PSM floor value double (8uA instead of 4uA) and eDRX values bigger ( 24uA instead of 19uA) compared to Product specifications using FW 1.1.1

[CommTech]

HI,

We are evaluation the nRF9160 in the devkit and we are trying to obtain the values specified in the nRF9160 Product Specefication v 1.1. After following the different software configuration and HW setups (ex: https://devzone.nordicsemi.com/nordic/cellular-iot-guides/b/hardware-design/posts/measuring-psm-idle-current-on-the-nrf91-dk), we are able to properly activate PSM and eDRX using different values. We are able to achieve pretty slow values but they are all bigger than in the Product Specification :

• PSM floor: 8uA instead of 4uA in the Product Specification
• eDRX cycle for 81.92 eDRX cycle: 25uA instead of 19uA (configured PTW = 1 (2.56sec.) and DRX=2.56)
• Test details:
  • Technology: Cat M1
  • Voltage: 3.7V
  • Current measurement tool: Otii from Quoitech (properly calibrated and achieving the expected low values in other modules that we are evaluating)
  • Modem Firmware: 1.1.1

Do you confirm that modem FW 1.1.1 is able to achieve the values in specification? Any other parameter non stated here that we should take into account.

Here you have some screenshots of the measurements.

eDRX:

PSM

Thanks for the help,

[Heidi-Irene Harvey Sollie]

Hi,

You should set the VDD_IO switch (SW11) to 3V. There is leakage through the RESET pin when VDD_IO is at 1.8 V.

The current consumption during an eDRX cycle is determined by the sleep current for the SIM card, and some cards pull more than others. So the increase is due to the SIM card you're using. 

You can measure the current from the SIM_1V8 pin from the nRF9160 SiP to find this value from the SIM card.

Best regards,

Heidi

[CommTech]

Thanks Heidi,

For PSM, the proposed change is allowing to reach the value... thx!

For eDRX, aren't you using the UICC deactivation feature ? We make the supposition that is what is used and we are not expecting any current leakage here as power should be cut. Are we wrong?

We have also made another test powering only the nRF91 and not powering the rest of the devkit (no external, no usb). Doing this, we have an increase of ~60uA in both PSM and eDRX. Is the nRF powering an external component of the devkit (due to internal pull ups,....)? Can we avoid this without connecting any of the possible external supplies?

Editing the post for an update: of the last point... Double checking the schematics and the PTS, we think that this may come from VDD_GPIO as per the comment " If VDD is supplied and VDD_GPIO is grounded, an extra current consumption can be generated on VDD". Are we correct? In that case, what is the extra current that is drawn in hibernate / PSM due to this Voltage? Will it depend on the state of the GPIOs or do we have any other considation to take into account?

Thanks

[Heidi-Irene Harvey Sollie]

Hi again, apologies for the long wait, we've had Easter vacation in all of Norway for the past five days.

So for your first question:

The SIM should turn off in eDRX mode, but it depends on the eDRX interval you have. The SIM card itself specifies a minimum interval, so if the eDRX interval you set is faster than this, the SIM will not deactivate. Try setting the eDRX interval to 655 seconds.

You can also measure the current on SIM_1V8. If it's 0V, then the SIM is turned off.

For your second question:

Yes, you can power the nRF9160 SiP, without the rest of the DK.
The nRF9160 has two power inputs: VDD1/VDD2 and VDD_GPIO. VDD_GPIO is powered from the same voltage domain as the rest of the DK. So when you don't have the USB cable plugged in, VDD_GPIO doesn't get powered.

You will need the USB cable connected to avoid current consumption leakage. And you can measure the current going into VDD1/VDD2 by connecting a power analyzer to the "current measurement" header.

Another reason to have the USB cable connection is to avoid leakage from the nRF9160 and out to RESET and SWD interface.

Best regards,

Heidi

[Thomas]

Hi Heidi, 

This leakage current - does it apply only to the devkit? Or is it an issue related to the nrf9160 itself? If not just the dev kit, what is the minimum voltage required for VDD_IO to avoid the excess current?

Kind regards,

Thomas

[Heidi-Irene Harvey Sollie]

Hi! The leakage only applies to the DK, not the nRF9160 SiP itself. 

Best regards,

Heidi