External supply power consumption is different from current consumption measurement using PPK2

[Ashishsult]

I am using NRF9160 (PCA10090, 1.0.0, 2021.7) for my experiment using the .hex file of UDP example provided here (https://devzone.nordicsemi.com/nordic/cellular-iot-guides/b/hardware-design/posts/measuring-psm-idle-current-on-the-nrf91-dk) at 3.6 V. 

There are two modes to measure the current consumption, namely Source and Ampere.

I tried both of them and found different results as with the Ampere meter setting I can see the device in PSM mode consume 8.91 uA but in Source mode, the device is not moving to the PSM state. What are the reasons?

Plus next, I tried to check the power consumption from the external supply pins P21, so that to know how much time a battery can last running a similar  NB-IoT application. But it seems that it consumes around 8.96 mA in PSM state which is too much. 

Therefore, my question is how to reduce this power consumption to a few uW.

FYI: if we add "while(1) k_cpu_idle();" at the very beginning of main(), the current consumption is 9.07 mA. 

[Håkon Alseth]

Hi,

 

How are you connecting the PPK2 to the nRF9160 when its in "source meter"?

 

Kind regards,

Håkon

[Ashishsult]

I did the setup as mentioned here: https://infocenter.nordicsemi.com/pdf/PPK2_User_Guide_v1.0.1.pdf

I re-attempted the test now the source and ampere mode have a similar result. 

But power consumption from the external supply pins P21, is still in mA. Do you have an idea if I need to somehow disable the peripherals to reduce the power consumption or there is any other solution?

[Håkon Alseth]

Hi,

 

Ashishsult said:

But power consumption from the external supply pins P21, is still in mA. Do you have an idea if I need to somehow disable the peripherals to reduce the power consumption or there is any other solution?

Are you connecting as described here?

Ie. the GND pin from P21, then the "VOUT" to the VDD_NRF:

 

Kind regards,

Håkon

[Ashishsult]

Yes, for the source meter mode, as I told you with this connection and also with the ampere mode, the device can now go to the PSM mode consuming only around 8 uA (my second attempt).

But if I connect Vout and GND (PPK2) to the P21 (nrf91 DK), the current consumption is in mA. This is my main concern because I want to power the device with a battery and want to increase its life using the PSM feature of NB-IoT.
As current consumption having USB connected shown numbers in uA which is good to know but in practice, if I can also achieve that then it's useful for me.

[Håkon Alseth]

Hi,

 

If you measure on P21, you are powering more than just the nRF9160 (switches +++). If you follow the "VSUPPLY" (which is the net P21 powers), you can see this clearly from the schematic/layout:

https://www.nordicsemi.com/-/media/Software-and-other-downloads/Dev-Kits/nRF9160-DK/Hardware-files/nRF9160-DK---Hardware-files-1_0_0.zip

 

Since features like LEDs etc are buffered, and there are connected lines between the debugger and the nRF9160 on the DK itself, you will have some leakage when powering only the nRF9160.

 

To measure the nRF9160 isolated, you can power it as shown in my former post (w/o usb connector present, you will get a lower consumption if USB is on). This will still have some leakage, but not in the high mA range:

 

Kind regards,

Håkon

[Ashishsult]

I understand that buttons, LEDs, Pins consume power. I worked a bit with nrf52840 Dk board where we can not initialize them and then the power consumption can reduce. Is it possible with nrf91 DK to do so? 

For me, this is important because I want to power the device with an external battery and keep the device in PSM mode for a long time to save energy. But with the current status of the application and hardware, I can't because the power consumption came out to be 8.96 mA in RRC idle state. I can tolerate few micro-amperes leakage currents but from uA to mA is too much. 

[Håkon Alseth]

Hi,

 

Ashishsult said:

But with the current status of the application and hardware, I can't because the power consumption came out to be 8.96 mA in RRC idle state.

As shown in the DK schematics and layout, P21 will power the whole board.

This is a development kit restriction, and will not apply when you create your custom board.

 

As I mentioned, if you power it as I showed, the leakage will be lower (but it will still leak through switches etc.)

Here is the description on how to measure the nRF9160 isolated:

https://infocenter.nordicsemi.com/topic/ug_nrf91_dk/UG/nrf91_DK/hw_measure_current.html?cp=2_0_5_5

 

It also states:

The use of a USB connector is not recommended for powering the Development Kit (DK) 
during current measurements due to potential noise from the USB power supply. 
Instead, the DK should be powered by connecting a power supply to connector P21.

 

Kind regards,

Håkon