STM32 VCAP pin problem

[fing]

 Hello,

I have designed a PCB with 3 STM32H725VGT6 MCUs, and assembled them according to my schematic:

(sorry for this resolution, this is from a converted pdf)

All 3 MCUs are the same, from this schematic.

R5 and R2 are just 0R jumpers to select the boot mode.

At first, ST-Link couldn't find any of them.  When I checked the voltage on the Vcap pin the MCU1 had 1.0008 V at both 2.2uF caps. STLINK still can't find the target. MCU2 and MCU3 had 0.0000V. on the Vcap pins.

All 3 MCUs have their:

- Orientation in the right place

- VDDA / VREF+ connected

- NRST at high 

- 3.3V at all the 100nF decoupling caps

- 2.2uF caps from Murata (ESR < 8 mOhm @ 1 MHz) GRM155R6YA225KE11D

So I was wondering maybe the ICs were bad (which is unlikely considering that they are from Mouser)

I checked the soldering and there were no bridges, but these symptoms are usually caused by some sort of soldering issue.  What else should I check?

[TDK]

Were the PCBs checked electrically before soldering? Who made the boards?

Continuity/resistance test between VDD/GND and VCAP/GND might show some issues.

I agree it's unlikely to be a chip issue--more likely to be a workmanship issue.

[CMYL]

• Can you compare the BOM and schematic versus the recommendations into the Getting started with STM32H723/733, STM32H725/735 and STM32H730 Value Line hardware development -
• Can you power-up only one MCU by STLINK VDD, isolate the 2 others and check if connecting to STLink ?
• How the 3 MCUs are powered together, if using only STLINK VDD, I think the current feeding the 3 MCU is not sufficient !!! external power supply is mandatory in this case.

Best regards

[fing]

Yes, JLCPCB made the boards and they passed the flying probe test. 

• VDD-GND: 3.998 kΩ

MCU1 (which has 1.0008 V on Vcap):

• VCAP-GND (2.2 uF connected to pin 97): 18.10 kΩ
• VCAP-GND (2.2 uF connected to pin 44): 18.09 kΩ

MCU2:

• VCAP-GND (2.2 uF connected to pin 97): 13.00 kΩ
• VCAP-GND (2.2 uF connected to pin 44): 12.98 kΩ 

MCU3:

• VCAP-GND (2.2 uF connected to pin 97): 20.68 kΩ
• VCAP-GND (2.2 uF connected to pin 44): 20.66 kΩ

I might rework the board and maybe replace the caps with fresh ones as well.

[fing]

Yes, it actually follows the LDO operation, where VDDLDO = VDD. They are powered externally with a power supply giving 3.3 V.

[fing · Accepted solution]

Update: 

After some investigation, I found that the main source of the problem was actually caused by incorrect settings in the .ioc file. 

These are the correct settings for LDO operation.

I can only upload these incorrect settings once. After that, the MCU spits out 0 V on the Vcap pin and it is basically bricked. I have read that if you manually inject 1.3 - 1.4 V on a Vcap pin, you can bring it back to life.

But this was not necessary for my case, because I was able to desolder my BOOT0 jumper and pull it high. Now, ST-Link can detect the MCU, upload and verify the code.

The current consumption went up a little, which meant that the MCU was up and running with 550 MHz.