I have an MP3 module connected to a Mega2560p.
It plays MP3 files from a USB stick, and upon requesting these files, you get a horrible 'click' from the amplifier.
The supply is all smoothed etc, but I did find that clipping a ferrite coil to the USB lead cut this noise significantly.
My question is, does a ferrite coil need to be a complete single piece to work?
The reason I ask, is the ferrite coils I have here are two halves that clip over a cable, but my brain tells me (rightly or wrongly) that they ideally would be one continuous piece of ferrite.
Does it make a difference? Obviously its a lot easier to clip these coils (or are they chokes?) over an existing cable, rather than unsoldering your cables to put them through a solid ring.
Well I lied..... it's back.
Just cannot get rid of this horrible raspy click when it accesses data from the USB stick.
The MP3 module (JQ8400) has onboard flash, which I have utilised for some constantly used sounds (and they sound fine).
But, the larger files need to be on the USB stick, and that makes the speaker click really badly.
I have a relay to bring the amp on after the systems has fired up. Not sure what I can do to clean up that noise.
I already have it 'capped' to death. Though a couple of ferrites and routed away from power supplies.
I was shocked that the ferrite helped at all. The ferrite beads works for high-frequency RFI (radio frequency interference). Usually it's there to meet regulatory requirements and most users don't see any difference with or without it.
The sound module may just be poorly designed, but I don't have any idea why the internal files would be different from your USB files. Have you tried a silent file?
One thing that can happen is that the DAC output is sitting/resting at 0V. Then when the audio starts it jumps-up to 2.5V (assuming a 5V power supply, etc.) so the audio signal can swing relatively positive or negative. An output capacitor is normally used to filter out the DC, giving you a signal that truly swings positive and negative. But, that doesn't fix the turn-on glitch. I have no idea if that's actually what's happening but it's one possibility.
The solution is to turn-on the output bias a power-up time but that would have to be built into the module and the bias has to be held-on when the sound is stopped/paused.
SteveRC2017:
My question is, does a ferrite coil need to be a complete single piece to work?
No, it just needs to be a good-enough magnetic circuit (any air gap small enough)
Does it make a difference? Obviously its a lot easier to clip these coils (or are they chokes?) over an existing cable, rather than unsoldering your cables to put them through a solid ring.
Any air gap due to imperfections in the polished surface of the ferrite halves will contribute some magnetic reluctance. If this reluctance is much less than the core's own total reluctance it won't make a difference. You can calculate reluctance of air gaps and cores very easily, the physics is simple and known, so long as you know the exact
ferrite material and thus its magnetic permeability.
DVDdoug:
I was shocked that the ferrite helped at all. The ferrite beads works for high-frequency RFI (radio frequency interference). Usually it's there to meet regulatory requirements and most users don't see any difference with or without it.
The various ferrite materials used each have a maximum useful frequency, but there's not a minimum frequency, so they have effect at low frequecies too. Too much current at too low a frequency can saturate a magnetic material though, so for high powers there is a minimum useful frequency due to saturation kicking in.
If the ferrite can only reduce common-mode noise, so it depends how things are wired as to whether the noise is common mode or not. Here I suspect you had a ground loop that the core was helping to block.
Obviously its a lot easier to clip these coils (or are they chokes?)
Neither, they are (toroidal) cores. The wire itself forms a 1-turn coil if routed through a toroid. The combination of wire and core is a choke. A (magnetic) core is something to wind a coil on.
I made a 3d printed USB housing, and put a PCB female USB socket inside that, and cut the pre-made end off a standard micro USB lead. Basically, made my own lead, with the ability to solder to the female socket.
I fitted a 1000uf and 0.1uf cap across the female socket, and that removed 99% of the noise from the USB stick when it accesses the music files.
It also seems to be different, depending on the make of USB stick. I tried several and some were worse than others.
The coil on the line also certainly helps. In the end, the best one was a soild ferrite coil, with the USB lead wrapped through it several times, and secured with a tie-wrap.
Obviously the load of the USB stick power requirements upsets the JQ8400 MP3 module.
Now I have cleared this issue, these MP3 modules are very good for the money.