2020-09-04 11:07 AM
CubeMX Version : 6.0.1.202008101643 com.st.stm32cube.common.feature.mx
Device: STM32H743
I believe there may be an issue when setting the VREFBUF output voltage in CubeMX. The definitions for VREFOUT in CubeMX are inconsistent with the definitions given in the stm32H7xx_hal.c file.
CubeMX uses the following in the dropdown:
Scale 0 - 2.5V
Scale 1 - 2.048V
Scale 2 - 1.8V
Scale 3 - 1.5V
However the function description for
HAL_SYSCFG_VREFBUF_VoltageScalingConfig()
in the stm32h7xx_hal.c indicates the following:
SYSCFG_VREFBUF_VOLTAGE_SCALE0 = 2.048V
SYSCFG_VREFBUF_VOLTAGE_SCALE1 = 2.5V
SYSCFG_VREFBUF_VOLTAGE_SCALE2 = 1.5V
SYSCFG_VREFBUF_VOLTAGE_SCALE3 = 1.8V
This issue leads to the incorrect voltage being present on VREF+ pin once the code is generated. I believe this is a similar issue to that reported on other STM32 devices. Setting of scale 0 in CubeMX results in scale 0 being used in the generated code, but the voltage is incorrect.
I don't know which set of scales is ultimately correct, but the scales listed in the stm327hxx_hal.c file do set the value of the CSR register correctly inline with method description.
My hunch is that the descriptions in the CubeMX dropdown are incorrect?
Kind Regards,
Simon
2020-09-07 2:38 AM
Hello @Simon Sq
Thanks for the feedback, the problem is in the description in the stm32h7xx_hal.c file not in STM32CubeMX:
/**
* @brief Configure the internal voltage reference buffer voltage scale.
* @param VoltageScaling specifies the output voltage to achieve
* This parameter can be one of the following values:
* @arg SYSCFG_VREFBUF_VOLTAGE_SCALE0: VREF_OUT1 around 2.048 V.
* This requires VDDA equal to or higher than 2.4 V.
* @arg SYSCFG_VREFBUF_VOLTAGE_SCALE1: VREF_OUT2 around 2.5 V.
* This requires VDDA equal to or higher than 2.8 V.
* @arg SYSCFG_VREFBUF_VOLTAGE_SCALE2: VREF_OUT3 around 1.5 V.
* This requires VDDA equal to or higher than 1.8 V.
* @arg SYSCFG_VREFBUF_VOLTAGE_SCALE3: VREF_OUT4 around 1.8 V.
* This requires VDDA equal to or higher than 2.1 V.
* @retval None
*/
So, issue reported to the tool development team that will plan the correction.
Best regards,
Nesrine
2020-09-07 2:40 AM
2020-09-13 4:14 PM
Thank you Nesrine.
I can work around it for now until the fix is realised at some point in the future.