Inspired by Isaac Moore's speaker concept on reddit using the gyroid infill as damping and locking, I have designed my own desktop 3d printed omnidirectional speakers. I would like to hear the Duevel Planets, but as I live in South America it's not that easy to have access to high end listening rooms.
Configuration used for the printing:
0.6mm nozzle
0.8mm wall x 4 (3.2mm total wall)
8% Gyroid infill
10 botton and top layers (2mm)
Fuzzy skin to finish
The printing time was about 25 hours per cabinet, and there's no extra damping material inside, only the gyroid infill which acts as infill as well as interlocking the structure. I can feel a little vibration from the walls when I really crank up the volume, but this does not happens on listening levels.
It's a bass reflex design, using the Tangband W3-2141 driver in an upfiring position, the port is downfirering and designed with the speaker, flanged in both ends and tuned to 70Hz. After a lot of 3D printed diffuser designs, this is the one that worked for me and it's a glass ball used in pet water fountains on a prited support that rests on the original fixed driver's phase plug.
I like what the omnidirectional design can do with the soundstage, the speakers really desapears. There's another one on the right side of the desk, nearfield listening as my ears are like 80cm from the speakers all the time in a 30º angle from the top. I have also designed and printed the magnetic levitating feed, using the repulsion of 2 magnet as isolators. This really helps to keep my desk from vibrating:
I'm currently using it as a 2.1 setup, with a Thonet & Vander subwoofer from 60Hz below that sits under my desk, amplified by a dedicated ebay amp with active variable crossover and lowpass filter. I use some EQ to push up the top end with a shelf filter about 6dB @9kHz.
Completing the setup, I have an ifi UNO DAC and the amp is a chinese integrated tube amp using EL84, with 3+3W single ended with RTC5654 on pre amp stage and 6P15P-EV on the power, same as the Decware Zen Triode amp recomends.
For now, I believe I have achieved what I was looking for, as I can listen to my preferred music all day long while working on the computer.
English is not my first language, sorry. The speakers were printed using 800g of PLA each.
Configuration used for the printing:
0.6mm nozzle
0.8mm wall x 4 (3.2mm total wall)
8% Gyroid infill
10 botton and top layers (2mm)
Fuzzy skin to finish
The printing time was about 25 hours per cabinet, and there's no extra damping material inside, only the gyroid infill which acts as infill as well as interlocking the structure. I can feel a little vibration from the walls when I really crank up the volume, but this does not happens on listening levels.
It's a bass reflex design, using the Tangband W3-2141 driver in an upfiring position, the port is downfirering and designed with the speaker, flanged in both ends and tuned to 70Hz. After a lot of 3D printed diffuser designs, this is the one that worked for me and it's a glass ball used in pet water fountains on a prited support that rests on the original fixed driver's phase plug.
I like what the omnidirectional design can do with the soundstage, the speakers really desapears. There's another one on the right side of the desk, nearfield listening as my ears are like 80cm from the speakers all the time in a 30º angle from the top. I have also designed and printed the magnetic levitating feed, using the repulsion of 2 magnet as isolators. This really helps to keep my desk from vibrating:
I'm currently using it as a 2.1 setup, with a Thonet & Vander subwoofer from 60Hz below that sits under my desk, amplified by a dedicated ebay amp with active variable crossover and lowpass filter. I use some EQ to push up the top end with a shelf filter about 6dB @9kHz.
Completing the setup, I have an ifi UNO DAC and the amp is a chinese integrated tube amp using EL84, with 3+3W single ended with RTC5654 on pre amp stage and 6P15P-EV on the power, same as the Decware Zen Triode amp recomends.
For now, I believe I have achieved what I was looking for, as I can listen to my preferred music all day long while working on the computer.
English is not my first language, sorry. The speakers were printed using 800g of PLA each.
Last edited:
That's a neat idea for the infill! And the magnetic repulsion feet also! And the fuzzy skin!
Look out that things don't fall down onto the speaker, as the cone will guide small items right down into the voice coil gap.
Do you have a high pass for it in the system? Super helpful for small vented designs. I've heard you can do one passively at line-level going into your amp - easy and removable. I haven't tried it though.
Look out that things don't fall down onto the speaker, as the cone will guide small items right down into the voice coil gap.
Do you have a high pass for it in the system? Super helpful for small vented designs. I've heard you can do one passively at line-level going into your amp - easy and removable. I haven't tried it though.
I'm now designing a cloth cover for it. Thanks!!
About the high pass filter, I don't have any right now. Do you have a link for it? Would it be a passive filter? I'm afraid of messing with the phase and adding more stuff in the way and loosing sound quality or dynamics.
