No announcement yet.

Help design a 2way crossover

  • Filter
  • Time
  • Show
Clear All
new posts

  • Help design a 2way crossover

    I am a beginner, am I on the right track or are there design errors?!

  • #2
    Make C1 about 6uf, and L1 about .3mH. See if it's better. Add a resistor, or two to the tweeter for padding.

    For padding try 6 ohms in parralel with the tweeter, and 3 or 4 ohms just after the xo.

    The woofer response looks OK, but it seems like a simpler filter would work. Try 3mH, and 12uf for example on the woofer.

    That tweeter is minimum order of 1000.


    • djg
      djg commented
      Editing a comment
      That means the tweeter is not in stock, and needs a minimum order of 1,000. Essentially unobtainable. Good for design practice I guess.

  • #3
    Thank you I will try....


    • #4
      Hello from Australia

      From the links provided, you have a 10"woofer and a 1" tweeter with a resonant frequency (Fs) of 1750 Hz.

      I don't know much about XO design, but from reading posts by experienced PETT members, the tweeter should be crossed at about double the Fs (3500Hz) and any 10" woofer might have trouble crossing that high. It looks from the supplied FR graph that it might be a stretch?

      Also, the off-axis response of the woofer isn't supplied - it may drop off quite a bit compared to the published graph which I would think is on -axis.

      Do you already have the drivers - the woofer is noted as out of stock.



      • khaled
        khaled commented
        Editing a comment
        No, I don't have drivers, just trying to learn the design method.
        I know you all will help me

    • #5
      So, I just used your XO w/the 4n(ohm) vers. of that same tweeter.
      I end up w/about a -10dB hole centered around 4kHz.

      W/the next suggestion (and adding a 10n series resistor "out front") at least the tweeter's level gets more reasonable w/only about a -4dB hole @ 4k.

      Your Zobel-like-thingy (on the woofer) doesn't do much due to the too small cap value.

      To get something even more reasonable (using the 4ohm horn - that you can buy in quantities of only 2) try this...
      2nd order HP on horn w/an "L-pad" (SR = 5n / PR(parallel resistor) = 2n): 6uF series resistor, and 0.30mH shunt coil (to grnd).
      Hey, that's exactly what rpb came up with! (The "L" rolls the tweeter off in a more controlled manner (near its Fs) than simple series resistance.

      2nd order LP (on 10"): 2.5mH series coil (low DCR - meaning "iron core"), and a 2uF shunt cap. The "Zoble-like-thingy" is 10n + 6uF.
      This yields a fairly flat 90dB system w/basically "full" BSC, although that woofer can't really get much below 70Hz in a vented box (0.5cf).
      Also, a 10" 2-way crossing near 3kHz (which this does) isn't the greatest idea due to the woofer "beaming" at freqs. above about 1.5kHz.
      Unless this is for some "special app", (IMO) you'd be better off w/a 7" woofer (which'll do 40Hz prob.) and a non-horn dome tweeter.


      • khaled
        khaled commented
        Editing a comment
        Thank you for your interest, I'll do just that

    • #6
      Hi Khaled - You picked some fairly challenging drivers for a beginner! Judging by the high sensitivity, I'm guessing you want to build some kind of MT for a PA system?
      Co-conspirator in the development of the "CR Gnarly Fidelity Reduction Unit" - Registered Trademark, Patent Pending.


      • #7
        Hi Khaled

        I see you're using Xsim, which is a good program and reasonably easy to use.

        However, as you don't have the drivers, I assume you're using either factory supplied data, or data obtained through fptrace or similar? Using design software will get you closer to a result than using an off the shelf XO but it has limitations: apart from drivers sounding different in situ, manufacturers' data isn't always accurate.

        Without measurements, Xsim can't take into account phase, baffle width, driver spacing and cabinet configuration. but it's really interesting to use Xsim to compare different drivers, mess around and see how different components affect the electrical response of a crossover.

        It's also interesting to compare a real world, measured project such as Paul Carmody's Classix II or Curt Campbell's Tritrix using the manufacturer's data in Xsim. PE's product pages and the designers' web sites have the in box measurements and the PE driver pages have the ZMA and FRD files for the DC28, DC130 and DC160. The BC25 tweeter page doesn't have those, but they can be found using the fptrace program, or I could upload it for you.

        I've attached my comparison for the Classix II, you can see quite a difference. The top graph is the Xsim, the bottom, real world. The simulation used the same XO components, of course.

        I'd consider Chris' suggestion - he really knows his stuff - and consider a smaller woofer and dome tweeter. Depending on what you want to achieve, your budget, how deeply you want to get into the hobby, listening requirements (room size, listening level etc) you could look at an existing proved project which has been well received, so you know you'll be happy.

        If you want to do your own project, we're here to help of course.

        Xsim v Real World.docx
        Attached Files
        Last edited by Geoff Millar; 09-15-2020, 08:30 PM.