Feature

3D Printed Bluetooth Speaker

Johann Wyss

Issue 11, May 2018

This article includes additional downloadable resources.
Please log in to access.

Log in

Making your own audio gear is almost a right of passage. Sometimes it's just because you can!

Sure, there’s plenty of Bluetooth speakers on the market, but as the DIYODE community knows, there’s an unrivalled satisfaction in building things yourself! Taking on the challenge to build a great looking 3D-printed speaker, using cost-effective parts, Johann has come up with a pretty awesome design. And we’ve got all the details...

The Beast v2.0 Prototype.

What made you start this project - were you after a low-cost Bluetooth speaker, or was it more about building something from scratch?

The inspiration actually came from my love of audiobooks. I listen to Audible while travelling to and from jobs in the car, and I’d often find myself sitting in the driveway, waiting for a particularly exciting chapter or series of events to unfold in the story. This prompted me to look into ways I could continue to hear the story outside of my car. Since the external speaker on my mobile phone was nothing special, I began to look at standalone external speakers. Whilst Bluetooth speakers are not expensive, I opted to build my own, just to get that personal touch and so I can prove all the money I spent on 3D printers wasn’t a waste!

After a visit to my local Jaycar store, I learned they didn’t sell any Bluetooth audio modules so the first iteration (electronics-wise) used the 2 x 2W class D amplifier module from Jaycar. This module uses the PAM8403 IC and due to no Bluetooth, the speaker required the phone to be directly connected. Whilst that worked fine, I decided I would prefer a wireless connection.

The final product came about as I searched for a Bluetooth audio receiver to use with the Jaycar amplifier module. I discovered on eBay an all-in-one solution, which contained a Bluetooth receiver as well as the PAM8403 amplifier.

Oh yes - we all need to justify our hardware! So this is actually V2.0. We agree Bluetooth makes connectivity better than a wired solution. Can you take us through each particular piece of hardware in this project, and why you selected them?

18650S: I choose to use the 18650 cells as they have, by far, the greatest energy capacity for their size; and because I have access to a vast number of recycled cells from faulty cordless power tools (thanking you very kindly, Joey)! The cells are generally 2.3A/h cells, which I have cycled several times recording the actual capacity. This ensures their suitability to be used together in parallel. Together they provide 3.2-4.2V potential and 4.6A/h or current.

BLUETOOTH AND AMPLIFIER BOARD: As stated earlier, this impressive board was found on eBay for only a few bucks including delivery! It contains a Bluetooth receiver as well as the same PAM8403 amplifier IC that was in the Jaycar amplifier, making it perfect for the project. It requires a stable 5V input and so was unable to run directly from the 3.2-4.2V provided by the lithium cells; therefore, I needed to boost the potential to 5V/DC.

Bluetooth receiver and amplifier.
Bluetooth receiver and amplifier.
Boost converter module to provide stable 5V.
Boost converter module to provide stable 5V.

BOOST CONVERTER: The boost converter takes the fluctuating potential (3.2-4.2V) from the lithium cells and boosts it to the nominal and stable 5V required by the amplifier and Bluetooth receiver. The 3A output capability of the module far exceeds the requirements of the circuit, so my original plan was to build my own boost circuit; however, the price these modules go for makes it significantly cheaper to use the pre-built module over sourcing the components and building it yourself.

LITHIUM CHARGING: Lithium cells are not the easiest cells to work with, as they require a precise charging circuit to charge them via a constant current around 1A to 4.2V. Charging a lithium cell outside of this range has the undesired potential to severely damage the cell, reducing the capacity and in extreme cases causing the cell to expand and/or even spontaneously combust. As this is something I intend to use inside my house, I decided to err on the side of caution and not risk the worst case scenario. At around 60c a piece on eBay it would be silly to attempt to build my own. Also the 5V micro USB input means you can easily use one of the countless unused old mobile phone chargers we all have lying around (c’mon, you know you do!).

