Solved

hey i am struggling to flash by custom onnx model on stm32N6570-DK . I tired using the stm32 cube ai cloud website but it only gives me option to generate c code and not a project or elf file . So i shifted to stm32cube Ai studio but it keeps me throwing this error -  (*) 'input'/'output' buffers are allocated in the activations bufferRunning the ONNX model...Running the ST.AI c-model (AI RUNNER)...(name=network, mode=TARGET)E200(ValidationError): TARGET: Unable to bind the ST.AI runtime with "network" c-model: [] E801(HwIOError): Invalid firmware - COM3:115200 can anyone help me to get started with the same

Hi ST Community, I am trying to deploy a YOLO-FastestV2 INT8 object detection model on the STM32N6570-DK. The model is successfully converted and runs on the board, but I am facing issues with the detection results.Model InformationModel: YOLO-FastestV2 (INT8) Number of classes: 6 Parameters: 235,120Model InputInput_0_out_0uint8 [1,1,320,320]Model OutputsQuantize_297_out_0int8 [1,12,20,20]Quantize_313_out_0int8 [1,3,20,20]Quantize_311_out_0int8 [1,6,20,20]Quantize_330_out_0int8 [1,12,10,10]Quantize_346_out_0int8 [1,3,10,10]Quantize_344_out_0int8 [1,6,10,10]Issues I am FacingThe bounding boxes are not displayed at the correct positions on the camera image. The detected bounding boxes appear to be shifted or incorrectly scaled. Even when the camera is not viewing any object (or the camera feed is unclear), random bounding boxes continue to be displayed. These false detection remain on the screen even though there should be no valid detection.Verification PerformedI also tested the quanti

Hi,I am working on a project using STM32H743ZITx and trying to deploy an ONNX model using X-CUBE-AI.My .ld file was doing just fine until I add the X-CUBE-AI in the middleware. It will split the BSS range across DTCMRAM and AXI‑RAM. The startup file’s BSS‑zeroing logic accesses invalid non‑hardware‑backed addresses between the two RAM regions and results in a HardFault right after Reset_Handler.My configuration is:MCU: STM32H743ZITx STM32CubeMX: 6.18.0 STM32Cube FW_H7: 1.13.0 STM32CubeIDE for Visual Studio Code: 3.9.0 Enabled X‑CUBE‑AI 10.2.1Files before and after the configuration of X-CUBE-AI are attached.

Hi all,i have a small question about Validation of a trained model.Atm i only work on PC with NEAI Studio Version 5.1.1.I used only data out of a *.csv fileI trained (benchmark) the model with around 170 Lines and 16 axis to get an accuracy of ~97%.I then validated the library with some test data with same axis count and data preparation. The learning file has the minimum amount of lines as suggested from the benchmark, i tried more lines with same result.The validation shows that even good data is treated as abnomaly so i only get an accuracy of max. 50%.Are there any hints what i am doing wrong? Many thanks in advance for answers.BR’sMartin

Hi, I ordered on rs-online memory M24M02E-UFMN6TP. The memory that I received has SMD marking 24M02FU (photo: https://ibb.co/gLCbX0vb). I was unable to lookup this marking on the st.com web nor in the M24M02E datasheet. The rs-online vendor also don’t know :). Can anybody verify that it’s the M24M02E, i.e. memory with the following datasheet: https://www.st.com/resource/en/datasheet/m24m02e-u.pdf ? I need 2 MBit, up to 5 ms write. IMHO the suffix 'E' distinguishes the 5 ms memories from the 10 ms and I need to be sure it’s the 5 ms version of the memory. Thanks.

When running an LCD display project, the screen fails to turn on after I cut the power and turn it back on.It only comes on when I short the NRST and GND pins.I have a 10k pull-up and a 100nF pull-down connected to the NRST pin; what should I do?

