Home › Forums › Forums › Technical Support for the OzzMaker SARA-R5 LTE-M GPS + 10DOF › Raspberry Pi – couple of questions
- August 14, 2023 at 6:33 am #18420
I seem to have this HAT setup and working on my headless Pi4 for my telescope mount but there’s a couple of things I can’t quite work out,
- Restarting the Pi with shutdown -r now results in the mux connection disappearing on reboot. /dev/pts only has 0 (when I ssh in) and ptmx on the reboot. How can I fix this? The only way I have found to get the mux back is to power down everything, which isn’t ideal! The /tmp entries are there at reboot but show as red as the links can’t be reached. Do I need to delete these at shutdown?
- I’ve updated the python script for python 3 (couple of small changes like the print uses brackets now) but I would like it to be able to differentiate between when the SIM card is available and when it isn’t. To get the normal GPS working with no assistance, I need to add AT+UGIND=1 to the list of commands but this isn’t needed if the SIM is available and the AssistNow connection and download works. Can the Python script be changed to realise when the AT+UGIND=1 is needed? My python skills are non-existent … running the script when the SIM card isn’t in the HAT results in the normal GPS not working (I think)
JimbobAugust 18, 2023 at 12:31 am #18435
1. You need to start gsmMuxd on every boot. You can use the rc.local to start it.
You also need to shut it down or kill in when rebooting the Pi. This disables the mux on both the 5G modem and the Pi itself. a simple
sudo killall gsmMuxdwill do.
We are working on a guide to create a systemd file to look after all this so it is automated
2.Even if you try assist online and you dont have a SIM in, it should still work, it will just fail to download the data and rely on normal data it receives from the satellites in the sky. AT+UGIND=1 is only used to report results. This command
AT+UGPS=1,1,71can be used to stat with local aiding only.August 18, 2023 at 5:54 pm #18436
Hmmm … interesting. I do have gsmMuxd starting via systemd and I assume it gets closed down as part of the shutdown (or reboot) process. My simple service file is attached and started before the python script and gpsd.
I *think* I’ve managed to change the python script to check if the SIM is present to fix question 2 … just need to check it actually works. Its working correctly without the SIM card inserted and need to check it works with the SIM and downloads the AssistNow data.
Thanks.September 25, 2023 at 7:24 am #18575
I can see you have updated the mux setup pages to help fix the missing mux on reboot … not tried it yet but have yet another question about this board. I noticed the attached on the SparkFun guide linked here for a firmware issue with the PPS setup with the version 2.06 firmware – attached is an image of the relevant text. I assume this is an issue with your board too or has it been resolved? One of the main reason for buying this board was the A-GPS capability but I also need to use the PPS sync … if this is impossible then what can you do to resolve this?
I purchased my board from PiHut in the UK and pretty sure they will augh if I try to return it now after the setup issues raised above.
Thanks!September 25, 2023 at 6:31 pm #18578richardpParticipant
Regarding sparkfun’s guide, I think that may be related to the use the LED.
We did test this in the past.. and I just tested now, it does work. The Pi can see the pin go high
On a Pi 4 i connected the TP pin from the SARA-R5 to GPIO21 on the PI.
I created a python script to see when this goes high;
import RPi.GPIO as GPIO GPIO.setmode(GPIO.BCM) GPIO.setup(21, GPIO.IN) while True: if GPIO.input(21): print (" high!")
I then configured the below commands on the SARA-R5
AT+UTIMEIND=1 AT+UTIME=1,1 AT+UGPS=1,4,67
The python script was able to see the pin go high every second.
Richard --OzzMaker.com --September 26, 2023 at 5:25 am #18583
Phew! Thanks for putting my mind at rest. I guess I’ll be digging out the soldering iron to attach some pins then.October 5, 2023 at 8:43 am #18597
This is definitely not working for me and the more I try to debug what is happening the more frustrating things become …
As mentioned at the start I’m using the python script to setup the GPS but when I check this script against the commands in the web page there are things missing – should these be in the python script? An example is the AT+CGACT=1,1 command. I also have to amend this script for it to actually run – the main example is the print statements need to contained in brackets plus the ser.write line needs the command to be changed to command.encode(). The UBLOX AT commands document is a bit of a bible and wading through this while flicking between windows is not ideal!
I now have nothing working … no gpsmon and the gsmmuxd seems to be as flaky as hell when using minicom or the device is just crashing … again, I can’t actually tell what is going on here. I have reverted to the python script on one of the GPS pages for this device and nothing.
It feels like I need to reset the device back to it’s factory settings and start again or just give up completely and write off the cash I have spent on this device. Like I said, frustrating …
Another thing I have noticed is there is no SHM entry for the GPS device like I see for cheaper GPS modules, usually I have chrony using the SHM for the GPS – is this normal?October 5, 2023 at 2:12 pm #18599
Hi, i am so sorry you are having issues.
What web page are you getting the python script from? is it this one?October 8, 2023 at 11:44 pm #18609
Yes, that is the script I’m using. The only changes I’ve made is a check to see if the SIM card is present, if it’s not then don’t bother trying to set up the modem to download the AssistNow data because it’s a waste of time. The AT+UPSDA seems to take longer when there is no SIM card in the HAT.
