cancel
Showing results for 
Search instead for 
Did you mean: 

STM32H747I-DISCO ground question

Tycho
Associate

Tycho_0-1712152017735.png

On STM32H747I-DISCO schematic, there are two different grounds. Why is SMPS ground connected to B6, C16 and T12?

1 ACCEPTED SOLUTION

Accepted Solutions
SofLit
ST Employee

Hello @Tycho ,

I'm getting back to you even after a while (receiving a feedback recently from experts) regarding this topic.

Two points:

  1. According to the schematics SMPS_GND and the ground are already connected via SB9SofLit_3-1718287297419.png
  2. According to the board designer the three VSS balls are internally close to some VDDIO33_LDO pins, and consequently, must be connected to SMPS_GND and not VSS as usually, for better signal immunity of LDO and he recommends to follow this rule in your design.

Hope I answered your question.

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.

View solution in original post

8 REPLIES 8
SofLit
ST Employee

Hello,

I'm checking internally and I'll get back to you if I have an answer.

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.
SofLit
ST Employee

Hello @Tycho ,

I'm getting back to you even after a while (receiving a feedback recently from experts) regarding this topic.

Two points:

  1. According to the schematics SMPS_GND and the ground are already connected via SB9SofLit_3-1718287297419.png
  2. According to the board designer the three VSS balls are internally close to some VDDIO33_LDO pins, and consequently, must be connected to SMPS_GND and not VSS as usually, for better signal immunity of LDO and he recommends to follow this rule in your design.

Hope I answered your question.

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.

@SofLit wrote:

2. According to the board designer the three VSS balls are internally close to some VDDIO33_LDO pins, and consequently, must be connected to SMPS_GND and not VSS as usually, for better signal immunity of LDO and he recommends to follow this rule in your design..


Is that information published anywhere ?

I don't think so.. could be a note to be added in the AN4938: https://www.st.com/resource/en/application_note/an4938-getting-started-with-stm32h74xig-and-stm32h75xig-mcu-hardware-development-stmicroelectronics.pdf

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.

Generally the reason for having TWO nets and bonding them at a single point, is to control the flow of ground/return currents.

With a flood-fill ground plane the electrons will take their own path of least resistance.

With distinct nets, the paths can be controlled and directed, so switching currents around supplies can be isolated and kept away from parts of the circuit that are going to be more sensitive to constant instability.

Tips, Buy me a coffee, or three.. PayPal Venmo
Up vote any posts that you find helpful, it shows what's working..

Yes,  I definitely think it should go in there - if not in the RM ...

FYI,

I've escalated the point internally and it was planned  to add a note in a coming revision of AN4938 regarding this point (internal ticket number 184166).

To give better visibility on the answered topics, please click on "Accept as Solution" on the reply which solved your issue or answered your question.

The balls on the STM32H747I-DISCO schematic make sense as they are next to the VDDLDO balls but it's not so clear on the UFBGA169 package as some VSS balls are close to both VDD and VDDLDO.  It seems they share the VSS balls in two of the 3 instances of VDDLDO balls.

Which pins, if any, do you recommend for the UFBGA169 package?

UFBGA169.png