MP3 Player with lm386

newplumber

Senior Member
Hi
When using the MP3 player with four lm386s...the sound is amazingly clear...when I was testing just using a picaxe/mp3/lm386/speakers...I was able to
kill alot of the amped corrupted noise output from the lm386 by adding a 10 Uf cap as close to the vcc/grd on the mp3 (not lm386) as possible.
Which worked great with almost perfect sound from the sd card ...then when I installed the mp3 player to my clock ...it started picking up the
noise from the pwm apa102 leds...I even unplugged the mp3 and the noise out of the speakers was the same (humming) ...my thought is somehow
the noise is getting to the lm386s ...i have twisted pair power to the lm386s...I was wondering if anyone dealt with a issue like this
I did try covering the circuit in aluminum and to ground with no change
I will get a picture/schematic (probably next week) to show whats going on.

FWIW I have the clock working perfectly with sound by turning off the power to the lm386s
using a mosfet and a picaxe pin and only turning power on when a sound file needs to play
 

techElder

Well-known member
Try to kill the noise at the source. I wouldn't fool with RF shielding in a low frequency circuit. Concentrate on power circuits.

Bypass the power supply as much as you can.

Investigate where the high current is flowing. Remember, high current through resistance equals voltage. Voltage that is low frequency will make noise in your audio circuit.

Bring your high current leads into one single point right at where the power leads come in.

Bring your power leads for your sound circuits into the same spot, but on their own power leads; separate leads. Don't route the high current through the audio circuit leads.
 

newplumber

Senior Member
Try to kill the noise at the source.

Investigate where the high current is flowing. Remember, high current through resistance equals voltage. Voltage that is low frequency will make noise in your audio circuit.

Bring your high current leads into one single point right at where the power leads come in.

Bring your power leads for your sound circuits into the same spot, but on their own power leads; separate leads. Don't route the high current through the audio circuit leads.
Thanks Texas ...I will try rerouting the power as far away as possible ...you got good points
I guess I wasn't expecting the leds power mains to be making noise but now that i think about it ....they are running a alot of mill amps


btw I still can't get over how extremely clear the mp3 pushes the sound out....of course when there isn't any other noise
I need to keep practicing this sound stuff....maybe I can get picaxe/mp3...to remind my none paying clients about their bill due ...every 3 minutes
 
Last edited:

techElder

Well-known member
maybe I can get picaxe/mp3...to remind my none paying clients about their bill due ...every 3 minutes
I brought my wife into our business for just that purpose! She had a few that needed a call every day; one of them got her into their office for four hours. They were happy to pay up just to get her out! :D
 

hippy

Technical Support
Staff member
it started picking up the noise from the pwm apa102 leds... I even unplugged the mp3 and the noise out of the speakers was the same (humming)
"Humming" sounds more descriptive of mains pick-up rather than high frequency PWM interference, but the LED updating or switching could be occurring at those sort of frequencies.

It's either noise being picked up on the op-amp input or being induced. That could be radiated EMI from the LED's or could be a result of noise going down the power rail.

You are probably going to have to experiment with what makes things better, try to find what the cause is from a correlation with what improves or alters things.

Check the noise goes away with the LED supply disconnected, then modify the LED software to only send LED off commands, change delays between each packet sent, add delays between each byte sent. Temporarily short the op-amp input to 0V on the amplifier board. Running the audio and LED's halves on separate power supplies, preferably battery supply for the audio.

Putting a scope on various parts of the circuit, including the power supply may reveal where the noise is greatest and may even characterise that noise which allows its source to be identified.
 

premelec

Senior Member
i agree with hippy that if you have hum it's likely from power line and high impedance pickup to amplifiers - just poke your fingers around circuitry and listen to how the hum changes... [on low voltage side!] - this might be cured by simply keeping mains AC away from other stuff using longer DC wires; picture of your setup would help...
 

