Connection to NB-IOT issues. How to get logs

[Nab]

Hello Everyone,

I am new with the pycom.
I am trying to connect to an NB-IOT network in belgium. I have bought a sim card from my provider ( proximus). But I cannot connect to the network following the script :

def init_nb_iot( blocking=True, timeout=15):
    lte = network.LTE()
    # lte.attach(band=20, apn="nb.inetd.gdsp") #changeme band is not mandatory
    lte.attach(apn="m2minternet.proximus.be") #changeme
    attach_timeout = timeout
    while not lte.isattached() and attach_timeout > 0:
        time.sleep(1)
        attach_timeout -=1
        print("lte attach_timeout : {}".format(attach_timeout))
    nb_iot_s = socket.socket()
    nb_iot_s.setblocking(blocking)
    return (lte, nb_iot_s)

The problem is that I cannot attach to the network. But I really don't know why and don't know where to look ? Can we get logs to see where the issue comes from ? Any debug procedure available I haven't managed to find one in the website.

[robert-hh]

@Nab The first attach may take VERY LONG, longer than 15 seconds.
May I also ask for:

• the version of the ESP32 firmware (push Ctrl-B)
• the version of the Sequans modem firmware (get it with)

from network import LTE
lte=LTE()
print(lte.send_at_cmd('AT!="showver"'))

that you are using.

[Nab]

Hi Robert,

Here are the different versions :

Pycom MicroPython 1.18.2.r7 [v1.8.6-849-df9f237] on 2019-05-14; FiPy with ESP32
Type "help()" for more information.

And here is the version of the sequans modem firmware :

print(lte.send_at_cmd('AT!="showver"'))

SYSTEM VERSION

FIRMWARE VERSION
Bootloader0 : 5.1.1.0 [39529]
Bootloader1* : 6.0.0.0 [41019]
Bootloader2 : NA
NV Info : 1.1,0,0
Software : 6.0.0.0 [41019] by robot-soft at 2018-12-13 21:41:15
UE : 6.0.0.0
COMPONENTS
ZSP0 : 1.0.99-3171
ZSP1 : 1.0.98-3170

OK

Thanks for your help :)

[robert-hh]

@Nab If you look at the bottom side of your FiPy, does it read r1.0 or r1.2?

[robert-hh]

@Nab And also, show the supported NBIOT bands with:

from network import LTE
lte=LTE()
print(lte.send_at_cmd('AT!="RRC:setDbgPerm full"'))
print(lte.send_at_cmd('AT!="RRC:showcaps"'))

[Nab]

@robert-hh it is written V1.2 in the bottom.

Also here are the to answer from the cmd :

print(lte.send_at_cmd('AT!="RRC:setDbgPerm full"'))

OK

print(lte.send_at_cmd('AT!="RRC:showcaps"'))

== CAPS config =============================
.Lock UE on SRV band : false
.MFBI support : true
.TM8 TDD support : false

== CAPS ====================================
. access stratum: R13
. catM : 1
. nb-IoT : 1
-- EUTRA bands --
. supported : 66/28/26/25/20/19/18/17/14/13/12/8/5/4/3/2/1
. board : 1/2/3/4/5/8/12/13/14/17/18/19/20/25/26/28/66
. admin :
. pending admin :
-- EUTRA carriers --
. admin :
. pending admin :

OK

Thx !

[robert-hh]

@Nab OK. So everything is fine. Newest board, most recent firmware, all bands supported. I assume also that the Antenna is connected properly.

You could try:

lte.send_at_cmd("AT+CFUN=1")
lte.send_at_cmd("AT+CSQ")

That will tell you, if there is a signal at the antenna.
You could also add band=20 to you attach command, because that is the band proximus uses, according to their web site. And wait a little bit longer than 15 seconds, like 15 minutes.

[Nab]

@robert-hh HI Robert,

I have tried the AT command and I had a result of 31,99. Which I guess means good signal at the antenna.

Something that I don't understant is that I have the same results with and without sim. What does this mean ? This signal is not related to my carrier provider ?

Also when I do a lte.ue_coverage() i got false has answer. But I am sure I do have network since I have other devices working on nb-iot.

I tried the following command to check networks : print(lte.send_at_cmd('AT+COPS=?'))

but I have no answers. Does this mean that my board has an issue ?

