Announcement

Collapse
No announcement yet.

LCR 2-way Crossover Design Help Needed

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • SubNRG
    commented on 's reply
    Thanks Chris. I'm going to give that a shot. I really appreciate it!

  • Chris Roemer
    replied
    These are the component changes you'd make to an MTM XO to get (roughly) the same response as for the TM - but the sensitivity will go up by about +6dB/v. Of course, the load drops (on the woofer section) to a Zmin of about 3.5ohms near 200Hz.

    HP: only the padding changes, to SR=9n (was 10) / PR=5n (was 2).
    Both these L-pads present about a 12n load to the pass filter.

    LP: since the MTM's impedance halves, the "equivalent" filters sees inductor and resistor values HALVE, and caps double.
    so, 1mH series coil (was 2), 12uF shunt cap (was 6), Zobel = 8n+10uF (was 15n+5uF).
    Box volume needs to double, and use 2 ports (or their equivalent) for the same tuning/F3.

    Leave a comment:


  • SubNRG
    replied
    Originally posted by Chris Roemer View Post
    Yeah, the horn (tweeter) makes a CC (horizontal MtM) a bit messy, since the horn (7" wide?) pushes the 2 midbasses an unusually large distance apart (and that affects horizontal lobing/comb-filtering).

    Could you run the basses (nearly) tight to each other w/the horn centered above them (so THAT baffle might be 18"w by 16"-17"h)?
    It looks like a 21" wide by 15" tall baffle will have the drivers about 1.75" apart. The baffle would be pretty much butted against the TV and fireplace with an inch or two gap, max on each side. Seems like that should be close enough, right? If so, any chance you can send me the crossover specs for this? I really appreciate your help!

    Leave a comment:


  • Chris Roemer
    replied
    Yeah, the horn (tweeter) makes a CC (horizontal MtM) a bit messy, since the horn (7" wide?) pushes the 2 midbasses an unusually large distance apart (and that affects horizontal lobing/comb-filtering).

    Could you run the basses (nearly) tight to each other w/the horn centered above them (so THAT baffle might be 18"w by 16"-17"h)?

    Leave a comment:


  • SubNRG
    replied
    Originally posted by Chris Roemer View Post

    If you can handle a 4ohm load, AND if you double your box volume and run 2 woofers in parallel (MTM), the CD can be unpadded by 6dB raising the system sensitivity to 96dB (which is pretty good for "near" Hi-Fi).
    Thanks Chris! I have a couple questions on this:

    1. Is it okay that it would have to be a horizontal MTM configuration for my center channel (I have a TV and fireplace limiting my available vertical space).
    2. If I can run the center MTM horizontally, would it be possible to send me the crossover specs for that configuration?
    Last edited by SubNRG; 09-02-2021, 03:49 PM.

    Leave a comment:


  • fpitas
    replied
    Yeah, the z offset is one of the tricky parts about horn speakers. If neither tweeter polarity works real well, I guess you have to go to unsymmetrical slopes.

    Leave a comment:


  • Chris Roemer
    replied
    Imagine parts drawn L > R (amp to drivers), as std. convention.
    2nd order HighPass w/ "L-pad" for about -18dB attenuation on CD:
    3uF series cap, 2.5mH shunt coil (to gnd. - DCR can be higher), SeriesResistor = 10n(ohms) / PR = 2n
    2nd order LP (on woofer) w/Zobel-like thingy to shape rolloff:
    2.0mH series coil (low DCR, so iron-core is indicated), 6uF shunt cap, Zobel is 15n + 5uF.

    Crosses near 2k. Essentially no BSC (for on/near wall), runs about 90dB sensitivity. Solid (above) 8 ohm load.

    Used data for B&C ME45 horn. Worst "blip" is driver (woofer) DROP just above 1kHz.
    Have NO idea on tweeter polarity (due to depth of waveguide/horn on CD). Try it both ways and see.

    If you can handle a 4ohm load, AND if you double your box volume and run 2 woofers in parallel (MTM), the CD can be unpadded by 6dB raising the system sensitivity to 96dB (which is pretty good for "near" Hi-Fi).

    Enjoy

    Leave a comment:


  • SubNRG
    replied
    Thanks, Chris Roemer . That would be great. I appreciate it!

    Leave a comment:


  • Chris Roemer
    replied
    The CD padding will be so severe (even w/NO baffle-step - for "on-wall") that to maintain 8n(ohms) the PR would need to be near 1 ohm.
    That CD DOES have F/Z data avail. (when mounted to their ME45 horn). I'll probably play w/an XO for you using THAT data, which should be substantially better than a "generic".

    Leave a comment:


  • fpitas
    commented on 's reply
    Be sure to design your CD pad to present 8 ohms to the crossover. You'll have to guess at the CD impedance to design the circuit, but the plane wave tube impedance should be close since you're not using a real horn. If you get DATS, that will measure the CD impedance in situ on whatever waveguide you choose.

  • Steve Lee
    commented on 's reply
    Excellent plan.

  • SubNRG
    replied
    Thank you all for chiming in. I don't currently have that kind of equipment, but I will order it. What I may do is order those drivers along with some pre-fab eminence 1,600Hz xovers and build everything out. That way I can take my time playing with the xovers and A/B them to the eminence ones as a learning experience. Once I get some true measurements, I'll pop back in with some better data.

    Leave a comment:


  • fpitas
    replied
    You'll definitely need measurements of each driver in the enclosure, as people described above. You might get by with the published data for the woofer (maybe, sorta), but the DE250 on that waveguide will be a complete unknown until you measure it. I also tend to agree with LOUT about using a bigger waveguide. Here's another appropriate one for a reasonable price: https://www.parts-express.com/Eminen...rew-On-290-555

    although you will need a screw-on adapter. Another nice one is the H290C by Pi Speakers (Google is your friend).

    Leave a comment:


  • LOUT
    replied
    Might be important that the DE250's claimed flat resonse and ~1.4khz recommended XO point are assuming you're pairing it with an exponential shaped horn that can load it down to ~1khz rather than the waveguide which doesn't reach as low. https://www.parts-express.com/B-C-ME...2-Bolt-294-622
    The lower XO point when pairing with a large, 8" woofer helps avoid beaming and can get a wider soundstage without gaps / lobes at certain frequencies.
    Without the horn's help down low, the tweeter will have to work really hard at its lower frequencies and you might get noticeable distortion or harshness.

    I'm pretty sure you'll ideally want to cross below 2khz, probably as near the 1.4-1.5khz as you can safely get the tweeter to reach (1.2khz would be really cool if it's possible, might be unrealistic to ask though).

    That said, I know very little about horns and there are folks here who ARE brimming with knowledge on the subject. So take what I said with a grain of salt.

    EDIT: that horn attachement in the link isn't the only one of it's kind...just happened to be the exact one the DE250 used in the spec'ed graphs/measurements. There are likely more affordable options like fpitas is showing in the post below.
    Last edited by LOUT; 08-31-2021, 01:19 PM.

    Leave a comment:


  • mattsk8
    replied
    Not sure what you mean by "reference levels", but that should be a fairly loud speaker if you plan on using a sub.

    When Kornbread asked if you can take .frd and .zma measurements he's asking if you have the equipment needed to measure the driver responses (frequency response and overall impedence).

    The short version for xover design is... Build your cabinet, then measure the frequency response of each driver, then measure the impedence of each driver. Then import those measurements into the xover design program of our liking and play around until we get a respectable simulated response. From there we assemble that simulated xover, and manipulate components until we're happy.

    Without response measuring equipment (like Omnimic and DATS for example), doing this successfully is very difficult.

    Leave a comment:

Working...
X