Closed BlitzCityDIY closed 3 weeks ago
testing with display_bus = displayio.FourWire(spi, command=tft_dc, chip_select=tft_cs, reset=None, baudrate = 9375000)
instead of spi.configure(frequency=)
and that actually affects display speed. display is staying up but the bottom 1/3 pixels on 240x240 disappear and show the REPL
I'm trying to reproduce this with a similar setup:
@BlitzCityDIY Are the boards on the tripler in the same place as in @jepler's photo?
The hardware setup I'm using is with a Pico 2, PiCowbell Camera and a PiCowbell Tripler.
I tested with a Pico 2, PiCowbell Camera, and the tripler. I used the 1.54" ST7789 240x240 display and a 100mm EYESPI cable. I put the Pico 2 in the same position as the EYESPI cable, and the camera in the middle.
I ran the test program in the original post for several hours, and saw no problems.
i did try rearranging the boards and had the same result. i think where no one else has been able to recreate it's okay to close this- it's probably some undiagnosed issue with my setup
Might be worth trying another Pico 2 if you have one, or another Cowbell doubler/tripler. Maybe there's an unsoldered pin?
CircuitPython version
Code/REPL
Behavior
The preview for the camera will show correctly on the display for a while, but then the colors will either invert or the display will go black. The code is still running because the camera FPS is still printing to the serial console. The hardware setup I'm using is with a Pico 2, PiCowbell Camera and a PiCowbell Tripler. The display is connected via the EYESPI connector. I've tried a 1.54" ST7789 240x240 display and a 2.8" ILI9341 240x320 display.
I have tried setting the camera size to
OV5640_SIZE_QCIF
,OV5640_SIZE_HQVGA
,OV5640_SIZE_240X240
andOV5640_SIZE_QVGA
(QVGA only when using the 240x320 display)Description
I have tried reducing the SPI frequency, which previously I didn't have to do for this example with the Pico. However, the issue still persists. I was trying increasingly lower frequencies all the way down to
spi.configure(baudrate=1000)
(which is an actual frequency of2288
when I print it to the console) and that seemed to keep the display alive for the longest. The default SPI frequency prints as18750000
and half of that prints as9375000
.Additional information
I have not been able to recreate the issue using a Pico RP2040. When displaying the REPL the display does not go out.