In Reply to: 2nd order 2 way crossover posted by MAS Tubes on August 19, 2024 at 08:48:55:
Not least of which is the unfiltered acoustic response, in cabinet, of the drivers themselves. The actual crossover slopes are a function of the electrical transfer functions of the filters COMBINED WITH the natural slopes of the drivers. The resulting component values may differ significantly from textbook calculations, especially once you factor in baffle step compensation (so you don't sacrifice bass and get a lean and shouty midrange) and impedance compensation (so the filters track properly). If the chosen drivers have any nasty response peaks, you'll also need notch filters, which add their own complications.
There is actually no such thing as 2nd order LR -- Linkwitz filters are by definition 4th order, consisting of two cascaded 2nd order Butterworths -- but you can certainly have a 2nd order filter with a Q of 0.5 (tapering gently down to the elbow frequency), that is down -6dB at the crosspoint. Butterworths of any order have a Q of 0.7 (maximally flat down to the elbow frequency) and are down -3dB at the crosspoint. The important thing is to achieve a summed response that is flat through the crossover region, for an octave or more either side, indicating good phase tracking.
I would no longer attempt such a design without proper software. There are many inexpensive or even free programs you can download, and if you get REALLY serious there is SoundEasy, the most advanced speaker design program available.
For fuller discussion, see both Ray Alden's Loudspeaker Design 201, and the Zaph Audio website. Zaph provides, for many of his designs, separate graphs of the raw driver responses, electrical transfer functions of the filters, and combined and summed responses, so you can really see how everything works together.
This post is made possible by the generous support of people like you and our sponsors:
Follow Ups
- Numerous variables to consider - Brian H P 16:35:31 08/19/24 (6)
- read Brian's 1st and 3rd paragraph a couple times nt - tomservo 06:49:14 08/23/24 (5)
- RE: read Brian's 1st and 3rd paragraph a couple times nt - seancuster71@gmail.com 15:22:57 08/23/24 (4)
- Semantic quibble - Brian H P 12:55:05 08/26/24 (1)
- RE: Semantic quibble - seancuster71@gmail.com 23:44:17 09/02/24 (0)
- RE: read Brian's 1st and 3rd paragraph a couple times nt - tomservo 16:07:24 08/23/24 (1)
- RE: read Brian's 1st and 3rd paragraph a couple times nt - seancuster71@gmail.com 18:15:07 08/23/24 (0)