EM7455B – GNSS sees 11 satellites (SNR 34–35) but never obtains a fix – firmware 02.24.03.00

EM7455B GNSS: 11 satellites with SNR 34–35, antenna bias enabled, but never obtains a GPS fix

Hello,

I am troubleshooting GNSS on a Getac V110 G5 equipped with a Sierra Wireless EM7455B.

At this point I have done extensive testing under both Windows 11 and Ubuntu Live. The GNSS receiver can see satellites and the active antenna is definitely being powered, but the modem never obtains a position fix.

I would appreciate help from somebody familiar with the EM7455 GNSS implementation, especially with firmware/PRI configuration.

Hardware

  • Laptop: Getac V110 G5
  • WWAN modem: Sierra Wireless EM7455B
  • Firmware:

SWI9X30C_02.24.03.00

r6978 CARMD-EV-FRMWR2

2017/03/02 13:36:45

The modem reports:

Manufacturer: Sierra Wireless, Incorporated

Model: EM7455B

Revision: SWI9X30C_02.24.03.00 r6978 CARMD-EV-FRMWR2 2017/03/02 13:36:45

The installed carrier/PRI configuration is:

!IMPREF:

preferred fw version: 02.24.03.00

preferred carrier name: VODAFONE

preferred config name: VODAFONE_001.001_002

current fw version: 02.24.03.00

current carrier name: VODAFONE

current config name: VODAFONE_001.001_002

The GNSS antenna is an active antenna connected to the dedicated GNSS connector on the EM7455.

Initial problem under Windows 11

My original objective is simply to have a working GPS position in Windows applications such as Google Maps, not necessarily to make Getac Manager itself display GPS.

Google Maps reported:

No precise location could be found.

Windows cellular data worked correctly through the EM7455 SIM connection.

Windows serial ports

Initially Windows reported:

COM1

COM2

COM3

COM4

COM5

The relevant Sierra interfaces were:

Sierra Wireless Snapdragon™ X7 LTE-A DM Port (COM3)

Sierra Wireless Snapdragon™ X7 LTE-A NMEA Port (COM5)

Sierra Wireless Snapdragon™ X7 LTE-A WWAN Modem

There was no AT Port exposed by Windows.

Getac GPS driver

The Getac GPS Location Driver was installed correctly.

The device appeared as:

Getac GPS Location Device

ROOT\SENSOR\0000

with:

Status : OK

Problem : CM_PROB_NONE

ConfigManagerErrorCode : CM_PROB_NONE

Present : True

Service : WUDFRd

Driver information:

gpslocationdriver.inf

Driver version: 1.0.2.13

Driver date: 24/06/2020

Manufacturer: Getac Technology Corporation

The Getac registry configuration was:

HKEY_LOCAL_MACHINE\SOFTWARE\Getac\GPSLocationDriver

DefaultComPort REG_DWORD 0x5

DefaultBaudRate REG_DWORD 0x1c200

SSN REG_SZ RM603V1502

Therefore:

DefaultComPort = COM5

DefaultBaudRate = 115200

Windows Location was also enabled:

lfsvc = Running

Location consent = Allow

The Windows Sensor services were:

SensorService Running

SensrSvc Running

SensorDataService Stopped / Manual

The Getac GPS device itself was nevertheless reported by Windows as completely healthy.

NMEA data under Windows

COM5 was found to be the actual NMEA interface.

Reading COM5 manually produced continuous NMEA sentences such as:

$GPGGA,0,*66

$GPRMC,V,N*53

$GPGSV,4,1,16,…

$GPGSV,4,2,16,…

$GPGSV,4,3,16,…

$GPGSV,4,4,16,…

$GPVTG,T,M,N,K,N*2C

$GPGSA,A,1,*1E

The important part was:

GPGSA,A,1

= no fix.

And:

GPGGA,…,0,…

= no fix.

And:

GPRMC,…,V,…

= invalid navigation data.

However, GPGSV showed satellites in view.

An example showed approximately 16 satellites visible, but many of the SNR fields were empty.

Google Maps therefore could not obtain a useful location.

Important discovery: GNSS antenna bias was disabled

At this point I booted Ubuntu Live to get direct access to the EM7455 AT interface.

Linux exposed:

/dev/ttyUSB0

/dev/ttyUSB1

/dev/ttyUSB2

The interfaces were identified as:

ttyUSB0 = USB interface 00

ttyUSB1 = USB interface 02

ttyUSB2 = USB interface 03

The USB driver was qcserial.

