I assembled a "Portable Bluetooth Speaker" with of the shelf Bluetooth/MP3 module and tiny stereo amplifier (supposedly 10W per channel) plus battery pack. Both devices need 12 volt.
When they are connected to independent power supplies (either different batteries or power adapters) the sound is perfectly clear. But when connected to the same supply, the module seems to inject noise trough the power leads, which is picked by the amp.
I connected a headphone in series with a 10uF capacitor to the power input of the module and sure enough, the noise was heard. By noise I mean clicks and ticks, like an old modem or the sounds some speakers make when a cell phone receives a call nearby. When connecting the headphone directly to the module's output, the music is heard faintly but the noise is not noticeable, that's why I guess this noise is produced in the power input side of the circuit.
What alternatives do I have to reduce or minimize noise while feeding both components from a single battery?
I tried connecting some capacitors in parallel with the module with the hopes that the noise would pass through them and thus "short circuit" while leaving the DC alone but that did nothing. (tried several values 10uF, 4.7uF, 2.2uF, None of them seemed to even reduce the noise level)
For this particular BT/MP3 module, when using USB the noise is very low and almost inaudible at mid to high volume (The amp has it's own volume control), but when using BlueTooth the noise is so loud that it renders the whole thing unbearable.
The construction is as simple as can be:
Side-dish question: Why does this happen and why some combinations do not suffer from this? why these modules seem to inject noise from digital operation into the power rail?
For example, using another module type, I have 3 of them and installed one in my car, one in a boombox and one in a small 4-AA Battery powered active speaker. The three modules are the same "model". (The cheap 10 Dollar ones from China).
In the Boombox I had the same "digital noise" (repeating ticks and clicks that vary when I change modes on the module or skip tracks, etc). It disappeared if the module has powered from a separate source. I was feeding the audio through the volume control potentiometer, as the boombox has no "line in", but the sound was clear if I connected a cell phone or the same module but powered separately.
In the car I connect it through the head unit's "aux in" and the digital noise is audible but so faint that at normal listening volume plus road noise is not noticeable, however, I have to use one specific 12v to 5v adapter (a phone charger) because any other, either branded or generic, introduces "alternator whine" (I call it that because the pitch rises when engine RPM increases).
In the 4AA battery operated speaker, using essentially the same schematic as above, the sound is perfectly clear in any of the modes of the module, it does not show the problem at all!
The goal of my question is to achieve more enjoyable sound in my hobbyist audio projects and learn electronics concepts in the process.
Edit May 27th, 2022. I installed a dc-dc isolator into another, similar contraption that had the same issue, and it worked wonderfully. Noise is absolutely inaudible. I think this is the solution if you have affordable access to the isolator and/or if your goal is to finish the project quickly. At the moment, the component did cost about US$ 2.12 per unit and all it took was to snip a pair of wires and four solder points plus heatshrink, a 20 minute job. I still want to try the discrete component proposed solution to learn and compare.
This is what the isolator looks like:
Edit 2: After too long trying to obtain components I tried the same isolator model on this BT Speaker and it reduced the noise a lot, but it is still perceivable. Since This component worked fine on other similar projects, I'm leaning to believe this MP3 module is not designed to isolate noise. (Among other things, it has common power ground and audio signal ground)