3D Printed Speakers: STOP wasting plastic on high infill percentage

Want to build this speaker yourself? Go to printyourspeak...
Instagram: / print.your.speakers
Parts Used:
Woofer: US: parts-express....
EU: rkn3.net/c/?si...
Tweeter: US: parts-express....
EU: rkn3.net/c/?si...
Dats V3 : US: parts-express....
EU: rkn3.net/c/?si...
About This Channel:
DIY audio has always had a steep learning curve of woodworking skills and audio engineering knowledge. The goal of this channel is to break down those barriers and make great DIY audio accessible to anyone with even a basic 3D printer. If you'd like to support this mission, like, comment, subscribe, and check out my website.
printyourspeak...

[@bigclivedotcom]

That was a very interesting test.

[@DonnyDonnMendoza]

This makes me want to try to print an enclosure in PETG with TPU infill! Gotta love the Aussie ingenuity!

[@austinschultz7698]

please try out the DIY Perks plaster and glue infil method if you do more testing, id be really curious how it compares!!

[@PrintYourSpeakers]

This is a popular request, definitely on my list!

[@PaperKhaos]

12:33 A: Told You So 12:54 A: A New Girlfriend 13:23 A: Yuzu 13:53 B: Told You So 14:15 B: A New Girlfriend 14:44 B: Yuzu 15:14 C: Told You So 15:36 C: A New Girlfriend 16:05 C: Yuzu 16:35 D: Told You So 16:57 D: A New Girlfriend 17:26 D: Yuzu

[@timreeves]

i've been designing and building 3d printed speakers for about 8 years now, and the conclusion I came to after extensive testing is that plastic on it's own just isn't very good. The best I found was filling cavity walls with a mix of PU resin and rubber dust (got it for free from a tire recycling company). Epoxy resin got too hot during curing, leading to warping and melting the print, and also got too brittle after a few months. backfill PU resin mixed with sand has worked well as a cheap(ish) and easy high performance filler, and casting cement (a fine cement with fibreglass fibres) has worked well as an incredibly cheap option, although it's performance was a little below the previous solutions i've mentioned (although still gave very comparable performance to MDF, similar amount of resonance but at a higher frequency). plaster of paris was too brittle and shrunk too much when I tried it. I've been tempted to try a silicon with a high shore hardness, but it's quite an expensive experiment if it ends up being shit! One thing that also kept going through my mind was creating an enclosure with a 3D sinusoidal wave for the wall shape, rather than just flat or curved surfaces, as this should be the stiffest surface you can get, it should also be very effective at breaking up standing waves inside the enclosure, leading to even less resonance. However when I've tried to embark on this project solidworks has brought my PC to an absolute standstill, and I've not mustered the energy to learn a surface modelling program like blender, that should have an easier time with it. drop me a message if you'd like to fire some ideas back and forth :)

[@lassikinnunen]

The complex wall/infill geometry seems like it would be the best reason to do it, like if you're printing flat surfaces you might just as well use some board. Or for just for internal structures, like trying to copy that wave trap thing kef puts on their speakers now

[@chrisbruce9497]

I was wondering if you would be willing to share your test results? I’m working on several projects at the moment and am looking for lessons learned.

[@timreeves]

@chrisbruce9497 unfortunately I did this all several years ago on a laptop that since died. I tested each method using a contact mic on a flat panel with 1.2mm PLA walls and a 10mm cavity. Using a weight on a string to strike the other side of the panel. Admittedly it would have been better to have it as part of an actual speaker, but this seemed like a suitable compromise for time and cost at the time, and mimicked the "knuckle rap test" as repeatedly and objectively as I could.

[@chrisbruce9497]

@@timreeves Ah! This is perfect! Thank you! It is the impulse response that I'm most curious about. I bought a few spools of woodfill and am curious to see if there are enough fibers to make a difference in damping.

[@MegaRockero20]

That dual printing between TPU and PLA sounds like a very interesting idea. That might be it!

[@chrisbruce9497]

I would love to hear a test with wood fill PLA!

[@DIYZONE101]

Really interesting video, I have a 3D printer but I like to build speakers the traditional way too much, it's fun!

[@ShaneCranor]

banger video! another simple test you can do is to hit the side of the speaker and listen to what sort of noise it makes. It would be cool to see how different infill shapes affect the speaker, it seems like cubic infill is the most structurally sound, but it might be interesting to see how that compares with lightning infill. Number of wall loops is also interesting. You could place the walls at varying distances inside the infill to increase stiffness and tune away certain resonances maybe?

