cancel
Showing results for 
Search instead for 
Did you mean: 

STM32H725VGH6 / Direct SMPS Supply / VLXSMPS Connection

SammExa
Associate

Hi,

We are using the STM32H725VGH6 MCU and are plannign to use the internal SMPS for VCORE.

With regards to the connections required for the VLXSMPS pin, I would just like to confirm that a 220pF capacitor is required on this pin in the "Direct SMPS Supply" configuration. It seems unusual to have a capacitor directly on the output of a Buck SMPS. It is also not show in Figure 2 of "AN5419: Getting started with STM32H723/733, STM32H725/735 and STM32H730 Value Line hardware development". However it is recommended in Table 2 & Figure 3.

Figure3.pngTable2.pngFigure2.png

 

4 REPLIES 4
M_P
Associate II

I've been wondering about that too and am very interested in any techincal reasons ST can provide.   220pF directly on the switch node is very unusual.  I observe severe rining on the node between the specified 220pF and 2.2uH inductor (Fresonant = ~7.5 MHz)

M_P_0-1698768517459.jpeg

 

M_P
Associate II

Linking a similar thread: https://community.st.com/t5/stm32-mcus-products/how-to-prevent-sporadic-ringing-on-vlxsmps-pin-in-quot-direct/m-p/80532

This thread explains the ringing when both switches are off in pulse skipping mode.  However, it does not explain why the 220pF is specified in the first place.  To a power engineer, the 220pF directly on the switch node looks like an error in the schematic.  It would be expected to cause additional switching loss (C*V^2*F) and high peak currents in the SMPS transistors..

SammExa
Associate

Hi @M_P

I did see that thread, and it does make sense why the ringing is there, but I agree that I'd like to understand the purpose of the 220pF cap as we flagged it in our review that this was an unusual architecture for a switching node. 

I have lodged a support ticket, so I will update you if I get an answer.

M_P
Associate II

I got this response from ST: 

"I'm confirming that this 220pF is indeed required for designs using LQFP/BGA packages as shown per table 2 in the forum post. 

 

I can see from an end customer's POV how this could be confusing as in Figure 2 there's no explicit connection of a 220pF cap, however the documentation team probably had left that out to avoid confusion for other customers who are not using a LQFP/BGA package. 

I think it would make sense to update the schematic in figure 3 with a note that says 220pF only if device package is of type LQFP/BGA. I'll work on escalating this point with the documentation team. 

 

Furthermore, if you look at our STM32H753I-EVAL's schematic, page 4 where the power supply is configured, you can see a 220pF capacitor connected on the VLXSMPS pin. This board embeds our STM32H753XI which is of package type BGA. "