STM32L462 Load Capacitors

[Remco Kleine]

Hello,

I am using an STM32L462 microcontroller with a crystal from Wurth: 

830 016 178

This crystal has a load capacitance of 6 pF

I calculated the two load capacitors CL1 and CL2 to be 4.3 pF

Now, everything is working OK, and the RTC is running fine...

But as soon as I measure the signal with an oscilloscope ( 10 M ohm probe 12 pF) the signal diminishes to zero)

Now, I understand that I add an additional 12 pF to the loop, but I am afraid that my solution is not stable.

The drive level in the firmware is set to low.

And if I turn it to high, it makes no change.

The traces to the micro are 4mm and the other 5 mm long.

In the reference manual of the STM32L462 they use also a chrystal with a load capacitance of 6pF ( with stopmode specifications they tell), but they have 6.8 pF capacitors

can anyone advise me?

[Danish1 · Accepted solution]

I assume you have read ST's AN2867 Oscillator Design Guide.pdf  - if not, read it now.

They suggest using a probe with an input capacitance of 0.1 pF. I don't have one of those, but I found a x100 probe worked for me.

[Tesla DeLorean]

Should always try to measure indirectly. The LSE should route through MCO or one of the RTC outputs (TAMPER?), or be measurable by one of the TIM counters.

Your values for the capacitors seem more reasonable than the reference design you cite.

You could presumably measure / record the start time, frequency, and do repetitive restart tests.

[Remco Kleine]

Aha, I read the applicaion note, and I calculated the two capacitors form there.But I didn't read the part about the special probe. 

I think it is normal behaviour what I see. The probe has just too much capacity.

Thanks for your tip!

[Remco Kleine]

Yes, if I measure indirectely on the tamper output, it show 512 Hz.

And also I can store the board, with the RTC running on a coin cell, and the time/date is correct after a week or so.

So, it is working.

I was just wondering if the oscillator was stable, but I think I overload it with my probe.

I am less worried now, after receiving these answers.