-
Notifications
You must be signed in to change notification settings - Fork 688
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Bug]: GPS on Portduino works on first application start, subsequent GPS checks fail indefinately. #3909
Comments
We've gotten several reports of this. The strange thing is that not every install has this problem. |
@jp-bennett Yeah, I saw, and also experiencing it, so thought id do the right thing and open an issue so it can be tracked :) |
How is the Pi wired up to the L76K GPS module? |
Could you post some of the data you see coming into /dev/ttyS0. |
Thanks for taking a look @GPSFan
Additionally i've just compiled portduino myself with some code changes in gps.cpp to ignore power off if ARCH_PORTDUINO is defined.... just doing some testing now :) |
@GPSFan UGH, Standby Switch - yup... that is... ON |
Thanks for the update, lots to discuss here. |
Sounds.. fun In the interim i'm attempting this to see if it keeps... updating :) |
If you can recompile, try turning on GPS debugging and Extra Verbose (you may have to edit the variant.h) |
If you point me where too for GPS debugging - im very happy to, didnt spot anything for it in gps.cpp ? |
Add these ti the variant.h file. #define GPS_DEBUG |
Done, i'll let it run for a bit and i can supply the tee'd logfile |
I may be full of horse pucky, all my u-blox parts are putting out GNGGA/GNRMC sentences, so that's not it. ;>( |
TinyGPS+++ takes care of the *GGA/*RMC handling. There was indeed a recent fix for that, but it's not in meshtastic but in our tinygps fork. |
Yeah, brain fart on my part ;>(( |
If it makes you feel any better, my idea has been blown out of the water by the looks by attempting to avoid powering it off....
|
No sweat, just pointing out the direction to save you some trouble looking :-) |
The line of D30CB0D0D30CB0D is puzzling, I'm recompiling my local code with debug & extraverbose to see my corresponding line. |
You get that from "Probing" to what i pasted above L76K GNSS init succeeded.
|
One Debug Logfile with GPS_DEBUG and GPS_EXTRAVERBOSE enabled. |
Reading through the log, I can't see where the Pi got an initial position, the positions reported don't seem to be coming from the GPS.. and all of them seem to be the same. |
this is what I'm seeing during normal operation: |
Just wanted to say I've been dealing with this same issue and I thought it was a hardware problem. Would be great if it wasn't. If you need any logs I dumped an hours worth of just the GPS stuff in discord in the Linux-native channel a few days ago. |
Thanks, I'm trying to get some sort of setup that I can run meshtasticd on. Might be able to borrow a Pi from a friend. I can then patch in a non L76K GPS to see what's going on. |
I've also got a neoM8n GNSS Module on order for testing. Will let you know (if I get mine first) how it does. |
I haven't done much with linux native as my plate is pretty full of DayJob (tm) and RealLife (tm) stuff, but it's on the list. |
I've seen similar message spam from other GPS units on non-pi targets. Never did track down the culprit, other than a GPS unit acting weird. |
@jp-bennett Linux native is a good candidate for tracking down this type of problem, I've got meshtasticd installed on my Rock64 debian Bookworm, but have yet to figure out how to start the daemon. |
Meshtasticd runs and doesn't crash. config is an issue more than likely. |
Make notes about what's missing, and we'll get it added. In this case, probably need a quick word about using systemctl to manage services. |
meshtastic/meshtastic#1259
Let me know if you want to remote into mine, ssh is available over a cloudflared tunnel :) |
OK, testing with a different GPS module ( Ublox Neo-M8N from here). I seem to be getting a similar if not the same issue. It seems that the GPS logging is a bit more verbose than for the waveshare module so maybe this log will help.
|
Very strange, I have a Pi3+ and waveshare hat coming on Wed. So I'll be able to see exactly what is going on, I will however need some help to configure the Pi & hat to duplicate your setup.. |
Noticed that Meshtastic found your u-blox at 57600baud, that is unusual unless you changed it, and saved the config to flash. And Meshtastic didn't print out the MON-VER info, and if it was a u-blox Neo-6 or Neo7 as your link shows, it shouldn't be configured for GPS+SBAS+GLONASS_Galileo. |
I just ran the latest 2.3.9 alpha release .deb |
But ive built the releases before using |
Ok, I'll start with that, I got it so it runs on my Rock64, but without a waveshare hat, no LORA and no GPS. I also have an issue with the SPI config on my Rock64, so I went for broke and got a Pi & waveshare hat on order. That way I'll have a duplicate of a known "working" system. |
Here's the bootup log for the GPS while using the USB interface.
|
What type of module was it sold as? Neo-6, Neo-7 or M8? |
So far I've been unable to download the reference materials for this module because they are behind a baidu link... |
That definitely falls into the counterfeit or fake category, the little 8 pin chip to the left is supposed to be an eeprom, the M8 series has internal flash and did away with that leftover of the Neo-6's. |
Well, counterfeit or not, it seems to be maintaining and updating GPS lock. I'm going to leave it for a few hours to see for sure. |
The internal 8030 chip is genuine, it's the surrounding parts that are substandard and not equivalent to the original. Not made by u-blox, made in some CN sweatshop by child labor, being paid in dog poop and being sold as real with a copied label. |
So, running standard install from the .deb from the assets page -> https://github.com/meshtastic/firmware/releases/download/v2.3.9.f06c56a/meshtasticd_2.3.9.f06c56a_arm64.deb |
URGHHHH Looks like serial login gets enabled by default on a fresh install..... once you have enable a serial port in /boot/firmware/config.txt raspi-config -> Interface Options -> Serial Port -> Or
Found it with Example of disabling it! Checking after 10 minutes i can see meshtasticd still has ownership of /dev/ttyS0 Be good to keep the bug open for another 24 hours and i'll report back if this keeps working, but so far (1 hour in) we are looking good!) Note to self: meshtastic/meshtastic#1259 (comment) |
That is what I had surmised, but without the exact hardware and OS I couldn't be sure. |
I can't believe I didn't think of this... I've been watching the RDX light blink while the Pi boots up without even thinking about it! |
Yeah, I think this is one of those face slapping moments. Docs need to be updated for sure. I guess it is a Pi thing to bring the console serial port out the GPIO header... |
Yeah, OrangePi does it too(with debug console)but unfortunately there is no easy way to disable it. (need to recompile Armbian and u-boot, still not entirely clear). |
Just u-boot requires a recompile from what I am reading - but that's a huge ooft. |
Closing as PR has been merged. |
Category
Other
Hardware
Other
Firmware Version
2.3.9
Description
Portduino (rPi 3 with Waveshare SX1262 LoRaWAN/GNSS HAT)
https://www.waveshare.com/sx1262-lorawan-hat.htm?sku=24654
On meshtasticd start, initial aquisition of GPS via /dev/ttyS0 is successful
Another report of this has appeared in Discord by 'madeofstown' - so not a single occurrence.
Relevant log output
The text was updated successfully, but these errors were encountered: