nrf5340 DK 1.0.0 draws more current than 2.0.0

[Frikkie Badenhorst]

Hi there,

We have recently acquired some new nrf5340DK's and we suddenly noted that the SYSTEM ON LOW POWER current increased from about 5 uA to about 25 uA. 

Attached below is an example project we received which we are using as reference. When uploading this onto the new DK 2.0.0 hardware, we are measuring a current of 24.25uA:

When uploading the exact same firmware onto the older 1.0.0 hardware, we are measuring about 5uA:

Could you please advise whether you can replicate this issue? Is this a known issue?

Our hw setup is as follow, with SB40 cut:

2063.button2 (1).zip2577.button2 (1).hex

[Frikkie Badenhorst]

Hi there,

Is there any update on this perhaps? We are able to replicate the issue on all 2.0.0 DK's, while the 1.0.0 DK's are still running fine with the lower power measured.

We suspect that the new 2.0.0 DK with an additional nrf5340 might add some leakage current to the measurement, but it is difficult to tell.

I really hope we can get a response soon as we are planning on deploying test devices in the field very soon, and it is thus important that we understand the current draw.

Regards,

Frikkie

[Priyanka]

Hi,

I tried reproducing the issue here, but there was not much a difference in the consumption and both showed similar results. Did you try reproducing this with another rev 2.0.0 DK? Also, are you trying in the Sourcemeter or the Amperemeter mode?

Best Regards,

Priyanka

[Priyanka]

Hi,

It seems the problem is with the setup when testing with the PPK. Could you try shorting the following pins?

What happens is that, when you change the supply 3.3V from the PPK, it only supplies to the 5340 and the rest of the board is at 3V which causes a leakage current. Now when you short the pins, the entire board is supplied. The fact that this problem arises only in the rev 2.0.0 could be because in the new revision, the SEGGER MCU has been replaced by the nRF5340.

Hope this helps

-Priyanka

[Frikkie Badenhorst]

Thanks Priyanka, we will investigate this and keep you updated. Either me or Kimberly will respond.

[Kimberly]

Thank you, Priyanka.

Shorting the above pins did reduce the current when running the 2577.button2(1).hex file. 

Without short, average 20.53 uA over ~2s: 

With short, average 3.74 uA over ~2s: 

Kind regards,

Kimberly

[Priyanka]

Glad to hear that it works

-Priyanka

[Frikkie Badenhorst]

Hi Priyanka, 

As Kimberly mentioned, the above works, however we do not really understand why this is working the way it is.

1) What is the VIO_REF pin used for? You mentioned that the rest of the board is running at 3V if we do not short this pin, however when measuring the pin voltage without the short we measure 0V? We are measuring 0V on both devkits. This does not makes sense does it?

2) Is this a hardware error by Nordic? (e.g. is this something that is going to be fixed in future dk revisions?)

3) Does shorting the pins on the 2.0.0 devkit allow for a TRUE reading of the current profile?

It is important that we understand the implications of shorting this pin and whether the current profile we measure is accurate and reliable.

Hope to hear from you soon.

Kind regards,

Frikkie

[Priyanka]

Hi Frikkie,

As I mentioned earlier, this might have been cause because of the replacement of the previous MCU with the nRF5340. We are looking into this. Connecting to VIO_REF basically takes it to the required VDD level.

You can also try by powering from PPK2 to P21 in sourcemeter mode. Then you might need a jumper on P22 and also set SW10 to ON. 

-Priyanka