Thank again for the support

[rskoniec]

@robert-hh Just wonder - here is the outputs from my FiPy, but my board is rev. 1.0, so by software it is capable using all of 17 bands, but by hardware it is not? Unfortunately here in Poland still hard to test either NB1 (T-Mobile) or M1 (Orange) - closed just for companies and pilot projects/tests. :/

>>> os.uname()
(sysname='FiPy', nodename='FiPy', release='1.20.0.rc11', version='v1.9.4-0a38f88 on 2019-05-14', machine='FiPy with ESP32', lorawan='1.0.2', sigfox='1.0.1')

>>> print(lte.send_at_cmd('AT!="showver"'))

SYSTEM VERSION
==============
  FIRMWARE VERSION
    Bootloader0  : 5.1.1.0 [39529]
    Bootloader1  : 6.0.0.0 [40343]
    Bootloader2* : 5.1.1.0 [41065]
    NV Info      : 1.1,0,0
    Software     : 5.1.1.0 [41065] by robot-soft at 2018-12-17 13:03:40
    UE           : 5.0.0.0d
  COMPONENTS
    ZSP0         : 1.0.99-13616
    ZSP1         : 1.0.99-12376

OK

>>> print(lte.send_at_cmd('AT!="RRC:showcaps"'))

== CAPS config =============================
  .Lock UE on SRV band    : false
  .MFBI support           : true
  .TM8 TDD support        : false
============================================
== CAPS ====================================
  . access stratum: R13
  . catM          : 1
  . nb-IoT        : 0
-- EUTRA bands --
  . supported     : 66/28/26/25/20/19/18/17/14/13/12/8/5/4/3/2/1
  . board         : 1/2/3/4/5/8/12/13/14/17/18/19/20/25/26/28/66
  . admin         :
  . pending admin :
-- EUTRA carriers --
  . admin         :
  . pending admin :
============================================

OK

[robert-hh]

@rskoniec The he hardware does not support all bands. You have to use the rma process in order to exchange it with a rev 1.2 board. The link is in one of the Annotations.

[robert-hh]

@Nab AT+CSQ-> 31,99 tells at least, that the RF section and antenna are ok. It does not tell whether there is a Nb-Iot service. I tried also to find information about the APN settings for proximus. I did not get more than marketing bullshit, and I cannot register without a BE mobile number.

[Nab]

@robert-hh Ok thx ! I havec contacted proximus to verify those settings. Will let you know :)

[rskoniec]

@robert-hh I thought so, thank you for confirmation. AFAIR on the forum there were some posts that the RMA process was quite long and with communication (email, phone) problems with Pycom.

[robert-hh]

@rskoniec It is this thread: https://forum.pycom.io/topic/3843/important-news-regarding-cat-m1-and-nb-iot-cellular-range

[andreas]

Hi there,

thanks again to @robert-hh for outlining different steps how to verify hardware and modem firmware for compatibility / band 8 support.

We can confirm we have "FiPy V1.2" hardware and are running a recent

Pycom MicroPython 1.20.0.rc11 [v1.9.4-0a38f88] on 2019-05-14; FiPy with ESP32

Inquiring the modem capabilities and firmware versions gives us

== CAPS ====================================

  . access stratum: R13
  . catM          : 1
  . nb-IoT        : 0

-- EUTRA bands --
  . supported     : 28/20/13/12/8/5/4/3
  . board         : 3/4/5/8/12/13/20/28

SYSTEM VERSION
==============
  FIRMWARE VERSION
    Bootloader0* : 5.1.1.0 [39529]
    Bootloader1  : NA
    Bootloader2  : NA
    NV Info      : 1.1,0,0
    Software     : 5.1.1.0 [39529] by robot-soft at 2018-09-28 16:07:40
    UE           : 5.0.0.0d
  COMPONENTS
    ZSP0         : 1.0.99-13616
    ZSP1         : 1.0.99-12355

While band 8 might probably be supported already (haven't been able insert an appropriate SIM card yet), we are humbly asking whether we should still upgrade to a more recent firmware version.

Until now, we believed from reading here that one would need different firmwares for the Sequans Modem in order to operate it in CAT-M1 vs. CAT-NB1 environments. However, some capability-dumps here show us that there might be modem firmwares around which support both variants already.

