STM32 MCUs

 I’m trying to get a USB Full-Speed Host Mass Storage example running on the STM32H753I-EVAL2 using ThreadX, USBX, and FileX.I’ve generated the base project from CubeMX, but I’m running into a few issues:VBUS Enable/Detection: VBUS is not being enabled correctly via the MFX (IO expander).Even when manually configuring, the stack doesn’t seem to handle VBUS properly in host mode.Memory Pool & Stack Sizes:Default memory pool size and system/stack sizes seem insufficient, causing failures with Mass Storage class. tried increasing but did not succeed and ended up in memfaultsI’m looking for:A working USB FS Host Mass Storage example for STM32H753I-EVAL2 with ThreadX, USBX, and FileX.Guidance on properly enabling VBUS via MFX or directly on the board.Recommended USBX memory pool and stack sizes for Mass Storage class.BSP package for STM32H753I-EVAL2Any example projects, tips, or guidance would be greatly appreciated.Thanks in advance!

Hi everyone, I'm struggling with a critical synchronization issue between TouchGFX Designer (4.2x) and IAR Embedded Workbench (9.60.2) on an STM32H563 project.Whenever I modify the UI (e.g., deleting a Button or adding a ToggleButton), TouchGFX correctly generates the new .cpp and .hpp files in the "generated/" folder.However, upon flashing the device, the UI is completely corrupted:- Images are replaced by random noise or other assets.- Texts disappear- Moving a widget "off-screen" instead of deleting it works, but adding new ones breaks everything again. I'm using Nucleo-H563ZI with Riverdi screen.I tried these steps without success:- Delete IAR folders: Obj, BrowseInfo and regenerated- Deleted TouchGFX/generated and TouchGFX/build and regenerated- Full chip erase (Internal + External Flash) via STM32CubeProgrammer- Forced "Rebuild All" in IAR Any idea on how to fix this issue?

Hi all.  I have been witnessing a rapidly shrinking amount of free space on my C drive.  On a regular drive clean up of non-essential data, I can generally remove 1 Gigabyte or so of temporary files.  This however, does not account for the rapid depletion of available free space. On further investigation, I looked into the 'Other' (Manage other large folders) section of the 'System -> Storage' properties of my Windows 11 C drive.  I was a little gob-smacked to discover that - between them - the ST related folders: stm32cubeide, stm32cubemx, STM32Cube and ST contain over 100Gb of data.Obviously allowing for software packs and the various STM applications in daily use, I found a considerable amount of data was held in several .jxbrowser folders across more than one location.  For example, the .jxbrowser-7.38.2 folder inside the stm32cubemx directory - alone - is at a whopping 17.5Gb.  I don't see any way of cleaning them or whether or not it is safe

I see in STM32CubeMX that a STM32G411 exists but it has the status "restricted". I'm curious why this type is not widely available. I'm currently using the STM32G431 in an application, mainly because of its raw clock speed of 170MHz. I don't actually need its analog functions but the flash size might become too tight. In that case, my only option would be to switch to the G491, which is more expensive and contains even more unneeded peripherals... So... why is the G411 not available to the public?

Hi Team,I'm recently using the STM32U5G9ZJT6Q part in one of my design. I couldn't find the package mechanical information in Datasheet. For now, I'm using the LQFP-144 form the STM32H573ZIT3Q part number. Will it be okay? Can you please help to verify it? Attaching the below Screenshots,Screenshot from the STM32U5 Datasheet. Not having the LQFP-144 details.Screenshot from the STM32H5 Datasheet. Having the LQFP-144 details

