Cellular radio firmware release (SQN3330) for both Cat-M1 and NB-IoT

[daniel]

Hello everyone,

I know many of you were expecting this. We have made available 2 new firmware files for the Cellular radio in the FiPy, GPy and G01 modules. All the details and instructions can be found here:

https://github.com/pycom/pycom-libraries/tree/master/lib/sqnsupgrade

Cheers,
Daniel

[arneme]

@daniel That is great news and an important step forward. Could you say more about the roadmap to support Huawei base stations and guard band deployments?

[bali]

@daniel Any chance why i get this? !

[daniel]

@bali can you try removing power, inserting a SIM card and trying again?

[bali]

@daniel I tried, i got the exact same thing unfortunately. The reason why i'm trying to update the firmware is because when i try to call lte = LTE() the GPy hangs.

[daniel]

Do you have SIM card properly inserted? Can you double check?

[bali]

@daniel I double checked it, still not working.

[daniel]

@bali have you ever been able to call lte = LTE() without hanging?

[bali]

@daniel with firmware 1.16.0.b1 yes, but i could never attch to cellular network because i’m on At&T and in 1.16 there was no way to configure for it.

[daniel]

@bali can you try changing the baud rate from 921600 to 115200 or 345600 and maybe 57600? Just in case somehow the default baudrate is different in your board...

[daniel]

@bali OK then it seems that some of the new logic is causing your device to hang. Are you able to build a custom firmware? Then we can disable the background LTE logic and simply run the upgrade...

[bali]

@daniel I did try at 921600, 345600, 115200, 57600 and it failed the same way at every baud rate. Unfortunately i can't build custom firmware. If you could send me one with the LTE disabled i could try it. I ordered the GPy but was never able to the the LTE on it :(

[Xykon]

@bali Could you please check what happens if you run the following commands on your GPy?

from network import LTE
lte = LTE()
lte.send_at_cmd('AT+CFUN=0')

[bali]

@xykon hangs right after issuing lte=LTE()

[Xykon]

@bali Please open a technical support request at https://pycom.io/contact as it seems your GPy is faulty.

[nathanh]

So I think this just rendered my entire FiPy useless....

I upgraded the CAT-M firmware yesterday without issue to see if I could connect more successfully to Telstra and besides not getting reliable connections the upgrade process seemed to go ok.

Today I tried to upload the NB-IOT firmware and it got stuck at 6% for over 5 minutes exactly like the description in the instructions (https://github.com/pycom/pycom-libraries/tree/master/lib/sqnsupgrade). So I power cycled as per the instructions and now I get stuck on this:

<<< Welcome to the SQN3330 firmware updater >>>

And nothing... I have let it run for over 10 minutes and power cycled multiple times. I also notice now I lose all UART REPL command/interface after about 1-2 minutes of powering up my FiPy and it is a race against the clock to even execute the update code. If I wait a while the REPL just freezes and won't accept any input, so essentially making the device unusable.

And as per @bali now when I can get in quick enough and try and call lte = LTE() my device hangs. So it looks like my device that was working is no longer functioning and I followed the instructions as directed...

[daniel]

@nathanh @bali I will contact each one of you directly to help resolve your issues.

Cheers,
Daniel

[vigmeister]

@daniel FWIW I seem to have run into exactly the same symptoms as @nathanh If and when a resolution is found, I would appreciate if you post it here. I would also not mind getting in on the debugging action if that would help.

[vigmeister]

@daniel @nathanh Additional hint from my observation. I looked at the TX buffer of the serial interface by inserting s.wait_tx_done(5000) after the s.write("ATE0\r\n") line in sqnsupgrade.py and I get "False" as the response. Will try moving downstream in the sqnsupgrade.py code later in the day and share my findings.

[vigmeister]

I tinkered a bit further and by setting the default timeout to 2000 instead of 20000 in sqnsupgrade.read_rsp , I was able to get my FiPy to get further along, but ran into @bali ‘s issue of AT+SMSWBOOT=3, 0 failed!

Looks like the issues are roughly the same, but that @nathanh and I have unstable REPL on top :/