[@alexallegria94]

printing myself an horn right now, for the closer part to the horn i am waiting to find a good material, great video btw it helped me a lot, about the resin, i did read myself that it could melt the plastic cause the chimical reaction of the resin, soo be careful, you might want to try different things, or test with an heat resistant material and try the resin, or you could test a lot of things like pouring liquid rubber inside with some steel balls , creativity is the limit!

[@johnrobinson3642]

Hello from a fellow Aussie! That was really useful info, thanks for the testing. I have printed a few speakers over the years, and always wondered about these exact questions, good to have some empirical data to reference in future.

[@vineethkumar6394]

Your videos are really good.. Keep going :) (from a fellow DIY speaker enthusiast)

[@roverdad]

This is awesome and the biggest question i had about 3d printed speakers.

[@mmmikem]

Expanding foam could also be an interesting infill

[@StarsAtNight1]

Maybe we can achieve better results with changing the orientation that we print the speaker in. Maybe printing using bricklayers could also help.

[@dantheman1998]

bro keep it up! this is exactly the info im looking for! I was also thinking about how you can print certain shapes to break up nodes. Like those um...sound diffusers inside of the box to eliminate peaks. This is just me spitballing because not sure if this would even apply for a box as apose to a room

[@BaddDukk]

I have been using 25-30% gyroid infill with 2-4 perimeter loops, all depending on wall thickness and if the walls are straight or curved. It has been working out alright so far.

[@andrewmcmillan8110]

Try a small % of infill and make the wall gaps bigger and pour in some liquid concrete or liquid rubber and run the test again. I also design and print smaller speakers enclosures and have always wanted to try this but never got around to it

[@philiprogers5772]

I've just had good success with another project using the following method. Slice your model using a 5% gyroid infill then remove the upper surfaces by reducing the top shell layers to zero. This will expose the top of the model and allow you to pour in a resin. Be careful though because all the top surfaces are exposed so use modellers putty to fill them in to prevent the resin draining out of the lower parts of the model. For my project I was also able to make pockets to take steel inserts.

[@SeanLumly]

TPU is plenty stiff with appropriate thickness. It's commonly used for mechanical gears.. Less flexible TPU filaments are also an option.

[@lobodelsur_]

Me gustó más el sonido del Speaker A. Que buen video para los que diseñamos algunas cosas en 3d como parlantes! Espero seguir viendo tu contenido!

[@rj7855]

It would be interesting the measure the effect on plastic type, different plastics have different densities and stiffness. Petg or a carbon fiber reenforced filament would be super interesting... Anyway subscribed

[@666nacirema666]

I print mine with a .8 - 1.2 mm nozzle usually do 4 or 5 walls walls 8 top and bottom layers usually and design all enclosures with built in bracing though I try to make it a bit more organic that what youre doing also no parallel or even flat walls in the interior seems to make the enclosure a lot stronger. I tend not to use more than 25% gyroid infill often less and if the need arises to have solid walls the good thing about gyroid is you can fill up the interior with just about anything that flows. In those cases the gyroid infill is mostly used for making prints easier to print and can get away with pretty low infill percentage. Gotta experiment with it more but theres this elastomeric polymer at the local big box hardware store and I think thats gonna turn out to be one of the more ideal and cheap options for filling the walls.

[@mmark666]

Great video! I can't hear any significant difference between the four with head phones.

[@someyoutuber7096]

Good stuff! Can't wait to print your speakers. It is in my queue of projects, already bought the files.

[@invendelirium]

my go to here is to have the entire volume filled with sparse gyroid, and just 2 perimeters as walls. Maxes out stiffness and internal volume. I do get occasional enclosure vibration problems with this approach though. And also... mass may not directly contribute to dampening, but a) reduces air-to-material coupling efficiency, b) indirectly improves dampening, because if the dampening rate is the same, it makes for a lower Q at the lowered resonance frequency.

[@TheRattleSnake3145]

What about the speakers Diy Perks printed and filled with a mix of plaster and pva glue?

[@blakefl8623]

Great video. Keep it up. Looking forward to the center speaker model.

[@PrintYourSpeakers]

Its coming!

[@GrehPinto]

maybe different infill patterns?

