3D printed Bluetooth speakers - LittleJet!

My goal is to design a 3D printed standalone speaker based on transmission line theory - info: Martin J. King -

Similar projects worth following
Hello to everyone! Today I am happy because I can finally introduce you to this project that I have been working on for many months now. These are 3D printable speakers connectable with a cable to a PC, or using the Bluetooth of your notebook or smartphone. They have a lot of acoustic power, pushing out without distortion 10 watts RMS (in general the computer speakers are 3 watts!), and thanks to the hi-quality full-range driver, they rock! They are also perfect to keep in the room to listen to music as you do with that little portable Bluetooth speaker, but I swear, these speakers sound much better.
Italian version here:

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.


“A transmission line loudspeaker is a loudspeaker enclosure design which uses the topology of an acoustic transmission line within the cabinet, compared to the simpler enclosures used by sealed (closed) or ported (bass reflex) designs. Instead of reverberating in a fairly simple damped enclosure, sound from the back of the bass speaker is directed into a long (generally folded) damped pathway within the speaker enclosure, which allows far greater control and use of speaker energy and the resulting sound.”
from Wikipedia.

Design process

The cabinet simulation was obtained with HornResp by entering the T / S parameters found on the woofer manufacturer datasheet.
Once I got the transmission line, I modeled it using Grasshopper for Rhino.
Since my transmission line had to be 3D printed I could introduce some innovation and work with much more complex surfaces; so the algorithm takes the parameters from HornResp and makes rectangular cross-sections in golden ratio, with an always correct parabolic trend of the surfaces of the duct during the folding. The folded TL is then used in Fusion 360 for the design modeling.

Baffle step crossover (not added)

For the baffle step crossover design I used XSim, and I created FRD and ZMA files using FPGraphTracer. You can download them from the attached files list.

Live recording

At the beginning I clap my hands to show the room acoustics, the sound level, and the background noise of the microphones. The recording is done with my smartphone, which has two microphones and records in stereo with a rather decent quality, but still nothing "ideal" for these things. Listen to it on headphones, but remember that what you hear is a sound captured by microphones and reproduced by the quality of your headphones, at best you can judge the richness of the sound.
The speakers were 90cm apart, and oriented towards the listening center at 160cm where the smartphone was positioned on a tripod.
Happy listening.

** The song is Bowsprit, from Costellations album, by Balmorhea. All rights go their respective owners **



1 - Probably I will release soon the "desktop speakers" version (without Bluetooth), based on the TPA3116 2 - The TPA3116 will need a baffle step crossover
3 - The YDA-138 board also would need a baffle step crossover, BUT, since the RLC filter built on this board is for an 8ohm load, we have already a "baffle step" on a 4ohm (from 2kHz). It's not a designed one, but enough to sound good here.
4 - The YDA-138 board doesn't have a power switch (d'ho!) so I had to add one.

BLUETOOTH 5.0 (yeah!)

1 - The only way was to add another external module, because the amp boards with built Bluetooth IC out there are just scam.
2 - I wanted a great SoC, then choose the CSR8675, but I have some regrets... Maybe a 4.2 would be enough here: older = cheaper.
3 - The analog output is BALANCED... This means that you have to unbalance it.
4 - The MAX97220 chip is a differential little amplifier, then I added this module, with gain 0, to unbalance the signal from the CSR8675.
5 - Since the power supply is shared, the chain "CSR+MAX" is really noisy.
6 - Then an isolated DC-DC here was mandatory, so I added the B1205S-1W (0.2A)


1 - The amplifier board doesn't have an input jack (d'ho!)... BUT
2 - The integrated one would be unuseful because we have to close the MAX97220 output when a line source would be connected.
3 - I then added an input female jack with a switch.

LittleJet-Bt_mk4 v42.f3z

f3z - 5.06 MB - 10/22/2022 at 14:45


SX_LittleJet_mk4 v33.f3z

f3z - 1.99 MB - 10/22/2022 at 14:45


peerless v5.f3d

fusion - 502.01 kB - 10/22/2022 at 14:45


peerless TC6FD00-04.7z

Here you can find the FRD and the ZMA for the woofer, and also the .txt for Hornresp.

7-Zip - 9.41 kB - 05/01/2020 at 16:16



The most important piece: the knob!

