[bombela]

When you turn off power, it might generate some unintended signals on the MCU pins that the screen has time to interpret before shutting down.

You could try setting the pins to Z state (floating) before turning off to see if it makes a difference.

If you have access to an oscilloscope you can also analyze what is going on in fine details.

[jrockway]

That's an idea that hadn't occurred to me. I can actually reproduce it by only removing the power for the e-ink display and not to the MCU. I think it's just a quirk of the carrier boards or something like that, but I don't know enough about the design to really debug it.

[shultays]

In my project I powered off my arduino (and the screen attached to it) completely by using latch and control the latch using a PIC mc. Even in sleep mode, arduino consumes quite a bit of power while PIC consumes relatively nothing while sleeping (datasheet says "20 nA @ 1.8V, typical" but probably more since you will at least WDT)

[jrockway]

Do check the quiescent current on your 1.8V regulator, though. That accounts for a lot of loss that goes away when you use something like a TPL5110 to disable it completely.

[shultays]

I don't use a regulator for PIC, it is directly connected to batteries