How to configure parallel compilation in cubeide, that is, "j-4" "j-8" and I encountered an error in the Windows 87 code, how to solve it, I changed the path and still couldn't do it··········································································································································00:10:38 **** Build of configuration Debug for project fishboat_2 ****make -j4 allarm-none-eabi-gcc "../Middlewares/Third_Party/FreeRTOS/Source/portable/MemMang/heap_4.c" -mcpu=cortex-m3 -std=gnu11 -g3 -DDEBUG -DUSE_HAL_DRIVER -DSTM32F103xB -c -I"D:/studystudystudy/Cube_32F1/fishboat_2/Middlewares/Third_Party/FreeRTOS/Source/portable/GCC/ARM_CM3" -I"D:/studystudystudy/Cube_32F1/fishboat_2/Middlewares/Third_Party/FreeRTOS/Source/CMSIS_RTOS_V2" -I"D:/studystudystudy/Cube_32F1/fishboat_2/Drivers/CMSIS/Include" -I"D:/studystudystudy/Cube_32F1/fishboat_2/Drivers/CMSIS/Device/ST/STM32F1xx/Include" -I"D:/studystudystudy/Cube_32F1/fishboat_2/Drivers/STM32F1xx_HAL_Driver/Inc" -I"D

I am running CubeMX in a WIndows on Arm VM on a MacMini, which is connected to a KVM. When I switch the KVM to another machine and come back, all of the windows have changed size (not unexpected). They all revert to their original sizes and positions, but the window for CubeMX does not redraw/scale it's content.For example, I see this before I switch away:And this when I switch back (note the double menu):Dragging the edge of the window to cause a (very small) change in size restores the expected view.CubeIDE and TouchGFX do not show this behaviour.

Greetings,I am in the process of integrating the OpenBootloader Middleware application onto a STM32C011 device.I got the code from the STM32Cube/Repository/STM32Cube_FW_C0_V1.4.0/Middlewares directoryI have the code in an STM32CubeIDE project, and it is not compiling?For example:The C file, Interfaces/Patterns/OPTION BYTES/optionbytes_interface.c, has code to write to aFLASH control register.                  WRITE_REG(FLASH->PCROP2ASR, (*(Data + 56) | (*(Data + 57) << 8)));Yes, the STM32C0 has not such field name PCROP2ASR in the FLASH_TypeDef structure? ../OpenBootloader/Interfaces/Patterns/OPTION BYTES/optionbytes_interface.c: In function 'OPENBL_OB_Write':../OpenBootloader/Interfaces/Patterns/OPTION BYTES/optionbytes_interface.c:162:22: error: 'FLASH_TypeDef' has no member named 'PCROP2ASR'; did you mean 'PCROP1ASR'?162 | WRITE_REG(FLASH->PCROP2ASR, (*(Data + 56) | (*(Data + 57) << 8)));| ^~~~~~~~~How can this

Hello.I'm currently working on the above cpu and trying to implement a custom ble p2p server app. I have 1 service and 2-3 characteristics and i want to add Freertos for time critical constraints managenments. Correct me if im wrong, but it seems that the Freertos stack is incompatible with a custom ble p2p server template. The only thing that i can choose to use when enabling Freertos is the heart monitor template. Is there any work around or do i have to ditch free rtos all together and go bare metal instead ? 

Hello,I'm attempting to work with the SAI using DMA on the H7S using the NUCLEO-H7S3L8, unfortunately there is no SAI example. There is an I2S example using DMA on the STM32H7S78-DK:https://github.com/STMicroelectronics/STM32CubeH7RS/tree/main/Projects/STM32H7S78-DK/Applications/USB_Device/Audio_StandaloneI've generated a project for the Nucleo using CubeMX, and then brought over the DMA code in the HAL_SAI_MspInit from that I2S example, and modified it for SAI. The result runs, but a user settings error is reported in the DMA interrupt:Here's the complete project:https://github.com/mojolabsdotnet/saiDmaH7SWhat is the right way to setup DMA for the SAI on the H7S?Thanks! 

