Announcement

Collapse
No announcement yet.

loudspeaker driver data sheet peeve

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

  • loudspeaker driver data sheet peeve

    In designing an enclosure from a modeled alignment, it is necessary to convert the net enclosed volume used in the alignment model to physical enclosure geometry in the mechanical design, and that requires some guesswork about the enclosed air volume displaced by geometry of the drivers, including some guesses about cone geometry, guesses about motor geometry and exposed former length, etc. The guesses are worse when working with the few measurements provided on a data sheet rather than useful physical measurements of a real driver.

    Variation in inaccuracies in that guesswork injects excessive error in realizing the acoustic response transfer function seen the alignment modeling.

    The data sheet should include the air volume displaced by the driver geometry behind a datum plane mating with the mating surface of the mounting flange. More calculations would still be needed to establish net enclosed volume in the application, but the effect of driver geometry would be known and the errors associated with guesswork would be removed from the design.
    "Our Nation’s interests are best served by fostering a peaceful global system comprised
    of interdependent networks of trade, finance, information, law, people and governance."
    - from the October 2007 U.S. Naval capstone doctrine
    A Cooperative Strategy for 21st Century Seapower
    (a lofty notion since removed in the March 2015 revision)

  • #2
    Originally posted by JRT View Post
    In designing an enclosure from a modeled alignment, it is necessary to convert the net enclosed volume used in the alignment model to physical enclosure geometry in the mechanical design, and that requires some guesswork about the enclosed air volume displaced by geometry of the drivers, including some guesses about cone geometry, guesses about motor geometry and exposed former length, etc. The guesses are worse when working with the few measurements provided on a data sheet rather than useful physical measurements of a real driver.

    Variation in inaccuracies in that guesswork injects excessive error in realizing the acoustic response transfer function seen the alignment modeling.

    The data sheet should include the air volume displaced by the driver geometry behind a datum plane mating with the mating surface of the mounting flange. More calculations would still be needed to establish net enclosed volume in the application, but the effect of driver geometry would be known and the errors associated with guesswork would be removed from the design.
    Modeling is good, but nothing beats an actual test enclosure of whatever alignment. What I mean is build a prototype and run your tests and measurements, then you'll have some real data to work with.

    Comment


    • #3
      Not too many manufacturers provide that info, not sure why. As enclosure size grows, the margin of error based on guessing driver displacement goes down - but it can dramatically alter the alignment on smaller enclosures. Indeed, "the measurement doesn't match the model" is a common refrain.

      I suppose a guy could measure displacement using rice or some other type of dry aggregate. David Weems recommended this in one of his books (I think - been a long time) when measuring Vas.

      I just add 10% to the model and go with it. My guesstimate is that covers drivers, bracing, and crossover. I fudge up a bit, maybe 15%, on smaller enclosures. I am usually pretty close on measured vs modeled.
      Don't listen to me - I have not sold any $150,000 speakers.

      Comment


      • #4
        Adding 10% has been a common practice since long before modeling software existed.
        www.billfitzmaurice.com
        www.billfitzmaurice.info/forum

        Comment


        • #5
          My take is, that for sealed boxes it doesn't matter, also - if you don't take your own T/S parms - it doesn't matter, also ...
          even when ported, for largish boxes (oh, about 1.5 cu.ft. or more) it won't matter much.
          Also, you can get very close just by estimating the motor (a cylinder, like a hockey puck), and a cone (shape). Also, I've often wondered how many guys ADD to the box volume for the cabinet wall removed (when front/flush mounting drivers)?

          Also, when I first got back into speaker building (about 20 yrs. ago), I used some equations from David Weems (book) for Vb, Fb, and F3. When I started using WinISD, I used Weems' formulas but modified different constants in them to hit numbers that WinISD spit out ('cause I liked to make up design "charts" for Vb & F3). I found that WinISD consistantly prompted box volumes that were larger (say, in the +5% to +10% range?) than David's equations. So ...

          I decided that Weems' equations (smaller Vbs) required you to account for all "internals" (drivers, XO parts, braces ...) so you could oversize the box a bit while not making it too small - to get the proper net volume; and that when using WinISD, I could basically ignore those (rather trivial) things and still hit my target box parameters. Never really looked back, really; and I've verified my box tunings using WT/DATS for the past 15 yrs or so. The only real caveat (as Johnny said) is when you're designing a "small" ported box (definitely under 0.5 cf) with a low-tuned, largish dia. port tube (which can consume LOTS of internal space) and/or using a driver w/a large motor (TB's W5-1138..."F" & W6-1139..."F" subs come directly to mind). ($0.02)

          Comment


          • #6
            Eminence lists Driver Volume Displaced on their spec sheets. You could use Eminence drivers or maybe pick a similar driver from them and use that number. After a few quick checks, 10% does seem pretty reasonable however.

            Comment


            • #7
              FaitalPro lists driver volume. Bass Box allows you to enter driver dimensions and will then estimate the driver volume. I can't comment on how accurate the estimate is, I've never used this feature.

              Comment

              Working...
              X