Measure current of nRF52840-DK with a connected secure element via I2C

[MarDor]

Hi,

I use the nRF52840-DK together with a secure element (Infineon Trust X) and want to measure current/powerconsumption of this whole setup during certain operations.

At the moment my setup looks like this:

• OPTIGA Trust X connected to the nRF52840-DK via Arduino-Pin headers
• Poweranalyzer connectet to P21
• SW9 -> VDD
• SW6 -> nRF only
• J2 not plugged in

But I am not sure about SW10 if I want to use Arduino-headers for powering and communication with the Trust X.
Can someone give me some input on this on how to setup up those switches correctly for this measurement?

With best regards,
Marc

[Stian Røed Hafskjold]

Hi, not sure if that's going to work. You need to set the SW10 to "ON" in order to isolate the debugger MCU from the nRF chip. Then the USB cable will have to be plugged in to power the debugger MCU, and the P21 will power only the nRF chip. But, If you use the Arduino headers for the Infineon board, it will be powered from the same domain as the debugger MCU on the DK, which is now powered from the USB, and thus measurements on P21 will not include the infineon board.

If you leave SW10 in the "OFF" position, you will measure the whole setup including the debugger MCU which will add several mA to the measurements.

I think the only solution is to remove the Infineon board from the Arduino headers, and do the wiring manually. Here's what you can try:

• Power analyzer connected to P21
• SW10 -> ON
• SW9 -> VDD
• SW6 -> nRF only
• J2 plugged in
• Add jumper to P22 if SB40 was previously cut
• Connect VDD_nRF on P20 to VDD on Infineon board
• Connect GND from DK to Infineon
• Connect I2C wires from DK to Infineon

[MarDor]

Thanks, for your support. I will try it with this setup.
If SB40 is not cut I do not need any additional jumper right?

[Stian Røed Hafskjold]

Ok, let me know how it goes. That is correct.

[MarDor]

The J2 must be pluged in for the measurement, even though when the debugger MCU is not powered?

I took a mseaurement with this setup. Average values were:

During Access of the SE: 20mA
During Zephyr's k_sleep(): 7mA
Complete Average: 8mA

Are there any experience on your side of power consumption during Zephyr k_sleep() on the nRF52840-DK?

[Stian Røed Hafskjold]

Yes, the USB cable has to be plugged in to power the VDD domain on the DK, this includes the debugger MCU, the IC switches, etc. If they are not powered they will be floating and you get unpredictable behavior, and you will get leakage through the SWD pins.

I tried a Zephyr example now, just putting the k_sleep(1000) in the main loop, and not initializing anything more and I get a current draw of 3 uA on the external supply header. With the same setup as I described below. Without the SE.