Electrons going mute? (modem)

Can't find much on a peculiar issue:

A couple units now have inexplicably gone offline, namely breathing white. Specifically this last one successfully connected to cloud only 2 weeks prior, was turned off 'til now, and upon reboot breathed white (I can private mgs you IDs).

Just in case of my 'operator error' I did the usual goodness of this post:
(Which I take is like a more complete master reset by wiping memory and starting over w/ OS 2.2.0)

Yet after this routine it goes right to breathing white (whereas some truly 'corrupt' units in the past, like with a dim d7 led have been successfully recovered).
One odd clue is that if I flash (just before doing 'particle keys doctor'):

particle flash --usb tinker@2.2.0.bin

It does in fact flash green like it's booting up. That said I don't 'hear anything' on 'particle serial monitor'. Anyway soon as I do 'particle keys doctor' it goes breathing white.

I've also tried the web interface 'particle device doctor', which does a flash routine (If I manually put in DFU mode), but it doesn't reconnect after flashing. (Web interface says 'lost connection', and I press reconnect, even after putting it back in DFU mode, but no dice).

Kind of makes me believe a keys / token issue? There's no way modems are just going bad, especially if it connected to cloud a couple weeks ago, no?

I don't believe this is a case of OS update / reversion sequence, since we're on 2.2.0 and beyond (not like the old days of 0.5.x right?).

@emcdee, what type of Electron is it? More specifically, what modem does it have on it (e.g. U270)?

Ok so built-in SIM (we didn’t change / touch), and LTE. Prob ~2021
I’ll have to venture over to give you exact deets but I recall chip stated ‘made in Austria’

@emcdee, you may want to try device doctor to get more info on the problem. You could also use Device Restore to reset everything to a know OS and tinker.

1 Like

Copy and thanks, will do.
Meanwhile what's this 'dead SIM' -

Not sure how to get the info here since there's no ID...
Suppose this is one of the two units that went offline? But still a valid question.

Model "ELC402EA"

Hope I'm not confusing too many things here...

So for this:

And then was wondering if this 'anonymous' ID is the recent 'breathing white' unit? Coz I don't see its ID in the Particle Product. Found it.

Along with that, this unit is acting funny, like it takes over 60s to get to blinking blue (listening mode), but only after it sorts of fritzes out and quickly flashes blue then quickly flashes red before going into blinking blue.
EDIT 1: Actually sometimes in listening mode it doesn't even return.. like:
Could not identify device: Could not get device info, ensure the device is in listening mode: Serial problems, please reconnect the device.

EDIT 2: One more, I did the device 'usb restore' and it breathed white, before going to blinking green, to now blinking blue.. but still no return on 'particle identify', nor anything heard from 'particle serial monitor'.

And then following the 'manual debugger flash' -

Not much, just breathing white.

Did a 'usb restore' after JTAG, and get this:
image

Ok and 'finally' after another web-based 'device doctor' it's now just breathing blue, then switches to blinking blue.. with no interaction.

@emcdee, you need to run Device Doctor and see what you get from that.

Sure we can recap (maybe I put too much up front);

Didi this -

While breathes white and doesn't reconnect... then goes to breathing blue, then blinking blue; but does not respond to 'particle identify' nor 'particle serial monitor'.

I'm not sure what info to gather from this, except that the summary says it's on OS 2.0.0-rc.2
despite that I manually (via 'advanced') specified to go to OS 3.0.0

@emcdee, from what you say, there are no serial messages from the device? @rickkas7, any thoughts?

It's weird that there are no USB serial messages other than Connecting to Particle. There should be a lot more messages after using Device Restore USB.

However I think the cellular modem failed. If you have an Electron LTE, it could be TAN001 or TAN004.

Copy that, much appreciate the awareness..

So sounds like we need to take some coding precautions (as outline in TAN004) to avoid any further corruptions, or can we just upgrade OS to avoid this?

And I didn't see, but is there a way to recover a corruption?

If it really is a failure of the modem, there is no way to recover it.

The newer devices like the B404X and Boron 404X use a newer model of cellular modem (u-blox SARA-R510 instead of R410) which is less likely to suffer from this problem.

The Tracker SoM T404 and M-SoM M404 use a Quectel BG96 or BG95 cellular which is not affected.

Ok. So to confirm, do we need a safe power down or just use OS > 3.0.0 to 'solve' this?

Else if 3.0.0 or lower, apparently needed the safe power-down like this before pulling power -

Cellular.off();
waitFor(Cellular.isOff, 30000);

Or prob a better question, what's the next best OS from 3.0.0?

For the Electron, just use the latest from the 3.x line, 3.3.1. It's highly unlikely that there will be any more versions of Device OS for Electron/E-Series.

1 Like