LVGL sample on a RAYTAC MDBT50Q

[mej7000]

Using the LVGL sample... The sample LVGL code works on nrf52840dk but not the RAYTAC MDBT50Q. I hope I'm just missing something obvious?

1. Flash blinking to nrf52840dk and RAYTAC MDBT50Q and both work (confirming jlink flash process/connections are good)
2. I have to manually repower/reset the board on the MDBT50Q after flashing, so I do that
3. flashing a LVGL sample, including led0 "on", to a nrf52840dk... works perfectly. 
4. flashing the exact same LVGL sample to a RAYTAC MDBT50Q (nrf52840) - doesn't work.
5. led0, which is part of LVGL sample - doesn't work on MDBT50Q
6. MDBT50Q appears to flash the LVGL sample correctly
7. command for both: west build -b nrf52840dk_nrf52840 -d build52840 -p -- -DSHIELD=waveshare_epaper_gdey0213b74
8. I couldn't get a 0.97" OLED to work over SPI on the MDBT50Q
9. I could get the I2c OLED version to work on the MDBT50Q.
10. checked MDBT50Q pwr and gnd (3v3)
11. I checked the SPI connections/reconnected many times
12. The sample code and images: https://github.com/mej558/lvgl_Raytac_Question.git
13. Included images and drawings in the repo folder /images.
14. I tried 3 different waveshare displays of the same model. All work on the nrf52840dk, but not the MDBT50Q.

led0 does not come on when I flash LVGL to the MDBT50Q. However, if I pull gpio1.10(display reset) the led0 lights. If I then plug gpio1.10 back in led0 stays lit. Not sure what that means...

Thinking that there was something weird about gpio1.10 I modified the Waveshare shield to use gpio1.07: 

        // reset-gpios = <&arduino_header 14 GPIO_ACTIVE_LOW>;  /* D8 <14 0 &gpio1 10 0>*/

        reset-gpios = <&arduino_header 12 GPIO_ACTIVE_LOW>; /* <12 0 &gpio1 7 0>*/

With this change I get the same behavior that I had with gpio1.10, but now on gpio1.07: If I pull gpio1.07 led0 lights. Plug gpio1.07 in and led0 stays lit... With gpio1.10 or gpio1.07 the display never shows any activity/data/flashing/life, etc... 

The gpio1.07 modification works perfectly on the nrf52840dk.

[helsing]

Hi Michael,

command for both: west build -b nrf52840dk_nrf52840 -d build52840 -p -- -DSHIELD=waveshare_epaper_gdey0213b74

I see from the images that you are using MDBT50Q-DB-40. Have you tried building for raytac_mdbt50q_db_40/nrf52840 ?

[mej7000]

I tried that after making this posting. Ultimately it didn't make any difference.

I ran into problems because the waveshare used the arduino_spi definition and the Raytac didn't define arduino_spi... Zephyr is relatively new to me, but I modified the waveshare_epaper_gdey0213b74.overlay file to use spi1 and test it on the RAYTAC.

As best as I could tell I modified waveshare_epaper_gdey0213b74.overlay correctly because I was testing the pin allocation changes on the nrf52840dk and nrf52840dk continued to work after being rewired.

That said, I expected that by using "nrf52840dk_nrf52840" it would work on the RAYTAC as long as the same pins were used on both the nrf52840dk and the RAYTAC. Is that a good assumption or am I missing something?

After this posting I used this same:  west build -b nrf52840dk_nrf52840 -d build52840 -p -- -DSHIELD=waveshare_epaper_gdey0213b74 on a custom pcb with a Fanstel module and got that display to work using the same arduino_spi default pins.

I would think there might be something off about the RAYTAC dev board, but once I pull the "spi reset" pin led0 starts flashing as expected, so something strange is happening that seems to be related to spi.