AnalogRead() external power causes fluctuations

Question

Hey guys!

Thanks for all the help so far in my previous questions. This community has been a great tool in my learning process.

I am building an automated door buzzing machine for my apartment. The project works by reading analog signals from a microphone sensor. With a positive reading from a sensor, the module will then move a servo motor to buzz the person in.

When plugged into USB, my analog readings from the microphone are stable around 24 consistently.
When plugged into my 5V 4A power supply, readings are typically 24 but often fluctuate up to 45.

Both readings are taken in isolation from any servo movement.

Given that this would make establishing a buzz correctly difficult, is there something that would be causing this problem?

The Arduino is powered via the 5v pin and common grounds across all components. The components i'm using also bridge off of the 5V pin.

The arduino connects to the microphone via 3 wires (+/-/A0) and I do not have any capacitors or anything extra.

I found a few forums online talking about reference voltage and Aref, but I was quickly lost in the terminology.

Thanks in advanced!

jc2048 · Answer

Looking at those boards online, it would appear that the board has a comparator. The potentiometer on the board is setting the threshold at which the comparator detects a sound when it goes above that preset level. The analog output is just the threshold and should be steady (it's there so you can measure the threshold that you've set in case you want to know what it is). The other output (the important one if you're interested in detecting sounds) is digital and will go either high or low when a loud sound exceeds the threshold (depending on which way round they've done the comparator).

beacon_dave · Answer

I found a few forums online talking about reference voltage and Aref, but I was quickly lost in the terminology.

There is a very lightweight introduction to AREF here which may help:

AVR Insights - Episode 17 - ADC Voltage Reference

https://www.youtube.com/watch?v=fN3J6wxijrE

There is also a block diagram of how it fits into the design on page 5 of this application note:

http://ww1.microchip.com/downloads/en/AppNotes/00002538A.pdf#page=5

As you can see the ADC is actually comprised of a DAC and a sample & hold comparator circuit.

More information on AVR ADC parameters in this application note:

http://ww1.microchip.com/downloads/en/AppNotes/atmel-8456-8-and-32-bit-avr-microcontrollers-avr127-understanding-adc-par…

rachaelp · Answer

pjacobsen wrote: The part that confuses me still even with this new information is why there is a large difference between stable readings between USB and external wallwort. The USB maybe supplies a relatively clean 5V but the wallwort possibly is not very well regulated (possibly unregulated) and maybe has a noticeable amount of ripple on it. Best Regards, Rachael

jw0752 · Answer

Hi Parker, It would be helpful to have a little more information. Is your 5 volt 2 A power supply linear or switching? Specifically what is connected to the 5V arduino pin. Can you draw a wiring diagram and post it? The solution may be as simple as a couple of decoupling capacitors but more information would be helpful. John

jw0752 · Answer

Hi Parker, This is a shot in the dark but it will be fairly easy to implement. Try putting a 100 nF (0.1 uF) ceramic capacitor between the ground and +5 volt line close to the microphone and another close to the arduino. Sometimes these little decoupling capacitors can solve these glitches. John

jw0752 · Answer

Hi Parker, In this situation the 1 uF caps will work fine. There are a lot of factors that come into play in choosing a capacitor for a given application. In a timing application for example it is critical to getting the frequency or timing that you are looking for. For decoupling a power supply it isn't as critical but other factors like the effect serial resistance come into play. Some of the real engineers on this site could give us a much more detailed list. Give the 1 uF caps a try, they can't hurt, and let us know what happens. Note your 1 uF might be polarized so you may have to watch their orientation. John

AnalogRead() external power causes fluctuations

Top Replies