newplumber

Senior Member
hippy said:
"Humming" sounds more descriptive of mains pick-up rather than high frequency PWM interference, but the LED updating or switching could be occurring at those sort of frequencies.
Thanks hippy ... I do have the led updating as fast as the 20x2 can push out to be able to change the colors, only stops updating briefly to read the RTC or send sound out. I will do some testing with my scope
when i have time. Just for clarification the humming noise seems to go quieter when leds are showing close to a single color and louder when its showing white (all colors)

premelec said:
i agree with hippy that if you have hum it's likely from power line and high impedance pickup to amplifiers - just poke your fingers around circuitry and listen to how the hum changes... [on low voltage side!] - this might be cured by simply keeping mains AC away from other stuff using longer DC wires; picture of your setup would help...
I tried moving my sound amplifier board around but didn't seem to make much difference ....My AC to 5vdc adapter is 6 ft away so I don't think thats the problem since it makes no noise when not changing the leds
and I will keep poking around to see what changes/no changes and thanks
btw the leds must change always... its a rule i have in my house....any clocks with digits over 12" must have flowing color changes because its cool:cool:
 

premelec

Senior Member
Are you distinguishing between 'hum' and 'buzz' - ? If so and you have 'hum' then possibly you are loading the DC power source to the point that it is putting an AC component on the DC by overloading the regulator in the DC supply. You can check this in part by putting a multimeter set to AC across the DC supply and seeing if AC appears on the DC line when the LEDs start up. [BTW the LEDs should just make buzz rather than hum... and note it's too early in ND for bees being very active... ;-0 ] . I hope you have considered using the color code to indicate the time [like resistor color codes... ] - using octal number notation perhaps to limit the number of colors you need to use ;-0
 

newplumber

Senior Member
Are you distinguishing between 'hum' and 'buzz' - ? If so and you have 'hum' then possibly you are loading the DC power source to the point that it is putting an AC component on the DC by overloading the regulator in the DC supply. You can check this in part by putting a multimeter set to AC across the DC supply and seeing if AC appears on the DC line when the LEDs start up. [BTW the LEDs should just make buzz rather than hum... and note it's too early in ND for bees being very active... ;-0 ] . I hope you have considered using the color code to indicate the time [like resistor color codes... ] - using octal number notation perhaps to limit the number of colors you need to use ;-0
Ok I will check for AC across the DC when i take it back down...(its on the plant shelf doing preflight checks)
I am using a 5 volt 2 amp plug type adapter....I have the led brightness running max 75 (0-255) so the max amps when showing time
at 10:58 on white was pulling 1.34 amps for short time then drops back to average of 1.15 amps
well I don't know how to explain it (buzz/hum) its not a bee flying by and you right it is to early for that
and its not a 4' fluorescent ballast ...its like in between so maybe like a fluorescent BEE :)
I didn't think to do color codes ..good idea ...but the kids seem to like the colors just flow one to another
I had all of them (kids) record many vocals like "good morning" etc... so every day its someone elses voice.
So I was able to get the kids to sit around staring at the clock within 5 minutes of a sound hour (7,12,5,9) just to see whos voice is going to say what
and since it comes in so clear ..I might just have them call out every hour from like 7 am to 10 pm ...it endless possibilities
 

premelec

Senior Member
Great project!... another 'cure' would be to silence the audio entirely unless an audio transmission is happening.. usual low level noise is not discernable or at least not annoying whilst sound is playing... in this instance the problem becomes killing the on/off 'click'... always something...;-0
 

newplumber

Senior Member
premelec said:
another 'cure' would be to silence the audio entirely unless an audio transmission is happening
thats what I have done so far ...and seems to work fine ...but I will keep playing around with it
one thing I will try from reading texasclods comment is turn up/down the brightness see if it creates more noise with more current


inglewoodpete said:
and later than 7 am on a Sunday morning!
Yes true ...and Saturday too ...and its a piece of cake with using the "days" on the ds3231 rtc
 

techElder

Well-known member
I know ... I know ....

At random intervals get it to say loudly, "Marco!" :D

PS. Then use voice recognition to determine who responded first. :D (Hey, its your money I'm spending! :D)
 

newplumber

Senior Member
Maybe someday I will be able to program "voice recognition" on a future picaxe 80X8 :)
OR
let me rephrase that......
Maybe someday a experienced picaxe programmer that knows nothing about plumbing might be able to code "voice recognition" on a future picaxe chip.
but no "Marco's" .....because it could (thru poor text to speech) sound like "mark knows" and it would embarrassingly be incorrect. :)
 
Top