I never use standard filters, i always use what is fit for the box, and that can differ a lot. Phase alignment is also done in the crossover, that is way more effective than a physical alignment (that is never perfect) or using allpass filters to try to match them perfect. It require some skill, but you can learn a lot by just doing exercise crossovers with frd and zma files you find online. They are a good theoretical excercise tool i think, i learned most of what i know like that.
And one trick, by putting a resistor behind the cap or coil in parallel with the driver you can tune the curve (phase and acoustic) also. Just make sure the resistor can handle enough current and that you don't make the impendance go to low.
And one trick, by putting a resistor behind the cap or coil in parallel with the driver you can tune the curve (phase and acoustic) also. Just make sure the resistor can handle enough current and that you don't make the impendance go to low.
A few 2c high-level comments on crossovers in general.
Acoustic crossovers need to be fully complementary, summing to flat magnitude response, and in phase with each other.
Electric crossover and filters only work to serve that purpose.
Which means, if driver sections being joined don't have the same orders of natural response roll-offs, asymmetrical electrical filters are required.
And also means if they do have the same order roll-offs, symmetrical electrical filters must be used.
In either case, the name of the game is nudging both sides towards some chosen order (whether an integer order or not), of an acoustic complementary crossover.
Juggling two sides of the electric crossover should not be used for fixed time delays, like needed geometric offsets.
The crossover has only one purpose, .....to provide complementary mag and phase summation.
Getting fixed time delays between acoustic centers correct, is a distinctly separate step. Keep the crossover out of it.
Likewise, all-pass filters are not substitutes for fixed time delays.
They can be used to help create better acoustic complementary phase matching, but their use needs to stay in the phase domain (which is part of the frequency domain.)
Only fixed delays accurately address fixed delay situations (like geometric acoustic center offsets).....
Kludge crossovers around to address fixed time issues....is well, just that...kludging around...
Acoustic crossovers need to be fully complementary, summing to flat magnitude response, and in phase with each other.
Electric crossover and filters only work to serve that purpose.
Which means, if driver sections being joined don't have the same orders of natural response roll-offs, asymmetrical electrical filters are required.
And also means if they do have the same order roll-offs, symmetrical electrical filters must be used.
In either case, the name of the game is nudging both sides towards some chosen order (whether an integer order or not), of an acoustic complementary crossover.
Juggling two sides of the electric crossover should not be used for fixed time delays, like needed geometric offsets.
The crossover has only one purpose, .....to provide complementary mag and phase summation.
Getting fixed time delays between acoustic centers correct, is a distinctly separate step. Keep the crossover out of it.
Likewise, all-pass filters are not substitutes for fixed time delays.
They can be used to help create better acoustic complementary phase matching, but their use needs to stay in the phase domain (which is part of the frequency domain.)
Only fixed delays accurately address fixed delay situations (like geometric acoustic center offsets).....
Kludge crossovers around to address fixed time issues....is well, just that...kludging around...
Yes. The point of my post is that once the physical design is completed, lobing is simply a matter of the phase relationship.
if you correctly dimension a butterworth filter and compare it to correctly dimensioned LR filter, the lobing patterns differ.
Either way it is still rather amazing how computer design.
Very much free computer design.
Allows diy hobbyist to quickly solve methods that otherwise would be difficult or impossible at home.
Software being so visual has helped confusing or overwhelming theory become understandable.
Fun thread @jayme thanks for posting your project.
Having a recent fascination with Technics Linear Phase line.
Have modeled a few with modern drivers.
Your thread sparked
interest in using All Pass Filters to test other methods.
Maybe part of the answer to your question. Why All Pass filters not seen in Diy so much.
Assume many might not be aware of it. Needing encouragement and introductions to explore
modern design tools for successful results.
Very much free computer design.
Allows diy hobbyist to quickly solve methods that otherwise would be difficult or impossible at home.
Software being so visual has helped confusing or overwhelming theory become understandable.
Fun thread @jayme thanks for posting your project.
Having a recent fascination with Technics Linear Phase line.
Have modeled a few with modern drivers.
Your thread sparked
interest in using All Pass Filters to test other methods.
Maybe part of the answer to your question. Why All Pass filters not seen in Diy so much.
Assume many might not be aware of it. Needing encouragement and introductions to explore
modern design tools for successful results.
Why this terminals instead of soldering?
Every not soldered contact on terminals have about same resistance as 1 m cable, on every terminal is 2 contacts.
Also this terminals are made for 230V supply, not for low voltages as on speakers.
If you are referring to WAGO connectors, they have lower contact resistance than alligators, and are simple to use in prototyping speaker passive xovers. Lots of people use these for this purpose.
As Wolf_teeth mentioned, the Wago connectors are for prototyping and ease of making changes so I can listen to any tweaks easily. Super nice to make make changes on the fly (while the song is paused) and then resume with the changed crossover.
I have had some very immediate responses of "nope!" after a couple of changes that looked fine modeled in the Boxsim software.
I cant seem to ever declare my design finished. I keep tweeking. Next job this summer is to rebuild the crossover boxes so that the components are on one side, wires feed thru the board to the backside for connections. I intend to keep the WAGO connections, but just hide them.