Unsatisfactory GPS reception on PinePhone?

[sog8b]

Hey Guys,

Thanks for your help.
I got things working in the end.
It seemed to help when I used the following command to start the GPS.
AT+QGPS=1,255,1000,0,10

Also I think I may have had two programs listening on ttyUSB1 which may have meant each only was getting half the data (however I'm not sure of this).

Thanks again!

[wibble]

Glad you got it working. I've not had reason to try the other parameters.

[calinb]

Hey Guys,

Thanks for your help.
I got things working in the end.
It seemed to help when I used the following command to start the GPS.
AT+QGPS=1,255,1000,0,10

Also I think I may have had two programs listening on ttyUSB1 which may have meant each only was getting half the data (however I'm not sure of this).

Thanks again!

UPDATE

Doh! I forgot that the phone must be enabled for the GPS to work. I read about it on a thread in these forums but I temporarily forgot. The first command works. On to acquire sats now.


I've had my Manjaro CE edition for about a month. I'm running Mobian from eMMC and I've seen no indication that the GPS is working at all on the two maps aps I've tried so I tried the first incantation from the OP:

Code:

mmcli -m 0 --location-enable-gps-nmea

and it produced:

"error: modem not enabled yet"

So I tried

Code:

sudo mmcli -m -0 AT+QGPS=1,255,1000,0,10

which seemed to run okay, but I still get the error.

I also tried it with Manjaro Posh on the latest multiboot SD and experienced the same error.

Mobian project says (https://wiki.mobian-project.org/doku.php?id=pinephone )

What's working with Mobian
GPS/GNSS Works with exceptions (2)

(2) GPS/GNSS only works with web apps using GNOME Web. (Editors note: Is this still true? I think it works in general now)

So it seems like the mmcli interface should be working for me. What am I missing? Do I need to risk a modem firmware update?

UPDATE

Doh! I forgot that the phone must be enabled for the GPS to work. I read about it on a thread in these forums but I temporarily forgot. The first command works. On to acquire sats now.

[calinb]

I'm still trying to get a fix from my GPS with Mobian. I've tried everything in this thread, I think, but have yet to to see a single satellite tracked. I referenced the secret decoder ring page on archive.org, because gpsinformation.org is down:
http://web.archive.org/web/2020013103103...e/nmea.htm

The agps scripts appear to be working (thanks alastair-dm) and I get three or four $GPGSV lines ("sentences") but the SNR field is always just a comma for all sats.

As an experiment, I disabled wifi Internet (I have no cellular service here at home) and ran the agps script to delete agps files. Then I rebooted for good measure and used mmcli -m 0 --location-enable-gps-nmea

and repeatedly
mmcli -m 0 --location-get

to monitor the arrival of satellite information. It never comes! The output of --location-get is perpetually devoid of any $GPGSV sentences, no matter how long I wait.

Given that the AGPS-seeded $GPGSV lines display no signal/noise ratio (SNR), could my GPS simply not be receiving satellite signals at all? My 10+ year old pocket Garmin works when booted from cold (after a couple of years of non-use) and my Moto smartphone GPSs work here too.

Does the diagnostic test in the factory Manjaro image contain a functional test for satellite reception? I could download and boot it.

Thanks for any ideas about how to proceed!

[spartacus]

I'm still trying to get a fix from my GPS with Mobian. I've tried everything in this thread, I think, but have yet to to see a single satellite tracked. I referenced the secret decoder ring page on archive.org, because gpsinformation.org is down:
http://web.archive.org/web/2020013103103...e/nmea.htm

The agps scripts appear to be working (thanks alastair-dm) and I get three or four $GPGSV lines ("sentences") but the SNR field is always just a comma for all sats.

As an experiment, I disabled wifi Internet (I have no cellular service here at home) and ran the agps script to delete agps files. Then I rebooted for good measure and used mmcli -m 0 --location-enable-gps-nmea

and repeatedly
mmcli -m 0 --location-get

to monitor the arrival of satellite information. It never comes! The output of --location-get is perpetually devoid of any $GPGSV sentences, no matter how long I wait.

Given that the AGPS-seeded $GPGSV lines display no signal/noise ratio (SNR), could my GPS simply not be receiving satellite signals at all? My 10+ year old pocket Garmin works when booted from cold (after a couple of years of non-use) and my Moto smartphone GPSs work here too.

