Announcement

Collapse
No announcement yet.

DIY Mini Bluetooth Speaker

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

  • DIY Mini Bluetooth Speaker

    Looking for a sanity check on a mini bluetooth speaker I'm looking to build. I'll be using a pair of Dayton DMA80-4 3" full range drivers in a 2.12L (each) ported enclosures tuned to 72hz for an EBS alignment with a PE3WBT amp (2x1.8w @ 4ohm). To save shipping cost I'm using manufacturer data to hopefully get a "pretty close" estimate so I only have to place one parts order. I used Dayton's 15 degree off axis curve and modeled the simulation in Bagby's Response Modeler at 15 degrees horizontal off-axis. I used the off axis because we'll probably never listen to this speaker on axis. It'll be sitting on a counter while we're up moving around, or even sitting on a desk we won't be on-axis with the driver.

    This is my first time flirting with a notch filter. I couldn't decide between a 1.5uF cap or 2.2uF, so I was just going to try them both.

    Other than the fact I'm not using measured data, anyone see any problems with the methodology/thought process or any glaring issues I'm missing with the filter? This is also my first time using VituixCAD so want to make sure I'm not missing a setting or something in the program.

    Thanks!
    Attached Files

  • #2
    I loaded PE's 15 degree frd file into VituixCAD and come up with pretty much the same model that you posted, except that my model does not have baffle step or diffraction applied. So my curve has the same overall smoothness, but a different slope which peaks up by about 6dB at 100Hz. If you could supply me with baffle dimensions, driver location, and roundovers, I could apply these to the frd and see if I come up with the same curve. Thanks.
    SideTowers: http://techtalk.parts-express.com/fo...corundum-build
    Totally Flat: http://techtalk.parts-express.com/fo...5-totally-flat
    Plumber's Delight: http://techtalk.parts-express.com/fo...notech-winners
    Linehopper: http://techtalk.parts-express.com/sh...Esoteric-build

    Comment


    • #3
      Originally posted by 4thtry View Post
      I loaded PE's 15 degree frd file into VituixCAD and come up with pretty much the same model that you posted, except that my model does not have baffle step or diffraction applied. So my curve has the same overall smoothness, but a different slope which peaks up by about 6dB at 100Hz. If you could supply me with baffle dimensions, driver location, and roundovers, I could apply these to the frd and see if I come up with the same curve. Thanks.
      Thanks. The baffle is 10.5" wide x 7.5" high. I had the driver centered at 2.625" along the X axis and 3.75" on the y-axis, measured from 6' at 15 degrees off axis. 3/8" roundover.

      Comment


      • #4
        Personally, I'd tune higher (80 or even 90Hz). I don't find an EBS alignment works very well on small drivers that don't reach that low anyway.

        Comment


        • #5
          I created a diffraction model in VituixCAD using your baffle dimensions and came up with the following. It is fairly close to your model, except that I reduced the series inductor from about .24mH to .05mH. I used VituixCAD diffraction, merger, and enclosure tools instead of Jeff Bagby's Response Modeler. I am not an expert on VituixCAD settings, so if you (or anyone else) spots an error, let me know.

          Enclosure tool six pack. Note that I used Chris's recommendation and tuned Fb a little bit higher (85Hz instead of 72 Hz)

          Click image for larger version  Name:	Enclosure tool 6 pack.JPG Views:	0 Size:	247.1 KB ID:	1482784

          VituixCAD diffraction model. Note that the horizontal angle is entered as 0 and the microphone is 6 feet on-axis with the tweeter because we are using Dayton's 15 degree off axis curve, which already accounts for off-axis roll off at high frequencies.

          Click image for larger version  Name:	Diffractrion model.JPG Views:	0 Size:	143.4 KB ID:	1482785

          VituixCAD merger tool model:

          Click image for larger version

Name:	LF align - diff - FF merger model.JPG
Views:	229
Size:	171.4 KB
ID:	1482791


          VituixCAD 6 pack model:

          Click image for larger version  Name:	Vcad 6 pack model.JPG Views:	0 Size:	274.3 KB ID:	1482790
          SideTowers: http://techtalk.parts-express.com/fo...corundum-build
          Totally Flat: http://techtalk.parts-express.com/fo...5-totally-flat
          Plumber's Delight: http://techtalk.parts-express.com/fo...notech-winners
          Linehopper: http://techtalk.parts-express.com/sh...Esoteric-build

          Comment


          • #6
            Originally posted by Chris Roemer View Post
            Personally, I'd tune higher (80 or even 90Hz). I don't find an EBS alignment works very well on small drivers that don't reach that low anyway.
            I can try different tunings.

            Comment


            • #7
              Originally posted by 4thtry View Post

              VituixCAD diffraction model. Note that the horizontal angle is entered as 0 and the microphone is 6 feet on-axis with the tweeter because we are using Dayton's 15 degree off axis curve, which already accounts for off-axis roll off at high frequencies.
              Thanks. I guess my thought here was that the diffraction itself would look different depending on the measurement location.

