GPS sample not getting a fix - indoors or outdoors

dajtxx

Hi,

I’m trying to get the GPS demo going and understand the AT commands necessary to initialise the GPS radio.

The demo cannot get a fix, either inside or outside. I have the antenna that came with the board connected to the GPS connector.

It seems to me the %XMAGPIO and %XCOEX0 commands won’t be issued because neither CONFIG_BOARD_NRF9160DK_NRF9160NS or CONFIG_BOARD_THINGY91_NRF9160NS are defined.

Could this be why the demo can’t get a GPS fix?

As it is, the radio is getting at %XSYSTEMMODE command to enable the GPS (I have disabled LTE for now) and a +CFUN=31 command, and that’s it. Is that enough for the Feather?

jaredwolff

dajtxx I’ve never experimented with getting a fix using AT commands.

dajtxx It seems to me the %XMAGPIO and %XCOEX0 commands won’t be issued because neither CONFIG_BOARD_NRF9160DK_NRF9160NS or CONFIG_BOARD_THINGY91_NRF9160NS are defined.

This very well could be your problem.

If you can set these guys in your prj.conf

# COEX0 is used to enable the GPS LNA, but it has to be configured to do so.
CONFIG_NRF9160_GPS_SET_COEX0=y
CONFIG_NRF9160_GPS_COEX0_STRING="AT%XCOEX0=1,1,1565,1586"

That will turn on the correct enable signal for the active GPS. (Are you using an active GPS antenna?)

dajtxx

Adding those lines to prj.conf didn’t seem to have any effect so I need to learn more about the zephyr build system.

However adding the AT command to the code made it work, so thanks for that 🙂

The sample code can also issue an XMAGPIO command, but that does not seem to be required given I could get a GPS fix without it! Do you know why? The sample code also implies (via the macro name) that the XCOEX0 command you gave me is for an “onboard” antenna, not an external one, but I am using the external active antenna you supply with the board. Do you know why the XMAGPIO command might be required, and why the macro name says the XCOEX0 command we use is for an onboard antenna?

jaredwolff

dajtxx Nordic likely developed these examples with the Thingy91 or the nRF9160DK in mind. So it may have something conflicting compared to the nRF910 Feather. The MAGPI pins aren’t used on the nRF9160 Feather.

Skalis

Hi,

I have a Feather V3 , I can get the GPS to work correctly (with a flexible passive antenna).

I bought last week on your website the new V4 version (and the Active patch antenna) , but on this version impossible to have a GPS fix (same location,same sim card) , whether with the active or passive antenna (I set COEX0_STRING) , do you have any idea of what could be the cause of the problem ? (modem FW 1.3.1 in both)
From schemetics the GPS circuit seems to be exactly the same.

Thank you

jaredwolff

@Skalis I’m not sure how a passive antenna would work for GPS. The board only supports active GPS antennas.

Are you using your own code or basing it off of a sample?

Skalis

I’m testing with the gps sample - nfed/samples/gps.

Yes I know but I didn’t have an Active antenna, I used a passive flex one (I just chnaged CONFIG_GPS_SAMPLE_AT_COEX0=“AT%XCOEX0”), and I was also surprised that it works.

But now with the V4 and with the active patch, I am not able to get a fix (I set back COEX0 to the default value).

Thank you

jaredwolff

Also are you doing this with or without AGPS Skalis?

jaredwolff

The COEX0 value should be

CONFIG_GPS_SAMPLE_AT_COEX0="AT\%XCOEX0=1,1,1565,1586"

Are you able to program the same code to your V3 and have it work with the same active antenna?

Skalis

jaredwolff Hello , I am testing with gps sample without agps .
I did a test outdoor with the same patch antenna with V3 and V4, and I analyzed NMEA data .

for hardware revision 3, C/n0 of the used satellites was typically in the 35-45 dBHz, while for hardware revision 4, it was more in the 25-40 dBHz.
when focusing on the reported C/n0 for the same satellite, the values reported by hardware revision 3 were usually better.

What could be the reason ? the new power IC ? Thank you

jaredwolff

New power IC may have something to do with it as its a little noisier (comes with the power improvements though!) I’ve set it to go into PWM mode during GPS activity to reduce noise.

I have noticed simply that some devices get a fix much faster than others. Any reason why you’re not using AGPS? I’d be curious your fix times with that on.