STM32G474CEU6 with 1.8V supply

[Sruthi1428]

Hi, The STM32G474CEU6 LQFPN48 was previously used in our projects with a 3.3 V supply, where we successfully operated a 25 MHz FL2500326 crystal using 6.8 pF load capacitors, and the crystal oscillation was stable.

In the current design, the same controller is powered at 1.8 V for all VDD, VDDA, and VREF+ pins, while using the same crystal and the same load capacitors. However, under this configuration, the crystal oscillator does not start, and no crystal output is observed.

As an alternative, we are currently using HSI clock (16 MHz) as system clock in our application. It is observed in other posts that HSI cannot be used for high speed peripherals like USB and Ethernet due to it's low accuracy, and also it varies with temperature. In our application, we are using 2 - UARTs, 2 - I2Cs and one SPI along with few GPIOs. 

Is it recommendable to run the STM32g474 with 1.8V VDD? Also can someone give advice on the HSI system clock mode?  

Please let us know how to solve this issue.

[TDK]

> In our application, we are using 2 - UARTs, 2 - I2Cs and one SPI along with few GPIOs. 

You can use HSI for all of those. The accuracy of UART in particular will be a little worse but should still be within allowable limits. The other two don't need particularly accurate clocks.

> Is it recommendable to run the STM32g474 with 1.8V VDD?

The STM32G474CEU6 can operate at 1.8 V for most things.

The crystal you're using should be working. Perhaps there is a bad solder joint or workmanship issue.

[AScha.3]

Hi,

did you set range1 ? ( on range2 it might be to close to max.frequ.)

btw

I dont want to play around with osc running or not - and use a oscillator;

for 24M something like this : from mouser , 24M , 1.8V..3.3V , about 50 ct ,

https://www.mouser.de/ProductDetail/Epson-Timing/SG-210STF-24.0000ML5?qs=Li%252BoUPsLEnsd33bROKUfZA%3D%3D