Does the diagnostic test in the factory Manjaro image contain a functional test for satellite reception? I could download and boot it.

Thanks for any ideas about how to proceed!

[spartacus]

I'm still trying to get a fix from my GPS with Mobian. I've tried everything in this thread, I think, but have yet to to see a single satellite tracked. I referenced the secret decoder ring page on archive.org, because gpsinformation.org is down:
http://web.archive.org/web/2020013103103...e/nmea.htm

The agps scripts appear to be working (thanks alastair-dm) and I get three or four $GPGSV lines ("sentences") but the SNR field is always just a comma for all sats.

As an experiment, I disabled wifi Internet (I have no cellular service here at home) and ran the agps script to delete agps files. Then I rebooted for good measure and used mmcli -m 0 --location-enable-gps-nmea

and repeatedly
mmcli -m 0 --location-get

to monitor the arrival of satellite information. It never comes! The output of --location-get is perpetually devoid of any $GPGSV sentences, no matter how long I wait.

Given that the AGPS-seeded $GPGSV lines display no signal/noise ratio (SNR), could my GPS simply not be receiving satellite signals at all? My 10+ year old pocket Garmin works when booted from cold (after a couple of years of non-use) and my Moto smartphone GPSs work here too.

Does the diagnostic test in the factory Manjaro image contain a functional test for satellite reception? I could download and boot it.

Thanks for any ideas about how to proceed!



I have been struggling with gps satellite acquistion for the past couple of months.  I got my first location earlier today.  I put my pinephone outside, ssh'd from a laptop, and watched the progress using gpsmon.  Took about 30 minutes before it began tracking a satellite.  Took 90 minutes to lock 4 satellites and report a location.  

[wibble]

@calinb The factory test image that came on my Breave Heart had no specific GPS test that I remember. I don't know what diagnostics they're providing with more recent models. It would be difficult to test in a factory without something transmitting a fake GPS signal, so it may not have a specific test.

I used an old Garmin Geko 201 as a cold fix reference before starting on the AGPS stuff. With the Brave Heart I was seeing sats in GPGSV messages even when I didn't have a fix, similar to the Garmin's sat view before fix. If you're seeing nothing in situations where the Garmin can get a fix then something's wrong. Have you tried accessing the modem control interface with a terminal (picocom or similar) when ModemManager is stopped? That may show error messages that aren't present in the NMEA feed - assuming the Quectel interface passes them on via AT messages. Beyond that you might have to go digging in libqmi to find out if there are any lower level GPS diagnostics from the Qualcom interface.

The other more invasive thing to check is the physical antenna connection. I think the GPS antenna is one of the printed ones on the inner frame, with spring contacts to the main board. It's not something I've looked into, so I can't give you more detailed pointers other than looking in the wiki for the mainboard swap instructions for inner frame removal.

[calinb]

I have been struggling with gps satellite acquistion for the past couple of months.  I got my first location earlier today.  I put my pinephone outside, ssh'd from a laptop, and watched the progress using gpsmon.  Took about 30 minutes before it began tracking a satellite.  Took 90 minutes to lock 4 satellites and report a location. 

Good to hear I'm not the only one having trouble (sort-of).

@calinb The factory test image that came on my Breave Heart had no specific GPS test that I remember. I don't know what diagnostics they're providing with more recent models. It would be difficult to test in a factory without something transmitting a fake GPS signal, so it may not have a specific test.

I used an old Garmin Geko 201 as a cold fix reference before starting on the AGPS stuff. With the Brave Heart I was seeing sats in GPGSV messages even when I didn't have a fix, similar to the Garmin's sat view before fix. If you're seeing nothing in situations where the Garmin can get a fix then something's wrong. Have you tried accessing the modem control interface with a terminal (picocom or similar) when ModemManager is stopped? That may show error messages that aren't present in the NMEA feed - assuming the Quectel interface passes them on via AT messages. Beyond that you might have to go digging in libqmi to find out if there are any lower level GPS diagnostics from the Qualcom interface.

The other more invasive thing to check is the physical antenna connection. I think the GPS antenna is one of the printed ones on the inner frame, with spring contacts to the main board. It's not something I've looked into, so I can't give you more detailed pointers other than looking in the wiki for the mainboard swap instructions for inner frame removal.