Problem I am seeing with the PPS commands is I cannot start the GPS with the AT+UGPS after I use the AT+UTIMEIND and AT+UTIME commands shown above. First I thought I might need the fully AT+UTIME command as 1,1,1,100 but this makes no difference whatsoever.
This is the output of my python script,
Oct 08 13:12:13 stellarmate systemd: Started Python script for Ozzmaker GPS.
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: Command sent: b’AT+CFUN=0′
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: response b’OK\r\n’
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: Command sent: b’AT+CGDCONT=0,”IP”,”hologram”‘
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: response b’ERROR\r\n’
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: Command sent: b’AT+UTIMEIND=1′
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: response b’OK\r\n’
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: Command sent: b’AT+UTIME=1,1′
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: response b”
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: Command sent: b”
Oct 08 13:12:26 stellarmate ozzmaker-gps.py: response b’OK\r\n\r\n+UUTIMEIND: 1,0,0,0\r\n\r\n+UUTIMEIND: 4,2,0,0\r\nAT+UGPS=1,4,71\r\r\nERROR\r\n’
Oct 08 13:12:26 stellarmate systemd: ozzmaker.service: Succeeded.
It doesn’t seem to matter what I enter for the AT+UGPS command as all the combinations I’ve tried so far result in the ERROR message. Neither gpsd or cgps show any data coming from the ublox module via /dev/ttyGSM2. The uBlox AT commands bible seems to suggest the PPS pin needs to be setup before trying to use it and I think this is pre-configured already.
Removing the UTIMEIND and UTIME commands results in a working GPS but no PPS sync. For the record, I have the PPS pin setup on the Raspberry Pi to use GPIO18 (previously used with my bog standard GPS modules) and have also tried to use GPIO21 from your example above.
I haven’t tried to solder a blob on JP1 and everytime I want to try a change I have to power everything down and restart. What am I doing wrong? It seems the UUTIMEIND reply above is trying to use the LTE modem, do I need the time commands after I’ve started the GPS messages?November 13, 2023 at 2:05 am #18679
Just picking this up again as it is still not working and I don’t particularly want to consign £60 (around AU$120) worth of kit to the dust bin …
Brand new install on a Pi Zero using the lite image burnt to an SD card via the Raspberry Pi Imager software. I have gsmMuxd built and working and I have also soldered the jumper to power cycle the HAT on reboot. Still encountering the issues mentioned above – using the information from post 18578, I get this,
Welcome to minicom 2.8
Port /dev/ttyGSM0, 14:49:27
Press CTRL-A Z for help on special keys
Doesn’t matter if I use ttyGSM0 or ttyGSM1, I get the same error message plus it doesn’t matter what the AT+UGPS command is as nothing works and everything results in ERROR. It seems as soon as I use AT+UTIME then I cannot start the GPS module at all. Is this a super bad bug or just something out of sequence? Is there an email address for support as the forum doesn’t seem to be helping much to get PPS working on this HAT.
Currently using my $5 u-blox module from AliExpress which surprisingly “just works” after I realised the interference from the USB3.0 connections on the Pi 4 were blocking the module / antenna – quick ferrite core suppressor added to the cable and it works.November 13, 2023 at 4:05 am #18680
Just found this thread on the u-blox forums,
Seems to be pretty similar to the issues I’m seeing and thinking this module is badly flawed …November 13, 2023 at 3:56 pm #18681
It looks like the GPS cannot be used for PPS when the GPS is streaming to the Cellular module.
I just tested with using LTE module as the source for time and it works
The only difference is that it is not as accurate, 500ns instead of 50nsNovember 14, 2023 at 7:53 pm #18686
Thanks for this. I’ve just tried using this via ttyGSM0 this morning and after an initial hiccup when I lost the mux connections the AT+UTIME=1,2 comes back with an OK but I still have no consistent pulse going into the raspberry pi from what I can see. I used the python code in this thread to check the pin I’m using (GPIO18) and it goes high a couple of times then pauses for a minute or two and then goes high. I think I’ll drop The Pi Hut a message and see what my options are as this isn’t providing what I need.
For example, I’m looking for the time pulse information to go into gpsd (like most GPS modules provide) and at first glance this seems impossible.November 15, 2023 at 5:55 am #18688
Small update, the gsmMuxd seems to crash or fail or something (I can’t tell because my Pi is running headless) after 2 hours and rebooting with the JP1 jumper fix does not bring it back (this is using the raspi-gpio commands on the mux setup page). No idea what is happening but a full power down for 5 mins or so fixes this. I do wonder about the quality of the u-blox firmware / implementation for this chip as it seems to be flaky as hell.
One question I have asked the Pi Hut is, could I have a faulty HAT? Mark, I’m sure you’re not seeing the issues I am with the gsmMuxd software and general stability for this module …
- This reply was modified 3 months, 2 weeks ago by Jimbob17.
- You must be logged in to reply to this topic.