But a VP is formulaic?
I was digging around in the original papers etc by Voigt. Are these just as applicable today?
My plan was this;
1. Build a decent outcome...ie if it takes a couple of tries, I have a fast prototype procedure and the mold material in that case is recyclable
2. Once that is finalised, to parametise the 3d model so that I can apply bespoke parameters and have the mold be custom made eg the current proto might be a 6m room but being able to simply drop in drivers, room type etc and have it create the molds as a response is where Im hoping it will go. This is the process I use for acoustic QRD/Helmholz solutions and it works amazingly.
Do you think that is too ambitious?
BTW...I did suspect this was your day job with level of reference. Thanks for taking the time to answer me 🙂
Other than the length calculation being wrong for the stated tuning, the offset calculation being sub-optimal, no account for driver volume requirements & no alignment derived it’s excellent in every way. 😉
Voigt’s patents are of historical interest but not a guide to creating a quality pipe design- the formulas above, in his defence, had nothing to do with him, thank goodness!
Voigt’s patents are of historical interest but not a guide to creating a quality pipe design- the formulas above, in his defence, had nothing to do with him, thank goodness!
lol! Glad I didnt spend any time there then...and glad you have a great day job!
So where do I start then? Is there no hope for parametric scaling (non linear)?
Last edited:
Yeah I initially had a good look there...love the gallery!
But no mention of cone but they have an FB
Well, no -there wouldn't be. Very few DIYers have the capacity or funds to produce 3d shapes of that size. As I said back in post 55 however, these are basically all standing-wave generators, and the difference between a 'cone proper' and a box with two parallel sides (technically parabolic expansion) is in most instances completely swamped by all the other factors (length, Vb, Zd, Zv, mass-loading, damping, driver production tolerance variations, local climatic conditions &c.) so if you assume a conical expansion, i.e. take an existing design & correctly adapt, preserving the critical design features, you'll be ~near enough, i.e. within a dB in most instances.
Thanks again for your succinct responses. Im pretty verbose...sorry...and yes I have the privilege of having left over resources from ex business to play with...a couple of 1m capable printers and in this case, one I made myself as 1100mm x 380mm delta (and its very fast with 30mm3 extruder @ 0.5mm so its perfectly suited to this type of shape.
Well that creates some confidence, so Im thinking of using the MAOP and the supplied PDF as a test project...would that be advisable? Then transpose some of those parameters if needed?
Of course its pretty easy to go expo or parabolic etc in shape as you just use formulas for the spline generation. Im thinking it might be better to model it all as a straight horn ie parabolic as it seems there are alternate compromises but cones seem to be the least favourable impedance wise) and then once its simulated, converting to analytic, creating the deformations eg curving the vertical and adding the "dressing"