Solved

I'm running into a pixel mismatch issue when trying to capture frames over DCMI from a [B-CAMS_OMV with an OV5640 sensor board attached. I've used CubeMX and the stm32-disco F746 development board to successfully collect RGB565 and JPEG image compression data in various formats and resolutions.Below is my CubeMX generated DCMI configuration:hdcmi.Instance = DCMI;hdcmi.Init.SynchroMode = DCMI_SYNCHRO_HARDWARE;hdcmi.Init.PCKPolarity = DCMI_PCKPOLARITY_RISING;hdcmi.Init.VSPolarity = DCMI_VSPOLARITY_HIGH;hdcmi.Init.HSPolarity = DCMI_HSPOLARITY_HIGH;hdcmi.Init.CaptureRate = DCMI_CR_ALTERNATE_4_FRAME;hdcmi.Init.ExtendedDataMode = DCMI_EXTEND_DATA_8B;hdcmi.Init.JPEGMode = DCMI_JPEG_ENABLE;hdcmi.Init.ByteSelectMode = DCMI_BSM_ALL;hdcmi.Init.ByteSelectStart = DCMI_OEBS_ODD;hdcmi.Init.LineSelectMode = DCMI_LSM_ALL;hdcmi.Init.LineSelectStart = DCMI_OELS_ODD; The OV5640 board is attached to an ST module, and I've written a driver heavily influenced by the [VISION-AI](https://www.st.com/en/emb

Facing issue with CAN baud rate change!1. The Master and slave devices are communicating at baud rate of 1Mbps.2. If I change the baud rate of Master device to any other baud rate(like, 250kbps, 500kbps,125kbps,50kbps) then master device contiguously restarting the device! Got to know that, it is restarting due to watch dog timer!3. To resolve the restart issue, added the timeout logic in the while loop where it is entering into intialize mode and normal mode.4. I gave CAN_TIMEOUT as 1000ms, If I add timeout logic, device is not communicating with slave and data is not exchanging due while timeout logic5. Tried with different logic by calling HAL_NVIC_SystemReset(); in the timeout logic. static UINT32 CanFunctin(void) { UINT32 tickstart = 0,tickstart1=0; // static UINT32 BaudRateChange=0; UINT32 error=0; // Exit Sleep mode and enter Initialize mode. CAN1->MCR &= ~CAN_MCR_SLEEP; CAN1->MCR |= CAN_MCR_INRQ; tickstart = FPGA_GET_FRT_UINT32; // Wait for Initialize m

Hi.I add the "x-cube-eeprom-v8-0-0" library to my project manually:I create folder Middlewares and copy sources the  folder as expected,I specify the paths in the project settings,I uncheck the "Delete previously generated files when not re-generated" on the Project Manager tab in Device Configuration Tool,but the folder is still deleted when regenerating after changing the any settings in ioc.What am I doing wrong?

I installed newer version of AutoDevKit studio from ST website and imported a project fromm samples.I tried to flash the generated elf but it is giving following error. 

Hello, I have the STM32N6570-DK kit. I can compile, upload and debug by following the instructions on the kit. I designed my own board but I can't debug. I'm reviewing all the possibilities and I noticed some problems. I couldn't find any explanation. https://community.st.com/t5/stm32-mcus/stm32n6-boot-rom-explained/ta-p/763648@B.Montanari  's  article says H(High) VDDIO. However, in the DK schematics and pcb, the left side is vddio and the right side is gnd. Also, the Boot1 pin is pulled to the right to load software on the DK. So Boot0 becomes important. What's the problem? Are the documents wrong? Also, on the board I made, I can load software when boot1 is connected to vddio. But on the DK, it allows me to load software when boot1 is GND. What's the explanation?    

