-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
3D Printed speakers - Unlock impossible designs compare to conventional
FWIW, I 3D printed the XMechanik Mechano23 speakers. I've designed cabinets before, but this was the first time 3D printing anything of this size. I built a test box to test out some materials and settled on PETG for the combination of strength and dampening. Dealing with the warping of the components was not fun, though. In the end, they turned out great and I'm really happy with how they look and sound. Getting there wasn't trivial though.
Any pictures?
Render here - https://audiosciencereview.com/foru...axis-measurements-included.41757/post-2229465
Print here - https://audiosciencereview.com/foru...axis-measurements-included.41757/post-2247946
My son got this first version -- we built it together. I'm currently working on another print to see if I can't mitigate that warping by printing 2 halves on the back where I really won't care and by trying to do a bit of work optimizing a few supports rather than avoiding them.
can you share the 3D files too?
Sure -- here's a link. Mind you, I'm still early in learning FreeCAD. This version fixes a few issues and is also setup to try to have the braces not need supports when printing. But, if anyone has any suggestions, I'd love to hear them.
cool thanks.
which PETG (brand) did you use? In my experience PETG behaves very different depending on supplier.
They look really nice in you picture.How to you surface finish them? Filler and lacquer coating?
This was bulk Sunlu off eBay. I had a lot of warping to deal with in those first ones. One was done in 4 layers and the other in 8 layers, which helped considerably. I've flipped the print direction now to have driver holes up so that it would only be the backs that would warp. I'll know in a day if it worked 
JB Plastic Weld epoxy for bonding and for some gap filling. Bondo and then the red bondo putty / glaze and finally filler primer and paint.
JB Plastic Weld epoxy for bonding and for some gap filling. Bondo and then the red bondo putty / glaze and finally filler primer and paint.
This was bulk Sunlu off eBay. I had a lot of warping to deal with in those first ones. One was done in 4 layers and the other in 8 layers, which helped considerably. I've flipped the print direction now to have driver holes up so that it would only be the backs that would warp. I'll know in a day if it worked
JB Plastic Weld epoxy for bonding and for some gap filling. Bondo and then the red bondo putty / glaze and finally filler primer and paint.
thx - really helpful
BTW, printing them in 2 parts with drivers up was the winning solution. Top half came out great! I've added some screw holes to mount the xover on the bottom half.
kemmler3D
Master Contributor
Too slow on this, but if you are set up to take measurements I'd be curious what impact doing a more organic-shaped cabinet might be - namely smoothing out the outside and also adding irregular structures on the inside to further combat resonances / standing waves.
I'd be willing to help model it if you were planning on building more, but otherwise nice work on the prints!
What printer are you using? (my guess is one of the Bambu units?)
LOL, the test box I did tried out PLA, PETG, and TPU and also had a 2D Gordon curve surface test. I'm not sure at all though that the variation I can reasonably get (say +/- a cm or so) will affect the standing wave frequency for anything not controlled by the tweeter anyway. It would impact the resonance of that panel though. I used a piezo type pickup mic to look at resonance of the panel and it might have helped a bit.
The front of the case has a bit of a quarter-round to help with the diffraction. It's not huge by any means. If we go much further, though, in terms of egg shapes or whatever, we'll need to rework the xover a bit and this is a super-optimized xover to begin with (see the thread on these speakers for more).
Oh, and yes, a Bambu A1. I am building a second pair now. Each half of a box is ~1.5-2 days to print and about a kilo of filament.
kemmler3D
Master Contributor
Hmm, good point. I was thinking egg shapes of some sort...
I guess it really depends on how big you're willing to let the exterior of the speaker get, which in turn probably means breaking it into more pieces, more difficult assembly, etc.
My broad idea is to have scattering surfaces inside the cabinet to break up standing waves, on the other hand, it's probably true that you can't do any internal structures big enough to matter much for woofer frequencies. Even 2Khz is 170mm.
Yes, to that end it's a lot easier to print elaborate bracing than it is to build with MDF or wood... just make sure everything's at 45 degrees or so and you can print it in any orientation you want.
somebodyelse
Master Contributor
- Joined
- Dec 5, 2018
- Messages
- 5,248
- Likes
- 4,821
I'd suggest looking at papers on metamaterial absorbers and diffusers as they work with panels much thinner than the target wavelength. They tend to use either Helmholtz resonators, or folded tuned pipes embedded in the panel similar to KEF's metamaterial absorber. The structures could potentially act as the infill too. This is of course much easier to say than do
For anyone interested, I put my model for the Xmechanik Mechano-23 speakers up on Makerworld
I've also started a video series on the 3D printed Mechano-23s. I've got the first 3 parts of the series up, on: 1) Why the Mechano23? 2) 3D printing? Really?, and 3) Crossover part purchasing. Zero idea if anyone is interested, but it's a fun distraction
For anyone daring out there, this might be an interesting print as well. It is quite a long print as well. I have not printed it, just stumbled upon it and thought it was interesting, since the drivers are also not to expensive.
