Supplying less than 5v to 5v pin?

Hello, ignorant question. I'm building a project that is powered by a battery and I wanted to minimize the current consumption as much as possible.

My configuration used to be like a 16v battery goes to a buck converter down to 3.8v (this is the lowest voltage I can supply the Arduino Nano without shutting down), to the Vin pin.

But I thought about moving the 3.8v to the 5v pin instead and it's running OK. I mean the Nano is working fine and doing its job without problems.

Is it ok to keep the wiring setup this way? Running a voltage lower than 5v to the 5v pin?

The reason I moved the supply from the Vin to the 5v pin is because I wanted to monitor the voltage drop of the battery source. And if I connect the supply to the Vin pin, I get incorrect voltage readings.

You need >6.5 to 7V to power via Vin so the 5V regulator can do it's job. 5V or less than that, connect to the 5V pin to bypass the regulator.

If it works, it works - for the moment. If you really want to run at the lower voltage, you might consider a 3.3v Nano or Pro Mini, and do the job properly.

I'm thinking of going around 3.3v. My Nano has a 3.3v pin and I'll experiment on powering it through that pin.

Thanks for all your responses.

One more question, (I think I know the answer to this, but I just want to be sure), bypassing the regulator is more efficient than powering the Nano to the Vin? Am I right?

Yes, more efficient. No power wasted as heat in the regulator.
At 16 MHz, Vcc should be >=3.8V per the 328P Speed Grade chart for all functions to be fully compliant (like Serial).
(maybe 3.78V if you do the interpolation math).

3.3V comes from the FTDI chip, it does not connect to the 328P's VCC pin. You'll have to add a wire for that. I don't know how the FTDI chip will be 3.3V being back-driven into it without 5V being applied to the chip. I bet not good.

Got it. I have a better understanding about this now. Thank you all.

The schematic is not a secret. This is my rearranged version of it.

I have been using three AA general purpose batteries in the remote controls of the scoreboards I make ( well used to make :-( ) using the 5v promini for the last 12 years, and they keep going until the voltage drops from 4v5 to about 3v3 off load. I do have a 470mfd capacitor across the 5v line to give the transmitter a kick when it sends score updates.

I tell customers to change the batteries every season ( the arduino goes into sleep mode when not transmitting ) but at least once a year I get some guy complaining that he has to walk half way across the field for the remote to update the 3 year old scoreboard. I ask him if they are new batteries and he says " what batteries ? "

Boffin3: I ask him if they are new batteries and he says " what batteries ? "

I guess that says it all, and marks you as a master of the black art of battery power.

Connecting to the 5v pin is fine (though you probably want to be using a Pro Mini to eliminate the baggage of the USB interface).

I think you also want to remove the 5v regulator - IIRC they burn some power even when backpowered with less than the output voltage (though almost all linear regulators are fine with being backpowered) - and you certainly want to remove the power LED, as that will consume about orders of magnitude more power than the chip itself when it's sleeping!

There is no 3.3 V version of the Nano - at least the original Nano with an ATmega328; the later variants with a quite different processor and peripherals in a basically matching form factor do use 3.3 V. This is because the USB interface operates at 5 V (though USB logic levels are 3.3 V :astonished: )

If you intend battery operation at 3 V, you need a 3.3 V Pro Mini which does not include the USB interface and uses an 8 MHz resonator and as DrAzzy says, you need to disable the regulator and pilot LED. Using the regulator when powering at 5 V or less makes no sense at all and the on-board regulator is always basically limited to powering the ATmega and very little else.