Hi, I am trying to use a Nucleo-F446re as master and another one as slave. But the master is probably being timed out (not sure). Can you guys please have a look at my code and configuration. Thanks in advance <3  My Wiring:1. I connected PB8 & PB9 of both master and slave via direct wire as I used then as SCL and SDA2. Connected GND of master and slave My code:Master code:   /* USER CODE BEGIN Header */ /** ****************************************************************************** * @file : main.c * @brief : Main program body ****************************************************************************** * @attention * * Copyright (c) 2025 STMicroelectronics. * All rights reserved. * * This software is licensed under terms that can be found in the LICENSE file * in the root directory of this software component. * If no LICENSE file comes with this software, it is provided AS-IS. * ***************************************************

Hello,I'm interested in understanding how much an NPU can accelerate a model. I reviewed the documentation for Neural Art and found that 38 TFLite operators are supported by the hardware:ABS, ADD, AVERAGE_POOL_2D, BATCH_MATMUL, CAST, CEIL, CONCATENATION, CONV_2D, DEPTHWISE_CONV_2D, EQUAL, EXPAND_DIMS, FULLY_CONNECTED, HARD_SWISH, LEAKY_RELU, LOGICAL_AND, LOGICAL_NOT, LOGICAL_OR, LOGISTIC, MAX_POOL_2D, MUL, PACK, PAD, PRELU, (RE)QUANTIZE, RELU, RELU6, RESHAPE, RESIZE_NEAREST_NEIGHBOR (with coordinate_transformation_mode=asymmetric and nearest_mode=floor), SPACE_TO_DEPTH (with same input/output quantization), SPLIT, SPLIT_V, STRIDED_SLICE, SQUEEZE, SUB, TANH, TRANSPOSE, TRANSPOSE_CONV, UNPACK.However, each operator has a different level of computational complexity. Therefore, the acceleration ratio provided by the NPU may vary depending on which operators are used. For example, a model that heavily uses CONV layers may benefit more from NPU acceleration than one that primarily uses ADD o

From the reference manue, I can find a term half-DMA, but this term is not well described, and I don't know what this half means, what is the difference bewteen a half-DMA and a normal DMA??? Is it just the size or the available target address is smaller??

Hi Team, We are facing issue in debugging like dissambly view is not showing correct flow even though we clean build the ws. code:  disassembly view:Tried methods:- changed workspace- clean and rebuild- changed HW- changed machineLet me know how it could be resolved ? 

I have just started to study STM N6 series MCU. I see a lot of functions named with ATON.Anyone can tell me what ATON stand for?thanks!

With new and fast N6 MCU comes new CubeMX package which is full of bugs and errors, I spending hours to find and solve such problems, now I cant trust cubeMX at all, every function must be checked after code generation !ADC rank must be 7, but cubeIDE generate 1: I spent a lot of time trying to enable SPI DMA, when I tried to choose DMA handle I got unknown errors in almost all enabled interfaces (I2C, SPI, UART... ), and in some way after some changes elsewhere, closing, opening .ioc file form cubeIDE, and after next try the error is gone and I finally enabled !I trying to enable LSE, but because of some error with frequency settings which I cant set, I cant enable ! There I cant find the way to enable DCMI DMA transfer, DMA handle for DCMI is not active, maybe I dont know some things there !?To get VDDCORE voltage and avoid errors every time after cubeMX generate code I added additional variable ADC_CHANNEL_VCORE, because cube generate ADC_CHANNEL_VCORE, but the correct v

Hi, I want to get data from an ADC connected externally through SPI DMA whenever data ready pin is set. I want to do it without any interrupt handler, since our firmware have already loaded with other stuffs. I have tried the synchronisation of SPI DMA with GPIO EXTI trigger. It works when I configure SPI DMA in circular mode only.Because synchronisation doesn't triggers the SPI DMA it just syncs the process. I can't find any options to trigger the SPI DMA with timer or any event.How to trigger SPI DMA through GPIO EXTI event? Is that possible in STM32G4 series?. Thank you!     

Hi ST community, Recently i am working on a project, home automation, and i would like to use STM32F429VIT6 MCU and 3.5 inch 320RGB x 480 LCD display on the project. The LCD display controller is ST7365P from Sitronix, with Frame Memory Size: 320 x 480 x 18-bit = 2,764,800 bits. My question is, is it possible to interface STM32F429VIT6 MCU with the TFT-LCD display through LTDC without using an external RAM just by using the On-chip Frame Memory? Attached are the TFT-LCD display and the ST7365 controller datasheet. Regards,

Hello everyone,I'm using Nucleo-wl55JC boards with the PingPong example code (which is working). I saw in the datasheet that we could have 2 different power supplies LDO and SMPS, I also saw that the SMPS mode was more economical than LDO, that's why I'd like to use it.I'd like to know if it's possible to activate this mode in the PingPong code? Is it already active?Best regardsKYZ.

