Go Down

Topic: Sound active without microphone (Read 403 times) previous topic - next topic

someuser

Hello,

I want to add a aux input to a nodemcu for making a sound active light. With a second jack so i can hook it up between the computer and the amp.

What is the best way to do this? i have read that the - voltage can damage the board.

Greathings

DVDdoug

You need a Y-Splitter.

Quote
have read that the - voltage can damage the board.
A line-level or headphone-level signal is just-about right for the 0-5V Arduino Analog input but the input has to be biased because the Arduino can be damaged by the negative-half of the AC audio signal and/or the negative half of the signal can be "damaged" (distorted).


There is a schematic for a bias circuit at the bottom of my World's Simplest Lighting Effect.


someuser

Thanks, i have all the item's that i need for testing, but this is for a arduino on 5V, i want to use a node mcu working on 3.3v.

DVDdoug

Everything should work the same.   But, if the signal is too hot for the 3V range add another resistor in series with the capacitor.    Or you can add a potentiometer (just like a volume control).

Note that for the audio signal the voltage divider resistors are in parallel (the current from the audio signal goes through both resistors).    So for example, if your voltage-divider resistors are 10K, adding a 5K series resistor will cut the signal in half.   

someuser

is it the same as using 3.3V instead of 5V?


will the polarity not damage the capacitor?

DVDdoug

Yes, it's the same but I'm not familiar with the nodemcu, so I don't know if it has a 10-bit ADC like the regular Arduino and I'm not sure if the ADC reference is Vcc.

The voltage divider with 2 equal-value resistors will give you a bias of 1/2 the supply voltage, so it's mid-way between zero and 3.3V instead of mid-way between zero and 5V. 

On Arduino (10-bit ADC and Vcc reference) the bias will read about half of the 0-1023 range which is about 512 with no signal.     With an "quiet" audio signal, the readings will jump-around near 512 and with "loud" audio the signals may jump-around between zero and 1023.

The numbers "jump around" and "appear random" even with a constant test-tone because you are sampling a constantly-changing that passes-through zero twice per cycle, etc.   (With the bias, it passes-through the bias-point twice per cycle.)     

Grumpy_Mike

Quote
will the polarity not damage the capacitor?
No.

Go Up