FEDEVEL
Platform forum

KA , 05-27-2025, 10:27 AM

Hi everyone,

I've been using passive crystals in my STM32 designs for a while now and was under the impression that this was the standard approach. Recently, I came across a schematic where an active oscillator module was used instead. This is new to me, and I'm curious to understand the pros and cons of each option.

I’ve started reading into the topic and found discussions around failure rates, EMI, startup times, and accuracy. I’d love to hear from those of you who have switched to using active oscillators:
- What drove that decision?
- Have you noticed any tangible benefits (e.g., reliability, EMI performance, layout simplicity)?
- Are there specific use cases or constraints where one clearly outperforms the other?

Any insight or experience would be really helpful. Thanks!

QDrives , 05-27-2025, 09:28 PM

I know the downside of the oscillator. Like you said -- **EMI**.
EMI to the level of failing compliance. Then it does not matter what the benefits are.
All I could do was go from a 48MHz device to an 8MHz device as that did reduce the signal rise time reducing the EMI to just below the limit.
It was not my design, not design decision to go for oscillator.

Robert Feranec , 05-28-2025, 05:46 AM

we used oscillators in some industrial designs to help with starting in very low temperatures. also we used to use oscillators with some chips (some old chips were very sensitive about capacitance on crystal pins and some board in some occasions didn't work), but I think now the chips are much more tolerant

KA , 05-28-2025, 06:27 AM

Interesting. Thanks for sharing the insights!