I need more PSRAM and so went exploring and found RM0486, which seems to show that I can have two PSRAM on one XSPI port (using EXTENDMEM, §28.4.24), but I cant seem to find more than that.Is EXTENDMEM with two octal DDR PSRAMs a silicon-validated configuration on the N6 XSPI, or is it documented-but-untested territory? RM0486 notes the automatic CS switch when an access "crosses the middle boundary of the external memory". Does this work transparently with DTR memory-mapped writes and MDMA/DMA bursts, or are there alignment/burst-length restrictions at the boundary I should design around?Has anyone run 128 MB total on one N6 XSPI port? The 256 MB address window and DEVSIZE math allow it, but I'd love to hear from someone who has it working on real hardware (or from ST whether an eval/reference design exists with dual PSRAM).Anything to watch on the RIF/RISAF side for the XSPI1 memory region when all masters run secure?Board: custom using the STM32N657X0HxQ.Planned setup:Two identical

Hello,I have an STM32G474RE Nucleo-64 with an X-NUCLEO-IHM16M1 driving a GB2208 motor with 7 pole pairs. My goal is to use an AS5048A absolute magnetic encoder for sensored FOC and position control.MCSDK already runs the motor correctly in sensorless mode, and I can reliably read the AS5048A. However, integrating it into MCSDK has been much less straightforward than expected.Current implementationI use TIM1 CH4 to trigger TIM2. TIM2 starts the AS5048A SPI transaction approximately 3 µs after the trigger. My assumption was that this provides a relatively quiet switching interval, avoids interfering with current ADC sampling, and reduces MOSFET switching noise. I am not certain whether this assumption is valid or whether encoder acquisition should instead be synchronized with a specific PWM/ADC event.After determining the encoder direction and electrical offset, I calculate the electrical angle as:electrical_angle = (pole_pairs * mechanical_angle + electrical_offset);I then inject the en

It seems like it's been over a year since reports surfaced that the ST-Link interface isn't working on Snapdragons.Are there still no ST-Link interface drivers available for Snapdragon CPUs?

I made a linux kernel driver for VL53L9 with ~ 10 FPS.Tested on Radxa Dragon Q6A with STEVAL-VL53L9https://github.com/momoisgoodforhealth/st-vl53l9-i2c-kernel-driver

Hi everyone,I am using an STM32H753ZIT6 on our custom PCB. Before this, I was successfully programming the STM32H753ZI on the NUCLEO-H753ZI board using the UART bootloader (system memory boot mode) with STM32CubeProgrammer and a USB-to-UART TTL converter. I have never used an ST-LINK for programming.I followed the same hardware connections and programming procedure on our custom PCB, but STM32CubeProgrammer always returns a timeout error and cannot connect to the MCU.Setup MCU: STM32H753ZIT6 Software: STM32CubeProgrammer Programming interface: UART bootloader (not ST-LINK) USB-to-UART converter connected to USART1 TX of USB-UART → STM32 PB15 (USART1_RX) RX of USB-UART → STM32 PA9 (USART1_TX) Manual boot mode selected (not using RTS/DTR auto boot) Programming Procedure Hold BOOT0 HIGH. Reset or power-cycle the board. Release RESET while keeping BOOT0 HIGH. Click Connect in STM32CubeProgrammer. This exact procedure works on the NUCLEO-H753ZI but fails on our custom PCB.

Hello ST Team,Good morning!I am currently working on integrating the VL53L9 sensor with the STM32N6570-DK development board. I have successfully obtained the STSW-IMG053 software package, which is designed for the STEVAL-VL53L9 evaluation board using the MIPI CSI interface.However, my target hardware configuration is different. My objective is to use the following setup:Software Package: STSW-IMG053 Host Board: STM32N6570-DK Sensor Board: X-NUCLEO-53L9A1 Communication Interface: I3CI would like to port the STSW-IMG053 software package to support the X-NUCLEO-53L9A1 evaluation board while retaining the STM32N6570-DK as the host platform.I would appreciate your guidance on the following points:Is there an official procedure or application note describing how to port STSW-IMG053 from the STEVAL-VL53L9 (MIPI CSI) platform to the X-NUCLEO-53L9A1 (I3C) platform? Are there any reference projects, example applications, or software packages available for the STM32N6570-DK with the X-NUCLEO-53L9

I'm trying to start working with Evaluation board based on VL53L9CX Time of Flight sensorIn I2C, I'm having a problem with identifying the component addressI find address 55 and not 29The connection looks like thisGND to GNDSDA to SDASCL to SCLXSHOUT to 3V33V3 to 3V3I removed resistor 25 for internal clockAccording to the datasheets, it seems that 55 is the EEPROM ID

