As usual, I measured the t/s parameters before committing to any particular design. The parameters of the driver were as follows:
The low Le was a bit of a surprise. It can actually affect the response in the pass-band, but this can be corrected if necessary with an external inductor wired in series with the driver.
After trying and disposing of several different variations, the images below represents the HornResp simulation that I finally decided to use. The result should be a small (4.3 cu.ft. net) box that's capable doing a decent 40 Hz to above 100 Hz, that could be built with one sheet of 4 x 8 plywood.
Coming up with a folding scheme that met the basic requirements of the simulation was somewhat challenging. I ended up creating an Excel spreadsheet that allowed me via the "Goal Seek" function to determine the optimum cabinet dimensions for a tapped-horn corresponding closely to the HornResp simulation parameters with a stepped and tapered path as indicated in the diagram below . I also configured the spreadsheet to give me the dimensions of each panel making up the box. I deliberately left out any bracing, so I could later objectively determine the impact of adding bracing to the box.
The following table indicates the panel dimensions provided by the spreadsheet, assuming panel thickness of 17mm (yes, my ply was just shy of an actual 3/4" thickness)
Not included in the above list is the panels required to brace the enclosure properly.
I then proceeded to build the box (see images below). Ingredients: one sheet of 4 x 8 3/4 ply, box of drywall screws, wood-glue, a few bricks (to act a temporary table), some mounting screws for the driver and of course the driver. I left the top and bottom panels as removable panels (only screwed into place, rather than screwed and glued), so I could easily add and removing damping and bracing to each section of the enclosure. My eldest daughter decided to help me with this build.
While it's a close match, a closer look at the measured impedance suggests that the horn's resonant frequency is around 43 Hz, a little higher than predicted.
The image below shows the measured frequency response, compared to the HornResp predictions. I used the close-miked method, which may not be accurate enough for this test.
Getting rid of the "notch"
The duct also shifts the resonance frequency of the horn down a little, from 43.7 Hz to 39 Hz. This is actually a good thing for this particular design, as it flattens the response a bit. It's possible that the addition of the duct will cause compression problems at higher power levels, but I need to perform more tests to confirm if this is actually happening.
Indicated in the graph below is
how the frequency response changes with the addition of the "dogfood duct".
And below is an image of the duct as I've implemented it in my P.O.C. tapped-horn: