Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Capacitive/inductive reactance in a series-parallel circuit

  1. Dec 2, 2015 #1
    I wired a capacitor in series with a 4 ohm tweeter speaker to serve as a high-pass filter. I used the capacitive reactance equation to figure my capacitor value to filter the speaker at 2.2khz.



    Solve for C and you get .00001809 Farads or 18uF.


    It sounds like something you'd buy from one of those big box stores. It sounds terrible. I'm running out of x-max on the tweeter & it's distorting at higher volumes. I want to turn my mediocre 1st order crossover into a 2nd order filter so I don't burn out the voice coil & destroy my tweeters. So, toss an inductor across the terminals of the speaker. That's the issue though. I don't know how to integrate (not literally) the inductive reactance equation...

    Can I just simply plug in 2200 for frequency, 4 for Xl, & solve for L?


    I'm asking because I'm not sure if the capacative/inductive reactance equations only apply when the capacitor or inductors are in series with the load. The inductor is "shunting" to ground here so...
  2. jcsd
  3. Dec 2, 2015 #2


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Hi there
    welcome to PF :smile:

    not sure what a big box store is ?

    18uF is a standard value, get one rated around 100V to be safe ( higher than needed is better)
    and it's going to be a non-polarised electrolytic capacitor most component parts stores should be able to supply
    or eBay or Amazon etc

  4. Dec 2, 2015 #3


    User Avatar
    Gold Member

    There is a lot more to crossover design than just cutting the lows to the tweeter. The audio out from the low and high end speakers add and the filter phase shifts cause notches and or peaks (terrible sounding, BTW). You need to match the high and low frequency drivers with properly matched filters that take into consideration the driver efficiency, speaker orientation, and so on.

    http://www.audioholics.com/diy-audio/building-a-do-it-yourself-loudspeaker-design/crossover-design is a random site from a search for "audio crossover design"
  5. Dec 2, 2015 #4
    Wal-Mart, Target, Best Buy, etc...

    The last place a professional (audio/recording engineer) would turn to for studio monitors.

    So the reactance equations apply to both seried & paralleled components? I.e. I wire an inductor in series with a subwoofer to block/suppress high-er frequencies & I wire an inductor in parallel with said subwoofer to ...not even sure how that circuit would model. Either case, you use the same inductive reactance equation?

    Reason I brought it up is because when calculating Rtotal in a purely resistive series circuit, you simply add the resistances up, but when they're paralleled it's 1/R + 1/R2 ...etc

    Didn't know if that, uh, inverse operation applied to the reactance equations now that we have an inductor paralleled -not seried to the load, the tweeter.

    Also, are there any rule of thumbs when it comes to crossover points? For instance, 4th order you want your mid 250 hertz down from your tweeter, 3rd order 500 down, 2nd order 1khz down...etc. It is suggested that I cross my mid 1.8khz. After 2.2khz, break up becomes a real problem.

    MeBigGuy, I just caught your post as I was typing this up. I do read/lurk several forums & to be completely honest w/ you, it is my opinion that 99% of the users on those forums have no idea what they're talking about. I called up tech support from two different companies -ones out of Springboro,OH & the other out of Wisconsin, and niether of the techs knew how to do it by hand. They all rely on software. If I keep using calculators, I'll never learn. End rant.
    Last edited: Dec 3, 2015
  6. Dec 3, 2015 #5


    User Avatar
    Gold Member

    Calculating filters is complex (pun intended). Inductors and capacitors are represented by complex numbers, which contains their "phase" information. So, any operations you do need to deal with complex numbers (also represented as laplace transforms)

    So, depending to the filter response you need (like butterworth, chebyschev, eliptical, gaussian, etc) you need to resort to synthesis tables to get polynomial coefficients, and then scale the filter to high pass, low pass, or bandpass. It's not something one typically does with a pencil and paper.

    You can look at some filter synthesis tutorials to understand why it is done with computer programs.

    And read about crossover basics, like crossover frequency and filter response for least phase distortion.
  7. Dec 3, 2015 #6


    User Avatar
    Gold Member

    It is very hard to obtain perfection except with good lab facilities. I suggest you reduce the capacitor to give a higher cross over frequency and to cut out the energy at lower freqs which is overloading the tweeter. You possibly only need the tweeter for freqs above, say, 10 kHz.
  8. Dec 3, 2015 #7
    I can't raise the crossover point. I'm limited by the upper end extension of the mid. That's why I needed verification on how to go about choosing the correct inductor value to turn my filter to one of 2nd order.

    Why are there two curves present?

    Regardless, both curves are pretty flat beyond 2khz which goes against what every user had to say about them. On the one, it doesn't look like it's starting to break up til 4khz

    See attached

    Attached Files:

Share this great discussion with others via Reddit, Google+, Twitter, or Facebook