A Multi Entry Horn that can play from 300 hz
A Multi Entry Horn that can play from 300 hz
I've started doing some tests while considering a 3d printed speaker build. The test box shape is meant to emulate what part of a speaker enclosure might look like with 19mm walls and a flat bottom with hollow space in the center.
Test setup: As suggested in one of the project links here I printed with a 7% gyroid fill, then poured a plaster/glue mixture into the hollow walls. From this project, I took the tip of adding some regular wood glue to the plaster mixture to make it stronger and (hopefully) less resonant. I used about a 1:10 ratio of glue to plaster+water (mix the plaster with water first to a nice consistency, then add glue). I just drilled a hole on one end to pour the plaster.
I have access to a CT scanner with my job so I took a scan of the dried plaster-filled box. This gives us a comprehensive view of the internal structure of a filled 3d printed enclosure. We can see if there are any air bubbles or areas where the plaster filler didn't get to. The CT also provides density information on the materials in the scan.
Results: The CT images are below. The plaster/glue mixture flowed well within the hollow walls and there were only a few small air bubbles. In the 3d view (top right) you can see the plaster volume and there are no big gaps or areas where the plaster/glue mixture did not reach (I didn't fill the walls up all the way so its irregular on top). The plaster/glue mixture averages about 850HU (Hounsfield Unit is a relative density scale used in CT. web search for more info). The PLA+ is about 700HU.
For comparison, Aluminum is about 1500-2000HU, regular water is 0HU and high quality Birch plywood averages -350HU (negative means its less dense than water). So the plaster fill at 800HU is quite dense and I'm pretty happy with that. I think this is a promising approach for a really dense speaker enclosure that was cheap and easy to make.
Below is a close up showing a few small air bubbles...
and here's a CT of some void-free birch plywood...
Test setup: As suggested in one of the project links here I printed with a 7% gyroid fill, then poured a plaster/glue mixture into the hollow walls. From this project, I took the tip of adding some regular wood glue to the plaster mixture to make it stronger and (hopefully) less resonant. I used about a 1:10 ratio of glue to plaster+water (mix the plaster with water first to a nice consistency, then add glue). I just drilled a hole on one end to pour the plaster.
I have access to a CT scanner with my job so I took a scan of the dried plaster-filled box. This gives us a comprehensive view of the internal structure of a filled 3d printed enclosure. We can see if there are any air bubbles or areas where the plaster filler didn't get to. The CT also provides density information on the materials in the scan.
Results: The CT images are below. The plaster/glue mixture flowed well within the hollow walls and there were only a few small air bubbles. In the 3d view (top right) you can see the plaster volume and there are no big gaps or areas where the plaster/glue mixture did not reach (I didn't fill the walls up all the way so its irregular on top). The plaster/glue mixture averages about 850HU (Hounsfield Unit is a relative density scale used in CT. web search for more info). The PLA+ is about 700HU.
For comparison, Aluminum is about 1500-2000HU, regular water is 0HU and high quality Birch plywood averages -350HU (negative means its less dense than water). So the plaster fill at 800HU is quite dense and I'm pretty happy with that. I think this is a promising approach for a really dense speaker enclosure that was cheap and easy to make.
Below is a close up showing a few small air bubbles...
and here's a CT of some void-free birch plywood...
Depending of what you have used, it might take weeks or months for the volatiles to go away and the shrinkage/cracks to appear. You don't mention how long have you waited before running the CT, but if not long, I would suggest to repeat after some time. Can you measure moisture level?
Good questions. The water is part of the chemical reaction in the plaster that makes it set, so there are no volatiles or off-gasses. The reaction is exothermic though so maybe pour in parts for a large enclosure in case it heats up enough to deform/melt the plastic.
I don't expect cracks to be an issue. The wood glue is added for the purpose of making the dried plaster less crumbly and its encased in the plastic walls with the gyroid fill in between to lend its support too. If you were to drop a dried speaker, then it might crack. I can CT again after "x" amount of time to confirm this theory though.