Testing ATI showed that ttyUSB2 is the AT port:

Manufacturer: Sierra Wireless, Incorporated

Model: EM7455B

Revision: SWI9X30C_02.24.03.00 r6978 CARMD-EV-FRMWR2 2017/03/02 13:36:45

The other two ports did not respond to AT.

The NMEA interface was confirmed as:

ttyUSB1 = USB interface 02

AT+WANT?

The most important discovery was:

AT+WANT?

+WANT: 0

OK

The active GNSS antenna was physically disconnected at this point and I measured 0 V at the antenna connector.

I then enabled the antenna bias:

AT+WANT=1

The modem returned:

OK

and the voltage at the antenna connector immediately became approximately:

3 V DC

Therefore the EM7455 was definitely able to provide the active antenna bias, but it had previously been disabled.

I then reconnected the active antenna.

After a reboot, I checked again:

AT+WANT?

+WANT: 1

OK

So the setting persisted across reboot.

This proves that the antenna bias is now enabled and remains enabled.

GNSS session

I used the following commands through the AT port:

AT!GPSSTATUS?

AT!GPSLOC?

AT!GPSAUTOSTART?

AT!GPSTRACK=1,255,1000,30,1

AT!GPSFIX=1,255,100000

The modem reports:

Fix Session Status = ACTIVE

but repeatedly:

Last Fix Status = FAIL

FAILCODE = 0

or, after one GPSFIX attempt:

Last Fix Status = FAIL

FAILCODE = 9

AT!GPSLOC? repeatedly returns:

Not Available

The modem’s internal time also remained:

Current time: 1980 01 06

while the Ubuntu system clock was correct:

Wed Aug 12 19:41:42 UTC 2026

Ubuntu reported:

System clock synchronized: yes

NTP service: active

I also tried AT+CCLK?, but that command returned an error on this modem/firmware.

I have therefore not yet established whether the 1980 date is a cause of the no-fix condition or simply a consequence of the GNSS engine never obtaining its first fix.

NMEA output under Linux

Initially /dev/ttyUSB1 did not output anything.

The important discovery was that this firmware requires the NMEA interface to be started using:

$GPS_START

I sent:

printf “\$GPS_START\r\n” > /dev/ttyUSB1

After that, /dev/ttyUSB1 immediately started producing NMEA.

For example:

$GPGSV,3,1,11,33,34,38,34,39,35,40,34*7F

$GPGSV,3,2,11,41,34,42,34,46,34,48,35*74

$GPGSV,3,3,11,49,34,50,35,51,34*73

$GPVTG,T,M,N,K,N*2C

$GPGSA,A,1,*1E

$GPGGA,0,*66

$GPRMC,V,N*53

So NMEA itself is working.

AT!GPSSATINFO?

This was probably the most important measurement.

With the antenna connected to the dedicated GNSS connector, I obtained:

Satellites in view: 11 (1980 01 06 …)

SV: 33 ELEV: 0 AZI: 0 SNR: 34

SV: 38 ELEV: 0 AZI: 0 SNR: 35

SV: 39 ELEV: 0 AZI: 0 SNR: 35

SV: 40 ELEV: 0 AZI: 0 SNR: 35

SV: 41 ELEV: 0 AZI: 0 SNR: 34

SV: 42 ELEV: 0 AZI: 0 SNR: 34

SV: 46 ELEV: 0 AZI: 0 SNR: 34

SV: 48 ELEV: 0 AZI: 0 SNR: 34

SV: 49 ELEV: 0 AZI: 0 SNR: 34

SV: 50 ELEV: 0 AZI: 0 SNR: 35

SV: 51 ELEV: 0 AZI: 0 SNR: 34

OK

This demonstrates:

  • 11 satellites in view.
  • SNR around 34–35.
  • Therefore the GNSS receiver is definitely receiving RF energy from satellites.

However:

GPGSA,A,1

GPGGA,…,0

GPRMC,…,V

continues indefinitely.

The unusual part is that all satellites reported by GPSSATINFO have:

ELEV = 0

AZI = 0

despite having SNR around 34–35.

Testing the AUX antenna connector

Because I found Sierra documentation indicating that GPSSEL=0 corresponds to the dedicated GNSS input and GPSSEL=1 corresponds to the shared/AUX input, I performed a physical test by moving the same active antenna from the dedicated GNSS connector to AUX.

With the antenna on AUX:

Satellites in view: 1

SV:255

ELEV: 0

AZI: 0

SNR: 15

and:

Fix Session Status = ACTIVE

Last Fix Status = NONE

This is clearly worse than the dedicated GNSS connector, where I had:

11 satellites

SNR 34–35

Therefore I returned the antenna to the dedicated GNSS connector.

