2.2V on VCAP pin on STM32H747BIT6

[bacaneriaslocas]

Hi, I started working on my own dev board for a flight computer based on the STM32H747BIT6.

At first, I struggled a bit to connect the board to the computer, only managing to do so using the hot plug mode in order to get some info from the microcontroller. My problem comes when trying to write something to the flash: it gets stuck on erasing internal memory sector 0. This means it is unable to write or modify the flash memory. Also, when I connect the board to the STM32CubeProgrammer, it shows me the flash memory, but every address is at 0, which makes sense given that I haven’t uploaded any code to the microcontroller.

I’m not really sure what this means — I’m able to read it, but not write to it.

The thing is, I started measuring the power source pins just in case I spotted something weird, and I sure did: my VCAP pins are at 2.2V, which is strange because they should be at 1.2V according to the internal LDO regulator.

My guess is that this may be the source of the problem — an incorrect power source for the flash memory — but I’m not sure.
Here I attach my squematic.

[bacaneriaslocas]

Btw, here is a picture from stm32cubeprogramer whith info of the mcu

 

[TDK]

VCAP at 2.2 V means there is a hardware issue (or you're not actually measuring VCAP). Probably a design issue or manufacturing issue. Include your schematic if you want. Be sure to follow the hardware design guide.

 

> every address is at 0, which makes sense given that I haven’t uploaded any code to the microcontroller

If the microcontroller is fresh, all flash is at 0xFF, not 0. Likely there is an issue actually reading the memory. Probably due to the underlying hardware issue.

[bacaneriaslocas]

Sorry, I thought I had already shared the schematic. Here, I attach it.

 

If the microcontroller is fresh, all flash should be at 0xFF, not 0. It’s likely that there is an issue actually reading the memory, probably due to an underlying hardware problem.

Speaking of this, something I have noticed is that sometimes the memory shows 0xFF when connecting, but I don’t know why or when it happens.

 

Thanks for the help, I will take a look at the hardware design guide.

 

[TDK]

I don't see anything critically wrong here.

 

VCAP pins should be connected to each other and VCAP1/2 should have a single 2.2 uF cap each. Might be the issue.

VBAT should be connected to VDD.

Why separate VSS and GND? A strong ground connection to ground at each VSS pin would be better. I suspect this might be the issue.

 

Hardware design guide:

Getting started with STM32H74xI/G and STM32H75xI/G MCU hardware development - Application note

[bacaneriaslocas]

Thanks for checking the schematic, I already corrected the VBAT and VCAP problems on my board but the problem of the 2.2V is still there.

Also, VSS and GND are on the same line, (not my best schematic design decision).

These are the logs that show up when I connect the board

 

and these are the logs when I try to erase the flash

I'm sending this because this was the source of my problem but I'm not sure if this has something to do with what I have comented about the 2.2V on VCAP.

Could this be asociated with a hardware problem or software?

[TDK]

This is a hardware issue. Correct the hardware issue and the software problems will go away.

Can you show your layout? Large in-focus pictures of the board with the fixes you made may help too.

[AScha.3]

Just a simple question: you write : STM32H747BIT6

but you use only LDO - why?  afaik it should be using smps + LDO (for max speed)

and vcap 2.2 uF ....not your many caps...

Try to change the vcap and check (if cpu still alive...).

And no, its not a fake. Maybe its a damaged cpu now.

[bacaneriaslocas]

"but you use only LDO - why?  afaik it should be using smps + LDO (for max speed)"
Correct me if I´m wrong but I think for the max speed is only necesary the LDO, but I´m not sure.

I also tried removing the VCAP capacitors leaving only two but no diferent result.

Do you think it could be worth trying to bypas the voltaje regulators and supling the 1.2V externally to se if the cpu is okey?

[AScha.3]

That should work....but is just a experimental attempt. LDO input should be...same, 1.2V .

vcap IS the core voltage.

Try with watching the current, while rising from 0 to 1.2V ; if current goes > 200mA ...forget it.

If lower...maybe still ok. Should connect then , VDD at 3v3, to st-link. Then you know...