half sphere open baffles

[turk 182]

i'm reminded of an old experiment regarding diffraction and baffle shape. being that rectangles are probably the worst ,truncated pyramids better and spheres being the best i wonder if anyone else has played with 1/2 spheres to create open baffles.

[mondogenerator]

nice idea, though you may create a cavity resonance, or extra reflection which may spoil or change the rear wave. Hard to say. Ive wanted to try something similar but im not sure im skilled enough to make it!

[turk 182]

well i think i'm going to try it
and i have just the toy to do it with a CNC millstation
once i have a chance to knock one off i'll measure it and post the results

[jrenkin]

I imagine it would model just like a u -frame. I use a u- frame box for subs, instead if H-frame and it is nice.

[turk 182]

jrenkin
how would the u or h frame reduce the diffraction caused by the flat surfaces of the baffle
H.F. Olsen would probably disagree with you

[jrenkin]

I think you misunderstand me. The baffle step would be different, since it is a sphere, but the resonant response of the box would most closely resemble a u- frame in practice, so that might be a reasonable place to start modeling, as opposed to trying to calculate the resonances if a half sphere, which would be pretty complex.

[Einric]

I thought about making spheres with a punching balloon and laying up fiberglass on the outside.
3/8" of fiberglass with that constant curve would be extremely strong.
You then would just need to split it in half and dampen the inside to prevent ringing.

[turk 182]

given that i have the cnc millstation at my disposal i'll have it cut a series of progressively smaller rings and glue them all together to form the half sphere then mill it smooth but i think i'll try doing a before and after as in just the stack of rings raw then milled and maybe the same for the inside just to see what the results are

[turk 182]

jrenkin
i have my reservations about modeling and design programs not all are created equal and some i have used have given me dubious results yes they make the math chores easier but sometimes when your pushing the normal envelope of things they poop out

[mondogenerator]

a lot of sense in the H frame analogy, i think. As far as front and rear radiation pattern, i think it would behave as a U or H frame, but the cut off would be governed by the front to rear path length. Im not a dipole expert though, so im just making semmi educated guesses. Would be interesting seeing the results. I think i recall a thread here where buoys where used as a form or mould for hemispheres.