulation: VituixCAD2 https://kimmosaunisto.net/
The actual measured FR will vary from the factory provided IEC-baffle FR, due to the baffle step, diffraction etc.
The measurements should be 2-channel, providing the correct time impulse.
Those are just the very basic steps, many more details are involved in the process.
Suggested software for a simulation: VituixCAD2 https://kimmosaunisto.net/