New sim card never connect (8h+) but connect instantly when forcing RAT to LTE

Hi,

We have FX30(WP7611-1) deployed in multiples markets and for most/all of our sim card/operators we have no problem.

Now comes the pickle, we also have sim cards from/for Magenta Telekom.

When we insert a “brand new sim card”, we never successfully connect (waited 8+ hours). But if we use a sim card coming from one of our other modem we support and we know get a connection, the modem get connectivity right away.

We toyed quite a lot to understand what was going on and eventually we decided to play with the RAT. Our default is Auto and it seems like forcing it to LTE result in a functional state (95% of the time).

Now, while this seems like a mitigation, we do not understand why that happens and thus we cannot just pray that it always works when we deploy devices around the globe. Which is why we are here.

We have a hunch that it is related to the fact that those sim card do not have preferred network out of the box since these are multi-carrier/operator (not sure if that is the right term). We also saw mentions that the modem might be scanning all freqs in that case.(FX30 with Verzon: Can't Connect to network - #8 by cchenry).

We do have special setup in for our other modems:
Quectels: AT+QCFG=“nwscanseq”
Gemalto: ^SCFG: “Radio/CNS” “Fast mode”

Is there an equivalent for FX30 ? (Had a look at the AT reference and the closest I got to i maybe force a “prefered” RAT for acquisition (!SELRAT) but it is unclear to me what acquisition means here.

Some info dump for context:

ATI3

Manufacturer: Sierra Wireless, Incorporated
Model: FX30(WP7611-1)
Revision: SWI9X07Y_02.37.03.00 73df45 jenkins 2020/04/08 10:59:14
IMEI: 352588320143405
IMEI SV: 14
FSN: ZW315200270610
+GCAP: +CGSM
ATI8

Legato Ver: 19.11.5.cdc94bb9_fa5a606a7dc8c397dd98e6abdb975bac_modified
Yocto Ver:  SWI9X07Y_02.37.10.02 2022-02-17_11:09:54
OS Ver: Linux version 3.18.140 (oe-user@oe-host) (gcc version 7.3.0 (GCC) ) #1 PREEMPT Thu Feb 17 11:02:54 UTC 2022
LK Ver: 1.3.0_482d4b1057
RootFS Ver: SWI9X07Y_02.37.10.02 2022-02-17_11:09:54
UserFS Ver: unknown
MCU Ver: unknown
AT!BAND?

Index, Name
00, All bands
AT+KSRAT?

+KSRAT: 0

XXX is redacted.

AT+CGDCONT?

+CGDCONT: 1,"IP","internet.XXXX.iot","0.0.0.0",0,0,0,0
+CGDCONT: 2,"IPV4V6","ims","0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0",0,0,0,0
+CGDCONT: 3,"IPV4V6","sos","0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0",0,0,0,1

The following are done with a “working” sim (after LTE RAT forced).

AT*PSRDBS?
*PSRDBS: 6917535626593305024
AT+CEREG?
+CEREG: 0,5
AT+COPS?

+COPS: 0,0,"Rogers Wireless",7
AAT+COPS=3,2
AT+COPS?

+COPS: 0,2,"302720",7
AT+CGACT?

+CGACT: 1,1
+CGACT: 2,0
+CGACT: 3,0
!GSTATUS:

Current Time:  43867            Temperature: 30
Modem Mitigate Level: 0         ModemProc Mitigate Level: 0
Reset Counter: 1                Mode:        ONLINE
System mode:   LTE              PS state:    Attached
IMS reg state: NOT REGISTERED   IMS mode:    Normal
IMS Srv State: NO SMS,NO VoIP

LTE band:      B2               LTE bw:      15 MHz
LTE Rx chan:   1075             LTE Tx chan: 19075
LTE CA state:  INACTIVE
EMM state:     Registered       Normal Service
RRC state:     RRC Connected

PCC RxM RSSI:  -89              RSRP (dBm):  -114
PCC RxD RSSI:  -88              RSRP (dBm):  -118
Tx Power:      --               TAC:         61EA (25066)
RSRQ (dB):     -11              Cell ID:     01075117 (17256727)
SINR (dB):     11.4

Unfortunately I do not have a “fresh” sim card right now but I will provide the same information with a “faulty” one ASAP.

Greatly appreciate any help that could get us out of the mud.

Thanks

The signal strength seems to be bad

How about using AT+cops in manual mode?

Which carrier are you using?

The aim here is to let the “modem”/“sim” decides since this is the big advantages of Magenta on our side, they have contract with a lot of carrier (network operator, sorry if I am confusing terminology here) in different markets. So the aim is to be as “agnostic” as possible on that front.

For example, right now I am in Montreal so the carrier chosen is Rogers but we have devices (fx30) in the field in Chicago that choose between ATT and T-Mobile depending on the signal at each locations.