Hi ST community,We are currently evaluating STM32WB05 vs STM32WB09 for an application involving continuous BLE data streaming from multiple sensors, and would appreciate guidance from experts who have worked with STM32WB0 series.The reason we choosing the STMWB05 over STM32WB09 MCU, STM32WB05 MCU comes at a lower cost.In brief our application sensor are2x IMU LSM6DSO sensors (high-frequency sampling)1x magnetometer1x analog force sensorA single BLE connection (no mesh or multi-node), with frequent sync occurring approximately every three minutes.Key Questions1. WB05 vs WB09 for continuous BLE streamingBoth MCUs seem to have similar BLE capabilities (Bluetooth 5.4, 2 Mbps PHY), but:Does WB09 (larger RAM/Flash) provide significant advantage for:Continuous streamingBufferingReducing packet drops?Has anyone faced RAM bottlenecks or BLE throughput instability on WB05, WB09 in such use cases?2. Single-core architecture suitabilityBoth WB05 and WB09 are single-core Cortex-M0+ with radio

I'm implementing I2C communication using an STM32WBA2 as the Master and an STM32U0 as the Slave.I've implemented a process to send one byte of data (0x75) from the Master to the Slave, with the Slave Address set to 0x68. However, when I check the SCL/SDA waveforms, it appears there's no ACK from the Slave.On the other hand, the Slave receives an event in HAL_I2C_AddrCallback() when it receives 0x68, but the value read by HAL_I2C_Slave_Receive() is empty. How can I receive 0x75 from the Master?I've attached the main.c files for both the Master and Slave.The CubeMX configuration is as follows:  

I tried building and flashing the sample project BLE_TransparentMode to verify Direct Test Mode on the Nucleo-WBA25CE1, but it doesn't work properly.When I connect using STM32CubeMonitor-RF, an error pops up and the application terminates.Sending 0x93, 0x94 (TX, 2440 MHz, PRBS9) from the terminal yields no response.Two COM ports appear, but connecting to either one results in the same outcome.How can I enable Direct Test Mode?  

Hi ST team,We are using the LSM6DSV80X SFLP quaternion output for a smart racket project. The quaternion is used by the mobile APP to control the game character posture.During our test, we found that when the racket is powered on in a flat/horizontal position, the SFLP Yaw output can stay close to 0°, which is expected. However, when the racket is powered on or reset in a tilted position, the Yaw value is affected by the current Pitch/Roll posture and cannot remain at 0°.For example, when the racket is powered on in a tilted position, the output is:Roll = -0.58°Pitch = -37.61°Yaw = -9.40°For our racket game application, the expected behavior is:When the game starts or when the user performs posture reset, the current racket facing direction should be defined as the game coordinate system Yaw = 0°. At the same time, Roll and Pitch should still reflect the real physical tilt of the racket.So the requirement is not a full orientation reset. We only need a Yaw-only reset:Yaw should be rese

Hello STM Community,I am trying to achieve live camera streaming using STM32H723VGT6, but I am unable to get the result for USB_UVC, Ethernet, using esp32 wifi, if there is any help available, please share it. Also, do share any GitHub link if available.  I am using the Weact mini STM32H723VGT6 board, and I have connected a camera module OV2640 to it.I have referred to the GitHub link below, which is for the H743 controller, but it hasn't shown any expected results. https://github.com/WeActStudio/MiniSTM32H7xx   Any guidance from ST engineers or Community members would be very helpful.Thank you in advance.

I’m trying to build a device that detects shaking using a LIS3DH, and then interrupts a Seeed XIAO RP2040 so that it can tell an OLED display to show numbers. I am using an LIS3DH in I²C mode. I want INT1 on the LIS3DH to wake a Seeed XIAO RP2040 from sleep when motion exceeds a threshold. Do I need to use INT1, SDA, and SCL, or could I just use INT1 to make this work? I’ve attached an electrical schematic of what I have right now, but I’m not quite sure if the connections between the LIS3DH and Seeed XIAO RP2040 are correct.  

