STM32HT43 Nucleo 144

[IAgui]

Hello All,

I have an intresting issue that came up. But I am not sure if I understand it. In the clock configurator, I gave the usart mux a 100Mhz clock for the uarts and set up the baud rate for 115200K. Once I started sending data, I noticed that the baud rate was off. I went back and gave the clocking mux 64mhz, and that seemed to fix that issue. I want to make sure I understand why this before I move forward. Any advice would be greatly appreciated.

[Tesla DeLorean]

>>STM32HT43 Nucleo 144

?? WHAT?

You mean an STM32H743ZI2 board ? The ZI2 board has a slightly off-frequency HSE clock, the ST-LINK/V3 is 8.33 MHz, not 8 MHz of the older ZI board

Check that HSE_VALUE is consistent with the source clock.

The 64 MHz HSI clock is likely closer to nominal.

[IAgui]

oops STM32H743ZI Nucleo 144 typo, sorry. The schematic is saying the HSE input clock is 25 MHz for that board. Should I measure the 25 MHz clock? Getting a bit concerned as I am getting ready to work with CAN FD. I am using PLL1Q for that which is derived from the HSE clock.

[Tesla DeLorean · Accepted solution]

Check the clock signal on the board as built, pretty sure it clocks at 8 MHz from the ST-LINK/V2-1

[Pavel A.]

When you update the ST-LINK firmware with a recent version, there will be two options for the clock to the MCU: 8.33... (HSE/3) and 8.0 (HSI/2 of 'F7)

[IAgui]

Please forgive me as I am confused about the stlink clock and the HSE clock on the MCU. They are two different devices. How would this affect the baud rate on the MCU?

[Tesla DeLorean]

To save parts the on-board ST-LINK can provide a crystal based clock source to the target MCU

Review schematic

[Pavel A.]

The ST-LINK MCU on the Nucleo board (STM32F7) has a 25 MHz crystal connected to its HSE. It's MCO goes to the target MCU's HSE in "bypass" (single-ended) mode. Several clocks can be muxed to the MCO, with programable divisor.