Modem only reached lowest power after 60 secs

[loumay]

Hi,

I try to understand the power consumption of the UDP example for NRF9160 with PSM.

I'm using the default configuration: RAT="00000000", RPTAU="00100001", no eDRX, no RAI.

Modem firmware version=1.2.3

Connect SDK=1.5.0

Board is NRF9160dk, with my own SIM card.

After sending the UDP package, the consumption drops to 50 [uA] for 55 [s]:

After this 55 secs, there is a 8.30 [mA] peak of 258 [ms]. Then the power consumption reached a floor current of 5.8 [uA].

Is it normal to have this 55 [s] at 50 [uA] before reaching the 5.8 [uA] floor current ?

Why the floor current is 5.8 [uA] and not 2.7 [uA] ?

Thanks,

Louis

[Heidi]

Hi!

No, this is a bit strange. It could be the SIM turning off, but it shouldn't take one minute to turn it off. Can you measure on SIM 1V8 to see if there's any voltage there, just in case?

Also, check AT+CEREG, AT+CEDRXRDP and AT%XMONITOR to see what values you're getting from the network. They're not necessarily the same as what you're requesting. 

A theory is that the active timer has been set to one minute and the eDRX interval has been set to something higher than one minute. If this is the case, the eDRX event will nver come but the SIM card will be turn on until the active timer runs out. 

Without knowing more about our application, the 8.30 mA spike looks like a timer that runs out after a minute? Could that be a possibility?

As for the floor current, what version of the nRF916 DK do you have? The floor current consumption you're seeing looks closer to the consumption seen on the older chips. If you're measuring through the current measurement header, you should use an external power supply and pull out the USB cable, Otherwise this causes some noise which leads to a slightly higher current consumption. 

Best regards,

Heidi

[loumay]

Hi Heidi,

Thanks for your feedback.

> Can you measure on SIM 1V8 to see if there's any voltage there, just in case?

The sim card is only powered off after the last peak.

> Also, check AT+CEREG, AT+CEDRXRDP and AT%XMONITOR to see what values you're getting from the network

+CEREG: 5,5,"0650","010CD706",7,,,"00000000","11100000"
+CEDRXRDP: 0
%XMONITOR: 5,"","","22801","0650",7,20,"010CD706",138,6300,37,28,"","00000000","11100000","01010011"

A TAU=11100000 means like PSM is not activated right?

> Without knowing more about our application, the 8.30 mA spike looks like a timer that runs out after a minute? Could that be a possibility?

I'm using the connect sdk udp example. So there should be no active timer in my understanding.

> As for the floor current, what version of the nRF916 DK do you have?
I'm using a PCA10090, 0.9.0

Using an external power supply drop the floor current to 5.28 [uA] only.

[Heidi]

If you're using the UDP sample in NCS unaltered then the only explanation for the strange current consumption is the network.

Could you please take a modem trace and provide it with the current consumption and log from the application?

As for the floor current, you have chip revision 1 where the typical PSM floor current is 4 uA (see nRF9160 PS v1.0). There are some leaks in the older DKs as well, so you should set the VDD_IO switch on the DK to 3V to avoid this. Then you should get closer to 4 uA.

To get 2.7 uA, you will have to buy a newer DK that has the newer revisions of the chip. 

Best regards,

Heidi