The reflector audio square might be closer to what you're envisioning. The coax in the Genelec makes the design pretty different from if you would use a compression driver I would think.
Yes the reflector is probably closer. But I would really prefer some domes (rarely existing for lf) instead of theese resonator-holes. In this aspect the genelec looks much nicer.
I believe there's an error in the documentation.
It states that "A short note on the Spacing item: In this case only the "top" and "bottom" values are used, the other ignored."
I have not found this to be the case. In my experiments with User Defined Enclosure Plans, the "spacing" parameter works the same way as it does when you DON'T have a User Defined Enclosure Plan.
In other words, you have to define the spacing for left, right, top, and bottom.
It's possible that Marcel may have been referring to the fact that the design in the manual is defined by two quadrants (not one) and in that case, the spacing WOULD be defined by the "top" and "bottom" and the left and right wouldn't matter.
But that's kinda confusing.
Just posting this in case anyone else runs into it, it took me some to figure out why my enclosures kept erroring out like this:
Info : Started on Fri May 09 23:42:21 2025
Info : Reading 'bem_mesh.geo'...
Error : Curve loop 211 is wrong
Error : 'bem_mesh.geo', line 523: Could not add curve loop
Error : Curve loop 212 is wrong
Error : 'bem_mesh.geo', line 525: Could not add curve loop
Warning : Impossible to recover edge 618 618 (error tag -1)
Warning : Impossible to recover edge 619 619 (error tag -1)
Warning : Impossible to recover edge 245 245 (error tag -1)
Warning : Impossible to recover edge 538 538 (error tag -1)
Warning : Impossible to recover edge 539 539 (error tag -1)
Warning : Impossible to recover edge 541 541 (error tag -1)
Warning : Impossible to recover edge 542 542 (error tag -1)
Warning : Impossible to recover edge 545 545 (error tag -1)
Warning : Impossible to recover edge 546 546 (error tag -1)
Warning : Impossible to recover edge 234 234 (error tag -1)
Warning : Impossible to recover edge 506 506 (error tag -1)
Error : Unable to recover the edge 677 (1/10) on curve 191 (on surface 52)
Error : The 1D mesh seems not to be forming a closed loop (2 boundary nodes are considered once)
Error : The 1D mesh seems not to be forming a closed loop (2 boundary nodes are considered once)
Error : ------------------------------
Error : Mesh generation error summary
Error : 11 warnings
Error : 3 errors
Error : Check the full log for details
Error : ------------------------------
error: gmsh call status = 1
It states that "A short note on the Spacing item: In this case only the "top" and "bottom" values are used, the other ignored."
I have not found this to be the case. In my experiments with User Defined Enclosure Plans, the "spacing" parameter works the same way as it does when you DON'T have a User Defined Enclosure Plan.
In other words, you have to define the spacing for left, right, top, and bottom.
It's possible that Marcel may have been referring to the fact that the design in the manual is defined by two quadrants (not one) and in that case, the spacing WOULD be defined by the "top" and "bottom" and the left and right wouldn't matter.
But that's kinda confusing.
Just posting this in case anyone else runs into it, it took me some to figure out why my enclosures kept erroring out like this:
Info : Started on Fri May 09 23:42:21 2025
Info : Reading 'bem_mesh.geo'...
Error : Curve loop 211 is wrong
Error : 'bem_mesh.geo', line 523: Could not add curve loop
Error : Curve loop 212 is wrong
Error : 'bem_mesh.geo', line 525: Could not add curve loop
Warning : Impossible to recover edge 618 618 (error tag -1)
Warning : Impossible to recover edge 619 619 (error tag -1)
Warning : Impossible to recover edge 245 245 (error tag -1)
Warning : Impossible to recover edge 538 538 (error tag -1)
Warning : Impossible to recover edge 539 539 (error tag -1)
Warning : Impossible to recover edge 541 541 (error tag -1)
Warning : Impossible to recover edge 542 542 (error tag -1)
Warning : Impossible to recover edge 545 545 (error tag -1)
Warning : Impossible to recover edge 546 546 (error tag -1)
Warning : Impossible to recover edge 234 234 (error tag -1)
Warning : Impossible to recover edge 506 506 (error tag -1)
Error : Unable to recover the edge 677 (1/10) on curve 191 (on surface 52)
Error : The 1D mesh seems not to be forming a closed loop (2 boundary nodes are considered once)
Error : The 1D mesh seems not to be forming a closed loop (2 boundary nodes are considered once)
Error : ------------------------------
Error : Mesh generation error summary
Error : 11 warnings
Error : 3 errors
Error : Check the full log for details
Error : ------------------------------
error: gmsh call status = 1
Attachments
I don't see how could be the width of the flat front baffle given by anything else than the x coordinate of P0.
Yes, the whole waveguide 🙂
There are far bigger concerns in the overall design than this. Like a crossover, for instance...
- BTW, just today, I moved all my prototype waveguides to another place, and again noticed how better are the lighter ones (at least they "feel" that way). This is again my recommendation - make the printed parts as lightweight as possible (i.e. with just a very sparse infill) and put some coating. As for the surface, I used two 0.6mm perimeters even on the largest ones and it seems enough. The overall thickness of the petals alone does wonders. (Just to be clear, this is for the freestanding horns. For the baffle mounted this could be different.)
Last edited:
It's all good, just took me a bit to figure it out, so thought I'd share. Things that are obvious to most people aren't always obvious to me 🙂
The curve P0 -- PB (the "side" shape of the enclosure) is fixed by the absolute coordinates set by the user, there should be no additional (automatic) arithmetics involved, as far as I remember. You only need to ensure that the flat baffle part (given by x_P0) is large enough to accommodate the waveguide.
No, I don't think it would work. The only chance would be an ESP, but so far I haven't seen a really sucessful implementation with the bigger drivers. The real wavefronts probably tend to be too random (which is a problem of every multicell horn).
Sure, the aim of an ESP is to improve the directivity at frequencies where the waveguide is no longer able to "bend" the flat wavefront (i.e. ~ wavelengths shorter than the throat diameter). Once it's spherical, there's suddenly no upper limit to constant directivity. At least in theory and simulation.
Last edited:
Just printed the Tritonia XS. Here it is with the -S and -M for size comparison.
Mickey
Mickey
The -S was printed with a Bambu X1C in Sunlu PLA+ 2.0. The -XS was printed with my new Bambu H2D in Sunlu silk PLA+. The -M was printed in PETG from a printing service. It did not turn out well, which drove me to purchase the H2D printer. I have a new -M printing as I type this. I will post pictures when it is done and I return from a work trip.
Mickey
Mickey
On his Instagram, @galucha shared some incredible results with PETG, with a long print time of 14 hours for a throat-piece of a MEH.
So X1C prints better than the new H2D? The S one in orange looks like it has been sanded already!
Yeah I've had some good luck with polymaker PETG-rCF08. Seems pretty strong too. Needs a hardened nozzle though.