Standard Tesselated Geometry - 61.61 kB - 05/01/2020 at 08:58


View all 8 files

  • 4 × *.stl files Print all the stl files attached
  • 2 × Peerless by Tymphany TC6FD00-04 2" Full Range Woofer Is the driver I based my design on, it MUST be this
  • 1 × Single RCA cable To connect the two speakers
  • 1 × 12v - 4A Power supply The general power supply
  • 1 × Stuffing material (polyester material) For reference: Dacron Hollofil 300gr

View all 12 components

  • more about WHUZY ZK-1002 | ZIZHI ZZ-1002 amplifier boards

    Davide Ercolano11/12/2022 at 15:39 5 comments

    Hello to everyone!
    On my research journey, I already talked about the WHUZY Audio ZK-1002 amplifier board but it is worth it to dive more deeply into that board and I bought it again lately.

    The new news is that probably you can't buy the "WHUZY" anymore: searching and buying "ZK-1002" I received the ZZ-1002 by ZIZHI Audio instead. Well, it's really the same board tho.


    Let me remind you about the problem with the AUX input stage of those boards: there is an RC filter with the wrong "C" value (10nF instead of 1nF) so the low-pass filter starts at 3.5kHz instead of 35kHz. I'm talking about C41 and C42, with R2 and R4 (4.65kOhm). To get back your hi frequency you must swap in the correct 1nF capacitors or you can remove them also; the remaining resistors will lower a small bit the volume (negligible), you could remove R2 and R4 too, and then bridge the pads.
    There is also a cut-out with a curve starting from 150hz down to 20hz due to the low value of the decoupling capacitors C37 and C39 (100nF): to fix this you must replace them with a 1uF instead (value from 0.47uF are good).

    Another problem is that the Bluetooth chip is always on (and the TPA3116 are on standby), but this turns out only annoying and not acoustically problematic.

    3. THE HIDDEN PROBLEM (yes, another one!)

    You have this board and when you rotate the volume POT you hear a scratching noise, but you are like "ah ok this POT is a cheap one but it's ok". Right?? you are WRONG, it's not "ok" and the scratching noise doesn't come from a cheap POT... the truth is much more insidious!
    The second stage before the POT is the Bluetooth one, with the IC by JieLi. This IC elevates the input audio signal (from AUX) and the output signal to the amplifier with 1.8v of DC current. Yes, DC current into the POT, that's the scratching sound you're hearing! Also, the fact you're hearing that, means that a DC current is going into your speakers! (tested it with a voltmeter and oscilloscope)
    Basically, the circuit design is wrong.

    4. HACK A FIX
    Well, electronically speaking it's just a matter of adding 2 more decoupling capacitors on the output of the Bluetooth chip, but to do so you have to cut 2 PCB traces. Here is my fix:

    As you can see I worked on the bottom of the PCB: is the only place before the POT with enough space.
    You must cut the traces, remove a bit of protective paint, and then solder on an SMD capacitor, I did it but I won't lie to you: it's a very hardcore fix, not for newbies.
    Did it work? Oh yes, it works now, smooth and silent! But unfortunately is not for everyone.

    And here we are again, I paid 14.70€ shipped for this board, but at the end of the day I can suggest to ANYONE to buy it, and it's really a shame; I'm so frustrated by this that I'm thinking to design one by myself... Should I? ;)

    Thanks for reading me and stay safe.
    Davide, baugrafik

  • And LittleJet! Desktop Version (finally)

    Davide Ercolano01/02/2022 at 11:35 0 comments

    Hello! Sorry for this long delay, but hey, better "later" than "ever".
    I finally shared the LittleJet! "desktop version", and you can download it from here:

    The BOM is:

    • x2 - Peerless by Tymphany TC6FD00-04 2" Full Range Woofer
    • x2 - RCA panel jack
    • x2 - Inductance 1.05mH or close value
    • x2 - Resistor 3.7Ohm or 4Ohm (from 3W are good enough)
    • DC panel plug
    • This amplifier module here (non-affiliated link)
    • 19v - 4A Power supply
    • Stuffing material (polyester material)
    • 3D print all the files once


  • Finally - Closed Project!

    Davide Ercolano05/01/2020 at 10:39 0 comments

    After a lot of work, learned a ton, designing and tinkering, I'm finally happy to present the LittleJet speakers!

    (Italian language, but you can skip to 5:39)

    (how everything fit inside)

    (the final back design)

    *** Sine sweep full spectrum, recorded with Spectroid APP on a Honor8 smartphone and built in microphone at 50cm far  (see the red line; not smoothed) ***

  • Final design is final!

    Davide Ercolano04/07/2020 at 15:08 0 comments

    Hello to everyone!
    Since my long waiting time for the needed electronics, I definitely closed the acoustic and aesthetics design! Yay! New frequency resonance for the TL (118hz instead of 165hz) and then a different TL shape, plus a front baffle added to improve the diffraction.