[@PcandTech]

You really went down the rabbit hole 😅

[@PrintYourSpeakers]

My curiosity won't let me stop, and even after all this, resisting the urge to keep testing was hard. If you look closely, my hair actually grows throughout the video because it took so long. I had to stop somewhere.

[@Beanpapac15]

Fantastic video. I'd be interested to see if theres a difference between a porous infill like gyroid vs a closed cell infill like cubic. Also I have some fairly large IDEX printers, so if you'd like dual material speakers for testing I'd be happy to print you some.

[@Dia1Up]

Thank you!!

[@slonlo364]

5:23 Thanks Bob.

[@mikecrane2782]

I think printing different infill patterns, gyroid might be worth pursuing

[@s0rc3]

What about silk PLA, at least for bambu this is a blend of PLA and TPU, you could use this as the infill if you have a multimaterial printer.

[@PrintYourSpeakers]

I actually didn't know silk PLA had TPU in it. I have to get my self a couple of rolls and have a play!

[@reddragonflyxx657]

​@@PrintYourSpeakersYou could also make the box as multiple components (inner shell, tpu, outer shell) which slot together like nesting dolls. That way you don't have to worry about mixing filaments in one part (for the tpu you can set 0 wall loops to make it pure infill). Also does it make sense to try and separate the inner and outer wall of your speaker boxes entirely? That's what they do in US construction (wood frame houses) for soundproofing: - drywall - sound deadening panels - framing with insulation in the gaps - air gap (minimized vibration conduction paths between the adjacent walls) - framing with insulation in the gaps - sound deadening panels - drywall ...and seal leakage paths, I think this can get you a STC above 60.

[@shanold7681]

Gotta look at what will stick to TPU Im pretty sure it wont stick to pla.

[@HolmgrenJensen]

Have you tried a fully PETG case? I suspect it will perform almost like PLA, but its inherent slightly lower stiffness could lend itself to a better dampening in the infill. I also can't help wonder if the open cell nature of gyroid infill might not also contribute a bit, dissapating some of the energy transferred in a more chaotic flow.

[@NoahHughes116]

Thanks for testing this! I've been printing all my speakers at 40% infill so I'm glad to know that I was right to not crank the infill up to ungodly high percentages. I think your tpu idea is great, if you're not printing the whole cabinet in one piece could you potentially print the cabinet as more of shell and then print a seperate tpu core and slide it in?

[@PrintYourSpeakers]

You definitely could print a tpu insert, another thing to my list of things to try!

[@AdeIino]

@@PrintYourSpeakers As we are dealing with 3D printed objects instead of woodworking, wouldnt it be possible to insert a rubber slab mid print instead? Or even other materials really, like polyurethane foam, maybe even as post print filling

[@PrintYourSpeakers]

@@AdeIino all of this is on my list to test!

[@mandryd]

Great video! Ive been printing my speakers at 6 walls and 45% cubic. Nice to see some data behind the numbers as i was just randomly choosing. My speakers are elliptical -- do you think that would help with the walls moving theory since force would always be symmetrical? Also ive been using paint on fuzzy skin for the interior of my speaker. The cabinets seem less echoey with the fuzzy skin, but i am just a dilettante playing with 3d printers so not sure what good it actually does.

[@pcmaster888]

I'd be very interested in the difference between PLA and PETG for speakers. Since PETG is a bit more flexible than PLA, but not "floppy" like TPU and hence unsuitable for printing a whole speaker out of it, it should be a great material, right?

[@fladder1]

Would the type of infill have influence? Because gyroid for example leaves all the pockets of air interconnected.

[@FuddButter]

I really need to get a printer so I can try things! Something like a liquid rubber or water based bitumen, something that has good vibration dampening?

[@Auggie246]

What if I use car sound dampening material on the inside of the speaker?

[@TheLukemcdaniel]

THANK YOU for acknowledging that doing listening tests over yt is fucking absurd. Not only due to what you mentioned, but also nobody seems to grasp the compression a/v goes through on yt. Maybe you could point out some anomaly or something, but there's absolutely no way to do a proper a vs b test over youtube. The same goes for tv/monitor reviews. A zoom in might show some differences, but the best they can do is a scaled up/down version since you're viewing it on a fucking tv or monitor yoruself.

