Strange gps lost warning

[msinger]

I would have landed as soon as possible. When your satellite count is high, the only other reason it'll auto switch to ATTI is if the compass data cannot be used for some reason. You don't want to fly when the compass is not working as expected.

 

[SteUk]

I would have landed as soon as possible. When your satellite count is high, the only other reason it'll auto switch to ATTI is if the compass data cannot be used for some reason. You don't want to fly when the compass is not working as expected.

Thanks for all your help bud, thats what i did must admit i did panic a little.

 

[TK P3A]

This has happened to me several times. I thought it might be due to solar flare crap but since this started I have had the Kp index on my phone and tried to fly with 4-5+ Kp readings and it made no difference. P3A with Tab S2 8" and android

 

[BudWalker]

Sounds like the same thing that has been reported multiple times since mid January around when the new app was released. Here are a couple of the threads that seem to be similar to what you experienced. Says no GPS, but still acts like it has GPS. As far as I know, no root cause has been determined.

15 GPS satellites, but GPS repeatedly drops

Non gps mode moments + log

"Safe to fly - Non GPS"

Safe to fly NO GPS problem

I'm going to respond here because you've gathered together all the relevant threads. When the message "Safe to fly - Non GPS" appears it actually means "There is an error but it's a Non GPS error" I.e. the error is not GPS related. The indicator that there is an error is that the color turns non-green. The text is just saying that it's not a GPS error.

There are several errors that can cause this. One comes from a field in the .DAT that has been labelled "compassError" in DatCon. It is being renamed "nonGPSError". If this field has the value YAW_ERROR_LARGE(6) or SPEED_ERROR_LARGE(5) then "Safe to fly - Non GPS" is presented in DJI Go (and probably the Litchi app). These are not GPS errors.

There are several non serious "errors" that can occur. Rather than report them in all their complexity DJI has chosen to simplify the reporting. Would it really work to have an error message like "Flight Controller Model predicted value doesn't agree with measured yaw from the magnetometers"?

Thanks to @AwfulPilot for helping me figure out how to extract from the DJI SDK so that it could be found in the .DAT A new version of DatCon will be released in the next day or two that incorporates these changes.

 

[alokbhargava]

GPS loss messages may also be related to app version 2.5.0 and above. I had never seen those messages before in my area but I see a different pattern now after SW upgrade. This is only an inference as I don't have the data to prove it.