Russ White wrote:...
Your filter coefficients may not be correct more likely or your manner of loading them is not.
As Brian says, there are not enough taps for serious crossover work unfortunately.
I know something must be wrong :-)
Here are the coefficients if someone would like to try:
-45367, -34066, 67693, 156872, 47470, -202765,
-223779, 162275, 485166, 126314, -629433, -647244,
443951, 1246729, 264377, -1549065, -1449905, 1102057,
2718525, 394292, -3333843, -2782583, 2469771, 5265790,
347618, -6501007, -4759303, 5066432, 9323373, -208365,
-11741941, -7414434, 9658321, 15393271, -1845571, -19986083,
-10664811, 17379784, 24102346, -5500456, -32599926, -14291924,
30056103, 36443755, -12783980, -52020668, -17954819, 51216090,
54605383, -26997242, -83967944, -21236435, 89946575, 85259451,
-57554167, -147667892, -23716063, 182876923, 160733938, -154407580,
-373865298, -25051736, 856746075, 1627086483, 1627086483, 856746075,
-25051736, -373865298, -154407580, 160733938, 182876923, -23716063,
-147667892, -57554167, 85259451, 89946575, -21236435, -83967944,
-26997242, 54605383, 51216090, -17954819, -52020668, -12783980,
36443755, 30056103, -14291924, -32599926, -5500456, 24102346,
17379784, -10664811, -19986083, -1845571, 15393271, 9658321,
-7414434, -11741941, -208365, 9323373, 5066432, -4759303,
-6501007, 347618, 5265790, 2469771, -2782583, -3333843,
394292, 2718525, 1102057, -1449905, -1549065, 264377,
1246729, 443951, -647244, -629433, 126314, 485166,
162275, -223779, -202765, 47470, 156872, 67693,
-34066, -45367
It was generated in Matlab. Since the filter is symmetric you only need to load the first 64 (which is the size of the first stage filter).
I did not have coefficients for the second stage, so I used the built-in second stage
Even simple crossover can be useful. For example the LP crossover for my speaker is a simple LC second order at 1000Hz.
Edited by user Wednesday, October 5, 2011 9:16:37 AM(UTC)
| Reason: Not specified