    Meanwile the electronics did arrive and I was able to listen thoroughly to the amplifiers... And then I finally chose and this changed a bit the final "final" design :D (spoiler: no more light hole and the jack hole in front baffle).
    Anyway everything is really going to be concluded, see you really soon!

  • The amplifier

    Davide Ercolano02/04/2020 at 16:58 11 comments

    Since this speakers are meant to be "stand-alone", I'm searching for a little amplifier with wireless connection and this features:

    • 15W rms power with lowest THD% with 12v
    • low heat working with 12v
    • bluetooth 5.0 (more on later)
    • volume POT
    • On/Off button
    • low budget

    I used to follow some YouTubers working in this field then I knew which IC amplifier to search, two of them actually: the Texas Instrument TPA3116D2 and the Yamaha YDA138-E.
    For the "budget" point, I'm searching on Aliexpress (a good platform indeed).
    Anyway, almost 2 months are elapsed now and I have some informations to share with you.

    FIRST BOARD: WUZHI AUDIO ZK-1002 - $13.09 shipped to Italy

    Really good PCB layout, pretty decent capacitors and inductances on the RLC filter and power line, volume POT with ON/OFF, two TPA3116D2 in single-channel mode... "Perfect!" do you would say... NO.
    The Bluetooth IC is a wild one, named AS19AP22702-25A4 from a company with a "greek Pi" as a logo (edit: is "JL" logo, by JieLi company); the input audio is managed by this chip, also the input from the AUX jack goes inside to this IC, then, is that bad?! Quite YES

    Using the AUX input  the amplifier is NOT linear and the signal get smaller from around 2.7kHz becoming really silly at 22kHz.
    Usually this behavior is caused by an incorrect RLC filter, but is this the case?

    Connected on Bluetooth you get a linear output till 19kHz... What happen later?...

    This! What is this? I really don't know, a noise I would say, but anyway it is crazy output therefrom 19kHz to 22kHz and up.

    So the Bluetooth IC it's not reliable and the AUX input is junk.

    Anyway, the TPA3116D2 in this single-channel configuration (two, one per channel) it's really good and actually listen to it by Bluetooth is nice, with hi-power and no distortion.

    SECOND BOARD: HAZY AUDIO XH-A354 - $14.86 shipped to Italy

    When I have chosen this board I was attracted by the YDA138 single-channel configuration, so to have more power on a 4ohm driver, like the one into LittleJet.

    The first impression on this board when it's arrived was so terrible: the codes on the Bluetooth IC have been erased, also on the controller for the SD and two more wild IC that I think are some OP-AMP (do someone of you know??). Also the volume POT doesn't have the on/off switch and the worst part: the inductances are ridiculously small. Finally the AUX input goes inside the Bluetooth chip, like the Wuzhi, then... Let's see...

    0.5 per division, almost 2v p/p and it is already REALLY distorted (1khz).
    And by bluetooth is the same...
    At full power the inductances get really hot and also the linearity (from 4kHz to 22) is a shame

    Measuring along the source path I found that the signal is good later on the decoupling filter (RC) and also later on the volume POT, then till the BL input, but it was absolutely distorted on the output of the Bluetooth IC.
    Listen to this amplifier is a pain, this board is simply garbage. DON'T BUY IT.

    Anyway I want to listen to a good YDA138 board, so I'm waiting for a new one (without Bluetooth connection).


    The actual integration of the standard by these boards (everyone you can find now) is wild and not reliable then I will adopt a different approach: amplifier board (TPA3116) + Qualcomm CSR8675 module. The budget will grow up a little but we will have more audio quality and reliability.
    Anyway now everything is going very slow in China because of Corona Virus, so I don't know when I will have this configuration. Meanwhile I will design the integration for the Wuzhi ZK-1002.
    See you soon!

  • First step: share it

    Davide Ercolano12/13/2019 at 18:34 0 comments

    Hello there, this is my first project here and I'm kinda excited for.
    I'm working on this by some months now and finally I have something to share. It is not a completed project yet anyway and I have some plans for it, infact my next step will be to implement (I already found it) an amplifier and a bluetooth connection and finally make the LitteJet a standalone speaker.
    I also want to verify my design for validation of the design process testing the speaker into an anecoic chamber.
    See you soon then!

    Davide Ercolano -

View all 6 project logs

  • 1
    3D print

    Download and print all the .stl attached files once.

    Layer height: the one you want, mine is 0.24
    perimeters: 4 (with a 0.4 nuzzle)
    bottom layers: 5
    top layers: 5
    infill: 15%
    speed: around 55mm/s

    It would need ~20h for print one cabinet.

