Issues with ESP32 deep sleep whilst using GDEP073E01.

[NickCole99]

Hello, whilst using GDEP073E01 and the Good display driver whenever I come out of deep sleep the colours on the epaper are slightly of and the display can takes a couple minutes to update. I was wondering if you knew the fix for this issue.

Thanks Nick

[G6EJD]

It seems like you're using partial refresh, is that correct? If so periodically the display needs to be cleared to enable the pigments to be realigned to their correct layers. There is no solution to colour merging in partial refresh conditions.

[NickCole99]

No not using partial refresh. I dont think the 6 colour display allows partial refresh. If I use deep sleep or restart the esp32 every update to the epaper afterwards takes 90 seconds and the colours dont look normal. Using lightsleep or running the code in the loop() method makes the epaper work well with 12 seconds update but I can't make use of low power deep sleep if I do that

[G6EJD]

Shouldn't be more than about 25-secs, odd. In situations like this I add diagnostic statements to the serial monitor with a time stamp (millis) so that the code location of the delay can be found, it might not be the screen refresh. Also if there's a clear statement before the screen painting, try removing that.

[ZinggJM]

and the Good display driver

Do you use the Good Demo code or GxEPD2?

If you meant the DESPI-C73 connection module, then please be more specific.
If your issue is with GxEPD2, then I would need diagnostic output from Serial Monitor.
And I need to know the display.init(...) call you use (parameters).

[NickCole99]

ok i'll give this a go when I get home . I didn't think it would take long to put the epaper to sleep though

[G6EJD]

It's the only difference between the two programme states. Either that or is a wiring or power issue.

[NickCole99]

@ZinggJM @G6EJD

Do you think it could be related to this persons issue coming out of deep sleep. and if so could someone explain in more detail regarding the solution That ZingJM mentioned

https://forum.arduino.cc/t/gxepd2-fuzzy-display-if-partial-update-after-deep-sleep-flawless-if-partial-update-without/1166952

[ZinggJM]

No. And you told you don't use GxEPD2.

You could consider to tell more exactly what you try, e.g. by posting your modified code, so that @G6EJD can continue helping.

[ZinggJM]

I just decided this topic is not related to GxEPD2, and the mixture of multiple responses that get hidden is confusing. I close this.

[ZinggJM]

Seems closing is not enough.

The original poster should start a new question, clearly indicating it is not related to this library.
And providing all information needed.

If this question is continued, I will use my last option and delete it.
I hope I do not need to disable discussions in this repository.

[NickCole99]

@G6EJD I have tried to hold the RESET pin down during deep sleep and it does work, when the epaper wakes up it will update within 12 seconds instead of 60. The problem then comes with the busy pin eventually becoming stuck. I've tried various solutions to get the busy pin unstuck but they all failed. Any ideas? It gets stuck in the waitWhileBusy method as part of the initlisation sequence.

In addition I have tried 3 different ESP32's (ESP32 Firebeetle, ESP32 devkit, ESP32 adafruit feather). The good display driver, the waveshare drivers, the gxepd2 driver. The waveshare 7.3inch epaper adapter board, the despi-073 good display adapter board. On every single combination I get the exact same problem with the epaper taking 45 seconds+ to update instead of the usual 12 after deep sleep.

[G6EJD]

This is all help to the e-paper community in solving display interface issues.

I'd add an external pull-up resistor say 10K is enough for the required input current and make sure the first thing the ESP32 does on wake is to release the reset pin by setting it high and maybe even releasing the sleep hold in that pin, but that should not matter. Then of course reassert before sleep again.

This must be a particular screen device driver chip attribute, as generally these measures are not required.

[NickCole99]

Ok I've ordered some resistors. I'll keep you updated on how this goes.
One thing to note is that if you click the Reset button on the ESP32 which I think cuts all power briefly it will almost always wake up and have a fast update on the epaper. I've tried recreating that in code but it's very difficult and doesn't cut the power for long enough.
There are also other people I know who have the exact same problem and I'm talking to Good Display engineers about a solution.

[G6EJD]

You could use the ESP32 internal pull-up resistors too (Arduino IDE):

pinMode(GPIO_PIN, INPUT_PULLUP); sets the internal pullup resistor

The ESP32 reset clears all pin states, leaves power connected, I've not seen a board that reset the power regulator. So actually, the reset gives you a clue as to why the busy state is stuck on, the SPI pins must be presenting a state that maintains the reset state or holds the busy line down. Resetting all the pins to inputs and removing the sleep hold state enable should do the same.