Solved

HiI imported simple CNN to STM32L462RCT using STM32CUBE-AI v5.1.2 ApplicationTemplateI found that compressing the model has no effect on inference time.The aiRun procedure runs for 115ms both in 8-bit compression and "none" configurations although the accuracy drops a bit.I thought compressing float network parameters to uint8_t would not only save the memory but also speed up the inference.So, is compressing the model supposed to speed up the inference?

I want to know the memory size of STM32MP151CAA3 that Coprocessor(Cortex-M4) can use.https://wiki.st.com/stm32mpu/wiki/Boot_chains_overview#STM32MP15_boot_chainthis wiki page says that coprocessor(Cortex-M4) program loaded in MCUSRAM(384kB).Does this means that the maximum size of "program and using data" of Coprocessor(Cortex-M4) program is up to 384kB?In this case, If I want to run a program with a size larger than 384kB on coprocessor(Cortex-M4), is there workaround such as using external memory?Best regards.H. Masuda

I'm bringing up a custom STM32MP15x-based design, and I'm encountering what looks like stack corruption issues in U-Boot (and I can't get a kernel to boot).Symptoms:I've enabled early debug printing, and get the following output:U-Boot SPL 2020.04-stm32mp-r1 (Aug 15 2020 - 13:11:31 -0500) Model: STMicroelectronics STM32MP157A-DK1 Discovery Board RAM: DDR3-1066/888 bin G 1x4Gb 533MHz v1.45 Trying to boot from MMC1At this point, the console hangs for a really long time. Looking at the UART on a logic analyzer, I see that during this time, it's printing seemingly gibberish characters at 1750 baud (which my USB-to-serial converter is ignoring due to the framing errors caused by the wrong baud). I did some brief JTAG investigations, and somewhere along the way, something calls printf() on a character buffer that's corrupted (so it doesn't encounter a null-terminating character for a very, very, very long time).Eventually, it will print something like this:▒▒*aQ▒rrr▒In: serial Out: serial Er

I have connected a sine wave of amplitude 3V with 1.5V shifted above the zero to 12-bit ADC channel by using dsPIC33FJ32MC204 controller and stored in an array. I want to detect the max value out of five samples because after the five samples my data is repeating, so please give me any suggestion on it. I am getting output as a zero.

Hi,is it possible to use the STPMIC1 on an I2C Bus where another slave does clock stretching? In the datasheet it says that the STPMIC uses the SCL to shift data. For my understanding the clock strechting of the other slave shouldn't bother the STPMIC cause it is not addressed when the clock stretching occurs.Does anyone has experience with that?

I have a custom PCB with a STM32G431 on board and STLINkV3ModsI wish to use PA9,10 for USART1 and PA2,3 for USART2 attached to VCP of STLINKV3MODS via 3K3 resistor.If I use USART1 only it is OKIf I use USART2 only it is OK. See Figure 1If I use USART1 and USART2. USART1 works OK. USART2 does not work. See Figure 2There is a correct pattern coming out of the Micro. But he STLINKV3MODS input pin stays high. This I assume must be being driven high by the STLINK.If I transmit a pattern before configuring the USART1 is also does not work.Figure 1 USART2 only configured Yellow USART2 Tx pin Blue via 3k3 VCP input pinFigure 2 USART2 only configured Yellow USART2 Tx pin Blue via 3k3 VCP input pin

I just started using STM32CubeIDE, and was able to get a blinking LED and getting some UART communication going in no time with a NUCLEO-L476RG eval board.I am trying to follow the examples in the book "Mastering STM32" by Carmine Noviello, but I am a bit stuck on the ARM Semihosting section. I have tried a couple of things, but I don't know how to get printf() to work for instance in the STM32CubeIDE. Any advice on getting these functions to work so the output is parsed to the UART2 interface would be really appreciated :)

Hello,I am a new STM32 member. In many previous projects from PIC, we often #define GPIOx as a bit. So we can manipulate some operator directly and easily as example below.Example: #define LED_AL1 LATBbits.LATB6 #define LED_AL2 LATBbits.LATB7 LED_A1 ^= 1;LED_A2 = ~LED_A2;So I think STM32 MCU will also parse to each bit such as other MCUs.If we want to parse a GPIOx to each bit, how can we do it ?If you have an idea, please share us your comment.Thanks in advance !