I looked at the list of test coverage in Martijn's test too:[url=https://gitlab.com/MartijnBraam][/url]  https://gitlab.com/MartijnBraam/factoryt...README.rst

It doesn't look like there's anything there that will help diagnose the problem. I get up to four $GPGSV lines of sats using mmcli, but only when I use AGPS. However, none of them indicate a signal (commas appear in all SNR fields), except yesterday I did briefly get three sats with a SNR entry, but all SNRs were around 20 or less and I never got a fix.

Despite my other Moto Android phones and cheap 10+ year old Garmin GPS working here at home, I never get any SNR on my Pinephone at home.  I live on a southern canyon wall so I drove to some very flat farmland nearby with no obstructions of the sky within sight. I got a 4G mobile signal and data too (no cell service at home), so I could "seed" with AGPS data. I spent about an hour waiting for a fix with the Pinephone tucked up under the windshield toward the southern sky and I also spent time standing in the cold outside. That's how I got my record three SNR readings. This capture only shows two SNR values, and I missed capturing the "record":

Code:

--------------------------
 3GPP |      operator code: 310
         |      operator name: 410
         | location area code: FFFE
         | tracking area code: 9103
         |            cell id: 066AA80F
  --------------------------
 GPS  |              nmea:
         |                    $GPGSV,3,1,11,05,14,314,16,09,73,144,21,02,,,,03,,,,1*69
$GPGSV,3,2,11,04,33,132,,06,,,,07,66,316,,08,31,113,,1*5D
$GPGSV,3,3,11,11,,,,14,18,215,,16,15,042,,1*6B
      |                    $GPGSA,A,1,,,,,,,,,,,,,,,,*32
      |                    $GPRMC,,V,,,,,,,,,,N*53
      |                    $GPVTG,,T,,M,,N,,K,N*2C
      |                    $GPGGA,,,,,,0,,,,,,,,*66

Sat PRN 05, SNR = 16
and
Sat PRN 09, SNR = 21

All others have commas in the SNR field.

So my Pinephone is receiving something (mmcli -m 0 --location-status has "status yes" too. Signals just seem to be very weak and infrequent.

Thanks for the suggestions!

I'll try 90 minutes sometime when the weather improves again. (It started snowing today so cloud cover attenuation is possible). I'll also look into picocom and maybe I'll need to run SSH remotely from my laptop from inside the car to the Pinephone on the car's roof. I'll stay warmer and perhaps get a stronger signal too. Finally, I guess I'd better at least do a visual inspection of the Pinephone antenna, if not try to "ohm it out" in some way. I have a good Fluke meter with a low ohms scale. I don't know what the resistive component of the impedance should be but I'll look into it.

[Benatti]

Estive tentando habilitar gps no meu pinephone 32G emmc / 3G RAM e não funcionou, observei que quando uso um "gps usb dongle" nele aparece /dev/ttyACM0 e /dev/gps0 e os programas marble e foxtrotgps funcionam e sem o "gps usb dongle" nós temos /dev/ttyUSB1 (gps) mas não temos /dev/ttyS1. Parece que o /dev/ttyS1 está fazendo falta para o gpsd, sendo que a saída de gpsmon é quase a mesma com ou sem "gps usb dongle".
Teria um comando AT para habilitar renomear ou criar /dev/ttyS1 e ver se resolve o problema?
Também encontrei diferença no funcionamento do foxtrot editando /etc/default/gpsd em DEVICES="/dev/ttyUSB1"
agradeço a atenção de todos.

[Benatti]

Estive tentando habilitar gps no meu pinephone 32G emmc / 3G RAM e não funcionou, observei que quando uso um "gps usb dongle" nele aparece /dev/ttyACM0 e /dev/gps0 e os programas marble e foxtrotgps funcionam e sem o "gps usb dongle" nós temos /dev/ttyUSB1 (gps) mas não temos /dev/ttyS1. Parece que o /dev/ttyS1 está fazendo falta para o gpsd, sendo que a saída de gpsmon é quase a mesma com ou sem "gps usb dongle".
Teria um comando AT para habilitar renomear ou criar /dev/ttyS1 e ver se resolve o problema?
Também encontrei diferença no funcionamento do foxtrot editando /etc/default/gpsd em DEVICES="/dev/ttyUSB1"
agradeço a atenção de todos.

esqueci de mencionar que estou usando mobian.