I tried querying and changing GPSSEL, but my PRI behaves differently from some examples found on the Sierra forum.

AT!CUSTOM? initially returned:

GPSENABLE 0x01

GPIOSARENABLE 0x01

IPV6ENABLE 0x01

SIMLPM 0x01

USBSERIALENABLE 0x01

It did not initially display GPSSEL.

However:

AT!CUSTOM=?

showed that GPSSEL is actually a supported custom parameter:

“GPSENABLE”

“GPSLPM”

“GPIOSARENABLE”

“GPSSEL”

“GPSREFLOC”

I tried:

AT!CUSTOM=“GPSSEL”,0

but the modem returned:

ERROR

I also tried querying it using:

AT!CUSTOM=“GPSSEL”?

which also returned:

ERROR

So at the moment I do not know the current GPSSEL value, and I have not been able to modify it.

GPSLPM test

I also found reports involving GPSLPM.

I changed:

AT!CUSTOM=“GPSLPM”,0

and restarted the modem.

The change did not solve the problem.

After reboot:

Fix Session Status = ACTIVE

Last Fix Status = NONE

and the modem still did not obtain a fix.

USB composition tests

The original USB composition was:

USB composition 8:

DM, NMEA, AT, MBIM

I temporarily changed the modem to composition 6:

USB composition 6:

DM, NMEA, AT, QMI

using:

qmicli --dms-swi-set-usb-composition=6

The command succeeded.

However, QMI access was not functioning correctly afterwards:

Couldn’t create QmiDevice

Operation timed out

CID allocation failed in the CTL client

The modem’s AT interface remained usable, but after the composition change the GNSS behaviour changed and GPSSATINFO returned:

NO SAT INFO

I therefore returned the modem to its original composition 8 before continuing.

The original composition 8 is:

DM + NMEA + AT + MBIM

Commands which do NOT work on this firmware

For reference, these commands produced ERROR or are not implemented on this particular EM7455:

AT!GPSSTART

AT!GPSCOLDSTART

AT!GPSGSV?

AT!GPSCFG?

AT!GPSNMEA?

AT!GPSXTRATIMEENABLE=2

AT+CCLK?

AT!ENTERCND

AT!HELP

AT+CLAC

More precisely:

AT!GPSNMEA?

→ !GPSNMEA: NOT IMPLEMENTED

AT!GPSSTART returns ERROR when a GPS session is already active.

AT!GPSEND=0 returns:

ErrCode = 12

OK

when there is no existing session to terminate.

AT!GPSFIX=1,255,100000 is accepted:

OK

but ultimately results in:

Last Fix Status = FAIL

FAILCODE = 9

or earlier FAILCODE = 0.

Current situation

At the moment the situation is:

Hardware

EM7455B

Active GNSS antenna

Antenna connected to dedicated GNSS connector

Firmware

SWI9X30C_02.24.03.00

PRI

VODAFONE_001.001_002

Antenna bias

AT+WANT?

+WANT: 1

Physical measurement:

approximately 3 V DC on the antenna connector

GNSS

GPSENABLE = 1

GPS session = ACTIVE

Satellite reception

Dedicated GNSS input:

11 satellites

SNR approximately 34–35

Position

Still:

Last Fix Status = FAIL

GPSLOC = Not Available

NMEA still shows:

GPGSA,A,1

GPGGA,…,0

GPRMC,…,V

Internal modem time

Still:

1980-01-06

even though the host system is correctly synchronized to:

2026-08-12

What I am trying to determine

At this point I believe the hardware RF path is basically working because:

  1. The active antenna is now definitely powered.
  2. The modem sees 11 satellites.
  3. SNR is around 34–35.
  4. The NMEA port works after $GPS_START.
  5. The GNSS session is active.

But the receiver never transitions from satellite acquisition to a valid navigation solution.

The most suspicious symptoms are:

1980-01-06 modem time

combined with:

ELEV = 0

AZI = 0

for all satellites despite:

SNR = 34–35

and the persistent:

GPGSA,A,1

GPGGA,…,0

GPRMC,…,V

I would particularly appreciate advice on:

  1. Whether GPSSEL can be read or changed on a module using PRI:

VODAFONE_001.001_002

  1. Whether this particular PRI expects the dedicated GNSS connector to be selected in another way.
  2. Whether the 1980-01-06 internal modem time is expected before first fix, or whether it must be explicitly initialized from the host.
  3. Whether ELEV=0 / AZI=0 / SNR=34–35 in AT!GPSSATINFO? indicates a GNSS firmware/PRI problem rather than a weak RF signal.
  4. Whether there is a known issue with:

SWI9X30C_02.24.03.00

and GNSS acquisition/fix on this PRI.

