STM32 MCUs Embedded software

  Hi everyone, I’ve been struggling with this for several days and could really use some guidance. I’m working on a project with the Nucleo-H743ZI, using LwIP’s built-in MQTT client over Ethernet. The basic MQTT setup on port 1883 works perfectly — I can publish and subscribe without any issues. But now I need to secure the connection with TLS (MQTTS), and I’m stuck. I haven’t been able to find a clear example or step-by-step guide to make this work. Here’s what I’m trying to figure out: How to run MQTT over TLS using LwIP + mbedTLS. How to load and manage certificates (CA/root, and possibly client certs). How to wrap or replace the default TCP socket used by the LwIP MQTT client with a secure TLS layer. If anyone has done this before, or has a working example project or even just the key integration steps, I’d really appreciate your help. Thanks in advance

I use the STM32CubeIDE Version: 1.18.1 and the stm32g030k8.I have a hardware that uses a firmware. This firmware correctly returns the values of the ADC. The ADC is linked to a DMA. Channel 4 == 2545 (this is a voltage divisor) 12.5 voltsChannel 10 == 1 (this is connected to a current shunt resistor) No current, then almost 0. Channel Temperature == 929Channel Vreference == 1506   I develop a new firmware with the very same hardware (not just same design but same device). I apply debug optimization. I use ADC1 with the DMA, first I use the ioc file to set up the ADC and then I copy the MX_ADC1_Initfrom the other stm32cubeide project. Yet It does not return the same value from the external pins. Why is that?  Thanks for all.      

I'm trying to build a STM32U5 (THREADX_ARCH = cortex_m33) project that use FileX (https://github.com/STMicroelectronics/stm32-mw-filex) with CMake.I encounter an error that say "FileX/ports/cortex_m33/gnu" is not an existing directory.Why Cortex m33 is not supported by FileX ?

when running the QSPI_Executeinplace example code on the stm32f769i_eval board,the program enters HardFault_handler() after executing GpioToggle().i just remove the QSPI_DummyCyclesCfg(),in GpioToggle() the GPIO Not Taking effect,so How can I solve this problem?      thanks

Hi all, I'm using ThreadX with NetX Duo and running a Modbus TCP server on an STM32 platform. The system works fine under normal conditions, but I encounter a hang when the following sequence occurs: A Modbus TCP client is actively communicating with the board. The client is closed or a Telnet session is opened. The client reconnects — at this point, the system hangs. From the ThreadX-aware debugger: Main Ip instance and Packet Monitor Thread are suspended, waiting on a mutex. The mutex is owned by my App Modbus TCP Thread, even though I don’t explicitly use any mutexes. It seems like the NetX internal mutex is being held too long. Can I trace or monitor the internal NetX mutex usage to better understand where the contention is happening?

Hello All,   Need your support regarding flushing of RX FIFO buffer in STM32H747.I am using a barcode reader which is connected to USB HS port . And i want to clear the received data from the buffer at particular instance. And after that instance want to read the buffer again. Can anyone please suggest how can i do it. I have tried multiple method ,and i have mentioned all of them below:- 1. Using Inbuild function in STM32 USB Library file. USB_FlushRxFifo(hhcd_USB_OTG_HS.Instance);   2.Masking the RXFLVLM bit from GINTMSK Register. hhcd_USB_OTG_HS.Instance->GINTMSK &=~ ((0x1UL << (4U)));   3.Disabling the GINTMSK bit in GAHBCFG Register hhcd_USB_OTG_HS.Instance->GAHBCFG &=~ ((0x1UL << (0U)));   4. Setting Core soft reset bit in GRSTCTL hhcd_USB_OTG_HS.Instance->GRSTCTL |= USB_OTG_GRSTCTL_CSRST;   5. Deiniting & initing the USB device using inbuild functions from STM32. USBH_DeInit(&hUSBHostHS); USBH_Init(&h

Hi,I am testing an application between multiple MCUs, I have an H573, an F756 and an F439.I am testing multiple things at a time, so the code generation is helping me.In the H573 I have the flag for using the webserver without FileX, and it will define "NX_WEB_HTTP_NO_FILEX"On the F756 I don't have that. I can define it manually in the file "nx_web_http_server.h", but at code regeneration it will disappear.I don't know where to define it manually, everywhere I define it is deleted at code generation or is not defined in time.Thanks for any helpcoso2 

