nRF9160 MODEM - What data does %XFACTORYRESET clear?

[Ivan Herrera]

Hi!

We observe that our nRF91 SLM's get "softbricked" at random times, and a lot of them at the same time (We feel this is because of network unstability). These get stuck in the registering phase to the network. We noticed that reflashing the modem (not the SLM application) fixes the issue, so we're fairly confident that there's something stored in flash that is preventing these modems to register. We found the %XFACTORYRESET command, and it seems like a good candidate to fix this issue in the field without having to manually reflash the modem. The documentation on this command is pretty vague:

https://infocenter.nordicsemi.com/topic/ref_at_commands/REF/at_commands/mob_termination_ctrl_status/xfactoryreset_set.html

A couple of questions:

* What kind of data specifically is cleared in both RESET 1 and RESET 0?

* What would the impact be on a factory reset every time we boot the device?

* Will the factory reset AT command become available for the Rev 1 modules in the future?n

[Ivan Herrera]

We tested our new application firmware (external MCU) on a modem that is stuck in registering, even after resetting/power cycling the modem. The %FACTORYRESET=0 command does in fact resolve our issue. Still it would be helpful for us to know what kind of information is actually being cleared, and if it is detrimental to execute this command every time we boot our device.

[Charlie]

Hi Ivan, 

Thanks for reaching us with your questions. After discussion with our development team, I got the following answers to your questions:

* What kind of data specifically is cleared in both RESET 1 and RESET 0?

We do not have a specific list of data types for these reset actions, because the flash is cleared with one action instead of cleaning data one by one following a list. The MFW is also close-sourced and we will not able to share detailed variables related to user configurations and network-related data used on source code. You could try and verify with other AT commands if you are interested in some specific "functionality" or "behavior".

Ivan Herrera said:

The %FACTORYRESET=0 command does in fact resolve our issue.

Flash wear(Min 10 000 for nRF9160) should be your concern. If your application needs to trigger factory reset frequently, this should take this into consideration.

* Will the factory reset AT command become available for the Rev 1 modules in the future?n

As long as you use mfw_1.3.x or later, factory reset is available. You should pay attention to that MFW 1.3.x release note mentioned that.

• Note that this modem firmware is targeted nRF9160 SiP Revision 2 
• It can be used for nRF9160 SiP Revision 1, but only for testing and development

-----------------------------------------------------------------------------------------------------------

We want to learn more information about your problem to see if there is a better solution than using %XFACTORYRESET.

1. Could you check if this registration problem is because of reset loop prevention? You can use %MDMEV command to get notification of that. The optimal way would be changing testing procedures to avoid reset loop detection.

2. If you do not get "%MDMEV: RESET LOOP", logs would be needed. +CEREG and +CNEC should be used, those will give a rough idea of the cell is not found or if the network rejects registration attempts.

3. You mentioned, "The %FACTORYRESET=0 command does in fact resolve our issue." Have you tested with %FACTORYRESET=1? Does it help?

Best regards

Charlie

[Ivan Herrera]

Hi Charlie!

Thank you for your reply. Reading through the documentation of the reset loop prevention and its behavior, we strongly suspect this is what is actually happening. Whenever the network is unavailable, our application MCU resets the modem since it can not register, this happens a few times and the reset loop prevention kicks in. Whenever the network comes back up, the reset loop prevention prevents the modem from registering ergo resets again, and it gets stuck in this loop. We will adjust our application accordingly.