Відео

Thank you

Can you provide more details about accessing the mapped flash via CubeProgrammer?

@@bluewat Hello, I get the error " Error: Data read failed" When I set the address to 0x90000000 and try to read the external flash module Is there any other information you would like in detail? EDIT: Apologies, not sure what changed but it seems to be working as intended now. Thank you for the informative tutorials regardless!

You're welcome. I appreciate your understanding of the effort I put into the translation.

Uri, the ILI9341 communicates via serial or parallel port, but in the video I'm not using (only) that driver: I am using a board mounting a display and the ILI9341. Some boards expose the ILI parallel port pins, while others expose the serial ones. There are advantages and disadvantages to using one port over the other. The main advantage of using SPI instead of FSMC is that the former requires (many) fewer pins, allowing the display (the driver) to be managed even with smaller uCs or when the pins are involved in other activities. I'm thinking to do some test even with parallel port handled displays.

@@bluewat but parallel mode is 16 times faster !( instead of sending 1 bit I send 16).

@@urichaplin1424 … advantages and disadvantages…

Paul, the above video should indicate all the steps you have to do updating the library you already have. A second possibility is analysing datasheets of ST7735 and ST7789 adapting the above library to your display.

you should find answers here: ua-cam.com/video/g1siKaPox88/v-deo.html

please watch the previous video to understand this video

@@bluewat hi i am able to display but touch is not working

There is no a pin dedicated to power measuring. PVD directly monitors VDD through an internal voltage reference: see description and diagram shown on minute 12:00 (ua-cam.com/video/AHBGlCDGqhE/v-deo.html).

Dziękuję Rafał 😯😉

@@mikejones-vd3fg 👍. Have fun!

Sembra quasi una forma di stalking

I agree: ST fonts shown on ILI9488, can appear too small for some needs. In Displ_WChar I added "size" parameter that, if set to 2, "doubles" character width. It should be easy also to double height. Be aware: You can see that there is no a buffer swap inside the WChar function. It means that you have to set buffer size able to store at least 1 char. eg. Font24 is 24x17 pixels per character. In RGB565 it is 24x17x2 = 816 bytes per character, RGB666 needs 1.2k. Doubling with buffer must be > than 2.4k. Doubling width and height it must be >4.8k. Maybe problem you had was just a buffer overriding.

CubeIDE and CubeProgrammer use THE SAME modules so you should easily detect the error: your EL is good (!), that's just a CubeIDE project problem. My 2 cents is: (first hypotesis) did you unselect "verify" in CubeIDE EL properties? If yes I see only (second one) a linker script problem. But it is not easy from here helping you much more than this.

@@bluewat Thanks for your help!!! Yes the verify is unselected. i'm using ST-Link-V3Set.

​@@bluewatOk after checking all, I could fix it. And now is working.

Well done!

i have the same problem is the touch working for u?

@@sharankuppuswamy8951 yes the touch is working properly, what's your problem exactly? It doesn't show anything sometimes???

@@Muhammad-qf9vl touch is not working

@@sharankuppuswamy8951 have you checked the communication between the uC and the touch? Or did you check the wiring?? Also , maybe it's about the calibration... Watch the video about calibration and set the parameters, it should work... At least it worked for me:)

I'm really glad to have been helpful! Let me know about your robot.

The "Definitive Debuging Method"? That's too much. I can try to show how I move debugging my projects... 1) HW is ok? I usually setup a simple project (a non-TouchGFX project) allowing me to verify that display, touch sensor, flash (if any), other devices are working. That usually running simple test functions you see in other videos here or, anyway, simple basic tests of the devices involved. I keep this project available to run it, testing the system, anytime I have a doubt about it. 2) Is TouchGFX running? a second simple project (maybe just involving a button and a counter as you see here) to see if TouchGFX is correctly set. I use that to check all the configuration (CubeMX and .h configurations) in case of doubt. ...or restarting from here... sometime. 3) So, it is the current project! If the two above tests are working, the problem (because if I'm debugging means I have a problem) is just inside the project I'm developing. Keep in mind that there is ALWAYS a bug somewhere. If it seems that your (my) project is working, it's only because the bug hasn't shown up yet! .

@@bluewat Thank you... but I'm askiing for a simple procedura .... The idea is .... where to set 4 or 5 breakpoints ...

Thank you

Glad to see that you like it!

I 'll take a look

Thank you

I had te same problem... I fix it so : #define ILI9488_V1 in z_displ_ILI9XXX.h...good luck..

Grazie Angelo! lieto di sapere che i miei video vengono apprezzati.

Thank you

Let me know the result of your tests!

Glad to see that you find useful my work.

Thank you for your contribution, Amarobarbosa.

what's the 3.5 resolution? As per what I see on datasheet, 9341 handles 240x320 displays: library should work without any changes. Anyway, if display has a different resolution, you should just set its size on z_displ_ILI9XXX.h around line 175

Hi Charumathi, I do not have an ILI9846 display, neither I'm able to find its datasheets herearound, online. Maybe you are referring to ILI9486? If so, I do not have even displays driven by ILI9486 but, after a quick look to its datasheet, if I'm not mistaken, differences mainly are in the parallel interface. On the SPI port ILI9486 should work as an ILI9488 on RGB565. Try to set library to ILI9488_V2. Let me know.

