Daer
usually we need minutes to lock the LTE cell and connect to the network. There is a way after the connection to save the data (NVM or flash) and on restart to use them?
Federico
Daer
usually we need minutes to lock the LTE cell and connect to the network. There is a way after the connection to save the data (NVM or flash) and on restart to use them?
Federico
Another minor detail I've never seen documented anywhere: even if you leave the modem in "AT+COPS=0" mode to let it auto-select carriers, it can still remember the most recent service it was connected to and use that to speed up the next re-connect. But, when you re-send the "AT+COPS=0" command to the modem it forgets that information, making the next connection search all carriers again.
In other words, leaving the modem in "AT+COPS=0" mode doesn't prevent the modem from saving connection info, but sending the "AT+COPS=0" command does.
Thanks, this is very interesting to hear. Could you tell me which modem firmware version has this "AT+COPS=0" behaviour? I'm currently having mfw_nrf9160_1.1.0 and it seems that the list of recent cells is not cleared. The attachment to the NB-IoT network takes only a few seconds even without any band locks.
I am trying compare the cell search times with different band lock combinations. Thus, it would be highly useful to erase the list of recent cells from the device memory. Do you know if there is any AT command for this?
Best regards,
Olli
We originally noticed this on 1.0.0 and 1.0.1 a long time ago. It is possible things have changed since then, but we still have our firmware query the setting using "AT+COPS?" at bootup and only send the "AT+COPS=0" command if that is our desired mode but not the response from "AT+COPS?".
A simple test we did to prove this behavior to ourselves, which you can use to see if it still happens:
In our experiments, the time for steps 5 and 8 from sending the CFUN command to getting online would be dramatically different.
Thanks for your help. I downgraded modem FW to 1.0.1 but it didn't seem to have any effect, i.e. the times in steps 5 and 8 were almost the same (in the order of seconds). The full frequency scan with at+cops=? takes minutes. Maybe the behaviour of remembering the visited cells is more related to how the memory is handled with HW. I'm having Thingy:91 PCA20035 1.0.0 2019.33.
Well anyway, a documented AT command to clear the visited cells and EARFCN would be nice to have for testing implementations for mobile IoT applications. I noticed that some other cellular IoT module vendors have included that to their AT command set.
Best regards,
Olli
Thanks for your help. I downgraded modem FW to 1.0.1 but it didn't seem to have any effect, i.e. the times in steps 5 and 8 were almost the same (in the order of seconds). The full frequency scan with at+cops=? takes minutes. Maybe the behaviour of remembering the visited cells is more related to how the memory is handled with HW. I'm having Thingy:91 PCA20035 1.0.0 2019.33.
Well anyway, a documented AT command to clear the visited cells and EARFCN would be nice to have for testing implementations for mobile IoT applications. I noticed that some other cellular IoT module vendors have included that to their AT command set.
Best regards,
Olli