I am working with the STM32U083RCT6 and I need clarification regarding the exact physical memory addresses of the option bytes.In the reference manual, I can see that the option bytes are located in the system memory area:0x1FFF7000 – 0x1FFF7FFFHowever, the manual only describes the FLASH option registers (OPTR, PCROP1SR, PCROP1ER, WRP1AR, WRP1BR) with register offsets from the FLASH peripheral base address. It does not clearly mention the exact non-volatile storage addresses inside system memory. I would like to confirm:1.What is the officially documented physical start address of the option byte storage in system memory for STM32U083?2. Is 0x1FFF7800 the confirmed start address of OPTR in non-volatile memory?3. If this information is not in the reference manual, in which official document can I find the exact byte-level storage mapping?4. What is the exact physical start address of Option Bytes in STM32U083?5. In which official document is this address explicitly mentioned?I am

Hi ST Community,During my application startup I am facing an issue where TIMx_UP_IRQHandler is constantly triggered while using HAL_Delay(). It seems that the Motor Control SDK is using TIM1 which periodically fires the TIM IRQ handler, interfering with my startup routine.To verify this, I commented out MX_MotorControl_Init() in main.c and my startup routine completed successfully.My startup routine must not be interrupted by the Motor Control tasks. Therefore I need to delay the Motor Control initialization until my startup is complete.Is there an official API to temporarily disable Motor Control Tasks during startup and re-enable them once the application is ready? ( I would like to avoid shifting the MX_MotorControl_Init() function in my startup code) int main(void) { ... /* Initialize all configured peripherals */ ... MX_MotorControl_Init(); MX_IWDG_Init(); MX_SPI2_Init(); /* Initialize interrupts */ MX_NVIC_Init(); // Disable IWDG if core is halted __HAL_DBGMCU

Split from How to program and debug the STM32 using the Arduino IDE.Hi,I am new to this forum having just bought a Nucleo STM32G431 which I want to program using Arduino 2.3.4 on MAC OS 11.7.11. On trying to upload the simple blink example I get the error:Warning: long options not supported due to getopt from FreeBSD usage.getopt: illegal option -- aTerminating...Failed uploading: uploading error: exit status 1Having read around this a little it seems that MACOS is using a BSD version of getopt which doesn't support long options. I believe the problem is in the Arduino core for STM32 MCU based boards (I have version 2.12.0 installed).  Do you know of a workaround for this? Do I need to install the GNU version of getopt and is this safe? I am not a MACOS expert. Where would I find that? 

Firmware Package: STM32Cube_FW_H5_V1.6.0 Product ProblemDA regression fails with timeout. I cannot recover the board. I was trying to run the STiROT Template example (STiROT_Appli) on the STM32H573I-DK board. 1. I ran a custom provisioning script (based on ob_flash_programming.sh + obkey_provisioning.sh)2. During image flashing step, I accidentally flashed an unsigned binary (appli.bin) instead of the signed image (appli_init_sign.hex)3. OBKey provisioning (DA_Config.obk, STiRoT_Config.obk, STiRoT_Data.obk) may have failed due to the MCU being in an unstable state4. I then set Product State to PROVISIONED (0x2E) Discovery output Discovery succeeds and shows- PSA lifecycle: ST_LIFECYCLE_IROT_PROVISIONED - ST provisioning integrity status: 0xf5f5f5f5 - ST HDPL1 status: 0x1 - Permission: Full Regression is listed as available Regression attempt Using default keys from the firmware package (not regenerated): - Key: DA/Keys/key_3_leaf.pem - Certificate: DA/Certifica

Hello Team,Refering to this Thread : https://community.st.com/t5/stm32-mcus-products/updating-firmware-through-usb-dfu-with-readout-protection-level/m-p/838635We want to achieve same , Can I upgrade the firmware using USB DFU, with RDP set to level 1? We are fine setting RDP1 -> RDP0 and Erase Flash, then Flash the Firmware.I tried with Cubeprogrammers and CLI , but unroftunately USB1 ( STM32 Bootloader in PC device manager ) not showing and connection does not happen. Tried making BOOT0 pin high when Reset.However , Actual plan is jump to system bootloader from User code through jump address. Regards,@STOne-32 @j_filipe 

