Power usage in deep sleep mode [solved]


#1

After calling Spark.sleep(SLEEP_MODE_DEEP, 60); i am measuring around 200uA, nowhere near the 3.2uA from the specs. am i doing anything wrong? i have nothing connected to any of the IO pins.


#2

@remcohn the 3.2uA is from the STM32 specs, you have to include the regulator, flash and other components on the Spark board. See this thread for more discussion -

200uA sounds correct.


#4

That’s a little higher looking at the data from: https://github.com/spark/docs-old/issues/307

What’s the code you used for deep sleep and any external peripherals?


#5

I see this thread is now quite old, but I too am trying to use a Photon in a battery-powered environment. I am measuring about 150uA on a brand new Photon when in deep sleep mode. This is feeding power in to the 3v3 terminal, with nothing connected to either Vin or Vbat. Some posts have said the switchmode regulator is drawing current, but looking at the schematics, unless it is “leaking” down from its output pin to ground, that should not be happening - the regulator is not in use when the 3v3 pin is the main power source.
I do not need ANYTHING to be running in deep sleep mode, not even the real time clock.
I could live with - say - 20uA. My alternative for this project is the Microchip module (previously Roving Networks, Wi-Fly), where the standby current is certainly below 10uA.
Do those of you more experienced with the architecture of this system know ways of further reducing current consumption in Deep Sleep, given that my only requirement is that it wakes up on the WKP pin?
My alternative will be to add 2 transistors as an external latch to disconnect the power when in standby. Much messier!