About the high pass filter, I don't have any right now. Do you have a link for it? Would it be a passive filter? I'm afraid of messing with the phase and adding more stuff in the way and loosing sound quality or dynamics.
The term is PLLXO for passive line level cross over. But you would only be interested in the high pass. It would go before your amp, and see the input impedance of the amp, rather than after the amp and see the impedance of the speaker. Since it works with high impedance stuff you can make it with low power / value / cost components. Making a high pass 70 Hz (venting frequency) at speaker level / impedance would involve some big components. The down side is that line level stuff may not have the grunt to drive the passive components. You could start with just a capacitor in line with your amp (value depends on high pass frequency and amp input impedance), and try going to second order after that. It could be in-line with your input cable, and removed if you don't like it. Higher tech solutions are probably better, but PLLXO has simplicity as an asset. It's basically an RCA f-mod (you could try just buying one of those).
https://t-linespeakers.org/tech/filters/passiveHLxo.html
https://www.sound-au.com/articles/pllxo.htm
I walk you entirely through it as I haven't used one. I have used DSP to high pass small vented speakers, and the savings in bass excursion (which isn't making useful sonic output below vent tuning) can be substantial.
Since you've got EQ, you could try just smashing all EQ bins below your tuning frequency to max attenuation, and then use the sub plate amp gain knob to bring LF level back up. This may also want some fiddling with the sub low pass frequency knob, since it'll be pushing around the FR going to that, changing the shape of the sub's (effective) low pass. But the idea is to keep low frequency content out of the satellite speakers.
https://t-linespeakers.org/tech/filters/passiveHLxo.html
https://www.sound-au.com/articles/pllxo.htm
I walk you entirely through it as I haven't used one. I have used DSP to high pass small vented speakers, and the savings in bass excursion (which isn't making useful sonic output below vent tuning) can be substantial.
Since you've got EQ, you could try just smashing all EQ bins below your tuning frequency to max attenuation, and then use the sub plate amp gain knob to bring LF level back up. This may also want some fiddling with the sub low pass frequency knob, since it'll be pushing around the FR going to that, changing the shape of the sub's (effective) low pass. But the idea is to keep low frequency content out of the satellite speakers.
I will start this approach right now, I believe I have understood what you've described. I'm playing right now with -6dB@50Hz with Q=0.9 and the response at 75Hz remains @0db. I have plenty of headroom on the sub amp (100W), so bringing back the LF will not be an issue. Thanks for the improvements tips, I'm always open to constructive feedback on my projects. 🙂
It's an honor have this comment from you, sir. I had a FH3 with Alpair 7.3, best system I had, but unfortunately I had to sell it when moving from Brasil to Uruguay.
FHXL on my plans for a W5-2143... I'm just waiting the right time to convince my wife.
That's the coolest use of 3D printing I've ever seen. The magnets are a clever idea however, I wonder about the use of a magnet spring in the vertical (up-down) axis where the sound is also being generated in the same axis. I think a rubber or oil damper (not a spring) would work better. Since you are really clever maybe you could design a copper damper?
Thanks Arthur! What I really like about this forum is that you guys always take an idea and take it to the next level! I was also worried about loosing some dynamics and SPL by decoupling the speaker from the desk using this "spring" system, so I have conducted some subjective + objective tests, searching for the best cost/benefit solution for my case.
As my desk is a common office desk with metal feet with a less than desired stiffness, when listening to the speakers directly coupled to the desk using 3 spikes, I got really strong vibrations at 150Hz and it's harmonics (75, 300, etc) using a swipe tone. This was really annoying and every light weight object over the desk as pens , keyboard and other things we keep around would rattle and amplify this. Using the magnetic springs has decreased the vibration on my desk as the graph below, reaching 2 to 3 (blue marks) on this scale to around 0.2 to 0.4 (green marks) when the magnets were active as I shifted from one state to the other just pushing down the speakes with my hands, so the upper feet gets coupled to the down feet:
The audible difference for me was negligible as I didn't perceived any lacks in dynamics or upper bass impact (the main difference was the absence of rattling things and the desk vibration itself when the magnets were working!).
The magnetic decouplers are around 2cm diameter and 2.2 to 2.5 cm high, using magnet neodymium cylinders of 10mm x ø5mm , and I'm thinking about the use of magnetic dampers with copper plates would require a much larger plate. Using a coil would keep the system with the same height, now trying figuring it out how to fit a coil on the upper section that may provide this damping effect and increase the stability of the system improving the general dynamics of it.
I'm also importing an RTA mic to get some calibrated measurements of the system so I can play around with the EQ a little more and have some nice graphs to translate what I hear in images so you can have a better idea of how things are going.