DIYing TAD TH-4003s...

[strader]

Normalizing to the y-axis shows there is a bit of a honk around 4khz. I wonder if this is the region the AFAST port is tuned too?

 

[strader]

field_vertical_plane.png showing dispersion at mouth of the horn
Field7k.png showing horizontal plane at 7khz
field16k.png shows what akabak starts doing once the mesh density is too low to accurately model 16khz.

 

[strader]

The fusion 360 project, STL files, Akabak project are included in this zip download for anyone interested: https://github.com/shawnrader/TAD-TH-4003-Replica/releases/tag/V1

 

[weltersys]

Normalizing to the y-axis shows there is a bit of a honk around 4khz. I wonder if this is the region the AFAST port is tuned too?

Seems likely, since your model shows a similar peak frequency to the "15 degree down" and null at "15 degree up" in the measured response shown in post #13.

By the way, most consider "honk" to be around 400 Hz, 4 kHz is closer to "quack" 😉 .

An AFAST port tuned to remove "honk" would be much larger than the one in the TH-4003.

 

[strader]

Made some progress to getting the sides and bottom horn sections done on my CNC. I used douglas fir construction lumber from my wood pile, which turned out to be a good decision because I had some sort of CNC screwup on each part. These should be good enough for initial measurements and to use as templates for future layups.