I’ve been attempting to formulate an empirical formula for finding Q-factor of second-order filter by using graphical method. Say we could find Q-value from the frequency response curve. And, finally, I obtained it, that is, Q = 10^(x/20) where x is the -dB at Fb (box resonance). I verified with theory and many examples and it worked.
For example,
if x = -3dB, it yielded Q = 0.707 or Butterworth slope
If x = -4.5dB, it yielded Q = 0.58 or Bessel slope
If x = -6dB, it yielded Q = 0.5 or Linkwitz-Riley slope
Etc.
Nonetheless, that formula can only be used for Q of enclosure. It wouldn’t work for electrical fIlter because I cannot find the corner frequency on the amplitude curve. So anyone advise me how to find corner frequency from the frequency response graph? Don’t suggest me to look at -3dB point because it would be correct only for Butterworth filter.
For example,
if x = -3dB, it yielded Q = 0.707 or Butterworth slope
If x = -4.5dB, it yielded Q = 0.58 or Bessel slope
If x = -6dB, it yielded Q = 0.5 or Linkwitz-Riley slope
Etc.
Nonetheless, that formula can only be used for Q of enclosure. It wouldn’t work for electrical fIlter because I cannot find the corner frequency on the amplitude curve. So anyone advise me how to find corner frequency from the frequency response graph? Don’t suggest me to look at -3dB point because it would be correct only for Butterworth filter.