Hi Is TouchGFX supported on MP1 or MP2 boards ?I found this article : https://wiki.st.com/stm32mpu/wiki/How_to_install_X-TouchGFX_MPU_AddonBut it looks like its only 1 board

My device uses an STM32H743VITx MCU.I need to use ADC3 at the beginning of my program to verify that three voltages are indeed at the correct values. Then I want to disable it completely before initializing ADC1 and ADC2 to avoid pin conflicts.Does the function HAL_ADC_DeInit(&hadc3) allow me to do this?

Everytime CubeMx downloads files but in the end shows the below error every time. Please tell me how do i resolve.

RM0486 describes CSI_PFCR bit 16 (DLD) as follows: Bit 16 DLD: Data lane direction of lane 00: Rx1: TxPer this description, configuring the CSI-2 host to receive from a camera sensor (i.e. lane 0 in RX mode) should require DLD = 0.However, in the shipped HAL, HAL_DCMIPP_CSI_SetConfig() explicitly sets DLD = 1 (via CSI_PFCR_DLD) when configuring the PHY for CSI-2 RX capture from a camera sensor:WRITE_REG(csi_instance->PFCR, (0x28U << CSI_PFCR_CCFR_Pos) |(SNPS_Freqs[pCSI_Config->PHYBitrate].hsfreqrange << CSI_PFCR_HSFR_Pos) | CSI_PFCR_DLD);The HAL source even contains a comment acknowledging the discrepancy:/* set basedir_0 to RX DLD 0 RX, 1 TX. Synopsys 1 RX 0 TX + freq range */This strongly suggests that the underlying Synopsys D-PHY basedir signal has the opposite polarity (1 = Rx, 0 = Tx) from what's documented for the DLD bit in RM0486, and that the HAL was written/validated against the actual behavior rather than the RM's bit description.Is this a mistake of myse

OpenSTLinux: Starter Package, plain image (FlashLayout_sdcard_stm32mp257f-dk-optee.tsv), version 5.0.17-openstlinux-6.6-yocto-scarthgap-mpu-v26.06.10M33 firmware: Zephyr RTOS (4.4.99), board target stm32mp257f_dk/stm32mp257fxx/m33, loaded and started from Linux via remoteproc.Goal: Drive a SPI device (Qorvo DW3000) from the Cortex-M33 over SPI6, which is exposed on the CN5 header (GPIO10/SPI6_MOSI = PC7, GPIO9/SPI6_MISO = PC4, GPIO11/SPI6_SCK = PF7).When the M33 firmware performs a SPI transfer, the GPIO chip-select line toggles correctly (verified on a logic analyzer), but SCK never produces any clock pulses & the line stays flat. Consequently no data is exchanged (MOSI/MISO flat). The transfer call itself returns; it just produces no clock.What I have already checked:-RIFSC firewall: SPI6 shows as NSEC, NPRIV, CID filtering disabled (from the RIFSC debug dump) — i.e. not blocked/assigned exclusively to the A35.-Clocks: clk_summary shows both ck_icn_p_spi6 (~200 MHz) and ck_ker_sp

Hi everyone,I am working on a project using BNO085 IMU sensor and STM32H7A3ZIT6Q. I have a single sensor working but stuck at 44Hz. The same sensor with SparkFun library on ESP32 achieves 1.2ms read time. I need help understanding why and how to match ESP32 performance on STM32.Hardware:NUCLEO-H7A3ZI-Q (STM32H7A3ZIT6Q, Cortex-M7 @ 280MHz)Adafruit BNO085 breakout boardI2C on PB8 (SCL), PB9 (SDA), PC0 (INT pin)P0 and P1 tied to GND (I2C mode, address 0x4A)Software:STM32CubeIDE with HALOfficial CEVA SH2 library from github.com/ceva-dsp/sh2REPORT_INTERVAL_US = 2500 (targeting 400Hz)SH2_ROTATION_VECTOR (9-axis full quaternion)STM32 performance achieved so far:I2C 400kHz polling: 28Hz, read_ms = 32msI2C 1MHz Fast Mode Plus: 44Hz, read_ms = 19msINT pin wait inside sh2_hal_read(): still 44Hz, read_ms = 19msESP32 performance with same sensor and SparkFun SH2 library:read time: 1.2msData rate: close to 400HzThis is a huge difference. 19ms on STM32 vs 1.2ms on ESP32 with the same SH2 library unde