Results 1 to 9 of 9

Thread: Scanspeak D2004 vs R2004 Tweeter Comparison

  1. #1
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247

    Scanspeak D2004 vs R2004 Tweeter Comparison

    I bought both these drivers to test and use the more appropriate one for my install. I wanted to see how the two compared so I did some testing. I do not have HD data on these, but to be honest, I don't need it. Either driver will be crossed above 3khz where distortion should be a non-issue. Additionally, I fear blowing them testing them at high levels, so I didn't... I might try to revisit this later using a filter in foobar to generate a bandpassed sweep in order to protect the drivers if I have the time. Until then, FR data is all I have along with some cool polars. Comparison time...The following test was done with Dayton's Omnimic system. Impulse response gated to approximately 11ms which is the first reflection time I get (I use a lot of treatment both on the floor and at the ceiling, and I have a large room).Thanks to Madisound for the special driver testing pricing!First off, mfg info:D2004:ScanSpeak Illuminator D2004/6020-00 Tweeter: Madisound Speaker StoreR2004:ScanSpeak Illuminator R2004/6020 Tweeter Small Ring Radiator: Madisound Speaker StoreIMPEDANCE AND T/S:
    D2004 Impedance:R2004 Impedance:
    FREQUENCY RESPONSE:
    Frequency Response of the D2004 at 0, 30, and 60 degrees:Frequency Response of the R2004 at 0, 30, and 60 degrees:
    POLARS FROM 0 TO 60 DEGREES:
    D2004:R2004:
    COMPARISON IN FREQUENCY RESPONSE BETWEEN D2004 AND R2004:Note: The Ring Radiator version is less sensitive than the Dome (d2004) by about 1dB, so to make it easy, the RR is always the graph below the top graph.
    0 Degrees:30 Degrees:60 Degrees:
    - Erin
    Your ears: The best tools you have... and they're free, too!

  2. #2
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247
    Driver Pictures:
    Your ears: The best tools you have... and they're free, too!

  3. #3
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam
    The bands in the polars are sort of odd phenomena but if I'm interpreting it correctly, it'd due to the oscillation from 3khz down in the FR measurements. At 11ms you should be good down to 90hz but it looks like the plot doesn't stabilize until 3khz. Wonder what's going on there.
    Last edited by Adam_MSS; 11-19-2011 at 10:13 PM.
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



  4. #4
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247
    Well, there's a few things going on here.

    One is the resolution and scale I'm using (1/24th octave, 2dB scale). As you know, the more data points you have, the more accurate the curve.
    Additionally the window itself being long should allow higher resolution down further. You typically don't see this in other data because these things may be presented differently by other testers.

    These two things coupled is what gives me the 'wave' response. I believe it's accurate data. When stitching NF/FF there's some overlap and when I measured these in the NF the result was practically the same below the Fmax (defined as 4911/driver diameter) which is what I expect with a larger window and a driver that rolls off naturally above 500hz (well above my window). If I'm measuring woofers, I'll be treating the test differently. I likely will from now on be doing NF/FF stitching to account for baffle rolloff* and to get more accurate data on the low end (nf allows you to set your window to a very high number because the S/N is so high).

    It's interesting to see just how measurements in NF and FF affect phase. Ultimately, it seems that the NF Fmax syncs up with with the FF measurement's beginning minimum phase. For example, these tweeters have an Fmax in the nearfield at about 6700hz. FF measurements show a phase swap (continuation from min to max given instaneously because it drops off the screen) at this point.
    I'll still have to measure more drivers of various size to see if this always is the case. Unfortunately there's just not much info on the effect of window setting and correlating high ressolution. Most everything I find simply discusses Fmax due to driver diameter and FF windowing due to (typically) the first reflection.



    *You'll note that I cut my measurement graph axis at 700hz on the low end. This is because at around this frequency, the driver rolls off naturally. Fs/Qts tell you this info so you don't HAVE to see it. Moreso, when I did this test I did it for myself to see how the two tweeter types compared in off-axis response. I didn't test these with intention on posting the results. I just figured I might as well after I did it.
    Last edited by ErinH; 11-20-2011 at 01:07 AM.
    Your ears: The best tools you have... and they're free, too!

  5. #5
    Devil's Advocate Adam_MSS's Avatar
    Join Date
    Aug 2010
    Posts
    1,578
    Real Name
    Adam
    Can you post the impulse response plot. I see the wave pattern often in my measurements prior to getting the gating adjusted so I'm just curious if there are any anomalies in the impulse response. Those oscillations are so smooth and sinusoidal that I have trouble believing they are anything but a mathematical or interference artifact, but I've been wrong before, so who knows.
    Last edited by Adam_MSS; 11-20-2011 at 01:11 AM.
    You don't use science to show that you're right, you use science to become right. - R.Munroe

    The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them. - W.L.Bragg



  6. #6
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247
    They could be a phase issue, for lack of better word. The fact that they're multiples is one thing that has caught my attention. Again, I've not found jack to really explain this. That's why I went playing around with the phase response and looking to see what the passband was for minimum phase response for both NF and FF.

    I'll have to run the test again to post an impulse response since I didn't save it off. IIRC, the first reflection was at around 11ms. I used treatment on the floor and above the baffle. Without treatment above, I'm at 7ms. Without any treatment at all it's around 3ms. All FF, of course.
    The higher the volume, the less any little ripples affect the response, but honestly, unless you're testing at mouse levels it won't show up. After a discussion here a couple months ago regarding linear distortion not being linear, I wanted to make sure what I said was right (that linear distortion is called linear distortion for a reason). Last night I did a few tests with other drivers and found that varying the volume only had an impact on the SPL level you see in the y-axis. ;)
    Unless I tested at ridiculously low levels where other noise was picked up, the results were the exact same and the impulse window would show you this since it's scale was affected as well via s/n. Obviously, the louder the input, the more rejection you get from the environment.
    Your ears: The best tools you have... and they're free, too!

  7. #7
    Big Daddy Chad's Avatar
    Join Date
    Aug 2010
    Location
    East-Central Illinois
    Posts
    1,570
    from a quick glance, other than beautiful pics... The myth about ring radiators is wrong... As I suspected.. Because I run them.

    Cheap ones though.

  8. #8
    Tester Extraordinaire ErinH's Avatar
    Join Date
    Aug 2010
    Posts
    1,247
    I think we're on the same page. I posted this on DIY:
    Quote Originally Posted by bikinpunk View Post
    What this shows me is something I had pretty much always thought: while a dome may have better response up high off-axis, the RR has a much smoother response and still isn't terribly bad off axis. In fact, the difference at 13khz is roughly only 3dB at 30 and 60 degrees after accounting for sensitivity.*This is the reason I provided the data the way I did. I wanted to see and show how the off axis response of these drivers differed. I think this may make some think twice before automatically discounting a RR based on the notion it can't perform well off axis.
    Your ears: The best tools you have... and they're free, too!

  9. #9
    Senior Member cvjoint's Avatar
    Join Date
    Mar 2011
    Location
    Northern California
    Posts
    801
    Real Name
    George
    You can fix the on axis response, you can't fix how many dbs is dropping with respect to on axis response. I dropped ring radiators behind a long time ago, not looking back. The Vifas are even worse.

Bookmarks

Posting Permissions

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