Skip to main content
gilberto-falker
Associate II
Oct 27, 2026 7:23 PM UTC
Solved

STM32U073CCT6 LSE disturbed By RF Field

Hi

We are using an STM32U073CCT6 as a host controller in a LoRaWAN device. We recently modified our design to incorporate an LSE (Low-Speed ​​External) oscillator to improve the precision of the RTC.

The LSE is functional, but it occasionally stops oscillating when the LoRaWAN module transmits a message. We have already checked the power supply, ground, and all associated components, but the issue persists. Is it possible that the STM32U0 is experiencing internal interference—perhaps through demodulation by its protection diodes—caused by the LoRaWAN transmission? The LSE remains stalled for the entire duration that the RF signal is active.

BR

Gilberto

PS. I can not edit the title. Please correct it.

gbm
gbmBest answer
Super User
Oct 28, 2026 6:26 AM UTC

There are few similar topics on this forum - search for them. PCB design details are VERY important for LSE. A s quick solution, you may try to set “drive strength” option in LSE config (if it is available in your MCU) - sometimes that might be enough. Generally - no ground plane unde LSE and its signal traces, keep them as short as possible and no other signal traces close to them/crossing them in other planes.

My STM32 stuff on github - compact USB device stack and more: https://github.com/gbm-ii/gbmUSBdevice
View in thread

4 replies

Ozone
Ozone
Principal
Oct 28, 2026 6:18 AM UTC

I’m not a hardware guy or PCB design expert, but I suspect either PCB-releated EMI issues or coupling due to physical proximity.

If you e.g. look closer at singleboard Linux computers like the Raspberry Pi, you will find the Wireless/BT modules are shielded.
I suppose for good reasons.

    gbm
    gbmBest answer
    Super User
    Oct 28, 2026 6:26 AM UTC

    There are few similar topics on this forum - search for them. PCB design details are VERY important for LSE. A s quick solution, you may try to set “drive strength” option in LSE config (if it is available in your MCU) - sometimes that might be enough. Generally - no ground plane unde LSE and its signal traces, keep them as short as possible and no other signal traces close to them/crossing them in other planes.

    My STM32 stuff on github - compact USB device stack and more: https://github.com/gbm-ii/gbmUSBdevice
    gilberto-falkerAuthor
    Associate II
    Oct 28, 2026 11:13 AM UTC

    Hi

     

    We have already tried all options for the LSE driver. The trace lengths for the LSE are similar to those on ST’s MB1932 board. The issue is that the power level for LoRaWAN is significantly higher than that of BT or Wi-Fi.

     

    BR

    Gilberto

    Andrew Neil
    Andrew Neil
    Super User
    Oct 28, 2026 11:47 AM UTC

    Sure, it could be an EMC issue.

    Have you done any EMC testing?

    You could try putting an independent LoRa transmitter, at similar power, near the LSE - see if that has the same effect.

    Can you adjust your LoRa TX power? Does it make any difference?

     

    As others have said, PCB layout would be key here - so it would help to post your schemaic, layout, and some good, clear photos.

     

    I can not edit the title. Please correct it

    See:

     

    A complex system that works is invariably found to have evolved from a simple system that worked.A complex system designed from scratch never works and cannot be patched up to make it work.
    Terms & ConditionsAccessibility statement