== CAPS ====================================
  . access stratum: R13
  . catM          : 1
  . nb-IoT        : 1

Also, this is the information presented by sqnsupgrade:

>>> import sqnsupgrade
>>> sqnsupgrade.info()
<<< Welcome to the SQN3330 firmware updater [1.2.5] >>>
>>> FiPy with firmware version 1.20.0.rc11
Your modem is in application mode. Here is the current version:
UE5.0.0.0d
LR5.1.1.0-39529

IMEI: {redacted}

So, we are effectively running firmware CATM1-39529, right? However, the firmware versions 40343 and 41065 are reference here in the forum and xx is also telling us about CATM1-39529 vs. another piece called NB1-37781.

When holding different FIRMWARE VERSION dumps from this forum against each other, we are getting different results from different people reporting about their firmware software versions.

• Software: 6.0.0.0 [37781] by robot-soft at 2018-07-05 20:21:02
• Software: 6.0.0.0 [41019] by robot-soft at 2018-12-13 21:41:15
• Software: 5.1.1.0 [41065] by robot-soft at 2018-12-17 13:03:40

Compared to that, the version on our modem is

• Software: 5.1.1.0 [39529] by robot-soft at 2018-09-28 16:07:40

In order to shed some more light onto that topic, we are humbly asking which path to follow to eventually be able to support both communication variants (M1 vs. NB1) within the same MicroPython progam - maybe even without reprogramming the modem firmware.

Thanks for your efforts and keep up the spirit.

With kind regards,
Andreas.

[robert-hh]

@andreas I have no experience i n running the actual ESP32 firmware with an older Sequans firmware. I used NBIoT 41019. I expect that this is required. Form what I see, the Cat-M1 and NBIot firmware are exclusive. So you cannot have them both on the device at the same time.

[andreas]

Dear @robert-hh,

thanks for sharing your observations and for giving more details about which variant you are actually using for your investigations and beyond.

Exclusive usage of mode CAT-M1 vs. CAT-NB1

We have been aware of the exclusive usage of both firmwares and just wanted to confirm that you don't know otherwise. As outlined above, we got curious about capability outputs like this guy scattered around the forum already.

== CAPS ====================================
  . access stratum: R13
  . catM          : 1
  . nb-IoT        : 1

This somehow indicates that both variants might be supported at the same time. We will let you know if we can find out something about that detail.

Firmware upgrade FTW

I have no experience i n running the actual ESP32 firmware with an older Sequans firmware.

We will be happy to upgrade the modem to the NB1-41019 firmware and to try the corresponding NB1 mode, which is obviously the better choice for usage within Germany others are telling us as well [1].

SIM cards from providers covering Germany

You are also using SIM cards from 1nce, don't you? For all people reading this: We will be happy [2] to learn about other providers offering respective SIM cards - maybe even for private persons in the future.

Thanks for your insights and with kind regards,
Andreas.

[1] https://community.hiveeyes.org/t/cat-m1-vs-cat-nb1-fur-bob/2320/12
[2] https://community.hiveeyes.org/t/welche-sim-karte-fur-cat-m1-cat-nb1-in-deutschland/2325

[robert-hh]

@andreas Yes, I am using cards from 1nce, which are in fact Telekom cards. I had a long discussion with people at several Telekom shops, after they announced their service. But they did not know about it, besides the commercial package (>100 cards, but only for companies). 1nce is also selling to companies, but they were kind to send trial cards. Vodafone and O2 also claim to have a service - again for companies only. While this is not really an obstacle - you will always find someone ordering it for you - it is not a friendly attitude. Especially for the package 1nce offers, which is more of sell & forget.

[securigy]

@andreas Can you please tell - me where did you buy FiPy 1.2 ?
Any site I try I see no information about exact version I will get. Instead there are pictures of the FiPy 1.0 or 1.0r
I already experienced the purchase of FiPy and got 1.0r instead of 1.2 - that's the reason that I am asking.
Seems I cannot get a streight answer from anyone, include Pycom sales which is quite non-responsive...

[andreas]

Dear @securigy,

I haven't bought any FiPy myself, just got it delivered to my house from people of the upstream project [1] we are working on.

Apologies for not being able to help you further. However, I believe all FiPy devices on the market right now should already be the newer ones.

With kind regards,
Andreas.

[1] https://community.hiveeyes.org/t/bee-observer-bob-uni-bremen-und-hiveeyes-werden-als-citizen-science-projekte-vom-forschungsministerium-unterstutzt/454