3D printing desktop speakers. Thoughts?

[ChargedCapacitor]

The video in post #2 has some measurements for a couple of infill densities with PLA vs. the same cabinet dimensions with a variety of woods. My thinking is like @Behrend's but that's informed guesswork rather than evidence based. I had wondered what a gyroid fill of the cabinet internal space would do, but found a paper suggesting a high-pass filter with lumpy pass band so probably not desirable.

I've done some extensive cabinet test prints on my P1P, and I have found PLA-tough printed at 40% cubic infill to have the best reliability, as well as price / time to performance. PETG at such a dense infill is almost impossible to print without serious gum-ups, and gyroid in general with any material at that density is time consuming, and prone to general de-coherence. With the PLA-T @ 40% cubic, my cabinet (which is also printed in such a way to increase structural rigidity) is as solid, or more solid, then a similarly sized wood cabinet.

Here's the beta version of the 3d printed speakers I've been using at my desk:

 

[ChargedCapacitor]

I have doubts about using any infill if attempting to fill up with any material after printing. I recognize the usefulness of filling up cavities, but I decided to go with 10 % Cubic infill and not do any fill ups at this point. This being my first 3D printed speaker I'm open for some adventure If the printed cabinets resonate, my first attempt to control that would be by damping them with such treatment that I've used with wooden cabinets before. I also plan to play with the cabinets internal geometry, this being a 3D print it should exploit the possibilities. I have yet to start printing, it turns out I need to have all parts at hand and measured before printing, because drilling holes to a printed cabinet is not an option.

-- Early version of the back plate internal surface. Attempting to diffuse internal sound waves to reduce cabinet ringing.

Cabinet ringing is caused by frequencies lower than what would be canceled out by your proposed geometry. What works best is non-uniform bracings and non-parallel walls. And stuffing, of course.