[@TheJohdu]

the brits have been right all along.... the body mic experiment kinda tends to hide that the contribution of the enclosure is really tiny because the amplitude is minuscule (even if the surface area is big)

[@TheJohdu]

only way that gets into hearable territory is when it produces higher order harmonics or intermodulation

[@PrintYourSpeakers]

@@TheJohdu I think you are on to something here. The distortion from sound transiting through the box does matter. I suspect this is what gives different materials their unique sonic character.

[@dakel20]

Oh heck, I literally just designed and printed speakers, and settled at 25%, lets hope I didn't screw up, lmao. Edit: I did ok! Maybe a bit low, but it's fiiiiiiiiine. Also, for future videos, I highly recommend sticking to the song for each speaker, as speaker B sounded SUPER hollow, but I think it was the recording. I've also noticed that youtube's psychoacoustic compression can really brutalize some tracks, so find something that plays nice with it.

[@t3chn0w1z4rd]

I liked D with headphones, but the bitrate (played, uploaded, or encoded) of the audio really ruined the fullness of those songs.

[@PrintYourSpeakers]

When I was doing my final play though before export I noticed compression artifacts, I'm not sure what happened as the tracks were played and recorded in lossless wav. This is only my third video and I'm still learning a lot, expect the listening tests to improve.

[@odeball22]

try concentric infill

[@anthonyortiz4254]

Why grid over gyroid?

[@s0rc3]

You mention the curves look the "same" but for me (as a scientist but not familiar with speaker measurement) its hard to know what you mean, can you repeat the measurement, re-installing everything in one of the infill density and moving the mic as usual and see how much they vary?

[@PrintYourSpeakers]

The footage of me measuring in the video is actually b-roll, the actual footage was really bad. Yes I really transferred the drivers between all of the speakers again just for 5 seconds of screen time. The actual process for on axis measurements was measure at 50cm on axis, carefully lift the speaker off the stand as to not bump or move it, swap drivers to the next speaker, carefully place the next speaker in the exact same place on the stand, measure again. This way mic positioning and speaker positioning are controlled. There may be a few mm of variation. I couldn't control room tempriture, humidity etc. Which all make small differences. This is why I don't feel that in good faith I can say 1dB is an important difference and not measurement error.

[@nicolas-seifert]

Instead of making a traditional hollow cabinet. Fill the entire speaker with an open cell infill.

[@chinchan9]

concrete infill

[@Reza1984_]

I see people pour cement into their prints

[@Programentalist]

Great video! From a quick not-that-critical listening test I would put B first, then D, then A and C in shared last place.

[@mmark666]

Sorry, can't hear any significant difference even after direct A-B switching with the same songs...

[@jonbourgeois8390]

Speaker B was clear and not as harsh as speaker A. C&D sounded more muffled. Just my first impressions. Curious to see further testing especially on infill patterns. Good stuff!

[@burner8959]

I agree that speaker B sounded the best and A was harsh. I swear I even heard some distortion up high. I wouldn't call C&D muffled, just not as clear as B. Maybe reserved? Not as revealing? IDK, maybe different words for the same thing.

[@jonbourgeois8390]

​@burner8959 Yes, not so much muffled but less clear or you could say warmer perhaps. I didn't mean a lot either. It was subtle, but yeah that's what I heard. B was tighter and balanced.

[@Luke-qs2cg]

The best way to test this is an accelerometer attached to the walls. On axis fr measured closer is irrelevant to cabinet resonances or how it affects the in room sound

[@PrintYourSpeakers]

Hi Luke, Both contact microphones and accelerometers that operate at audible frequencies function in a very similar way because they both use the piezoelectric effect. The main difference is that an accelerometer relies on the vibration of its entire body to stimulate the piezoelectric crystal, whereas a contact microphone just picks up vibrations from the surface touching the speaker. Resonances can appear in on-axis measurements as peaks and dips in the response, especially when enclosure resonances pass through the speaker cone. The on-axis measurement shown in this video is gated to 5 ms, so any room effects are completely filtered out. You would see the same response in an anechoic chamber.

[@tdurham5918]

Great video. Have never printed speakers. Did you have a link for the crossover or is it multiple parts that are assembled?

[@1stWorldProblemsSolved]

D

[@anzoXanzo]

Firrrrst!?

[@celstark]

SHAPE. 3D printers give us this as a free parameter. Why straight walls? Maybe you want the outside as boxes, but the inner surface can be whatever you want. A flat plane is designed to resonate. Don’t be flat.