Hi Team,Currently I am working with STM32U575ZIT6 SOC and I am using Timer for counting external clock pulses,Below is my requirement :1. I have a requirement to count external pulses using timer for 6 channels, All those channels are connected to spin rate of a physical system; a wind-turbine RPM or something similar.2. The maximum frequency of the incoming signal To ensure the timer can handle it without missing counts is up to 100 kHz and probably closer to a max of 3600.3. The counter will roll over and is expected to reset on reboot.4. Also noticed that a single timer can support only one counting element to count its pulses, Can anyone suggest How I can add a support to count 6 external pulses coming from wind turbines ? Thanks and Regards,Ram Jan

Hi all, I try to use the HAL lib in Zephyr. Anyone knows, why I cant select the hal_st modules/drivers? Best regards Lukas

Hello ,I am working with nucleo-H743Vitx. I have created Shtml pages which has different tags. When I executed the code, it is working fine.I want to display some more dynamic data . So, I added few more tags to the shtml file and I included these tags in ssi handler function as well as in ssi tags like I did for working ssi tags. When I run this code, I can see my project is working fine but the webpage is not loading and it is giving error but my other hardware is working fine. I checked the console window but there is no error related to shtml file.Then, I commented the another tag  which was working and executed the code, Then the code with new tag  is working fine. Again I tried with all the tag and the webpage is not displayed. I found all the tags are working fine indepenedtly but not together.I checked the tag length and it is fine . I think the issue is with number of tags or memory. Do we have any limitations on adding number of tags? if yes, Could yo

Hello everyone! BackgroundWe are working on STM32G474VET6 based custom board for some specific motor control application. Project demands to have a custom bootloader on FD-CAN with 500kbit/s arbitration baudrate and data rate with 2Mbit/s with BRS option turned ON. Bootloader is done and working fine.Board has a support for three CAN buses, therefore using FDCAN1, FDCAN2 and FDCAN3 periphery. All CAN buses has the same bus configuration and are not inter-connected. Everything works perfectly with running only application on uC. As soon as we add bootloader that runs on FDCAN1, that same application (adjusted to be bootloadable ofcourse - offseted in flash, vector table adjusted) that was running all three CANs, now hangs if any message is received/transmitted on eihter FDCAN2 or FDCAN3. If there is no traffic on FDCAN2 or FDCAN3 everything is working as expected. Beside CAN periphery we're running also ADC1&2 with injected and regular channels, triggered by cent

Split from this thread, as it's a different question.Hi @ASEHST,the example projects are made for the NUCLEO-N657X0-Q board. I tried to adapt one of that for the STM32N6570-DK board but keep on getting stuck during the DHCP resolution. Checking the schematics I realised that the pin assignment is different. But even changing the pins does not fix my problem and I keep getting stuck in the following section of "app_nextduo.c". printf("Looking for DHCP server ..\n"); /* wait until an IP address is ready */ if(tx_semaphore_get(&DHCPSemaphore, TX_WAIT_FOREVER) != TX_SUCCESS) { /* USER CODE BEGIN DHCPSemaphore get error */ Error_Handler(); /* USER CODE END DHCPSemaphore get error */ } I checked the network traffic and can confirm that the DHCP server does not receive discover package. Do you have an idea what might the problem be and if there are any known issues regarding ethernet/the tcp stack?Kind regards

Hi everyone,I'm currently working on a custom board featuring an STM32H747, and my goal is to implement a firmware update mechanism via the I2C interface. Since I'm a beginner when it comes to bootloaders (especially with dual-core microcontrollers), I'm reaching out for some help.I don't have physical access or the ability to control the BOOT0/BOOT1 pins of the MCU. This means I must rely exclusively on a custom bootloader residing in the internal Flash. This bootloader will need to handle receiving the new firmware via I2C and subsequently programming the Flash.Can anyone help me out? I'm not sure where to start.Thanks a lot for your time and support!

