Results 1 to 3 of 3

Thread: 3-way with 2235H, 2123H and SEOS-12

  1. #1
    Junior Member
    Join Date
    Oct 2007
    Location
    US
    Posts
    17

    3-way with 2235H, 2123H and SEOS-12

    I've built a pair of cabinets with curved sides and the aforementioned components. All panels except the sides are made from 3/4" MDF; the baffle is two 3/4" MDF panels laminated together. Each side was made from 6 laminated layers of 1/8" hardboard, and required no kerf cuts. I CNC-machined the baffles on both sides: rebates, dadoes and flush-mounts for all drivers. I will be rounding over the cabinet corners with a 3/4" bit when they are next back in the shop.

    Cabinets are 44" tall, 18" wide (front), 12" (back) and 15" deep. 44" places the waveguide at, or just slightly above, ear-level when seated in any number of desk chairs, which is what I usually have handy, and also allowed the 2235H to be raised a bit higher off of the floor. Distance between each driver is 1.25". Sealed cabinet of approximately 4.8 cu ft for the 2235H after subtracting bracing and separate boxes for 2123H and SEOS-12. I chose to isolate the waveguide from the back-pressure waves of the 2235H in order to minimize the chance that they could stress the waveguide walls. The compression driver is a Denovo Audio DNA-360.

    Cabinet is fully stuffed with fiberglass and braced with 2X4's.

    I have very little experience in the acoustical/electrical parts of building speakers. I would like these to be a fully active setup, preferably with as little complexity as possible. I find the capability of software crossovers very intriguing, in particular the combination of Equalizer-APO and rePhase. When combined with an HDMI connection that can carry 6 separate channels of audio, I would only need one A/V 7.1 Receiver to do all of the amp work. I have listened to this combination of drivers using an Ashly XR-4001 and separate Yamaha amps, and liked it very much; however, I would like to learn how to do a proper crossover design rather than depend on my ears.

    I would very much value input into several things:
    • Measurements: impedance and gated impulse response, particularly in as much detail as possible. (as in, I need an English-language description of how to set up amps, mic preamps, software, cabling etc.) I do not own a measurement system (LEAP, Omnimic, CLIO or similar), and so planned to use either HOLMImpulse or Room Equalization Wizard along with a mic preamp.
      • Physical placement of microphone relative to the front baffle when executing impulse response measurements - distance from baffle, and x/y location relative to baffle, especially the y-axis for the waveguide and 2123H.
      • Measuring individual drivers versus measuring combinations of drivers with the same signal:
        • Why would I choose to measure drivers individually, versus together? Is there a good book or series of articles that anyone can refer me to?
        • If I measure all drivers with the same signal, how do I prevent damage to the compression driver from receiving low-frequency signals?
        • If I measure each driver independently, how do I account for the vertical offset of drivers (I believe VituixCAD has the ability to incorporate these offsets; I'm learning how to use that software).

      • Level-matching between all three drivers:
        • I assume that it is better to pad down to the 2235H, rather than up to the compression driver.
        • Matching levels may be more functionality in VituixCAD than I am currently aware of. If anyone on the forum has experience with this software, any pointers would be greatly appreciated.

      • I will likely be forced to measure these in a small room, which will probably limit my usable far-field impulse response to about 400hz. I will probably need to merge both near- and far-field measurements for the 2235H in order to get valid low-frequency data.
        • Anyone here have experience performing these two types of measurements?

      • I plan to crossover the 2235H at about 250 hz, where both it and the 2123H should still be radiating omni-directionally. I assume that I need not be concerned about a directivity match between these two drivers, but must absolutely be concerned about it between the 2123H and SEOS. From the polar plots I have seen of the 2123H and SEOS, it appears that approximately 1700Hz would be appropriate.
        • The 250Hz frequency is basically arbitrary. The 2235H seems to have some "fatness" at frequencies above 250Hz, although I am aware that the 4430/4435 used it up to 1000Hz. My subjective observations here likely are biased due to room effects, poor level-match between the 2123H and 2235H, and who-knows-what-else. Is there a solid reason why I would raise this crossover frequency?


