DIY Bluetooth Speaker Upgrade
by diyboomboxlab in Circuits > Speakers
2311 Views, 15 Favorites, 0 Comments
DIY Bluetooth Speaker Upgrade
This Bluetooth speaker has been with me for a full year and survived every trip without a single issue. Loud, bright, and reliable — everything you’d expect from a simple DIY build. But one thing kept bothering me: the bass. It was… nothing special, and I knew it could go much deeper.
https://www.youtube.com/watch?v=bCdhqXvEh1s
Inside, it’s a straightforward design: a 2-way system with a passive radiator, two 3" woofers, a TPA3110 amplifier, and an SP-107 RGB music processor. A big battery pack keeps it running all night long.
A simple project — but with a lot of untapped potential.
So I decided to take the scientific route: REW measurements, WinISD simulations, and a precise adjustment of the passive radiator’s mass.
After the upgrade, the speaker finally “opened up” — the bass became deeper, tighter, and surprisingly mature for such a small box.
Here’s a brief explanation of how I built the enclosure
(the full build video is available on my YouTube channel ):
I cut all the enclosure panels from 3 mm hardboard — three layers for the walls and four for the front baffle. Hardboard cuts cleanly with a utility knife, so no table saws were harmed in this process (and I don’t like them anyway).
After cutting the panels, I drilled all required holes according to the blueprint — carefully. The enclosure was assembled with PVA glue, then sealed and primed with the same glue. After a 6-hour cure, I sanded the surface smooth.
Once the vinyl wrap was applied, the enclosure was ready for installing the drivers and electronics.
Specifications:
- Supply Voltage: 4S(14,8V), 9 Ah;
- Max output Power (Amp TPA3110):
- THD+N = 1%: 2 x 15W;
- THD+N = 10%: 2 x 20W;
- Frequency response: 50Hz-20kHz;
- Wireless: Bluetooth 5.0;
- Woofers 3 Inch, passive radiators 4 Inch, tweeters 2 Inch;
- Enclosure's Dimensions: H=160mm; L=430mm; D=163mm;
- Weight - 4.6 kg.
Supplies
Parts List:
LED RGB visualizer
WS2812B WS2811 LED Strip Light Bluetooth/Music Controller SP107E
WS2812B Pixel Smart RGB LED Strip Light WS2812 Individually Addressable 144Leds DC5V-
LM2596 DC-DC adjustable step-down power Supply MINI560 5V
Translucent Black Plastic Sheet 3mm
Audio
AIYIMA 4Pcs 2 Inch Tweeter Speaker 6 Ohm 30W
KYYSLB 25W 3 Inch Speaker Drive 8 OHM
Speakers 2 Way Crossover
ZK-502L MINI Bluetooth 5.0 DC 5-24V Power amplifier Stereo board 50Wx2
Speaker Passive Radiator 4 Inch
Power supply
Hot Sell 2X 18650 Battery Holder -
Bms 4s 18650 Li-ion Battery Protection Circuit Board
XL4015 5A DC to DC Lithium Battery Step down Charging Board Module
TZT 4S 4.2v li-ion balancer board
5.5*2.5mm Socket DC Power Supply
4S Single 3.7V 18650 Lithium Battery Capacity Indicator
6mm Metal LED Indicator 12V Red Blue White -
Volume Button Knob Handle 30x17mm
128mm Simple Vintage Cabinet Door Pull Handle Black
Handle Suitcase
Carbon Steel With Black Hexagon Socket Cap Head self tapping M3, 16mm
Pan Head Hexagon Socket Self Tapping Screws Round Button M4, 25mm
Tools
Drill Stand BG6117, Drill Carrier Bracket 90 Degree Rotating Fixed Frame Workbench Clamp -
XCAN Forstner Drill Bit 15-35mm Carbide Drill Bit Set -
Hole Saws Kit 19mm-127mm
Digital T12 Mini Soldering Iron DC 12-24V 72W
Welding Wire 1.5MM 63/37 FLUX 2.0% 45FT Lead Wire Melt Rosin Core Roll
XCAN Drill Bit 13pcs HSS
What’s Wrong With This Wireless Speaker?
To avoid guessing, I measured the system in REW — and got a brutally honest result: below 110 Hz, there’s basically nothing.
https://youtu.be/bCdhqXvEh1s?t=17
The speaker needs a different passive radiator tuning, no excuses.
A quick note on passive radiators:
They look like regular speaker cones, but inside they’re completely empty —
• no magnet,
• no voice coil,
• no wires.
When the woofer pushes air inside the enclosure, the passive radiator moves in sync. At its resonant frequency, it’s supposed to boost the bass.
https://youtu.be/bCdhqXvEh1s?t=36
Sounds like a neat trick, right? But the real question is: does it actually work?
The Experiment — Finding the Resonance
To check whether this is magic or physics, I set up a small lab-style experiment.
I mounted the main driver in a test box, so it could pump air into the passive radiator. Then I glued a tiny mirror to the radiator and pointed a laser at it.
When the radiator vibrates, the laser dot jumps — showing the amplitude clearly.
Sweeping through the frequencies, I found the sweet spot: at 79 Hz the movement peaks. That’s the radiator’s natural resonance.
Then I added 10 grams of mass — small magnets attached to the cone.
Ran the test again.
Result? The resonance dropped to 61 Hz.
Not magic.
Not guesswork.
It works — and it means the bass can go much deeper.
https://youtu.be/bCdhqXvEh1s?t=55
Simulation in WinISD
Here’s the trade-off: the heavier the passive radiator becomes, the lower it goes in frequency — but the quieter it gets.
To find the sweet spot, I built a proper model in WinISD:
- Enter the woofer parameters.
- Add the passive radiator.
- Adjust the added mass to shift the resonance downward.
I simulated 10 g, 18 g, and 24 g.
18 grams turned out to be the best compromise — deeper bass without killing efficiency.
The frequency response curve for 18 g showed exactly the result I needed.
Tuning the Passive Radiator
To reach that target mass, I added weight directly to the passive radiator cone.
And yes — wine corks combined with screws, washers, and nuts worked perfectly.
The manufacturer didn’t include any built-in system for adding mass, but the cork fit so snugly that it became the ideal base for adjustable loading.
Everything is held together with a nut and a drop of instant glue.
It’s simple, solid, and easy to fine-tune.
Important:
The final mass must match your WinISD simulation exactly.
Miss the weight, and the whole tuning falls apart — no magic, just physics.
Final Measurements of the Upgraded Version
Everything is ready — time to measure again.
Low-frequency measurements in a room are a mess: wall reflections ruin the data and make the graph lie.
But there’s a solution.
I measured in the near field, placing the microphone right next to the woofer and the passive radiator.
This way, the room acoustics stay out of the picture, and REW can later combine both curves into a clean, accurate frequency response.
And here’s the new result: the response now reaches down to about 50 Hz.
For a speaker of this size — that’s impressive.
And here’s how it looked before.
Yeah… the difference speaks for itself.
Conclusion
This upgrade was absolutely worth it — and I’m genuinely happy with the result. Nothing complicated: a few measurements, some simulations, and one precisely tuned mass on the passive radiator.
And the effect is great. The speaker finally sounds the way it should.
If you want to repeat this upgrade or check out my next projects, follow my profile and visit my YouTube channel. There’s a lot more coming.