my UART settings are as follows:void USART_Configuration(EUartCom uartCom) { USART_Config(uartCom, COM_BAUDRATE[uartCom]); } void USART_Config(EUartCom uartCom, uint32_t baudrate) { USART_DeInit(COM_USART[uartCom]); GPIO_InitTypeDef GPIO_InitStructure; // DE: GPIO_InitStructure.GPIO_Pin = COM_DE_PIN[uartCom]; GPIO_InitStructure.GPIO_Mode = GPIO_Mode_OUT; GPIO_InitStructure.GPIO_OType = GPIO_OType_PP; GPIO_InitStructure.GPIO_PuPd = GPIO_PuPd_DOWN; GPIO_InitStructure.GPIO_Speed = GPIO_Speed_50MHz; GPIO_Init(COM_DE_PORT[uartCom], &GPIO_InitStructure); GPIO_ResetBits(COM_DE_PORT[uartCom], COM_DE_PIN[uartCom]); // TX: GPIO_InitStructure.GPIO_Pin = COM_TX_PIN[uartCom]; GPIO_InitStructure.GPIO_Mode = GPIO_Mode_AF; GPIO_InitStructure.GPIO_Speed = GPIO_Speed_25MHz; GPIO_InitStructure.GPIO_OType = GPIO_OType_PP; GPIO_InitStructure.GPIO_PuPd = GPIO_PuPd_NOPULL; GPIO_Init(COM_TX_PORT[uartCom], &GPIO_InitStructure); // RX: GPIO_InitStructure.GPIO_Pin = COM_RX_PIN[uart

STM32H563 on custom board. Using NVIC_SystemReset() to reset. It resets consistently.I expect LL_RCC_IsActiveFlag_PINRST() to return true after the MCU comes out of reset. It consistently returns false.NVIC_SystemReset() should result in pin reset, yes?LL_RCC_IsActiveFlag_PINRST() returns false after a restart with STM32CubeIDE debugger also.Could something on the board hold NRST high?

I have created a project in MotorControlWorkbench version 6.4.0 with both no RTOS and FreeRTOS (by clicking the radio button in MotorControlWorkbench) and gotten it to work successfully on my custom hardware based on the EVLServo1 eval kit.  However, because I use he Keil toolchain I would prefer to use RTOS2 RTX5 for my RTOS instead.  I have managed to create the project with MotorControlWorkbench (with no FreeRTOS) and then when I open it in Keil I go into the Manage RunTime Environment and enable the RTOS2 RTX5 RTOS.  This adds a number of files to the project.  I have worked through a number of compilation errors and am now getting one last error:Error: L6200E: Symbol SysTick_Handler multiply defined (by irq_armv7m.o and stm32_mc_common_it.o).I understand this to mean that RTOS2 RTX5 has a SysTick_Handler of it's own and so does the HAL libraries created by MotorControlWorkbench and I need to switch one of them over to use a hardware timer like TIM6 instead.&nbs

I am testing the Timer and PWM with different combinations to understand it better on NucleoF446RE board. I have configured as belowTimer1 as Master with trigger output as update event, PWM channel1 output on PA8 (interrupt). Timer2 as slave in trigger mode and output PWM Channel1 on PA5 same as LED based on interrupt.Problem is Timer2 Channel1 output does not toggle on PA5 but if i place break point in tiimer2 interrupt is triggered.void TIM2_IRQHandler(void) { /* USER CODE BEGIN TIM2_IRQn 0 */ /* USER CODE END TIM2_IRQn 0 */ HAL_TIM_IRQHandler(&htim2); /* USER CODE BEGIN TIM2_IRQn 1 */ /* USER CODE END TIM2_IRQn 1 */ }Please guide to solve the issue.