    • Crossover slopes
      • With linear-phase filters, where (to my knowledge) I need not be concerned about the phase-shifts inherent in minimum-phase crossover typologies, I have never found a convincing argument to be made favoring a steeper crossover slope over a shallower one (or vice versa), other than to reduce the amount of sound being broadcast by more than one driver simultaneously, usually for power-handling reasons. Any insights, and/or reference articles?



    I know this is a lot, so If you've read this far, I both commend you and appreciate it! I am aware that the SEOS guys over on diyaudio and avsforum have done quite a bit of crossover work with this waveguide, and I'll likely be asking over there for help also.

    A few photos of the build:

    Name:  CAD Assembly.jpg
Views: 383
Size:  34.1 KBName:  Baffle toolpath #1.jpg
Views: 364
Size:  37.1 KBName:  20170715_174610.jpg
Views: 389
Size:  73.6 KBName:  20170909_194904.jpg
Views: 366
Size:  82.8 KBName:  20170909_194913.jpg
Views: 365
Size:  84.5 KB

  2. #2
    Junior Member
    Join Date
    Oct 2007
    Location
    US
    Posts
    17

    Additional photos

    Name:  20170923_175211.jpg
Views: 363
Size:  76.5 KB
    Name:  20170923_175237.jpg
Views: 360
Size:  65.1 KB
    Name:  20170923_175250.jpg
Views: 363
Size:  49.6 KB
    Name:  20171015_225903.jpg
Views: 407
Size:  74.4 KB

  3. #3
    Senior Member Ian Mackenzie's Avatar
    Join Date
    Apr 2003
    Location
    Australia
    Posts
    7,100
    You have already made a good start

    Itís easy to over think this stuff until you do some practical work but l will try and answer your points:

    If you can take the enclosure outside

    Put it on your smooth concrete drive on itís side

    Put the mic on the same surface 2m from the box

    Tilt the box forward slightly

    This is by far the best and easiest good measurement unless you can haul the box up 3 m in the air

    If you can only do Inside do near field 2 cm from the cone

    But splicing the curves together not that easy and you donít see the integration of the crossover point

    There are some good text on boxes, l am away so l will post in the NY

    Use a20-30 uf capacitor in series with the compression driver and do not go below 400 hertz

    Unless your soft ware can process the mls data with minimum phase conversion and tail adjustment worries about offset are pointless

    Pad down the compression driver not the woofer

    You need to acquire a calibrated mic and at a minimun studio6 or one of the packages from Parts Express before any of this becomes worthwhile. You just canít do the it otherwise

    Take you enclosure outside as per my comment above and the levels can be precisely

    Take it to a friend who has a paved area otherwise and in one Saturday afternoon you will know exactly where you are.

    Talk to the guys on the seos forums about your drivers and the seos

    I un clear about your crossover plans however:

    The the 2123 has a hump in the response and requires specific equalisation working the filter

    The best example is the 4344mk11 schematic

    The seos need a form of CD compensation and a min 12 dB -18 dB filter

    I would not take the 2123 above 1300 hertz

    Phase compensation between drivers is a complicated subject

    The intent is to avoid spl null in the vertical power response through the crossover region and within your listening window

    You also need to optimise the on axis phase at crossover point

    Point blank: without investing in test equipment capabilities you are a non starter and should look hard at an existing Jbl design.

    I appreciate diy Audio is a lot of fun BUT the reality is your project is a one of a kind

    You have chosen some good drivers. On your own it could be 1-2 years of learning and empirical adjustments before do are in the ball park. Many give up after buying the best drivers like the Jbl 1501 /0435 drivers and sell it off because they are dissatisfied. My advice is leave it to the experts who will charge a fee for their time or diy a clone of a known design

    Itís a bit like building your own ocean yacht without any experience.

    You will be a lot happier

    The scientific side of this is a world of knowledge and specialist experience that might look straight forward from bypassing the purchase of a commercial product but itís not

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Similar Threads

  1. What should I build, 2225h,2234h,2226j,denovo dna360+seos 15
    By Mrmky in forum Lansing Product DIY Forum
    Replies: 15
    Last Post: 12-29-2017, 02:08 PM
  2. JBL system w:2235H,2123H,2425H+xover?
    By Russellc in forum Lansing Product DIY Forum
    Replies: 2
    Last Post: 07-25-2007, 06:25 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •