Auto power off for NANO

Please refer the sketch below and the comments on it. Basically it works as expected when S1 is pressed momentarily. But i notice when the NANO is not powered there is a voltage of about 2.98 V across its Vin pin and GND pin. Why is this ? When in OFF condition the MOSFET is supposed to be open …right ? No ?

Mosfet_As Switch.jpg

Is anything else connected? If there is, watch out for phantom powering the Nano. That's why it's pretty useless to really power down the Arduino. Just turn all leds off and put it to sleep.

Vin is not the right input to use with a 5V supply. There is no VCC pin on a Nano, either, only 5V.

There's a problem with that design - you forgot about the protection diodes. In order to prevent damage to the chip, all pins have internal diodes on them to Vcc and Ground to keep the voltage on a pin between Vcc and Ground (as required by the datasheet, see "Absolute maximum ratings". So you disconnect ground from the nano - now ground (as far as the nano is can see) should hop up to ~= Vcc, so the signal line should do the same - BUT IT'S CONNECTED TO THE GATE OF THE FET - (and also connected to ground through the pulldown, which would backpower the nano anyway, which is bad too)....

I think putting a small P-channel fet in, and inverting the logic of the power hold signal would solve this?

But as noted - you can get insanely low power consumption using sleep (though you have to remove the power LED and regulator from the nano board to really get the power usage down); that's what people usually do to turn a microcontroller "off".

OK let me address all mistakes ( goof ups ?)

  1. Nano does not have a Vcc … it has only VIN and 5V . In my case i had actually connected top the 5V pin. Anyway this will be corrected as i now have a 7.5V DC supply and can connect it to VIN.

  2. If it were only the Nano, then power down is not a requirement and i can enter low power mode. But in the final set up there are other items like a RFID reader and RF24 module and i was wanting to totally switch off 5V power input to all.

  3. I was wondering why the power LED on the Nano was glowing dimly always… DrAzzy explained it well. I am now clear thanks. In fact i have another version with more components on it, which works. The Vcc is 5V going to the MCU ; and the base of the SL100 transistor is connected to D2 of MCU. The Mosfet was a NDP6020P ; For shut down the MCU goes to sleep and also effectively shutting down all 3.3V peripherals.

Any issues with this setup ??
5v and 3.3V.PNG

Mogaraghu:
2. If it were only the Nano, then power down is not a requirement and i can enter low power mode. But in the final set up there are other items like a RFID reader and RF24 module and i was wanting to totally switch off 5V power input to all.

I assume with RF24 module you mean a NRF24L01+.

There is no reason to cut its power

nrfpower.png

Whandall:
I assume with RF24 module you mean a NRF24L01+.
There is no reason to cut its power
nrfpower.png

Thanks for reminding that aspect. I am using this library : http://tmrh20.github.io/RF24/classRF24.html

I can of course use the powerDown() and powerUp() functions and save lots of power. But just a query apart from this …instead of powering down, if i just use the read() or write() functions only when required , does the RF24 module enter the Standby_I mode when not executing either of these ?

Standby I is entered after begin() or stopListening(), and after a write() completion.

states.png

Now let us know what RFID reader you've got and we can see if it has a low power mode too.

Don't forget that if you turn off power to peripherals and not the Nano, you can sometimes have the same issue with parasitic power through the protection diodes, only this time the diodes are in the peripherals not the Nano.

Whandall: Standby I is entered after begin() or stopListening(), and after a write() completion.

yes I have seen that State Diagram but wasn't too sure .. thanks for the confirmation. Sounds good.

Jiggy-Ninja: Now let us know what RFID reader you've got and we can see if it has a low power mode too.

This one : MFRC522 reader.

Jiggy-Ninja:
Now let us know what RFID reader you’ve got and we can see if it has a low power mode too.

Seems so.

mfrc522.png

Section 8.6 on page 33 lists the power saving modes, how to enter and exit them, and what implications they have on functionality.

https://www.nxp.com/docs/en/data-sheet/MFRC522.pdf

@Whandall @Jiggy-Ninja

That posts by you set off a search process and I landed on this GEM

Seem to be the exact one I needed. Since I use the MFRC522 I was earlier searching for a Soft Reset function but could not find one. Looks like there is an easier option of Hard Reset by pulling the RST pin low.

Thanks to you guys, looks like there is no need for power switch at all ....atleast for now.

Mogaraghu: Thanks to you guys, looks like there is no need for power switch at all ....atleast for now.

Most digital CMOS chips should have low power modes built into them since it's so easy to do. In a steady state condition, a CMOS logic gate only draws leakage levels of current. A large pulse of power is drawn only when something changes (like when an inverter changes from HIGH to LOW). Power reduction comes from disabling inputs and clock signals to reduce the number of gates changing state as much as possible.

In contrast, BJT based logic gates are harder to conserve power with since they always draw power in their steady state to bias the transistors. Compare the supply current specs for a 74LS125 and a 74HC125. The LS, made with BJT transistors, has a maximum supply current of 20 mA. The HC version of the exact same chip, made with CMOS, has a maximum supply current 1000x less.