nRF9160 SIBA unable to connect to NB-IoT network and resets every ~10 seconds.

[elephant]

nRF9160 SIBA on custom board is unable to connect to NB-IOT and resets after about 10 seconds of operation.

Without a call to lte_lc_init_and_connect the reset does not happen.

The power, as suggested in other topics is at 2.6V from dc power supply with limit set to 1A.

Side note: this setting is the only one allowing for the programmer to reach the SiP. At planned 3.3V the programmer is not able to connect to the chip. For now just a missing power IC -> power supply is used.

[JONATHAN LL]

Hi,

Can you provide a log ?

What application are you using? 

Does this happen with the AT_client ?

What MF do you have?

Regards,
Jonathan

[elephant]

I am curious about two things by now. Is pin 57 really important for RF part of chip to work well? (it is MIPI VIO, not connected in my design)

And is antenna really mechanically symmetric, but electrically not? (As stated by Fractus antennas)

[JONATHAN LL]

If XMONITOR = 2 or 4 then it is not connected


So it might be something with the sim card, what sim card and provider are you using ?


Edit:  I think there is a problem with the antenna, i would try to contact Fractus antenna and ask for a review and simulation of the board. 
With the CESQ it seems that there is 255 – Not known or not detectable. 

Alternatively you can create a new case, make it private, call it antenna layout review nrf9160 and add the HW files and schematic. And we will do our best to guide you.


Regards,
Jonathan

[elephant]

This very sim card is working in Thingy:91, so no problem with it itself.

Is there a way to know nRF9160 accepted it ok or with errors?

AT+CPIN? returns "OK"

AT+CIMI returns some valid-looking digits

any other tests I could do?

[JONATHAN LL]

I thin the issue is with the antenna, if the SIM works on a different device in the same location then the RF layout and antenna is probably the issue. 

I think there is a problem with the antenna, i would try to contact Fractus antenna and ask for a review and simulation of the board. 
With the CESQ it seems that there is 255 – Not known or not detectable. 

You can create a new case, make it private, call it antenna layout review nrf9160 and add the HW files and schematic. And we will do our best to guide you.

Regards,
Joanthan

[elephant]

Good news: it actually gets to about cesq = 99,99,255,255,24,40 and xmonitor = 2 after ~20 seconds of delay since at+cfun=1 command. But still no connection. Going to give it still more time.

Edit: after about 5-10 minutes no connection.

[elephant]

Just to conclude this. There appears to be some sort of a circuit break inside nRF9160. Most probably due to mounting procedures.

This is because antenna and matching network togather, connected to thingy:91 do really work. It connects ok just like with its own antenna. And thingy behaves just like my custom board when the wire is just touching a pad without proper soldering. This thing occurred spontaneously during experiments with the two boards.

[JONATHAN LL]

Good to see that you where able to find the issue, and thanks for updating. Are you able to test if re-soldering the SiP could fix the issue or are you cairten that it is some malfucing with the SiP and not the solder joints?

Regards,
Jonathan

[elephant]

Now waiting for the second board to come. Will test that next week.

[elephant]

Ok, everything turns out to be modem firmware. Matching network from antenna supplier is just fine. Soldering is ok also. But fw v.1.3.0 DOES NOT WORK IN RUSSIAN CELLULAR NETWORKS (NB-IoT way). The two newly assembled boards were upgraded to version 1.2.3 and are doing just fine. Connection and data flow - check!

To all my russian collegues: do use modem firmware version 1.2.3 and stay well.

[JONATHAN LL]

Good to se that you have a working solution but i would like to know if you could provide a modem trace, you can create a new private ticket called NB-IoT, Russia and modem firmware 1.3.0 does not work. And we will investigate further. 

Regards,
Jonathan