Announcement

Collapse
No announcement yet.

The SideTowers - An F6 Corundum Build

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

  • 4thtry
    replied
    Time to fix my 4.6kHz on-axis boo-boo.

    Olson's classic AES paper (Google: Direct Radiator Loudspeaker Enclosures) suggests that the worst shape for a baffle is a circle (fig 8), while the ideal shape is a sphere (fig 6). My OmniMic testing bears this out. I was able to obtain the best overall measurements by converting my facets into spherical roundovers with a continuously variable radii. Kind of a cross between standard facets and standard round-overs.

    So, I decided to grind the facets down with my belt sander and rebuild them with flexible material that could be fashioned and/or re-fashioned based on measurements. I used OmniMic's polar protractor tool (see OmniMic's help file) to design and make a foam rubber replacement skin to fit over my poorly designed facets.


    Leave a comment:


  • 4thtry
    replied
    Time for an update. Need to solve some problems and get these things finished. Here is my solution to the banana jack problem:

    The Peerless corundum tweeters have banana jack type connectors on the back (see pics). So, to insure a solid, trouble free connection, I picked up a set of PE's 091-3608 locking banana plugs (4 in each package) (see link below). These plugs have an insulated thumbscrew that spreads the banana plug tines, which locks the plugs firmly into position. In addition, the insulated thumbscrew and polycarbonate outer shell on this model help to prevent any possible shorting as I push the wires into the cabinet and mount the drivers.

    Please note that I did not attempt to use PE 091-3604, because this particular model does not have have insulated thumbscrews. As a result, they completely short out this tweeter when the two plugs are inserted. Since the jacks are only spaced 1/2" on center, completely insulated plugs are a must.


    Link: https://www.parts-express.com/angled...insu--091-3608

    Leave a comment:


  • jhollander
    replied
    Nice work Bill.

    Leave a comment:


  • 4thtry
    replied
    After several days of testing, I am starting to reach some conclusions as to what is causing the 4.6kHz response dip. I taped a temporary cardboard template over the facets to flatten them out to see what happens. I also removed the woofer and taped a piece of fiberboard over the woofer hole to eliminate the woofer flange reflection from my measurements.

    Switching back and forth with the OmniMic sine sweep running, the dip seems to be mainly caused by the proud, surface mounted woofer flange, especially the part of the woofer flange and surround located closest to the tweeter. The faceting also appears to be enhancing the 4.6kHz dip, but only by about 1dB or so. Most of the dip appears to be coming from the proud woofer flange. To confirm this, I ran a test with the woofer removed and replaced by a flat surface and the dip almost completely disappears (no change to the facets).

    Also, if I cover the baffle, up to the edge of the woofer flange, with 1/16" foam rubber and then cut a small 1.625" hole to expose the tweeter dome, wrapping this foam tightly around the facets, the dip at 4.6kHz almost completely disappears, both on and off axis. I can also eliminate most of the dip by placing two 1/16" thick foam rings around the tweeter dome (no other changes).

    Attached are a couple graphs showing my results so far. Looks very promising.

    Leave a comment:


  • jhollander
    replied
    It looks like an asymmetric baffle would minimize the diffraction lumps in the higher frequencies. The SB Acoustic Rinjani and Ara kits are using a unique faceted baffle. http://www.sbacoustics.com/index.php...-kits/rinjani/

    Leave a comment:


  • PWR RYD
    replied
    Yeah looks like it. The starting edge of the facets lie almost equal distant from the dome around 2/3 to 3/4 of the tweeter's circumference. So there is a pronounce diffraction effect just like John's sim shows.

    Leave a comment:


  • Wolf
    replied
    So the facets are doing more harm than good here...
    Wolf

    Leave a comment:


  • jhollander
    replied
    off axis, note 1 db scale

    Leave a comment:


  • jhollander
    replied
    From Edge

    Leave a comment:


  • 4thtry
    replied
    Originally posted by jhollander View Post
    Looking good Bill! You can always do a near field on the tweeter then back away to see where the dip occurs.
    Thanks for the tip, John. Here is what I get by following your suggestion:


    1) Mic 6" or less from the tweeter protective shield, on axis with dome: At this distance, the dip at 4.6khz does not exist. Response is very flat from 3-15kHz. And no dip can be seen as I move the mic up or down or side to side 0-30 degrees.

    2) Mic approx 8" away. A very slight depression (approx 1dB) begins to appear at 4.6kHz on axis. But this slight dip goes away as I move the mic up or down or side to side 0-30 degrees.

    3) Mic approx 12" away. The dip on axis is now approx 3dB centered at 4.6kHz. The depression begins at about 3kHz and ends at about 7kHz. As before, the dip goes away as I move the mic up or down or side to side 0-30 degrees.

    4) Mic approx 20" away. The dip on axis is now approx 5dB centered at 4.6kHz. Again, the depression begins at about 3kHz and and ends at about 7kHz. As before, the dip goes away as I move the mic up or down or side to side 0-30 degrees.

    5) Mic approx 30" away. No change, same as #4.


    Seems promising. Will continue to experiment until I solve the problem.

    Leave a comment:


  • jhollander
    replied
    Looking good Bill! You can always do a near field on the tweeter then back away to see where the dip occurs.

    Leave a comment:


  • 4thtry
    replied
    Originally posted by Billet View Post
    Cool looking design and craftsmanship, I expect they will image very well...

    Can you hear that dip in the tweeter? If it is only on one axis, it may not be an issue at all.
    Thanks, Bill. At a distance of about 6 feet, as I move my head from side to side (or up and down) by about 10 to 15 degrees, the overall tonal balance remains the same. The speaker does not, all of a sudden, sound dull in the 3-7kHz region when I listen directly on-axis. But I can detect a slight change in the imaging, kind of a "phasy" type quality, which seems to do something to the soundstage. Still working with pieces of felt around the edges to see if I can clear up the problem.

    Leave a comment:


  • Billet
    replied
    Cool looking design and craftsmanship, I expect they will image very well...

    Can you hear that dip in the tweeter? If it is only on one axis, it may not be an issue at all.

    Leave a comment:


  • Wolf
    replied
    That's what I had to do in my Cecropia build. The faceted ridges needed the felt badly, and it did fix the issues.

    Later,
    Wolf

    Leave a comment:


  • AdamM3
    replied
    Originally posted by Navy Guy View Post
    If you look at the sim, you can see the red line doesn’t drop below the black, which means it’s not a cancellation. It’s in the actual tweeter measurement, not just the system response. It looks like Bill has the woofer surface mounted so my guess would be reflection off the woofer frame causing diffraction. This would also be why it disappears off-axis.
    Good point, I stand corrected.

    Leave a comment:

Working...
X