Mag field mistake in auto analysis since Arducopter 3.4.x?

I and a few other people I have made 3.4.X flights with different ships observed systematically in the auto analysis of their logs an indication of this type (this is one example):
Max mag field length (702.43) > recommended (550.00)

We are all experienced Arducopter builders and we know how to setup correctly an external compass module far from any magnetic disturbances.
We have tried to setup our compass modules as far as 20 cm on non metallic stands, still getting this high max mag field warning.
So the questions coming to mind:
-why is this happening since version 3.4.3 ? (same setup does not produce same results on older firmware). Is there thus a mistake on the mag field logging in 3.4.X or another way ta calculate it in the latest mission planner ?
-What does this mean ?

thx for help

1 Like

You have raised an interesting and, at the same time, troubling issue with the current code.

We have FC’s now with multiple onboard Mag’s, GPS’s with Mag’s, and individual Mag’s we can mount where we wish.
The Mag calibration screen caters for up to 3, but there is always a discrepancy between Mag’s that causes errors, and generally all but one has to be disabled to fly.
In essence, this concept of redundant Mag’s is a false one that does not work.

We fly a lot in power stations where redundancy is a MAJOR factor, and the Mag, as a critical sensor for any auto flight, is the weak link.
The new machines I am building with PH2.1 cannot run the latest firmware if I cannot run multiple mag’s.

I also would like this discussion to continue.

Anyone?

1 Like

Does the log actually contain values that high? Should be simple enough to check, and if it does then this post would be horrendously misnamed :slight_smile:

Yeah, I have the same problem, I am confusing too.

Here the same, with AC 3.2 no Compass problems, in addiction since AC 3.3 we see persistent problems from Compass on Mission Planner EKF Compass Graph, often peaks raise the highest RED level on every build, and its very hard to evaluate if a real problem is there and when or what make it happened .I really liked to see what QGroundControl implement recently as post calibration MAG graphs, that starts to gives some feedback to users in order to understand a little better IF, and when and where problems happen.
Right now I’m involved in micro quadcopters and not 20cm but 10 cm distance is not an option for placing the Compass away from interferences :frowning: so I think a robust MAG ecosystem understanding post AC 3.3 is really WELCOME :slight_smile:

Noticed same thing, even when compass/motor calibration didn’t show anything dangerous.

1 Like

I’m having issues with the mag field aswell, as I’m trying to use my quad indoor with pozyx beacon system. The Auto-Analyses tells me the mag field is 680 and 550 is recommended. I’m using the internal compass from the PixHawk 2.4.8…
Does anyone found a solution on how to reduce the mag field to the recommended value?