SPEAKERS: Arguably the most important part of the design is also the simplest! I found a set of 3W 4Ω 52mm speakers on eBay, again for just a few bucks! Whilst they are not great and came very poorly packaged they are more than suitable for this task. I did however go through a bunch of different speakers trying to find a quality pair; interestingly, there does seem to be a lack of retail stores selling them.

A fairly sensible collection of hardware! Everything you need, nothing you don’t! The 3D-printed case looks great. Was this the first design that stuck, or did you go through various iterations?

The oval design was largely inspired by my Logitech desktop speakers, and quite oddly (for me at least) the speaker enclosure didn’t change much through the design process. The front panel had a few significant changes with the original design having an LED and rotary switch, compared to the final version having a toggle switch and LED. Other than this, it has remained largely unchanged. With that said, a few changes are in the works; currently the enclosure is held together by hot glue (for sealing), so I intend to change that to become screwed together, and with a 3D-printed rubber gasket for sealing.

We did wonder about plastic rattling. In reality, hot glue is probably just as good as a rubber gasket unless you plan on taking it apart and putting it back together a lot. Twelve hours of playtime is impressive! You’re using a boost converter to step up voltage, so are you running your 18650 cells in parallel?

All components wired before mounting.
All components wired before mounting.

Yeah, the cells are in parallel now, but originally they were in series. Last year we had a group assignment which I built a power supply for. I used two 18650 cells in series, and stepped the voltage down to 5V using an LM2576 switch mode regulator. Intending to use the same circuit for this project, I compared the efficiency of this module stepping the potential down with the off-the-shelf boost converter and stepping the voltage up, and I discovered that having the cells in parallel and boosting the potential was moderately more efficient than bucking the potential with the cells in series. Given the device is battery operated, I opted to go the most efficient route, and so the cells were placed in parallel and a boost circuit was used.

Interesting result on the efficiency front. Parallel does make charging simpler too. If you add a high pass filter, you’ll reduce low-end distortion, but surely miss out on some amazing bass. Perhaps you could add a wireless subwoofer to the mix?

Given the device was, for the most part, intended to be used for audiobooks I’m not so sure bass is all that necessary. However it wouldn’t be at all difficult to have a switch for a bass boost mode or similar! As if I needed more motivation for feature creep! Although, now you mention it, a subwoofer attachment could be really fun to build.

Very tidy mounting in the custom case.
Very tidy mounting in the custom case.

Absolutely! What’s not fun about a subwoofer? With the knowledge gained during this build, what would you do differently, if you did it all over again?

I would create the entire circuit on one PCB, with mounting studs designed into the enclosure. This would make it significantly more rigid, easier to construct, and a significantly more professional item overall. However it would also make it significantly more expensive.

As it stands currently, the entire device costs less than around $25, with the online sourcing of components (including the lithium cells) being the most significant portion of the cost. Creating the PCB and sourcing components would triple that cost, making it fiscally non-viable, albeit still worthwhile as a learning (and fun) experience.

Other than that, I would like to have used the F-6188 Bluetooth audio receiver board. It can comfortably work with the voltages directly from the lithium cells, thus reducing the load on the boost circuit, and therefore increasing efficiency. It’s also lower in cost and seemingly sounds better when matched with the PAM8403 amplifier.

That makes for a rather cheap Bluetooth Speaker by anyone’s standards! What awesome project are you working on at the moment?

I have two projects in the works: the first is my 0-15V and 0-3A linear power supply, which I have been working on in my down time for nearly half a year, making it a somewhat slow process. Apparently getting a linear regulator to drop potential without getting hot enough to melt solder isn’t a trivial task!

The second is a project with my 12-year-old son. It’s an app-controlled RGB LED lamp, which will allow him to switch it on/off, and change the colour of the output using a mobile phone. He is keen to build stuff and hopefully a simple project like this will get him even more interested in programming.

A render of the main case.
A render of the main case.

Those both sound like great projects! We look forward to seeing the results! Thanks for taking the time to share your Bluetooth speaker, and create a DIYODE branded 3D-printed shell - it looks fantastic!

Johann Wyss

Johann Wyss

Electrical fitter and Electrical / Electronics Engineering Student from QLD, Australia.