|
Forum Index : Microcontroller and PC projects : Picomite - lcdpanel console white-out
| Author | Message | ||||
| ebbandflow Regular Member Joined: 31/08/2023 Location: United StatesPosts: 48 |
Hi all, I am hitting a roadblock and thought I'd ask here because I must be doing something wrong. I'm setting up one of these newer waveshare RP2350-Touch-LCD-3.5 boards, and am having an odd issue with the BG color. I noticed it when trying to list the pins. The BG color of the display goes white after the initial print fills the screen, and then each new line is correct (white text on black BG) but turns white on white as soon as it scrolls up. This happens even on a fresh boot with no program loaded. (edit: updated to add a gif) > option list PicoMite MMBasic RP2350B V6.02.00 OPTION SYSTEM SPI GP18,GP19,GP16 OPTION SYSTEM I2C GP34,GP35 OPTION FLASH SIZE 16777216 OPTION COLOURCODE ON OPTION PICO OFF OPTION CPUSPEED (KHz) 200000 OPTION LCDPANEL CONSOLE ,,,, 60 OPTION DISPLAY 26, 60 OPTION LCDPANEL ST7796S, RLANDSCAPE,GP20,GP23,GP21,GP22,INVERT OPTION LCD BACKLIGHT 60 OPTION TOUCH FT6336 GP25,GP24,,50 Edited 2026-05-06 06:51 by ebbandflow |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 4316 |
Just so you know, imgur is blocked in some countries, such as UK John |
||||
| ebbandflow Regular Member Joined: 31/08/2023 Location: United StatesPosts: 48 |
Thank you for the heads-up John, I did not realize that! I tried to upload here but it rendered as a still image. Do you recommend any specific services that are more accessible? Thanks, James Edited 2026-05-06 07:22 by ebbandflow |
||||
| ebbandflow Regular Member Joined: 31/08/2023 Location: United StatesPosts: 48 |
This may be a workaround for those at their computers. Here is the .gif file, zipped up for uploading. white-out.zip |
||||
| mozzie Guru Joined: 15/06/2020 Location: AustraliaPosts: 304 |
G'day James, Been caught out by this one before, try: OPTION LCDPANEL CONSOLE 1,,,60,NOSCROLL If I have that correct it is: OPTION LCDPANEL CONSOLE Font,F-colour,B-colour,B-light,NOSCROLL NOSCROLL works for slower serial LCD's or for those that cannot be read from (ILI9488 etc). The other option is to use one of the buffered drivers. Regards, Lyle. Edited 2026-05-06 15:49 by mozzie |
||||
| ebbandflow Regular Member Joined: 31/08/2023 Location: United StatesPosts: 48 |
Good morning Lyle, Appreciate the reply. I just tried setting NOSCROLL, and that prevents the white-out, but it scrolls the 'list pins' command so quickly that I cannot see the first x number of pins as they immediately scroll off the screen (no prompt to 'press any key...). Very odd. Shouldn't it work without NOSCROLL and simply be slow instead of white? The manual says that this display type is able to handle transparency (ST7796S), but I wonder if there is a hardware hack required to get that working? Thanks, James Edited 2026-05-06 20:00 by ebbandflow |
||||
| ebbandflow Regular Member Joined: 31/08/2023 Location: United StatesPosts: 48 |
edit... It appears that MISO is not connected to a GPIO pin Still working on a solution. Edited 2026-05-06 21:43 by ebbandflow |
||||
| mozzie Guru Joined: 15/06/2020 Location: AustraliaPosts: 304 |
G'day James, No MISO connection is going to be a problem A quick look at the schematic shows the ST7796 has its own SPI connection to the RP2350, perhaps try the buffered LCD driver instead (ST7796S_BUFF) this is detailed on PG68 of the V60201 manual. I looked at these displays for an idea but opted for the display without the embedded RP2350, allows more flexibility at the expense of creating a PCB. Regards, Lyle. |
||||
| robert.rozee Guru Joined: 31/12/2012 Location: New ZealandPosts: 2515 |
have just been thinking... for a 60 x 26 character screen, that is 1560 character cells in total, and with each cell consists of one 8-bit character code and two 4-bit colour codes (foreground and background), just 3120 bytes of the RP2350's RAM would be required to hold a complete representation of the screen contents - provided that the screen is being used for text display. this pretty much negates the need for the MISO line to be connected - the data coming back from the LCD can (if indeed needed) be reconstructed from the above 3120 byte buffer when combined with character pattern maps and the RGB values from the palette table. as far as i can see, technically the only time the MISO line should be needed is when there is graphical data being displayed and when there is a need to scroll that graphical data either horizontally or vertically. someone might like to give Peter a hand by rewriting the LCD drivers so they make use of an internally held character/colour attribute map when displaying just text ![]() cheers, rob :-) |
||||
| Mixtel90 Guru Joined: 05/10/2019 Location: United KingdomPosts: 8816 |
...or just rewrite the entire screen when you update it. That's what it would do anyway. SPI is serial so there probably wouldn't be much speed loss. Without MISO there can never be data whatsoever from the LCD. I just looked at the schematic. It's a good board to avoid. :) They've skimped on pins and sacrificed MISO as it's not needed in many cases. As usual they've used GPIO pins for the SD card that they don't need. Mick Zilog Inside! nascom.info for Nascom & Gemini Preliminary MMBasic docs & my PCB designs |
||||
| mozzie Guru Joined: 15/06/2020 Location: AustraliaPosts: 304 |
G'day, Testing has shown the ST7796 LCD works ok with the ST7796S_BUFF buffered driver with NO MISO connection, however as Robert.Rozee points out, the screen buffer does use a chunk of memory if this is important in your application. Regards, Lyle. |
||||
| robert.rozee Guru Joined: 31/12/2012 Location: New ZealandPosts: 2515 |
from the photo of the board here: https://www.waveshare.com/rp2350-touch-lcd-3.5.htm https://www.waveshare.com/img/devkit/RP2350-Touch-LCD-3.5/RP2350-Touch-LCD-3.5-details-inter.jpg ![]() it looks like the RP2350's flash chip, labelled (2), is in a largish (SOIC-8?) package with the pins readily accessible. perhaps one could just piggyback a PSRAM on top of this (with CS routed to a suitable GPIO line - 0, 8, 47 are available on the I/O header) to provide space to hold a full 320 x 480 screen buffer - 300k required if my maths is right. it is a pity the designers didn't include an unpopulated footprint for the PSRAM. cheers, rob :-) Edited 2026-05-08 02:35 by robert.rozee |
||||
| The Back Shed's forum code is written, and hosted, in Australia. | © JAQ Software 2026 |