Just wondering because my chinese knock-off Nano does it !
Plugged into my PC USB, I am getting:
4.80V on Vin
5.04V on +5V
3.28V on +3.3V
Either it has a really fancy 5V regulator, or the terminal holes are not connected to the right places ??
or you have a not very precise voltmeter 
those values are good enough for most use (unstable reference will mess with the analog readings)
1 Like
I've just tested several atmega32u4 boards, they give sensible results eg
Vin= 4.64V
5V = 4.61V
And then I tried a ATTINY88
Vin = 4.02V
5V = 4.69V
My serious question: If I am powering the Nano or ATTINY boards from an external supply, which pin do I squirt the power in at? The one labelled Vin, or the one that looks (to me) like it is attached to the supply when running from usb ??
PS. I am going back to basics while troubleshooting an analogue to digital project that is giving me some trouble.
Vin would be the safe one but it needs to be between 7.5V and 12V
You can feed a Nano with 5V on the 5V pin. You should not feed 5V into Vin, the onboard voltage regulator drops the voltage to a lower value (the input voltage will be out of spec).
Further the internals of the onboard voltage regulator might feed back into the Vin and there might be losses there.
Attached a possible schematic of a Nano clone (only God knows how many different versions of that board exist). The schematics for official Nanos and Pro Micros can be found on the sites of Arduino and SparkFun.
nano_ch340_schematics-rev1.pdf (851.9 KB)
2 Likes
Thanks, that schematic is useful! I had assumed the 5V or whatever from the USB would be fed through the same regulator as power applied at Vin. Now I know it's applied direct to the chips and out the 5V terminal.
And I guess the 5.0V usb power goes backwards through the regulator and gives 4.8V on Vin ???
Anyway, it looks like I've been using the pins correctly. (edit: No I wasn't, I was attempting to feed Vin with 5V)
But for my AtoD project, I'm now thinking that I might be too reliant on the quality of the 5V USB power regulation.
And I've also noticed that the CH340 3.3V out is rated at 25mA, some of my project configurations have been trying to draw 100mA from there. (edit: actually my nanoclone has a 3.3V regulator with possibly 150mA capability)
The voltage regulator on the nano has an accuracy of +/-1%, most people find that good enough as the ADC reference even without calibration.
If you need more accuracy you should use a high presision external reference.
Thanks, jim-p. For simplicity I wanted to use USB power, and I've just learned that that doesn't get regulated.
I might be confusing 328P and 32U4, but I remember something about different reference voltages, so I'm hoping I can use a lower than 5V reference value.
Not possible.
There is something wrong with your measurements, Perhaps at which point you are measuring.
Not being familiar with the Nano I suspect there is a diode between the USB and Vin, but not to the regulator.
For the 328P you can use any voltage between 1V and Vcc for the ADC reference.
Not sure about the 32u4
So many different Nanos... So many different Arduinos!!!
The built-in regulator (5V reference) should be pretty-stable as long as the load is constant, and as long as you are feeding-in 7V or more. The internal 1.1V reference is super-stable but not necessarily accurate so you might need a calibration adjustment.
Or you can supply your own external reference.
Of course if you are using USB power, the 5V reference is only as accurate as the 5V you feed-in.
This is complicated.
I did my calibration runs with my nano plugged into USB. So it was being powered with nano Vcc= 5V. And the 5V pin was feeding 5V to my IR photocells and 3.3V pin feeding my IR emitters. I used the serial monitor to check the analogue reads and algorithms.
Then I used it powered from the output from a 32u4 board (4.7V) feeding the nano Vin pin. This results in a nano Vcc of 3.6V and 3.6V out of the 5V pin to my sensors. The 3.3V pin gives a good 3.3V and on my clone it comes from a regulator with possibly 150mA capability so that's probably OK.
The fact that it worked is down to the code working with relative levels not absolute, and maybe the self calibration helped.
My problem is actually intermittent glitches, I will find out soon if sorting out the power supplies fixes it.
My question is, what happens when Nano Vcc is less than 4.5V, did it switch itself to 10Mhz operation? Or did it somehow hobble along at 3.6V 20MHz
This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.