Important update regarding LTE modem updates

[danielm]

@peterv @robert-hh

This is Sequans modem version info using "showver" command:

FIRMWARE VERSION
Bootloader0 : 5.1.1.0 [33080]
Bootloader1* : 5.1.1.0 [38638]
Bootloader2 : 5.1.1.0 [38638]
NV Info : 1.1,0,0
Software : 6.0.0.0 [37781] by robot-soft at 2018-07-05 20:21:02
UE : 6.0.0.0
COMPONENTS
ZSP0 : 1.0.99-319
ZSP1 : 1.0.99-302

Today I will test the conenctivity again to be sure. NB-IoT network coverage is not nation-wide yet (should be until end of 2018) so I will travel to different city to try it. In the meantime I installed my development FiPy board into custom PCB and housing.

[Jiemde]

@tuftec Here is the sequence I've used for connecting the Gpy to the Orange network in Belgium.```

from network import LTE
import pycom
import time

pycom.heartbeat(False)
​
lte = LTE()
lte.send_at_cmd('AT+CFUN=0')
# print("OK line 1")
lte.send_at_cmd('AT!="clearscanconfig"')
# print("OK line 2")
lte.send_at_cmd('AT!="addscanband band=20"')
# print("OK line 3")
lte.send_at_cmd('AT!="disablelog 1"')
# print("OK line 4")
lte.send_at_cmd('AT+CGDCONT=1,"IP","nbiot.iot"')
# print("OK line 5")
lte.send_at_cmd('AT+CFUN=1')
# print("OK line 6")

pycom.rgbled(0xff0000)
	
while not lte.isattached():
	print('Attaching…')
	time.sleep(0.25)

pycom.rgbled(0x0000ff)

lte.connect()

while not lte.isconnected():
	print('Connecting…')
	time.sleep(0.25)

print("lte is connected")

pycom.rgbled(0x008000)

# Print line are just there to verify if the AT command are executed or not

[arneme]

@arneme I can confirm that we have managed to connect the FiPy to our guard band NB-IoT Huawei eNodeB based RAN (Telenor Norway). We do however see some strange signalling behaviour when the FiPy tries to connect to the network. It would help if the documentation could be updated (so that we might be able to get the FiPy to behave correct) or if we could get someone in Pycom or Sequans to discuss this with. We are officially launching a country wide NB-IoT (and Cat M1) network in a few weeks so it would be great if we could get this sorted out asap.

[Xykon]

@arneme As a network operator please email bettina(at)pycom.io as we can provide special hardware & software to allow access to the DCP port.

For everyone else, NB1 firmware 38729 is now available for download at https://software.pycom.io/downloads/sequans2.html with the same updater.elf file as all the other packages.

[robert-hh]

@xykon said in Important update regarding LTE modem updates:

For everyone else, NB1 firmware 38729 is now available for download at https://software.pycom.io/downloads/sequans2.html with the same updater.elf file as all the other packages.

@Xykon Do you know what has been changed? And did you succeed in testing this version?

[Xykon]

@robert-hh All I've been told is that it contains various bug fixes compared to the previous version. I haven't yet had any success with this firmware... I keep getting LWM2M/ audit >| lwm2m: error while getting IP addresses and I'm getting the same message in any firmware and both with Vodafone Netherlands and Amarisoft's NB-IoT software. I've prepared trace logs that I've sent to both Vodafone and Sequans for analysis.

[robert-hh]

@xykon There is some difference in behavior to the previous version: it now shows RF activity. It still does not attach, but at least there is some communication.

[arneme]

@xykon OK, I will try again to reach her :-)

[tuftec]

@robert-hh That at least sounds positive.
I will try the bug fixed version in the next few days.

[tuftec]

@xykon I have tried the latest firmware release now and I still can not see any RF activity in band 8 and can not get the unit to attach. Vodafone here in Australia have also confirmed that they can not see any attach activity on the network.

We all need a bit more assistance from Pycom and Sequans.

Peter.

[danielm]

I just want to confirm that I am able to connect to Slovak Telekom NB-IoT network with both FiPy and G01 running Pycom FW release 1.18.1.r1 and Sequans modem firmware version 37781. I did not try version 38729 yet.

[robert-hh]

@tuftec The RF activity I see is only once ~40 seconds after a hard reset, and consist of a burst of short transmissions in a time frame of 200ms. With a spectrum analyzer you might miss that.

[kevin]

@tuftec Not sure if you are using FiPy or GPy, but at least for GPy it appears the currently supported bands are 3/4/12/13/20/28. I (and others) have asked for an update on band support throughout the forum, but no clarification so far.

[tuftec]

@kevin the FiPy specification clearly indicates that it supports band 8. That is why we selected the device for our application. If the firmware does not cureently suuport it, that might explain my problem.

@administrators can you please ensure that Band 8 is supported in the next firmware release. Please advise when this will be available. This is holding up development.

Peter.

[tuftec]

@kevin we are using FiPy.

[tuftec]

@robert-hh should be able to pick up 200 ms bursts.

[robert-hh]

@tuftec The RF burst itself are much shorter, like 10ms. They just span over a time period of 200 ms.

[tuftec]

@robert-hh Ohhh! That might be hard to detect on my spectrum analyser.

[robert-hh]

@tuftec I picked that up with an oscilloscope, using a Diode at the Probe tip for rectification. Se picture. The diode I use is a 1N5711 or BAR11 (fast Schottky). A 1N4148 works as well. The Oscilloscope Probe is a standard 10:1, ~3 pF one. The capacity has to be low. The bandwidth of the oscilloscope can be low. The signal you get is pretty high, like 200mV - 4V. Works for LoRa as well.

[tuftec]

@robert-hh That's worth trying. Thanks