cancel
Showing results for 
Search instead for 
Did you mean: 

stm32u5 vdda abnormal current consumption

junichi ohashi
Associate II

I set several unused pins in stm32u5 to floating as hardware and set them to low output by pushpull for ESD protection.
The microcontroller was heating up abnormally when I looked at it with a thermo camera, and I found that the current consumption of VDDA was mysteriously high, and the current consumption changed within a few seconds.

When the unused pins were set to analog input, the current consumption of VDDA dropped by about 30 mA, and the current consumption became the same all the time.

It is very strange to me that the VDDA current consumption increases instead of the VDD current consumption increasing.

Is there any possible cause for this?

1 ACCEPTED SOLUTION

Accepted Solutions
junichi ohashi
Associate II

Sorry for the delay in replying.

We have found the cause of the problem and are successfully moving toward mass production.

The cause was a bug in our software, which supplied the clock to the MCU's built-in opamp mounted on a pin that was set to Low as an unused pin, even though it was not used, and also set it to enable state. (Perhaps the code got mixed up when someone was debugging on another board)

Because the opamp was set to Enable and the pin was dropped to Low, current flowed from the opamp output pin to GND up to the limit of the Opamp, which caused the abnormal current consumption of the VDDA.

This was completely a glitch in our software.
Thanks a lot for your support.

View solution in original post

9 REPLIES 9
TDK
Guru

It sounds like a hardware/layout issue. Possibly those pins are not floating but are tied to VDD, so output low just sinks current? Which pins/chip in particular? Does current stay high if you configure them in input mode?

If you feel a post has answered your question, please click "Accept as Solution".
STOne-32
ST Employee

Dear @junichi ohashi ,

To debug this extra consumption, please provide the package / part number / PCB and also the firmware. Most probably a conflict either in PCB or software. Here are the I/O types - the ones with _à are connected to VDDA . Hope it helps you 

IMG_3794.jpeg

Billy OWEN
ST Employee

Hi @junichi ohashi 

 

The forum moderator had marked your post as needing a little more investigation and direct support. An online support case has been created on your behalf, please stand by for just a moment and you will hear from us.

 

Regards,

Billy

AScha.3
Chief

To get the lowest power consumption, hal/cube has options to set "all unused pins to analog" , so no need to gamble around with other settings and wonder about different effects.

If you feel a post has answered your question, please click "Accept as Solution".

Sorry for the delay in replying.

I think you are right if the VDD current increases, but I am wondering because the VDDA increases.

In the design (on the circuit CAD), Pin is indeed NoConnect, and I was worried about shorting solder balls etc. in the mounting, but all ()

Yes, I understand that.
However, I had dropped unused pins to Low in reference to AN4899 to increase ESD immunity.

This caused this phenomenon, so I set the input to analog floating and the abnormal current consumption disappeared, so I found that the cause was caused by setting low in output pushpull.

However, I am wondering why the current consumption of VDDA has increased by more than 30mA, although I can understand if the current consumption of VDD is increased by this setting.

I would like to know what logic is causing this phenomenon.

 
 

 

 

The part used is STM32U575QGI6.
As for the PCB and firmware, they are our assets and I cannot give them to you at my discretion. I am sorry. I can give you a fragment of information.

I will talk to my boss and try to do the same experiment with ST evaluation board.

I am not familiar with analog switch circuits,
If there is an analog switch, does the current flow from VDDA to GND when the PushPull setting is set to Low output?


From the schematic in the reference manual, I thought that even if there is an analog switch, the current consumption of VDDA would not be abnormally high due to the VDD port setting.

 

stm32U5

junichiohashi_0-1689313622415.png

stm32l4

junichiohashi_1-1689313803431.png

 

 

TDK
Guru

> An increase of 30 mA

Sure sounds like a short. The current limit of GPIOs is about that.

One possible exercise:

  • Set the pin to input with pullup enabled, verify the level is 1.
  • Set the pin to input with pulldown enabled, verify the level is 0.

If that's not happening, there's a hardware issue.

You're right that current should be drawn from VDD rather than VDDA, but perhaps investigating the problem more would provide insight.

If you feel a post has answered your question, please click "Accept as Solution".
junichi ohashi
Associate II

Sorry for the delay in replying.

We have found the cause of the problem and are successfully moving toward mass production.

The cause was a bug in our software, which supplied the clock to the MCU's built-in opamp mounted on a pin that was set to Low as an unused pin, even though it was not used, and also set it to enable state. (Perhaps the code got mixed up when someone was debugging on another board)

Because the opamp was set to Enable and the pin was dropped to Low, current flowed from the opamp output pin to GND up to the limit of the Opamp, which caused the abnormal current consumption of the VDDA.

This was completely a glitch in our software.
Thanks a lot for your support.