+ Post New Thread
Page 3 of 3 FirstFirst 1 2 3
Results 41 to 47 of 47
  1. #41
    Advanced Member level 2
    Points: 3,727, Level: 14

    Join Date
    Nov 2013
    Posts
    570
    Helped
    124 / 124
    Points
    3,727
    Level
    14

    Re: cascading filter after a 2 poles sallenkey

    Quote Originally Posted by FvM View Post
    But Butterworth doesn't meet the linear phase requirement in post #18. A 40 kHz Bessel filter would, but it needs f0=60 kHz Q=0.5.
    I don't see this criteria there, however it makes sense, KlausST mentioned phase shift earlier maybe, but Butterworth doesn't seem too bad I think. In the pass band the group delay is not so dramatical if the output signal's bandwidth is much smaller than 40kHz, and what is placed down now on the PCB I found it as a trade-off.
    "Try SCE to AUX." /John Aaron/



  2. #42
    Advanced Member level 5
    Points: 40,206, Level: 49
    Achievements:
    7 years registered
    LvW's Avatar
    Join Date
    May 2008
    Location
    Germany
    Posts
    5,764
    Helped
    1722 / 1722
    Points
    40,206
    Level
    49

    Re: cascading filter after a 2 poles sallenkey

    Quote Originally Posted by FvM View Post
    But Butterworth doesn't meet the linear phase requirement in post #18.
    Yes - the same applies to Chebyshev (as in my example) - however, a special equalizer stage could improve the phase response considerably.



  3. #43
    Super Moderator
    Points: 251,556, Level: 100
    Awards:
    1st Helpful Member

    Join Date
    Jan 2008
    Location
    Bochum, Germany
    Posts
    43,808
    Helped
    13322 / 13322
    Points
    251,556
    Level
    100

    Re: cascading filter after a 2 poles sallenkey

    I don't see this criteria there, however it makes sense, KlausST mentioned phase shift earlier maybe, but Butterworth doesn't seem too bad I think. In the pass band the group delay is not so dramatical if the output signal's bandwidth is much smaller than 40kHz
    The OP claimed the filter should "not change the phase till 25-30kHz", see the discussion in my previous post #21. If we take this serious, Butterworth won't be acceptable.



    •   AltAdvertisment

        
       

  4. #44
    Advanced Member level 2
    Points: 3,727, Level: 14

    Join Date
    Nov 2013
    Posts
    570
    Helped
    124 / 124
    Points
    3,727
    Level
    14

    Re: cascading filter after a 2 poles sallenkey

    Hmm, I should be more careful with reading, thank you by the way. My last recommendation is a 6th order "Linear Phase with Equiripple Error of 0.05°" characteristic. It has a stage with Q=0.55 and a normalised Fc=0.9807, not too far from the ready filter stage....maybe. The group delay is very flat up to Fc. However I have never designed this one.
    See page 42 for the characteristic and page 53 for the design table. https://www.analog.com/media/en/trai...n/Chapter8.pdf
    "Try SCE to AUX." /John Aaron/



  5. #45
    Advanced Member level 5
    Points: 40,206, Level: 49
    Achievements:
    7 years registered
    LvW's Avatar
    Join Date
    May 2008
    Location
    Germany
    Posts
    5,764
    Helped
    1722 / 1722
    Points
    40,206
    Level
    49

    Re: cascading filter after a 2 poles sallenkey

    Quote Originally Posted by frankrose View Post
    My last recommendation is a 6th order "Linear Phase with Equiripple Error of 0.05°" characteristic. It has a stage with Q=0.55 and a normalised Fc=0.9807, not too far from the ready filter stage....
    Frank - did I overlook some information? How do you know that the existing 2nd-oder stage has a Q in the range of 0.55 ?



  6. #46
    Advanced Member level 2
    Points: 3,727, Level: 14

    Join Date
    Nov 2013
    Posts
    570
    Helped
    124 / 124
    Points
    3,727
    Level
    14

    Re: cascading filter after a 2 poles sallenkey

    LvW, from post#1 it is a Sallen-Key, where R1=R2, so Q is determined by the ratio of C1 and C2: Q=0.5*sqrt(C1/C2), and they are equally 100pF, so Q is 0.5 of post#1 filter. 0.55 maybe isn't so far from this I think.
    (I used S-K equation from here: https://www.daycounter.com/Filters/S...quations.phtml)
    Last edited by frankrose; 15th February 2019 at 16:04.
    "Try SCE to AUX." /John Aaron/



  7. #47
    Advanced Member level 5
    Points: 40,206, Level: 49
    Achievements:
    7 years registered
    LvW's Avatar
    Join Date
    May 2008
    Location
    Germany
    Posts
    5,764
    Helped
    1722 / 1722
    Points
    40,206
    Level
    49

    Re: cascading filter after a 2 poles sallenkey

    Quote Originally Posted by frankrose View Post
    LvW, from post#1 it is a Sallen-Key, where R1=R2, so Q is determined by the ratio of C1 and C2: Q=0.5*sqrt(C1/C2), and they are equally 100pF, so Q is 0.5 of post#1 filter. 0.55 maybe isn't so far from this I think.
    (I used S-K equation from here: https://www.daycounter.com/Filters/S...quations.phtml)
    Frank - merci for clarification.



    •   AltAdvertisment

        
       

--[[ ]]--