              Comment


              • #8
                Originally posted by Impious View Post

                Thanks. I guess my thought here was that the diffraction itself would look different depending on the measurement location.
                This is probably true. On VituixCAD's diffraction tool screen, I initially tried to enter 15 degrees instead of 0 degrees in the "Axis - Angle Hor" field. But when I did this, the high frequencies rolled off dramatically. If you enter 15 degrees in Bagby's Response Modeler, the high frequencies develop a small amount of ripple, but they do not roll off dramatically. There is definitely a difference in the way the two programs handle microphone position and off-axis angles that I do not fully understand.
                SideTowers: http://techtalk.parts-express.com/fo...corundum-build
                Totally Flat: http://techtalk.parts-express.com/fo...5-totally-flat
                Plumber's Delight: http://techtalk.parts-express.com/fo...notech-winners
                Linehopper: http://techtalk.parts-express.com/sh...Esoteric-build

                Comment


                • #9
                  Originally posted by 4thtry View Post

                  This is probably true. On VituixCAD's diffraction tool screen, I initially tried to enter 15 degrees instead of 0 degrees in the "Axis - Angle Hor" field. But when I did this, the high frequencies rolled off dramatically. If you enter 15 degrees in Bagby's Response Modeler, the high frequencies develop a small amount of ripple, but they do not roll off dramatically. There is definitely a difference in the way the two programs handle microphone position and off-axis angles that I do not fully understand.
                  Me either. But I've heard some say that Bagby's is more accurate than VirtuixCad, thats why I went that route.

                  Comment


                  • #10
                    In terms of power handling, tuning to Fs is always your best bet. From the impedance model, it looks like Bill's tuning is pretty close to this and will yield the most control in the lower range.

                    In terms of the notch applied, I am guessing it is to hit the bump at 3.9kHz. I would also think centering on that bump a bit closer to either side will lessen the dips to either side of it and result in a flatter overall response. Granted I have not modeled it, so take this with a grain of salt. I would try a 0.6mH and a 3.3uF cap just eyeballing it, and likely a 10 ohm resistor. Then use the tilt coil (your 0.25mH) as your baffle step to tilt down the overall response, and maybe a 3.0uF bypass cap across it to get some treble back.

                    From another perspective, if you used the on-axis response for the driver, and provide the amount of tilt *up* in response that would typically yield flat at 15* off-axis, you'll end up likely at about the same result.

                    Let me know if you wanna mock this up, as I might have some parts handy if you already have the drivers. I'd be game for a voicing day...

                    Wolf
                    "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
                    "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
                    "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
                    "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

                    *InDIYana event website*

                    Photobucket pages:
                    https://app.photobucket.com/u/wolf_teeth_speaker

                    My blog/writeups/thoughts here at PE:
                    http://techtalk.parts-express.com/blog.php?u=4102

                    Comment


                    • #11
                      Originally posted by Wolf View Post
                      In terms of power handling, tuning to Fs is always your best bet. From the impedance model, it looks like Bill's tuning is pretty close to this and will yield the most control in the lower range.

                      In terms of the notch applied, I am guessing it is to hit the bump at 3.9kHz. I would also think centering on that bump a bit closer to either side will lessen the dips to either side of it and result in a flatter overall response. Granted I have not modeled it, so take this with a grain of salt. I would try a 0.6mH and a 3.3uF cap just eyeballing it, and likely a 10 ohm resistor. Then use the tilt coil (your 0.25mH) as your baffle step to tilt down the overall response, and maybe a 3.0uF bypass cap across it to get some treble back.

                      From another perspective, if you used the on-axis response for the driver, and provide the amount of tilt *up* in response that would typically yield flat at 15* off-axis, you'll end up likely at about the same result.

                      Let me know if you wanna mock this up, as I might have some parts handy if you already have the drivers. I'd be game for a voicing day...

                      Wolf
                      Thanks! I'm using the notch to flatten the rising response due to baffle step. Couldn't both correct properly for the baffle step and get that 3.9khz peak knocked down, so this was the trade-off.

                      For tuning, since I only have 1.8w on tap and running FR, I was trying to extend bandwidth, accepting the associated trade offs. I can cut a few port lengths and try them out.

                      I got excited and bought all the parts :D Just waiting on one inductor and the weather to break so I can make some sawdust

                      Comment


                      • #12
                        if you used the on-axis response for the driver, and provide the amount of tilt *up* in response that would typically yield flat at 15* off-axis, you'll end up likely at about the same result. techzpod.com mobdro

                        Comment


                        • Steve Lee
                          Steve Lee commented
                          Editing a comment
                          This ^ bot just copied some of Wolf's posting # 10 and claimed it as its own.
                      Working...
                      X