Hello ST Community,I’m currently developing a project on the STM32U5G9J-DK2 board and working to implement a FileX + LevelX-based file system on the external NOR flash via XSPI.I found the example project Fx_NoR_Write_Read_File in the STM32Cube firmware package, and while it’s functionally suitable, it does not include the .ioc file. This has made it challenging to fully understand and replicate the setup, especially regarding the XSPI (OctoSPI) peripheral configuration and the use of the lx_stm32_xspi_driver_glue.c for flash communication.Why This Is Important:The CubeIDE (latest version) does not provide this driver (lx_stm32_xspi_driver_glue.c) and configuration by default, making manual recreation risky and error-prone.Without the .ioc or clear documentation, it’s difficult to ensure correct XSPI timings, GPIO assignments, and memory-mapped mode settings.My current development progress is blocked until I can correctly mirror the peripheral and middleware configuration from this exa

Hello,I am trying the sleep example for the development board B-U585I-IOT02A, called PWR_SLEEP. I decided to add a channel from ADC1 module and I want to test, but I do not see any header file on the project. I enabled ADC1, channel 1 on pin PC0 in the .ioc file. Is it allowed to modify and add modules there?How do i see the complete code? Furthermore, I see the file from hal_adc contains #ifdef HAL_ADC_MODULE_ENABLED and currently the whole file is in grey as if the ADC module has not been enabled. I am not sure it should be like that.Regards,Gergana

Hi everyone,I faced a frustrating issue with lwip_connect() hanging forever when trying to use an STM32 as a TCP client (FreeRTOS V1 + LWIP stack). 🧠 Problem SummaryCalling lwip_connect() inside my STM32 TCP client code caused the system to hang indefinitely.It was getting stuck at this internal LWIP line: sys_arch_sem_wait(LWIP_API_MSG_SEM(msg), 0);This only happens when the TCP/IP stack is waiting for the connection to complete, and never gets a response.🧪 Observations- I could ping the server (PC) from STM32 successfully.- My STM32 TCP server mode worked perfectly.- But when STM32 tried to connect to Hercules TCP server on port 503 — it hung forever.- LWIP debug logs showed that:- tcp_connect() was called- But the TCP state machine never moved forward- The stack kept calling tcp_slowtmr: polling application LWIP Debug Logslwip_socket(PF_INET, SOCK_STREAM, 0) = 0 lwip_connect(0, addr=192.168.1.1 port=503) tcp_connect to port 503 tcp_slowtmr: processing active pcb tcp_slow

Hello,Could someone please help me?After writing the following code to my custom board using the STM32H725, I can no longer connect to CubeIDE or CubeProgrammer.I believe the issue is due to accidentally enabling the MPU in CubeMX.I would like to recover the MCU by initializing the flash or using another method.Thank you in advance.#include "main.h" #include "stm32h7xx.h" #include "stm32h7xx_hal_flash.h" #include "stm32h7xx_hal_flash_ex.h" #include "stm32_hal_legacy.h" void SystemClock_Config(void); static void MPU_Config(void); static void MX_GPIO_Init(void); int main(void) { MPU_Config(); HAL_Init(); SystemClock_Config(); MX_GPIO_Init(); while (1) { /* USER CODE END WHILE */ /* USER CODE BEGIN 3 */ } } void SystemClock_Config(void) { RCC_OscInitTypeDef RCC_OscInitStruct = {0}; RCC_ClkInitTypeDef RCC_ClkInitStruct = {0}; /** Supply configuration update enable */ HAL_PWREx_ConfigSupply(PWR_DIRECT_SMPS_SUPPLY); /** Configure the main internal regulator output volt

I have a u5 dev board running a usbx msc reference design , which works fine. I'm trying to disable the usbpd side of things within the reference example. I've disabled the usbpd midlewear and peirphal port for the 2 pd pins aswell as disabling the adc input.  The design compile fine, after removipng the bsp however the usb msc doesn't work. Is there something I'm missing? There was one post mentioning about removing threadx messageing ques, I did try tweaking some of the threadx code but couldn't get anything working  Any ideas on how to fully remove the pd code as I'm just using usb+and usb- ldata pins.

Hello, I am making a project where i'm using usart2 as a virtual com port and where i'm making a CLI based HMI using teraterm , so there's a lot of diplay and hal_uart_transmit functions , it all worked fine until I added a DMA in the uart1 in order to read gps data.What happened is that the HMI menu doesn't work as it used to but when i removed the gps reading data part from dma receive to idle it worked like it used to.Someone told me that using DMA in receive mode like that + Hal_Uart_Transmit causes a block.Since HAl_uart_transmit is a blocking function, it will cause problems for dma receive ,And among the problems is that my functions in HMI don't work anymore like they used to, but at the same time i can implement dma with adc and it works fine, the problem is in the uart Can anyone please explain to me the logic behind it? The issue reminded me of the Hal_delay() block when used inside an interrupt.You'll find in the first picture how the normal behaviour should be an

