STM32H755 Supply Config Modes 2 & 3 on Custom Board

[bramble]

Hi,

I'm designing a PCB using H755, and the corresponding Nucleo's reference design is a helpful resource that I've been looking at closely.

I'd like to support supply config modes 2 & 3 on my board but would prefer to do this using jumpers rather than solder bridges, adding/removing resistors or changing capacitor values: all these being needed on the Nucleo. Although these component changes are simple enough, I'd like an approach that's more convenient and less likely to result in damage to the board if the cycle is repeated a few times.

Referring to the schematic MB1363 and Nucelo reference manual UM2408, I've highlighted the changes that need to be made when switching from config 2 (SMPS only) to config 3 (SMPS cascaded with LDO):

My proposed variant on this is:

• Replace R35 with a jumper, placed as close as possible and with short traces to VCAPn and VFBSD.
• Replace R34 with another jumper, similarly using short traces and placements.
• Replace SB75 & SB76 also with jumpers, short traces etc..
• When changing the VCAP capacitors, always leave the 2x100nF in place but use a jumper to add another 2x2.2uF in parallel with them.

Does this seem OK? I'm wondering if there's extreme sensitivity to parasitics that are the reason why ST didn't use a jumper scheme like this, or maybe I'm overthinking this?

Thanks in advance!

[Fred_ST · Accepted solution]

Hello,

Sorry for this late reply.
Your proposal is feasible, it should work, but you may not achieve the desired performance.
Because adding a jumper on the feedback will add an antenna and modify the behavior of the SMPS.
The same comment applies to jumpers on capacitors.

Regards

[bramble]

Hi @Fred ,

Thank you, this is exactly the advice that I was looking for - I had a feeling that there was a risk with parasitic effects such as this. I think that I will proceed and I then have the option to not fit jumpers, but instead use tiny pieces of wire as zero-ohm links, which isn't ideal, but should be markedly better than a conventional jumper, whilst allowing me to determine whether my more convenient jumper arrangement can indeed work.