Hi The terminal printing information is as follows. This TF-A will start properly in EMMC. In the docs/plat/stm32mp1.rst file, I saw that USB is not supported. May I ask how to modify it so that it can be burned via USB Regards User.44

Device Reset after every data conversion Value of all CTRL registers = Power On RESET valuedevice initialization stepspower ondelay 20 msread device id reset devicedelay 5 mswrite CTRL1 0x30write CTRL2 0x0Cwrite CTRL6 0x60Read samples in polling mode

There appears to be a bug in STM32CubeMX where some settings are not loaded when reopening the app and project, even though they were previously successfully written to the .ioc file. Default settings are always used instead, and then re-saving the project (without making any changes yourself) will write these undesired settings back to the .ioc file. These settings are:Project Manager > Project > Firmware Location (custom path reset back to default path)Project Manager > Project > Add library files as reference (reset back to copy)Steps to reproduce:Create a symlink to `STM32Cube/Repository` which is relative to the intended project workspace. This enables sharing projects among multiple developers, since CubeMX uses a relative path to the FW repo.ln -s ~/STM32Cube/Repository ~/projects/demo/cube_repoLaunch CubeMX, and create a new project with the following settings:Save and close CubeMX. This produces a .ioc file.Track the .ioc file with version control (this gives us mo

Hi, I recently migrated from CMSIS_V1 to CMSIS_V2 ( in STM32Cube_FW_H7_V1.7.0). That came with a significant increase in the number of defined thread priorities ~ around 56 priorities. However, CubeMX sets the FreeRTOS MAX_PRIORITIES parameter to 32. The setting is also disabled if CMSIS_V2 is enabled so it can't be changed. Even the CubeMX comment says it should be set to 56.The consequence of this defect is that setting any thread priority above 32 they will all be truncated to same setting of (32 -1). To workaround the issue, I undefined configMAX_PRIORITIES and reset it to 56.Can ST please confirm this is a defect and fix it if required. Or does anyone know a better way to fix it?CheersDavid

We are working on a project that uses ST motor control library. As part of our project requirements we have to do a FEMA for the software. Does anyone know if there is a FEMA document for ST motor control library?

In the STM32F413/423 reference manual ( RM0430 Rev 8), there are two UART8_RX options, but no UART8_TX option. This seems like a typo, since all other UART instances have both a TX and RX mapping.Also, CubeMX lets you assign UART8_TX to stream 0.

Compiles OK.I set the GFX screen to blue.Loads the black screen into the STM32F746G-DISCO.Why?

Bug: CubeMX STM32F4 Keil project generation does not work with Java 14, where it works with Java 11I have a project that generates fine using CubeMX 6.0.1 STM32F4 1.25.1Now I upgraded to Java 14.The project generation has problems on the Keil Project file generation.When I go back to Java 8 it works. Same problem on ST32F4 1.25Please ST fix this bug.Adib.

Is there any technical data regarding when the FIFO becomes functional after power up? I have noticed I2C comms and registers can be accessed quite quickly but if the FIFO is initialised too early (e.g. going into FIFO mode), it never starts storing data, despite the relevant configuration registers being perfectly correct. My guess is that it is something to do with the need to come out of bypass mode and if this is done too early after power up, it doesn't work. Have any ST support personnel any details?

