I'm using a Nano Every for measuring temperature with NTC. I first use USB port to supply the Arduino. When I convert the analog signal with 10 bits ADC of the Arduino the digital value is very stable (+/- 1bit) everything fine till then...
But when I supply the device with a 12V power supply passing by the onboard 5V regulator (MPM3610) the digital value becomes more instable (+/- 10bits).
I can see on my scope the ripple of 5v supply is more than 100mv !
I tried using a very precise laboratory power supply with same result.
When I supply the arduino directly with 5V power supply the problem disappears...
The onboard 5V regulator seems very noisy ! The datasheet of MPM3610 shows a maximum output ripple of 10mv for 1A load, my circuit needs only 100mA !
It may be a faillure on this particular Arduino or anyone already had this kind of problem ?
I've not experienced this situation. I would start trouble shooting by verifying the regulator has the proper input and output capacitors. I it you measured the ripple with a scope, can you post the frequency?
The problem doesn't come from the 12v power supply, it's providing clean signal checked with my scope. Anyway when I connect this power supply (reduced to 5V) directly to the 5V pin of the Arduino there is no problem. Moreover I tried to supply with 12V AC/DC converter or 9V battery and the problem remains every time I pass by the onboard 5V regulator.
Not sure if related, I used Nano classic and Nano Every with the same sketch to process analog input, Nano Every appeared to be too sensitive, the fluctuation of the input value was great and it was not easily settling on 0 when the signal was minimal.
I will try to post a screenshot of my scop as soon as I could. It's not a pure sinus ripple, the lowest frequency is about 500Khz with higher harmonics...
I ordered a new Arduino Every to see if the problem appears with another board.
I tried many reference voltages for the ADC : default, internal 4V3 and external there is no change.
I have to check with my scope the NTC signal entering in the ADC cause the NTC is powered by the 5V output of Arduino.
The Nano Every has a switching DC/DC converter on the V-in pin, and the 500kHz ripple you're seeing could be from that supply.
Try a 100n ceramic capacitor between the 5volt and ground pins of the board.
Maybe you should just power the Every with a 5volt cellphone charger, connected to the USB socket.
Leo..
Because of my many tries I did a mistake with pins and burned my Arduino !
I will receive two new ones on tuesday so I could see if I have the same problem with those devices and take screenshots of my scope...
There are many capacitors already between 5V and ground on the Nano's board (4.7u, 22u and 100n), I tried to add bigger until 2200u, it reduces the ripple voltage but not enouth. There must be a problem somewhere else...
The measured temperature is about 40°C (+/- 5°C) I have connected a 6K resistor (and 100n capa in //) to the CTN witch provides a signal around 3V to the A0 pin with good linearity for this gap of temperature.
I have about a dozen Nano's and they all have the AMS1117-5.0 device in a SOT 223 package. Mine are about two years old so I guess revision and manufacturer enters into it.
A Nano Every has very little to do with a classic Nano.
The designers stuck the word "Nano" to a whole range of new-fangled Arduinos, just to bloody confuse us.
Leo..
Good answer, that is why I ask for links for all hardware devices. I wish more would do that. I also try to download schematics for each device I get. Most of the time I do find one.