The author of this how to get CANopenNodeSTM32 working video is very good (The length of the video is 3:31:33; The part that is pertinent to this post 57:05 - 1:18:22).  Unfortunately, this video targets FDCAN (but I don't have this available; my chip just supports plain CAN) and this video also targets bare metal (but I don't have this available; FreeRTOS is what I am using).  I tried to follow all of the author's steps but was unable to.  When I ran I got a HardFault with the CAN node never even coming online.Executing this line is what causes the Hard faultcanopen_app_init(&canOpenNodeSTM32); From the instructions in the video, there are numerous steps that are different between FDCAN and CAN and I don't know what the equivalent (plain) CAN steps are which I believe may be the root of the problem.      Under the "Parameter Settings" tab"FDCAN1" and "FDCAN2" are in the video at 1:00:55. Mine doesn't have these.  Mine jus

I'm using LL_RCC_DeInit within a boot loader so that the clocks are in the reset state when the application is launched. However, LL_RCC_DeInit also clears the reset cause flags as its last action: /* Clear reset flags */ LL_RCC_ClearResetFlags();The application needs to use the reset cause, so I will need to either:Modify the LL_RCC_DeInit routine (not ideal, as there are no user code sections to retain the changes); orClone LL_RCC_DeInit and remove this part (also not ideal as I would need to keep on top of any LL HAL changes); orFind some other way of passing the reset cause into the application. I would rather not do this, as I want to keep the boot loader and application completely decoupled.Resetting the flags within LL_RCC_DeInit does not seem to be compatible with its stated behaviour ("Reset the RCC clock configuration to the default reset state"), as the flags could be seen as being part of that state.For now, I will use one of the TAMP->BKP?

Hi Mike, Thanks again for your previous support on the SPI interface.I have since moved to using the SDMMC interface instead of SPI for better performance, but I am still facing issues.   Current Issue Summary:I can successfully mount the 32GB SDHC FAT32 card and read the capacity using FatFs. However, f_write always fails. After debugging, I found that BSP_SD_WriteBlocks fails when writing to LBA 32768. I also verified this with direct HAL_SD_WriteBlocks, which fails at LBA 32768 but works on lower LBAs like 0 or 1. This suggests the problem is not in FatFs logic, but rather in high LBA (large address) block access stability.   My Questions:Is this a known limitation with STM32H7 SDMMC or HAL drivers when writing to higher LBA regions on large-capacity SD cards (like 32GB)? What is the commonly tested and verified capacity for STM32H7 SDMMC + FatFs reference designs? Do most STM32 applications use 4GB / 8GB / 16GB cards, or is 32GB commonly used and verified? Any additiona

Hi ST Community,I’m working on a JPEG encoding pipeline using the STM32N6570 Discovery Kit, and I’m running into an issue when storing both my input and output buffers in PSRAM.Setup:Both JPEG Encoder input (RGB_ImageAddress) and output (jpgBuffer) buffers are placed in PSRAM.I’m using RGB888 input format, the same as in the official JPEG example here: JPEG_EncodingFromOSPI_DMA Linker script (.ld) snippet:MEMORY{  AXISRAM1_S (xrw)      : ORIGIN = 0x34000400, LENGTH = 1023K  PSRAM (xrw)           : ORIGIN = 0x91000000, LENGTH = 16M}.psram_section (NOLOAD):{   . = ALIGN(32);   *(.psram_bss)   . = ALIGN(32);} >PSRAM Buffer declarations:__attribute__((section(".psram_bss")))__attribute__((aligned(32)))uint32_t jpgBuffer[160 * 160]; __attribute__((section(".psram_bss")))__attribute__((aligned(32)))uint32_t RGB_ImageAddress = {0}; // RGB888 image Problem:After running HAL_JPEG_Encode_D

HAL code for interrupt-driven SPI transmit attempts to fill a transmit queue, which uses up the 10 bytes I am transmitting.  But, a TX interrupt is also triggered which grabs an extra byte past the end of the buffer and decrements the count which wraps.  There is code to turn off TX interrupts in the handler, but it checks after transmitting a byte rather than before.  Has anyone else noticed this?

I have set up an I2S interface. The idea is to send it an initial HAL_I2S_Transmit_IT() request.In the ISR of SPI2 I'm sending the next HAL_I2S_Transmit_IT() and this way the transmit should be continously kept running.What happens is, that the SPI ISR is entered once. The HAL_I2S_Transmit_IT() issued within the ISR, doesn't fire the subsequent IR. Why?