Can I power classic nano with external regulated 3.3V power supply connected to 5V pin?
Thanks!
Not unless you boost the 3.3V supply to 5V
Why would you think that it would work as you suggest ?
Some Uno like boards have a switch to swap between 5v and 3v3. This although running the ATmega328P at 16MHz while powered at only 3v3 is not supported according to the data sheet.
You may also have to lower the default brown out voltage also using say AVRDUDE but it could work. YourDuinoRoboRED2 (Upgraded Arduino UNO Compatible)
Yes , I have done it and it works.
Due to the 16Mhz resonator stability problems may occur.
Just try it and see how long it will work.
The Pro Mini uses the same Atmega 328 , the 16Mhz one's is 5V and the 8Mhz one's is 3.3V.
No.
Read the documentation for the board and learn the proper way to power it
Agree !
Whilst it might work , it might be unreliable .
And if you look at the data sheet for the 328 processor it will say it can run at lower voltages , but only if the clock speed is reduced . So you could make your own simple board with a 328 and 8 MHz crystal
It might work, but the Uno is relatively complicated as compared to later boards like the Nano or Pro Mini. You have the barrel jack, and all the USB stuff that might complicate things. If possible, you might consider using the 3.3V 8MHz Pro Mini instead, along with a UART adapter.
Yes you can do it, but do not expect to work properly. My recommendation is to use a boost converter, adjust for 8-9 volts our and connect to the Vin. This will also give you a little more filtering.
Also realize the 5V is the reference for the A/D so your code accessing the A/D's will need to be updated.
What is the maximum output load current on the Nano powered with 5 volts?
What is the current capacity of your 3.3V external supply?
According to the datasheet the ATmega328P is guaranteed to work properly down to 3.78V at 16MHz. It's not like a door that is slammed shut at that point, your device may function down to lower voltages, especially at room temperature. But at some point the chance of 'falling bits' starts increasing and you may not notice this if it's just a calculation producing a different result.
If you gamble on your device working properly at 3.3V, use the 5V pin to feed your supply voltage into your board. BUT you need to watch out and disconnect when uploading a program, the USB power will try to feed back into your 3.3V power source. (A diode may solve the feed back problem but will lower the supply voltage to the board even more.)
@vakulenko
Don't even think about it. It will never work at 3.3V.
If you wanted to go to all the trouble you could burn the fuse bits to start up with the clock divided by 8 and then switch to divide by 2 in setup. That will cause you to have to do some additional timing changes or perhaps burn a bootloader for an 8 MHz board.
I’m not sure if the 16MHz oscillator will work at 3V3 so that suggestion might not work.
Won't help, will never work at 3.3V
Everyone here is overlooking all the other components on the Nano board. The FT232 usb chip will never work at 3.3V so there is no point in trying to make the processor run at 3.3V if you can program it.
Correct: the FT232 will not work properly at 3.3V supply. BUT that's irrelevant unless a USB connection is present, in which case there will be 5V available. (With the possibility of feeding back into OP's 3.3V powersupply.)
Not quite, even if 5V is USB is still will not work and you will probably blow out he USB port.
There is no point in even considering using 3.3V.
Connected? Not connected?
No, the 5V pin expects about 5V. If you feed it 3.3V the Nano may not run reliably. If you have a regulated 3.3V supply, connect it to the 3.3V pin instead.
Even if it did there is no point. See post #15
There seems to be quite a lot of dogma in this thread.
Of course, if you are designing say a life support system it would be very risky to go out side the parameters specified in the manufacturer's data sheet; in this case the minimum voltage for an ATmega328P running at 16MHz. However, for hobby type projects, some testing under the anticipated use conditions may be all that is required to confirm it "works". The data sheets anyway will have a high degree of CYA and will attempt to anticipate even the worst edge cases.
In post #3 I gave an example of a commercial product which relies on ATmega328P running at 16MHz with a (selectable) 3.3v power supply. This is sold by a long term forum member @terryking228 . Maybe he can say something about known problems with this configuration. I have also two "Freaduino v 1.8.1" Uno clones from Elecfreaks (dated 2012) and these work fine also when switched to 3v3. Infact, these are better that the Uno in that they also have a real crystal and a switch mode regulator. It is similar to these: Freaduino UNO new version with new features is now online - micro:bit Accessories Store | ELECFREAKS
Adafruit also have this in their trick box: 3.3V Conversion | Arduino Tips, Tricks, and Techniques | Adafruit Learning System