Hello,I have been trying to flash firmware from one MCU to another MCU, and everything until actually jumping to the memory address where the firmware has been flashed to is working so far.Here's what I do in my code:I have the "Application" project, which is basically the firmware to be flashed. I build this project in STM32CUBEIDE, and then I take the application.bin file and convert it to a C array using some software. Then, MCU-1 uses that C array and flashes it to MCU2. MCU2 has 3 applications slots in its flash memory, and it flashes the bytes it receives to application slot 2 (0x08080000). Once it finishes, it jumps to the starting address of the second application slot using the following function:void BOOT_JumpToAddress(uint32_t address) { printf("Jumping to 0x%08X\r\n", address); // Function pointer to reset handler void (*app_reset_handler)(void) = (void *)(*((volatile uint32_t *)(address + 4U))); // Disable RCC, HAL, reset SYSTICK, and remap vector table offset HAL_R

Hello,I have 2 STM32 Nucleo F746ZG boards, and I am trying to use one board (MCU 1) to program the second board (MCU 2) through its default bootloader through code. To do so, I used the following command in the MCU2 to start its default bootloader:#define DEFAULT_BOOTLOADER_ADDRESS 0x1FF00000 BOOT_JumpToAddress(DEFAULT_BOOTLOADER_ADDRESS); void BOOT_JumpToAddress(uint32_t address) { printf("Jumping to 0x%08X\r\n", address); // Function pointer to reset handler void (*app_reset_handler)(void) = (void *)(*((volatile uint32_t *)(address + 4U))); // Disable RCC, HAL, reset SYSTICK, and remap vector table offset HAL_RCC_DeInit(); HAL_DeInit(); SysTick->CTRL = 0; SysTick->LOAD = 0; SysTick->VAL = 0; SCB->VTOR = address; uint32_t msp_value = *((volatile uint32_t *)address); __set_MSP(msp_value); // Jump into application app_reset_handler(); } Then, on the programmer, I send the start byte 0x7F. When I receive ACK, I send the bytes for the "Go to applicat

Hello,we are using a STM32F429.The bootloader is located in the first 64kB. The firmware in the remaining sectors.It erases the remaining sectors of the flash and writes the new firmware.Now, 3.6.5 Read-while-write (RWW) of the reference manual states that "This feature allows to perform a read operation from one bank while an erase or program operation is performed to the other bank."Does this mean that reads from the same bank as a sector is erased is not allowed?If so, how should the bootloader update the firmware in Bank 0, where the bootloader sits itself?Many thanks in advance for your answer! 

Greetings,We've found several STM32 MCUs with what appear to be poorly executed laser engravings. Is there anyone I can contact to get an official opinion on these chips?We are not the firm that designed and built the boards. We are reviewing other firm's boards and do not have access to supplier or similar information.Thank you.-David

Hello,i'm trying to implement firmware upgrade in my Windows application via USART for stm32f103 following AN3155. My issue is with the Write Command. More specifically:- if i send 1 byte of user data at time it works, it's just slow cause i have to introduce 1ms delay or shorter between each write (or it stucks and i get no message from the bootloader)://PSEUDO CODE: for (i=0 to N_byte){ USB_Out_Buffer[0] = FirmwareBytes[current_chunk_num, i]; USB_write(1); //sends 1 byte delayms(1); }- if i send the whole 256 byte block (or lower) it stucks after sending the checksum and i get no message from the bootloader://PSEUDO CODE: for (i=0 to N_byte){ USB_Out_Buffer[i] = FirmwareBytes[current_chunk_num, i]; } USB_write(N_byte+1); //sends 256 bytesI'm using USART on PA9, PA10. i tried everything, reduce baudrate, adding delay etc. Any idea what could be the issue?