With the attached ioc, CubeMx 6.17.0 freezes when I click on "Generate Code" on macos 14.8.3 . If I start it from the command line I can see this error in the console:2026-03-09 17:20:52,902 [INFO] UtilMem:75 - End SaveConfig() Used Memory: 550133736 Bytes (1459617792)2026-03-09 17:20:52,976 [INFO] UtilMem:75 - End SaveConfig() Used Memory: 580273760 Bytes (1459617792)2026-03-09 17:20:53,182 [INFO] ProjectBuilder:603 - [PM] Start Project Generation Thread2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plugin: RCC2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plugin: I2C2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plugin: I2C2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plugin: USART2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plugin: MEMORYMAP2026-03-09 17:20:53,207 [INFO] IPsOverviews:2785 - set Specific Code input for plug

Hello,I have been using CubeIDE for some time to flash the STM32H755ZIT6 microcontroller with the J-Link Base. After the latest update, I started getting this error: Error in final launch sequence:Failed to execute MI command:target remote localhost:2331Error message from debugger back end:Truncated register 16 in remote 'g' packetFailed to execute MI command:target remote localhost:2331Error message from debugger back end:Truncated register 16 in remote 'g' packetTruncated register 16 in remote 'g' packetI have already updated J-Link to the latest versionIs there any method to resolve this issue?Thank you.

Hello, I am using below the libraries with my project.STM32Cube_FW_H7_V1.12.0 (STM32H7xx HAL 1.10.0)- LwIP V2.1.2, FreeRTOS V10.2.1, 2017 - ethernetif.cAnd I designed the HTTP Server with STM32H743ZIT6. It is very well operate.And I updated the libraries such as STM32H7xx HAL 1.11.5 recently.- LwIP V2.1.2, FreeRTOS V10.3.1, 2022 - ethernetif.cAll normal operations are functioning properly. (TCP, UDP)However, an errors is occurring in the HTTP Server.I implemented it by referenceing the LwIP_HTTP_Server_Socket_RTOS example from the STM32H7 HAL repository.When approximately two files are requested simultaneously in a Web Browser, LwIP becomes unresponsive.(Ethernet Event Handler is running when LAN cable Unplug and Plug) (capture packets in old version) (capture packets in new version)There are no changes to the application. (httpserver_socket.c)But, these error continue to appear. I would appreciate any helpful advice. Best regards,&nbsp

Hi, I'm using STM32H745ZI implemented ethernet with Lwip it's working fine STM32CubeIDE(ver-1.90.0) but not in IAR(ver-9.30.1) like any changes that I need to do in .icf  file of IAR because it's working fine Cube IDE

Cube MX v6.17.0 and even previous version v6.16.1 have a problem generating SystemIsolation_Config for STM32N6.I made a test IOC file. There are 3 SECURE checkboxes checked in RIF -> RIF Aware IPs -> EXTI1:Line 13 for  EXTI13Line 17 for RTC secure-wakeupLine 62 for Wake-up PIN3Screenshot: Generated SystemIsolation_Config(void) in main.c of FSBL contains: /* set up EXTI configuration */ /* EXTI13 line */ HAL_EXTI_ConfigLineAttributes(13,EXTI_LINE_SEC|EXTI_LINE_NPRIV); /* RTC secure wake-up line */ HAL_EXTI_ConfigLineAttributes(17,EXTI_LINE_SEC|EXTI_LINE_NPRIV); /* Wake-up_PIN3 line */ HAL_EXTI_ConfigLineAttributes(62,EXTI_LINE_SEC|EXTI_LINE_NPRIV);These calls fail in this assert in stm32n6xx_hal_exti.c: assert_param(IS_EXTI_LINE(ExtiLine));Full assert is turned on in IOC so it lands in assert_failed.The assert condition check macro is very messy, but it appears like it expects type and register mask with it so I added them (based on table 138. EXTI line connec

I have an STM32H7 project (created by TouchGFX) that uses FreeRTOS.FreeRTOS uses the systick, so the HAL uses TIM2 for it's timebase.However, HAL_IncTick is never called, so calls to HAL_Delay never return.HAL_InitTick() is being called during startup and this results in TIM2 counting and its update interrupt flag being set. However, the interrupt handler never gets called.