STM32WL5MOD Questions

[DWalk.3]

I'm working on a new design with the STM32WL5MOD and ran across a few questions from the datasheet.

• The datasheet indicates, "Filtering capacitors on power-supply pins and components for the SMPS are already integrated into the module." The schematic that I've seen for MB1779 shows decoupling caps. Are the decoupling caps required, recommended, or not required?
• The layout for MB1779 shows vias under the device. Is copper allowed under the LGA92 package?
• I assume the LGA92 is shielded, so is an external shield not required?
• There aren't any specifications of the LSE oscillator in the datasheet, but I did notice that the high LSE drive level is required. It would be nice to have a specification on the LSE frequency accuracy (short- and long-term) in the datasheet.

[DWalk.3]

Also, what type of 32 kHz RTC crystal is used? Is it an AT-cut crystal, or is it a tuning fork oscillator?

[PLaid.1]

Excellent questions. I also need to know this.  I hope someone (who knows) answers this.

My observation is that the MB1779B has pads for the bypass capacitors and they are marked as "DNP" and indeed they are not there.  Can I safely design a board without these caps? 

[DWalk.3]

@PLaid.1 I’ve learned (with some degree of pain) that the best way to get answers to your questions is to submit a support ticket. The developers and product experts don’t always read this forum. 

FWIW, my design with the module was done without the decoupling caps and it is working fine. I guess I believe them when they say they are included in the part. My board, of course, uses a 100uF cap for bulk decoupling. I’m pretty sure I also had visa under the part, but I did try to minimize them. I never got answers to my other questions (I never submitted a support ticket), but I suspect the LSE is a tuning fork type crystal because it’s pretty small.

If you do get answers from a support ticket, post them back here! Curious minds want to know!

[PLaid.1]

Thanks for reply.  Mine will use a SMPS with very low quiescent current for applications that have long sleeps in remote areas. I'll post any interesting findings for sure.