Bonjour ! Je travaille avec une carte NUCLEO L496ZG. Maintenant, j'aimerai intégré le STM32L496ZGT6U dans mon projet mais j'ai besoin du schéma de la carte NUCLEO pour étudier certains points. Où puis-je trouver ce schéma ?Merci d'avance !Google translates:Good morning ! I'm working with a NUCLEO L496ZG board. Now I'd like to integrate the STM32L496ZGT6U into my project, but I need the NUCLEO board schematic to study some points. Where can I find this schematic?Thanks in advance !

Hello,I'm using NucleoU575 with CubeMx and I want to make a file system on my SD card.I want to use FreeRtos but there are no Fatfs in CubeMx and the FileX cannot work with FreeRtos (image). Is there anyway i can use Fatfs with STM32U5 ?Thank you. 

I am looking if there are any usage and layout examples for the STWLC38JRM as a 15W receiver. Thank you so much!

Hello,I am working with stm32 h757 riverdi kit. I get this error in stm32cubeidearm-none-eabi-g++ -o "Riverdi_70STM32H7_CM7.elf" @"objects.list" -l:libtouchgfx-float-abi-hard.a -mcpu=cortex-m7 -T"C:\TouchGFXProjects\AKUSTEK_V3\STM32CubeIDE\CM7\STM32H747XIHX_FLASH.ld" --specs=nosys.specs -Wl,-Map="Riverdi_70STM32H7_CM7.map" -Wl,--gc-sections -static -L../../../CM7/Middlewares/ST/touchgfx/lib/core/cortex_m7/gcc --specs=nano.specs -mfpu=fpv5-d16 -mfloat-abi=hard -mthumb -Wl,--start-group -lc -lm -lstdc++ -lsupc++ -Wl,--end-group C:/ST/STM32CubeIDE_1.18.0/STM32CubeIDE/plugins/com.st.stm32cube.ide.mcu.externaltools.gnu-tools-for-stm32.13.3.rel1.win32_1.0.0.202411081344/tools/bin/../lib/gcc/arm-none-eabi/13.3.1/../../../../arm-none-eabi/bin/ld.exe: Riverdi_70STM32H7_CM7.elf section `.text' will not fit in region `FLASH' C:/ST/STM32CubeIDE_1.18.0/STM32CubeIDE/plugins/com.st.stm32cube.ide.mcu.externaltools.gnu-tools-for-stm32.13.3.rel1.win32_1.0.0.202411081344/tools/bin/../lib/gcc/arm-none-eab

I'm trying to download SR6G7 elf code into internal Flash. So far this MCU seems to be not directly supported from the StellarStudio (?).Is there a way of programming my board using StellarLink from the command line? I found a working solution for other chip (like SR5E1):C:\StellarStudio-2.0\openocd\bin\openocd.exe -d0 -s C:\StellarStudio-2.0\openocd\scripts -f board\sr5e1_evbe7000p.cfg -c "program blinkled_portE13.elf reset exit"However, there are missing SR6G7 configuration files in the latest StellarStudio_rev4.1 and StellarLink releases. The one defining target board I could prepare myself, yet the other describing SR6G7 chip and FLASH_ERASE algorithms are not available.Can someone worked already with that chip and could share the proper config files (chip and algo)?  

Hi,The hardware seems to support 16 bits activations, but this seems to not be usable according to the ST Edge AI documentation.We have a model where at a particular tensor between a Conv and Sigmoid, the precision loss due to the quantization cause non-negligible accuracy loss of the output of the model. The output of the Conv layer is between -90 and 2 mostly, then it goes to Sigmoid. But the low values cause the quantized output to loss too much accuracy, while being needed for an accurate Sigmoid output (which drives audio related ratio where accuracy is important).I've tried to use these parameters with ONNX quantization, but ST Edge AI fails:conf = StaticQuantConfig( calibration_data_reader=dr, quant_format=QuantFormat.QDQ, calibrate_method=CalibrationMethod.MinMax, activation_type=QuantType.QInt16, weight_type=QuantType.QInt8, per_channel=True)The error message is:NOT IMPLEMENTED: Unexpected type for constant input of Dequantize layer (SIGNED, 16 bit, C Size: 16 bits Scale