[Bug]: Separate U-Blox Neo-6M (v14) GPS init from M8 (v15) init and recover bitrate

Max-Plastix commented 1 year ago

Hardware

T-Beam

Firmware Version

2.1.18

Description

In some cases, the Neo-6M GPS found on LILY TTGO T-Beam v1.1 units with SMA does not initialize correctly.

Many of the U-Blox commands sent to it are rejected because they match a later version of the protocol (v15) instead of the v14 that is supported.

If the power-on bitrate is not 9600, the module might never sync correctly or have the correct sentences configured.

I would propose making these changes:

Speed u-Blox interface from 9600 to 115200. Messages arrive faster, with less ongoing data stream burden.
Add a baudrate search (try them all) if the module does not connect at 115200 on powerup
Detect firmware & protocol version to differentiate M6 from 8M
Add protocol-specific initialization to configure messages and save settings
(maybe) Add a more thorough "factory reset" fallback to zero out settings and reconfigure.

Relevant log output

DEBUG | ??:??:?? 1 WANT GPS=1
WARN  | ??:??:?? 3 No response for class 06 message 3E
WARN  | ??:??:?? 3 Unable to reconfigure GNSS, keep factory defaults
WARN  | ??:??:?? 4 No response for class 06 message 39
WARN  | ??:??:?? 4 Unable to enable interference resistance.
WARN  | ??:??:?? 6 No response for class 06 message 23
WARN  | ??:??:?? 6 Unable to configure extra settings.
WARN  | ??:??:?? 7 No response for class 06 message 01
WARN  | ??:??:?? 7 Unable to disable NMEA GGL.
DEBUG | ??:??:?? 7 GxGSA NOT available

Max-Plastix commented 1 year ago

I'd welcome any discussion around these proposed changes, as I'm new to Meshtastic but not new to the T-Beam or GPS configuration.

I'm happy to make a test & PR if this sounds like a move in the right direction.

RandMental commented 1 year ago

Agree this needs to be done

jurriaan commented 1 year ago

Agree! I had some issues with the Neo-9M (Beitian BK-357) which uses a 38400 baud rate. With no way of configuring the baud rate in meshtastic I had to use the U-center software to set the module to 9600 baud.

jp-bennett commented 1 year ago

Working on better GPS detection, including a baudrate scan, here: https://github.com/meshtastic/firmware/pull/2752

Would love to get more eyeballs/ testers on this.

kokroo commented 1 year ago

Working on better GPS detection, including a baudrate scan, here: #2752

Would love to get more eyeballs/ testers on this.

How exactly can we help? I have a unit with a 6M and another one with M8N.

jp-bennett commented 1 year ago

How exactly can we help?

Particularly if you have a device where the GPS is set at a different speed, try an image from https://github.com/meshtastic/firmware/actions/runs/6029724453#artifacts

davcx commented 1 year ago

Particularly if you have a device where the GPS is set at a different speed, try an image from https://github.com/meshtastic/firmware/actions/runs/6029724453#artifacts

I have 2 Tbeam 1.1 with neo6m stucked. I uploaded firmware-tbeam-2.2.4.3d2c419 and after almost 20 reboot gps emerged with first sentence [GPS] NMEA GPS time 2023-08-31 ...

Now downgraded to 2.2.3 both Tbeams with 6m gps still working (finger crossed).

esp-web-tools-logs (3).txt

jp-bennett commented 1 year ago

after almost 20 reboot gps emerged with first sentence

Super interesting results, thanks!

jp-bennett commented 1 year ago

@Max-Plastix I think we're partway there with #2752, which just merged. Feel free to test. I don't have a T-beam with the 6m to try it on. https://github.com/meshtastic/firmware/actions/runs/6066814336#artifacts

davcx commented 1 year ago

@jp-bennett by my POV the issue has been solved. Thx

meshtastic / firmware