Wouldn’t forcing the AT+cops go against that ?

Or are you suggesting that the reason it takes time for a “new sim” to “get going” is that the signal is poor ?

Them seems you are using generic pri as it is using roaming network

Yes, not sure if the bad signal makes the problem

Btw, you might also check at+cpol for the preferred operator list

AT+CPOL?

+CME ERROR: operation not allowed

AT+CPOL=?

+CME ERROR: operation not allowed

I guess that this is not what we expect…

Btw thanks for helping.

here is my test:


ati3
Manufacturer: Sierra Wireless, Incorporated
Model: WP7608
Revision: SWI9X07Y_02.37.06.05 b15b59 jenkins 2022/09/27 07:54:33
IMEI: 352913090521435
IMEI SV: 18
FSN: XG207430712345
+GCAP: +CGSM,+DS

OK
at+cpin?
+CPIN: READY

OK
at+cpol?
OK
at+cpol=?
+CPOL: (1-16),(0-2)

OK
at+cpol=?
+CPOL: (1-16),(0-2)

OK
at+cpls?
+CPLS: 0

OK
at+cpls=1
OK
at+cpol?
+CPOL: 1,2,"45400",1,0,1,1

OK
at+cpls=2
OK
at+cpol?
+CPOL: 1,2,"45400",0,0,0,1

OK

Just so we have a comparison:

at+cpin?
+CPIN: READY
OK

at+cpol?
+CME ERROR: 3

at+cpol=?
+CME ERROR: 3

at+cpls?
+CPLS: 0

OK

Sorry did not see there was other commands there, here are the rest:

at+cpol?

+CPOL: 2,2,"0001650"
+CPOL: 3,2,"166500"
+CPOL: 4,2,"015165"
+CPOL: 5,2,"1650000"
+CPOL: 7,2,"0001650"
+CPOL: 8,2,"166500"
+CPOL: 9,2,"015165"
+CPOL: 10,2,"1650000"
+CPOL: 12,2,"0001650"
+CPOL: 13,2,"166500"
+CPOL: 14,2,"015165"
+CPOL: 15,2,"1650000"
+CPOL: 17,2,"0001650"
+CPOL: 18,2,"166500"
+CPOL: 19,2,"015165"
+CPOL: 20,2,"1650000"
+CPOL: 22,2,"0001650"
+CPOL: 23,2,"166500"
+CPOL: 24,2,"015165"
+CPOL: 25,2,"1650000"
+CPOL: 27,2,"0001650"
+CPOL: 28,2,"166500"
+CPOL: 29,2,"015165"
+CPOL: 30,2,"1650000"
+CPOL: 32,2,"0001650"
+CPOL: 33,2,"166500"
+CPOL: 34,2,"015165"
+CPOL: 35,2,"1650000"
+CPOL: 37,2,"0001650"
+CPOL: 38,2,"166500"
+CPOL: 39,2,"015165"
+CPOL: 40,2,"1650000"
+CPOL: 42,2,"0001650"
+CPOL: 43,2,"166500"
+CPOL: 44,2,"015165"
+CPOL: 45,2,"1650000"
+CPOL: 47,2,"0001650"
+CPOL: 48,2,"166500"
+CPOL: 49,2,"015165"
+CPOL: 50,2,"1650000"
+CPOL: 52,2,"0001650"
+CPOL: 53,2,"166500"
+CPOL: 54,2,"015165"
+CPOL: 55,2,"1650000"
+CPOL: 57,2,"0001650"
+CPOL: 58,2,"166500"
+CPOL: 59,2,"015165"
+CPOL: 60,2,"1650000"
+CPOL: 62,2,"0001650"
+CPOL: 63,2,"166500"
+CPOL: 64,2,"015165"
+CPOL: 65,2,"1650000"
+CPOL: 67,2,"0001650"
+CPOL: 68,2,"166500"
+CPOL: 69,2,"015165"
+CPOL: 70,2,"1650000"
+CPOL: 72,2,"0001650"
+CPOL: 73,2,"166500"
+CPOL: 74,2,"015165"
+CPOL: 75,2,"1650000"
+CPOL: 77,2,"0001650"
+CPOL: 78,2,"166500"
+CPOL: 79,2,"015165"
+CPOL: 80,2,"1650000"
+CPOL: 82,2,"0001650"
+CPOL: 83,2,"166500"
+CPOL: 84,2,"015165"
+CPOL: 85,2,"1650000"
at+cpol?

+CPOL: 1,2,"23203"
+CPOL: 2,2,"040230"
+CPOL: 3,2,"350083"

OK

I don’t quite see there is 302720 in the list

I remember the module will first try to register HPLMN and then the operator in +CPOL.
After that if it still cannot register network, it will try to register other PLMN with received high quality signal, lastly it tries to register other PLMN in order of decreasing signal quality.