  1. Whether I should change the PRI/firmware to another known-good configuration for GNSS operation.

My ultimate goal is simply to make the EM7455 provide a valid GNSS position to Windows so that applications such as Google Maps can use it.

I have intentionally not flashed new firmware yet because I would prefer to understand whether this is a GNSS configuration/PRI problem before changing the modem firmware.

Any advice on the correct next step would be very much appreciated.

I can provide additional command outputs or logs if required.

Have you tried other newer firmware ?

https://source.sierrawireless.com/resources/airprime/minicard/74xx/em_mc74xx-approved-fw-packages/

Other user can make it work:

Thanks. Yes, I am currently running SWI9X30C_02.24.03.00 with PRI 001.001_002 (Vodafone).

I have not tried a newer firmware yet.

I have already done extensive GNSS testing under Ubuntu. The modem can see 11–16 satellites with SNR around 34–35 dB-Hz, the active antenna is powered with approximately 3 V (AT+WANT: 1), and NMEA output works correctly. However, the receiver never obtains a fix:

GPGSA,A,1

GPGGA,…,0

GPRMC,V

and AT!GPSLOC? returns Not Available.

The modem GNSS time also remains at 1980-01-06.

I am now back on USB composition 8 (DM, NMEA, AT, MBIM).

I see that Sierra currently lists newer EM7455 firmware, including 02.33.03.00, 02.38.00.00 and 02.39.00.00.

Which firmware/PRI combination would you recommend for the EM7455B if the main objective is to get GNSS working correctly under Windows 11?

I would prefer to use a recommended/approved combination rather than flashing a random carrier image.

Also, regarding your comment about Skylight registry keys: could you please provide the exact keys/values that enabled the hidden Location/GNSS functionality?

I have already confirmed that the Getac GPS Location Driver is installed and the Windows Location service is running, but Windows still does not obtain a position.

Thanks.

Which firmware/PRI combination would you recommend for the EM7455B if the main objective is to get GNSS working correctly under Windows 11?

Maybe try the latest two firmware?

I have tried the version 02.38.00.00 before , and it can get fix in NMEA frame

Thanks. That’s very useful information.

My current firmware is SWI9X30C_02.24.03.00 with VODAFONE_001.001_002 PRI.

Since you have personally confirmed that 02.38.00.00 obtains a GNSS fix and outputs valid NMEA frames, I would like to try it.

Would you recommend 02.38.00.00 or the newer 02.39.00.00 for my EM7455B?

Also, did you use the Generic PRI 002.082_000 with 02.38.00.00, or did you keep your existing carrier PRI?

My LTE connection currently works correctly with the Vodafone PRI, so I would like to avoid changing the carrier configuration unnecessarily.

i used this before:


ati3
Manufacturer: Sierra Wireless, Incorporated
Model: MC7455
Revision: SWI9X30C_02.38.00.00 rE3F3E55C03AE94 jenkins 2022/01/13 05:16:00
MEID: 35907206012435
ESN: 12800001234, 80001234
IMEI: 359072060012345
IMEI SV: 22
FSN: LQ530100512345
+GCAP: +CGSM


OK
at!impref?
!IMPREF:
 preferred fw version:    02.38.00.00
 preferred carrier name:  GENERIC
 preferred config name:   GENERIC_002.082_000
 preferred subpri index:  000
 current fw version:      02.38.00.00
 current carrier name:    GENERIC
 current config name:     GENERIC_002.082_000
 current subpri index:    000

OK

Thanks, that is very useful.

So you used:

SWI9X30C_02.38.00.00
GENERIC_002.082_000

and GNSS was able to obtain a fix and output valid NMEA.

I currently have:

SWI9X30C_02.24.03.00
VODAFONE_001.001_002

Since my modem is an EM7455B and I am also seeing satellites with good SNR but never obtaining a fix, I will probably try 02.38.00.00.

Before I flash it, did you have any problems with LTE/network operation after switching from a carrier PRI to the Generic PRI? Did the modem work normally with a SIM from a Vodafone network?

Also, would you recommend 02.38.00.00 over 02.39.00.00 for GNSS on the EM7455?

no idea, probably you can make a test to verify

Actually you can go back to Vodafone firmware after testing