  • 2
    Mount the driver

    1 - Break the last bottom layer (MK4_DX_LittleJet.stl) to free the canal (look inside the cabinet to see where he is)
    2 - Push the 2 driver wires from the bottom, through the canal (you would need around 250mm of wires), into the cabinet.
    3 - Pull out the 2 wires from the cabinet through the driver hole.
    4 - Connect the 2 wires on the driver (there is a grey dot on the driver connector meaning the "hot pole", then connect there the +)
    5 - Fill the cabinet with the polyester fiber
    6 - Orient the driver with the connector looking down and force the wires to be close each other then:
    insert the magnet in the driver hole starting from the top side, tilting it
    then push the down side in, using the notch in the driver hole to let the wires pass easily
    7 - Screw with four 3x10 or 3x12 screws

  • 3
    Mount the electronics boards on the base

    You need:
    4x - 3x5mm screw
    2x - 3x5mm till 3x10mm screw

    Screw in place the amplifier board with four 3x5mm screws; then screw in place the CSR8675 module with just one screw because you need to rotate it later (to connect the output wires)
    Slide in the MAX97220 module, in the little slot on the left.
    Push in the squared slot the B1205S converter.

View all 9 instructions

Enjoy this project?



Romain wrote 05/22/2020 at 14:30 point

This is a super nice project!
Would you mind sharing the source Fusion360 file? I would like to try and modify the design of your speakers while keeping the transmission line!

  Are you sure? yes | no

Davide Ercolano wrote 05/25/2020 at 23:10 point

Thanks Romain!
For now I'm not sharing the sorce Fusion360, but I will in future :)

  Are you sure? yes | no

Romain wrote 05/28/2020 at 15:17 point

Let me know when you do! I'd love to play with it!

Thanks a lot anyway for the rest of the information! I'll be sure to keep an eye on your project for the update!

Congrats again on your excellent work!

  Are you sure? yes | no

Davide Ercolano wrote 10/19/2022 at 16:06 point

Hi Romain, it's me again! The desktop version is out so, as I promise, I could share the F360 source file, are you still interested?

  Are you sure? yes | no

Romain wrote 10/21/2022 at 11:49 point

Yes definitely! Thanks for the reminder, I'll get my hands on the speakers to get started!

  Are you sure? yes | no

Davide Ercolano wrote 10/22/2022 at 15:06 point

Done, you'll find them in the files list

  Are you sure? yes | no

RobG wrote 05/15/2020 at 13:58 point

Awesome work! Well done. I've used the TPA3116 boards a few times, like in this one: drop me a line if you need any help, though it looks like you've got it covered!

  Are you sure? yes | no

Davide Ercolano wrote 05/15/2020 at 14:53 point

Thanks Rob! In the end I choose the Yamaha YDA-138 for some reasons, as you can find in the "details". Anyway I do have the TPA3116D version here too, then I will share it soon (will be the "desktop version" without the bluetooth connection, but with a baffle step crossover)

  Are you sure? yes | no

Davide Ercolano wrote 05/15/2020 at 09:03 point

Hi Phil! Thanks for this trick! I would like to try, could you please link me some magnets that could do the job?

  Are you sure? yes | no

Lightning Phil wrote 05/21/2020 at 09:25 point

Pretty much any magnet will do something.  Measuring the ones in the drivers and get something similar sized is usually a reasonable first step.  Overly strong neodymium pucks are good too (better perhaps as they're stronger), but more expensive.

At a guess, these may work:

Have used various different types over the years.  Lowering the Q of guitar speakers started it all off.  Some of the ones with weedy magnets make for quite a pronounced jump in volume for whichever note is closest to its resonance.

  Are you sure? yes | no

Lightning Phil wrote 05/14/2020 at 22:52 point

Looks like a fun project.  I've built some bass reflex speakers using the full build volume of an Ultimaker for 2" drivers and a tiny guitar cabinet using a 5" driver.  Yours look cooler :O)

Could I suggest adding a backing magnet to the drivers to increase the field through the coil a little.  It increases the BL product, lowering Qes and thus Qts.  By flattening the peak, it could potentially make the output a little more even and extend the bass further.  As a bonus, it makes driers a little more efficient also (on guitar speakers usually by 1dB per W, perhaps a little more in inexpensive speakers with a low starting BL).

It's not a magic solution, but is easy to undo by removing the magnet.  Further, by attaching the magnet such that it's initially repelled by the driver's magnet before it gives up and sticks to the steel, BL is increased.  If the driver attracts the magnet, then it will decrease the BL and have the opposite effect.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates