No announcement yet.

MicroSat compact 2 way for use with a sub

  • Filter
  • Time
  • Show
Clear All
new posts

  • MicroSat compact 2 way for use with a sub

    Original build thread

    Self Hosted write up
    Compact 2 way satellite speakers for use with a subwoofer.
    Design Goals

    The primary goal for this design is to provide accurate +-3dB response from 20kHz down to 100Hz. The primary purpose was not to be a stand alone full range speaker, but to pair with a subwoofer to handle all content below 100Hz, specifically my Indy 8 powered sub. This had to be done in a compact enclosure to fit as easily as possible into a tight room. The design was intended to use against a wall with minimal response adjustments, but also be able to operate on a stand without losing too much mid-bass. Aesthetically the design had to look as good as I could make it with a hybrid real hardwood dark stain and gloss black combination.
    Driver Selection

    Targeting the Denovo mini speaker knock down cabinets available from Parts Express gave me an enclosure volume to work with. Many woofer models were ran in the small enclosure, ultimately the DA115-8 was chosen for a good sealed alignment that will get close to 100Hz when assembled. The ND16FA-6 tweeter was chosen for the top end due to its smooth response and compact installation footprint.
    Enclosure Design

    Enclosure Assembly

    The Denovo box was remarkably easy to glue up. Simply applying Titebond glue and using painters tape to hold it all together to dry was sufficient to get the box together without gaps. Clamps would cut down a little tolerance wiggle from the joints, but were not required. The baffle was intended to be removable, so a set of 45 degree corners were glued in flush with the baffle edge in order to accommodate threaded inserts. The threaded inserts were installed right in the corner meeting points of the cabinet edges and the 45 degree corners.
    The baffle was made from a poplar board. I used some machine screws that I cut and grinded to a point to provide reference holes for the mounting bolts. Using an outline of the cabinet, I measured and marked the center points for the woofer and tweeter.
    Pro tip
    Once together, the drivers were test fit and some breathing room was opened behind the woofer using a simple hand rasp and files.
    The finish is in 2 parts. The MDF cabinets were primed, painted, wet-sanded, and polished to a gloss black. I had never done a wet sanded gloss black finish before, so this was a learning experience. It took several rounds of paint and sanding to quit sanding through and into the primer again.
    The Poplar baffles are stained a deep expresso and finished with a semi-gloss poly. The finishing part took quite some time, as I had not tackled these types of finishes before. Ultimately it took about 4 rounds of finishing, sanding back down, refinishing, and repeating the whole process to get a satisfactory result.
    Electronics engineer, woofer enthusiast, and musician.
    Wogg Music

  • #2
    Crossover Design

    The crossover was planned using the Dayton FRD and ZMA files before the drivers were even purchased. Using free tools including Response Modeler, WinPCD, and XSim the speakers were modeled with a 2 way 2nd order electrical filter plus a notch filter on the woofer to suppress the breakup of the aluminum cone. The tweeter padding was added both before and after the filter, I found that adjusting values on both sides allowed me to better match the phase between the drivers.
    Once the cabinets were finished, the drivers were installed and tested using a breadboard version of the crossover outside of the cabinet and measured with ARTA. The rear of the cabinet was left with a hole waiting for the terminal cup, but this would affect the low end primarily leaving the crossover region of 3-4kHz unaffected for measurement. The actual measurements led to a minor tweak to the capacitor values and padding resistors to bring the response within to my +-3dB requirement and improve the reverse null measurement that indicates the phase is well aligned between woofer and tweeter.
    With the values finalized, the crossovers were built up on pegboard and mounted in the rear of the cabinet.
    Tips and Tricks

    Most of the learning involved in this build revolved around the finishing portion of the cabinet build. Tips include:
    • For gloss black finishes using enamel, ensure the paint layer is very think to avoid wet sanding back down to primer. Do not do this in layers of wet coats, instead keep coats reasonably thin, and apply repeated layers with 24 hours in between to ensure the enamel is cured. Otherwise, the paint is too thick to cure properly and takes weeks or months to properly harden.
    • For stain and poly on Poplar, do not over sand the bare wood. Leave some roughness for the stain to soak in. Use fine steel wool or 600 grit paper to smooth prior to poly application. Use high quality brushes to apply your poly and minimize bubbles. Use the same 600+ grit or 000 steel wool to smooth between poly coats.
    Ultimately, the best tip for finishing is to be patient, read directions, and research on line to hear what others have already experienced.

    With these in place and powered by the Indy 8 subwoofer / 2.1 amp
    Electronics engineer, woofer enthusiast, and musician.
    Wogg Music


    • #3
      Nice project. Very nice write-up.


      • #4
        Originally posted by Chris Roemer View Post
        Nice project. Very nice write-up.
        Thank you sir! Sent from my SM-G920P using Tapatalk
        Electronics engineer, woofer enthusiast, and musician.
        Wogg Music


        • #5
          great write up -- looks like a fun project

          for the XSim crossover, what speaker offset / mod delay (inches) did you use?

          using the PE sim files, when I enter the values from your crossover design, my XSim graphs look a bit different.


          • #6
            I didn't use an offset in Xsim, that was mostly for a simplified schematic. In WinPCD I used X = 0.025, Y = -0.070 and Z = -0.010 for the woofer relative to the tweeter. Those are in M at 1M, so you'll have to do some trig to get that into Xsim. Also, those are total guesstimates

            Wait... I lied. I looked at the tweeter in Xsim, which of course is the reference and at 0". Whoops... the woofer is set with a 0.5" mod delay in Xsim. Which was calculated from the WinPCD entered measurements.
            Last edited by wogg; 12-18-2017, 08:25 AM.
            Electronics engineer, woofer enthusiast, and musician.
            Wogg Music


            • #7
              That is some serious "shoehorning" work.
              Most impressive.