|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
165.225.68.50
In Reply to: RE: Danger! posted by Jon Risch on October 08, 2015 at 18:13:59
Yes, I have used the nominal value 8 Ohms in my previous calculations.
I am using is the Radian 750Pb (2" compression driver), attached to a 200Hz Tractrix horn, which I am intend to x-over at approx. 325Hz (2Žnd order).
If I can get the impedance chart from Radian (I will try).
Where do I go from there, since the impedance vary widelly trough the whole audio band?
Ps. IŽll try to keep things simple, and I am not skilled enough using sw-simulation tools
Follow Ups:
The electrical behavior of the crossover depends strongly on the impedance - both resistive and reactive parts - over a wide range around the crossover frequency. Typically, in my experience, a range of 2 octaves either side of the crossover (total four octaves) for second order crossovers.
There are things that can be done to make the load impedance more nearly constant and resistive, but they are neither simple nor easy to figure out.
In addition, you need to include the acoustic response in the design, unless it is essentially flat over that same range. No reasonable horn and driver is going to be flat over the range 80-1300Hz (325Hz +/- 2 octaves).
Without having both impedance and acoustic response (magnitude and phase) plus either a good understanding of filter design or a good optimization software, the only reasonable course is to invest in an acoustic measuring setup, and experiment with design topologies and component values until the net response looks good. It is usual to end up with a net acoustic response that is higher order than the electrical filter. As an extreme example, I once made a fourth order highpass for a horn driver that used only a single capacitor. That was pure luck, more than clever design; you just have to be open to such unanticipated possibilities.
Radian can't help, the impedance changes on the horn.
There will be several impedance peaks near where you want to cross, each one will need an LCR trap.
Ok,
I do have an XTZ room analyzer. Will an acoustic near field measurement give me the information about how to design such a filter?
Does DSP hold a solution? Even filters of 96db/octave can be designed. Some of the software includes as many as 5 parametric EQ (PEQ) channels per INPUT and OUTPUT. Those outputs can be for a 2-way or greater speaker with sub.
I don't understand the math, but the intent would seem to clearly indicate use of some form of Digital Filter would work.
One such, the FIR filter is phase flat.
Too much is never enough
I initally used a minDSP for various x-overpoints,filter orders and time-alignment settings. But to my ears, even the simpliest passive x-over (47uF Bi-polar electrolytic cap (approx 470Hz/6dB)) was far better.
Perhaps is just me, but something sounds dead when processors and software gets in my signal chain.
If you ever get to an audio show, one of the big ones, try to hear the Linkwitz Labs stuff. Some of their speakers use the MiniDSP gear for crossovers and such.
And don't forget, if you previously tried the entry level 2x4, nothing but improvements have been made in computing power and capability SINCE than.
I think by the time you construct an all-analogue solution it will be VERY complex and may not sound as good as your simple capacitor.
Too much is never enough
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: