Return to Home Failure

[KCVAPOR]

Please retrieve and provide the .txt for this flight as requested by @msinger . I suggest that you also retrieve and provide the .DAT for this flight which is on the Mavic itself. (There may be a .dat on the tablet but that is not the right one). Since the app crashed the only way to know what happened during that interval is through the .DAT. Look here for info about retrieving the .DAT. It'll be large so you'll need to Dropbox or GoogleDrive it and provide a link here.

From your description it seems unlikely that there was a compass issue. But, if there was it's likely the .DAT will provide the necessary data.

With regard to the comments made about compass calibrations I've done some experiments and research over the last 18 months or so. My observations:
1) It's very difficult to get a bad compass calibration. I did an experiment with my P3 where I attempted to calibrate next to a large metallic object (my pickup). If the P3 was too close the Go App would complain about interference and reject the calibration. After several attempts I knew how close I could get and still have the calibration accepted. That "bad" calibration wasn't bad enough to cause a flight problem.

The other method I used was calibrate with an allen wrench attached to an adjacent leg (the compass is in one of the legs). After the calibration the wrench was removed which yielded a P3 with a bad calibration that probably would caused an incident had I flown it.

I suspect that under just the right circumstances it may be possible to unknowingly obtain a bad calibration. So why risk it.

2) With one exception I have never seen an incident caused by a bad calibration supported with data or some other compelling evidence. I've been looking at compass related incidents for well over a yer and that one exception happened last week. Actually, that incident was likely the result of a hardware change or partial failure that got "fixed" with a calibration. Take a look at
Fly Away
if interested.

3) It wan't stated explicitly here but some may believe that a calibration is necessary at a new location in order that the geomagnetic declination can be determined. The calibration procedure for the Mavic, P3, etc does not and can not determine the geomagnetic declination. The only thing that can be determined about the local geomagnetic filed is it's strength. I suspect the P3 was more sensitive to the geomagnetic strength than is the Mavic.

BTW, I've never calibrated my Mavic's compass or IMU.

If anybody is still reading this and is interested you might be interested in the MagData player in CsvView.
http://www.datfile.net/Doc/MagDataPlayer.pdf
Basically, it presents magnetometer data in a way that would expose a bad compass calibration. It's eye candy for geeks, but it's good eye candy.

Thank you. You are much appreciated and well informed on these topics.

 

[KCVAPOR]

I ended up going with ipad pro 10.5 and a mavic mount and anker cable (plus 4 32gb cards to swap with each battery I have if I don't have time to back up to the ipad storage). I also rolled the firmware back because if it ain't broke don't fix it.

If anyone is looking at ipads, holy crap they are officially awesome in ipad pro 10.5/12.9" 2017 version. I couldn't even buy a used mini 4 after picking up a pro and seeing the difference in power, cameras, and pencil features (i do a lot of sketching).

The mini 4 was definitely nicer for people on a budget, and has higher pixel density etc but the processor in that thing is years dated at this point. Hopefully djgo likes the new CPU class processing and iOS better than my s7. Its also insanely bright when needed which is nice over the mini which is bright, but not as many nit. It also compensates somewhat for bright situations with automatic color changes for different conditions... better on battery than more nit to accomplish more visible screen.

Its a neat way to have 512gb of storage for backing up footage as I go too, nice perk over my phone which has a slow mSD card for memory in it.

 

[KCVAPOR]

Hopefully none of this RTH (in app failure) happens again, nor a loss from such a short range that my micro quads can handle lol.

I also realized I did have a giant piece of metal right next to me. About 1000lbs of iron/magnesium alloy... and another 2000lbs sheet steel in corrugated, high carbon, and other forms...

My 4400lb VW Eurovan MV was right next to me when this happened. It is a large strange van with fiberglass, iron, wiring through out containing inverters and multiple batteries solar panels etc.

It is also completely lined with reflectix insulation... every square inch of it...

So perhaps even though I had a short range, direct, properly oriented transmitter, the VW Eurovan MV w/a million things going on including wifi/4g booster and a laptop crunching some numbers in heavy color grading, burning about 240w/h at the time could have done this.

I fly right next to the van (sitting on the roof usually) all the time but this is my first delve into HD signals this powerful so who knows maybe it matters.

I still find it strange that landing assistance wasn't working upon return as well as the complete reboot of phone that was necessary.