Not for RTOSs

VL53L1X I2C address is 0x52 and you can program it. Note that the programmed address is lost on power loss, so, if you have multiple devices connected , turning the system on all device have the same address. You have yo use XSHUT to select them individually on startup, at least to set their own address. (that means that in a multiple-device project all devices must be xshut-connected: being so I let device with their original address and select them (even on runtime) through XSHUT)

Glad to see you like my videos, Tiwag

Hi, the easiest part of the answer is that you can be sure that there is something wrong. Not easy is to say what is wrong. That could be also just the structure in Dev_Inf.c. Maybe just a too long string as device name?

In my case, an additional section was created in RAM before the loader functions. STM32G0B0CET6

Hi, in the playlist you can find a video about the usage of an external video in a TouchGFX project

I'm not sure, but it shouldn't be difficult to adapt the library to another type of communication. As you can see, Loader_Src.c makes calls to a library handling SPI communication through simple functions: Flash_Read(), Write(), Clear(),...

I agree on that, I don’t see why it should not work. I will try!

Keep in mind that you must avoid interrupts (so, DMA also) as they are disabled exiting the external loader functions

👍Keep on working

Now it is public. Thank you!

I'm sorry, I'm not using ESP32 since a long time.

Hi. I am also facing the same issue. External loader cant write more than 128kb data. Have you solved problem? Or anyone have solution?

I had a project having about 1MB loaded onto a flash without problems, I don't remember if sometimes I had something needing more than 1M. I checked that project right now: an F103@64MHz (a bluepill) with a w25q64 on spi1@2Mbps. If you like a help, you should provide at least some information: what's your system? How you say you can't write data over 128k? You mean you get an error? what's the error? ...

I guess you have an error READING/VERIFYING flash (via CubeProgrammer). If so your EL is ok, flash is (hopefully, if you haven't a mistake somewhere) correctly programmed, it is a limit of CubeProgrammer. Try to disable verify option (even on CubeIde), and see what happens. I'm close to publish a new video explaining (among other topics) even this case

That was an F411. STM32F1 doesn't support TouchGFX. You can find a list of uC that can run that graphic system here: support.touchgfx.com/docs/development/hardware-selection/hardware-components/hardware-selection-mcu

There is an Application Note from ST about EEPROM emulation on the internal uC flash memory: www.st.com/resource/en/application_note/an3969-eeprom-emulation-in-stm32f40xstm32f41x-microcontrollers-stmicroelectronics.pdf Keep in mind that emulating EEPROM on a flash memory should only be done under certain circumstances. Generally, it's not a good idea to use a flash memory as if it were an EEPROM. Perhaps the introduction of this video can help understanding differences between Flash and EEprom: ua-cam.com/video/KlG2doCkREM/v-deo.html

Thank you.

Oh you are right! I see only now that on 43:30 I said I added the capacitor to CC1. That's a slip. I added it to T1, as said few seconds before. As per USB-C standard, only a 5.1 resistor can be (should be) added to CC1 and CC2. After this video, using this project as a power stage for a board, I also changed inductor with another one having the same inductance value and having a lower RDC . I adopted C5339514 (25mohm RDC) if you refer to Jlcpcb catalog. The c285742 inductor I used here (87mohm) tends to raise in temperature sinking 2A or more. Nothing critical, but trying to have a long live board... My intent in this video was (1) to show the work I made in understanding USB-C PD documentation to use USB port as a power supply without using a USB-C controller chip and (2) to update information about using Jlcpcb services I shown time ago in a video on Arduino. That's why I haven't published (on GitHub) an updated design of the board.

Maybe, it is a calibration problem? ua-cam.com/video/020k2vbLAPc/v-deo.html github.com/maudeve-it/ILI9XXX-XPT2046-STM32/tree/main/4-CALIBRATION

@@bluewat I calibrated it in the near past. Maybe, a problem is with SPI settings? - reading touch data is to slow or to fast?

Actually I remember I've seen that there is a range of touch SPI speed where the sensor still works and replies to queries but giving low precision results. Try to drastically slow down touch spi speed and see the results. Keep in mind that touch sensor needs a communication of just a couple of bytes: you will see that a slow touch-SPI port doesn't affects the global TouchGFX project performance. Let me know

That's strange. Writing cycle starts as soon as /CS is driven high and autonomously resets busy bit at the end.

@@bluewat you are right but in the function Flash_WaitForWritingComplete(); ---> infinite loop at while (buffer[0]& SR1_BIT_BUSY ); . This buffer never receive S1 register value and causes a loop infinity

@@bluewat is library support w25q64?

is library support w25q64?

@@bluewat now working... my chip was faulty i replaced with new one and program is working