September 29, 2012

S3 progress

Here is a sneak peek at the S3 throat:

In S2, this aspect wasn't modelled very carefully because I built it manually, using a combination of filling with bog and smoothing with a circular file and sand paper. S3 will be sculpted with a CNC router so it has to be right.

The profile is critical in this design.

I have a couple of goals here:
  • avoid diffraction
  • create the smoothest possible transition from the exit tube (circular cross section) to the horn where the midrange units come into play (rectangular)
  • keep the machining cost sane
  • retain constant directivity performance


  1. That resembles a bit with an oblate spheroid profile a la Geddes.

    If that's the case, it would be the first attempt to use such a transition to a four-sided conical.

  2. The OS profile was made popular by Geddes but he doesn't actually own it as such, although I did come across a patent of his for applying it to an elliptical cross section.

    This isn't quite as unique as you think. There is some similarity with JBL progressive transition waveguides although some of them appear to use diffraction around the throat, which I have avoided. The work of DB Keele also comes into it with his CE horn which joins an exponential throat to a conical mid section with an added mouth flare. However, the CE horn is narrow coverage vertically and a different profile with pretty major pattern flip. I've used the same profile vertically and in this design, being multiway there is no need to get a single driver to run so low. So the longer expo throat becomes undesirable and in this design it would make crossover work more difficult.

    Essentially I've had to come up with a new throat that is unique but not without similarity to what others have done.

    This isn't an OS profile. This design transitions a little more gradually from the exit angle to the coverage angle. I've also changed the design since the iteration that you see above.

  3. I should also mention that this is not the final design, but one iteration of many that didn't meet every design goal.


All comments are moderated.