Hi all!,I'm, triyng to program using CMSIS in order to learn how registers works.If I use the SystemClock_Config() generated by STM32CubeIDE all works properly, but if I use my own rutine, the TFT wired to SPI3 dont show the fonts properly (I think it must be something about activate/set some missing clock)This is the SystemClock_Config():void SystemClock_Config(void) { RCC_OscInitTypeDef RCC_OscInitStruct = {0}; RCC_ClkInitTypeDef RCC_ClkInitStruct = {0}; RCC_PeriphCLKInitTypeDef PeriphClkInit = {0};   /** Initializes the RCC Oscillators according to the specified parameters * in the RCC_OscInitTypeDef structure. */ RCC_OscInitStruct.OscillatorType = RCC_OSCILLATORTYPE_HSI|RCC_OSCILLATORTYPE_HSE; RCC_OscInitStruct.HSEState = RCC_HSE_ON; RCC_OscInitStruct.HSEPredivValue = RCC_HSE_PREDIV_DIV1; RCC_OscInitStruct.HSIState = RCC_HSI_ON; RCC_OscInitStruct.HSICalibrationValue = RCC_HSICALIBRATION_DEFAULT; RCC_OscInitStruct.PLL.PLLState = RCC_PLL_ON; RCC_OscInitStruct.PLL.PL

HiI'm trying to use my STM32F407-DISC1 board's onboard ST-LINK to program a STM32 black pill (STM32F103C8), but can't get it working for the life of me.I've connected the two boards as follows:Discovery SWD Pin 2 (SWCLK) -> Black Pill SCLKDiscovery SWD Pin 3 (GND) -> Black Pill GNDDiscovery SWD Pin 4 (SWDIO) -> Black Pill DIODiscovery JP1 -> Black Pill Vcc (Powering the black pill instead of the STM32F407). I've also tried powering the black pill through usb and connecting Discovery SWD Pin 1 -> Black Pill VccI've updated the Discovery's ST-Link to the newest firmware.Using the st-info utility, I'm able to detect the ST-Link on the Discovery board correctly, but st-info --descr either returns nothing or "unknown board". When I set the black pill's jumpers to pull BOOT0 high, I can see the correct output from st-info --descr, but when I use Atollic TrueSTUDIO to upload the program to the black pill while BOOT0 is high, it uploads successfully but immediately loses the deb

Hello, i'm a new about this embedded systems, i am working with stm32f4 discovery card with HAL Library.İ looked the GPİO topic and i want to create a program like :When i clikt the button 3 times red led will open and when i clikc the button 5 times green led will open only. i create "int count" and i used conditions but couldnt be success about it.Here is the my code , can anyone help me to show right algorithm?edit : Finally it worked :

Thank you in advance for this program file and for the answer.best regardsAndrzej Danielewicz

I have configured STM32h743ZI with below Settings: ADC1 with trigger set to tim1 update event DMA to transfer from ADC to memory Disabled continuous conversionProblem:I want the ADC to sample every 50us (TIM1 configured to 50us)but conversion starts with first event and then continuously  happening without timer event

I tried to find the frequency that the digital ports of Nucleo-F303k8 can accept data (example: the arduino works with 10kHz on these pins), but i couldn't find anywhere. Someone can help me? I need to use some sensors and I need this information to but them.

I have a problem with running the software on the Cortex-M4. Probably the default device tree configuration does not start the m4 core. Has anyone used a SoMlabs board and tried to run software on the Cortex-M4 core?

Hi,I need to calculate DSP speed for STM32F407, but I can't find ARM documentation on DSP extension of Cortex-M4 nor any documentation of DSP extension implementation in STM32F4.I know that Cortex-M4 have pipeline with 3 stage + branch speculation.So, if we disregard pipeline, worst case time it takes for one instruction to complete is 4 periods of master clock. Right? On average, due to pipeline, this time approaches to one period of master clock. Right?But, is the same truth for DSP instructions? - I think that it's not, but I'm not sure.ARM site only says, quote: "Single cycle dual 16-bit MAC". But what is one cycle in this situation? Is it one cycle of master clock, one DSP instruction cycle or something else?How is this "Cycle" defined?Basically, I want to know how much 16bit MAC instructions per second can be executed by STM32F407 working at certain clock (master clock). By the way, on which bus is DSP extension connected in STM32F407?Thank you very much.