Current peaks in low-power udp example

RE: Current peaks in low-power udp example

Achim Kraus — Tue, 05 Sep 2023 10:39:47 GMT

Great.

RE: Current peaks in low-power udp example

hypn0 — Tue, 05 Sep 2023 09:06:49 GMT

Dear Achim,

Thank you for your time and answers. I really appreciate it.

I updated the modem firmware to the version 1.3.5 and carried out a new measurement:

The current peaks are gone, so I guess the update did the trick.

I will close this thread as solved.

Thank you and best Regards,

hypn0

RE: Current peaks in low-power udp example

Achim Kraus — Fri, 01 Sep 2023 14:25:42 GMT

> How can I lock the PLMN?

AT+COPS=1,2,"xyzAB"

(left question from your previous post:)

> And how is this period determined?

That's configured on your SIM card by your provider. Before you spend too much time in trying to circumvent a HPPLMN search, check if this is really the case.And check your IMSI with "AT+CIMI". HPPLMN search is just my first guess, when someone comes with charts like yours.

RE: Current peaks in low-power udp example

hypn0 — Fri, 01 Sep 2023 13:50:39 GMT

> For "low power cellular" this is poison for your battery.

True. I am currently trying to convince my customer that NB-IOT can be suitable for ultra low power applications as well. However, if I show him the measurements above, I am not so sure, if this will convince him :-D

I will update the modem firmware and carry out another longterm measurement.

I hope I wont need it, but just in case:

How can I lock the PLMN?

RE: Current peaks in low-power udp example

Achim Kraus — Fri, 01 Sep 2023 11:48:45 GMT

> it is not possible to turn off this search?

You may lock the PLMN, but that comes also with other drawbacks.

> What is the purpose of this search?

In my opinion, this is an historical artifact, which helped to reduce expensive fees for roaming. If you're abroad, you're mobile usage may be charged extra when using other operators as the one your provider is preferring. With that HPPLMN search that classic mobile phone use-case was optimized.

For "low power cellular" this is poison for your battery.

Anyway, the main pain was the searching for global PLMNs (that's the 9xxxx IMSI), where, why ever, Nordic believed and told, that this could be found everywhere.But in reality, it's not found at all, and so the modem search repeating for very long to find nothing.

It tokes then a year, and I don't what makes Nordic change their position, but with 1.3.5 the modem doesn't longer search for 9xxxx. What is left, that it may search for a HPPLMN in your country, if it's connected to an other PLMN. That doesn't harm too much, because it is much likely to find that real PLMN. There may be only very rare cases left, where in a region such a PLMN could not found. If you statically operate the device there, then I use the PLMN lock.

RE: Current peaks in low-power udp example

hypn0 — Fri, 01 Sep 2023 10:04:05 GMT

Dear Achim,

Thanks for the immediate reply.

I will try to read out the SIM parameter and update the modem firmware.

According to your post some time ago it is not possible to turn off this search?

What is the purpose of this search?
My hardware is connected to the network and can send the data without problems. During the measurement shown above, I did not move the hardware to another location, the environment remained the same. Why does the hardware have to search for the network every 2.5 h?
And how is this period determined?

Best Regards!

RE: Current peaks in low-power udp example

Achim Kraus — Fri, 01 Sep 2023 08:18:10 GMT

I guess HPPLMN search ;-).

Or some other kind of network search.

You may read out your SIM card to see the parameters for that.

See modem - read SIM

If you use a mfw version between 1.3.2 and 1.3.4, an update to 1.3.5 may help in the case, that you use a global SIM (IMSI 9xxxx).

RE: Current peaks in low-power udp example

hypn0 — Fri, 01 Sep 2023 08:10:26 GMT

Hi Michal and Achim

Thanks for your answers!

I made a 'longterm' measurement in order to get some more current data.

I used the same configuration shown above, but set the transmission time to 30 minutes.

The measurement data show high power consumption, which occurs regularly every 2.5 hours.
As far as I can see, I have switched off eDRX, so I am not sure now what exactly the hardware is doing during this time.

Do you have any ideas?

The following figure shows a zoom of on of the high current peaks:

Have a great day!

RE: Current peaks in low-power udp example

Michal — Mon, 28 Aug 2023 14:18:26 GMT

I would recommend zooming into those peaks first, but I think I may know the answer.

This may be caused by the regulators running in the refresh mode. You can find some information (and measurements) here: Getting started with current measurements on the nRF9160.

Best regards,

Michal

RE: Current peaks in low-power udp example

Achim Kraus — Sat, 26 Aug 2023 15:07:37 GMT

Beside the academical general benefit, did you try to get the amount of energy spend in that peaks? Sometimes such peaks are that small, that investigations do not really pay off.