I’ve just built a new long range quad, its my first of this size and power. I fitted the mag my usual 50mm from power wires and used the Setup > Optional Hardware > Compass/Motor Calibration, only to find I was getting off the scale interference. Several GPS booms later and it took 120mm to get the interference down to 32% at 75% throttle as the Ardupilot documentation says.
Seems a crazy long boom, is this normal ?
Large Li-Ion batteries in close proximity can be problematic as you found.
Thanks. The boom hasn’t added to the vibes, so other than looking silly, I can live with it.
Have you been able to run Magfit on this quad yet? That will give you better results than the MP tool.
I have the same frame, but… LiIon pack on bottom, power leads in the back, and GPS+Compass above camera.
Compass calibrated, and after magfit works just fine.
Also, second Mario8 with 6s1p LiIon on top+front connector and gps/compass mounted in original TPU piece - after magfit works without problems.
Oh, I didnt think about that, I used the Motor/Compass Interference just to check how much the motors are interfering, that interference is less than 30%. I had to extend the boom to get to 30%. I then used Magfit (amazing tool) afterwards.
Whats the best approach? Sounds like I mis-understood the tools and got the order wrong.
Now that you mention it, maybe the Magfit tool also does motor compensation if I select one of the two options below.
Thank you
The best approach is to Read the entire documentation, and make sure you know all the webtools that exist, and know how to operate them.
Or… you can use AMC and it will tell you what tools there are and where their documentation is and in which order to use them.
Happens to must users. That is why ArduPilot methodic configurator got created. To avoid that exact issue.
The casing material of cylindrical batteries is usually strongly magnetic.
All I tend to do is get a good enough calibration on the ground that I can arm and fly normally. Then once I can fly around and get some figure 8’s I’ll run the log through Magfit. If you have enough data and can use the “Battery 1 Current” option to get the COMPASS_MOT values. (ideally Offsets and Iron)
It sure looks normal to me.
In case anyone is interested, I posted my boom to Printables.
It sure looks like my boom is close to the size of yours.
I was lazy and designed the boom from within the slicer. You can see from this screenshot my boom is pretty close in size to yours. The small grid uses 10mm spacing so the larger grid is spaced at 50mm. The center of the GPS/compass is 130mm from the rear standoff of the drone.
I didn’t perform any experiments to determine a good length for the boom. I just based on my past experience using compasses on ground robots.
Since the part was designed in the slicer, it shouldn’t very hard to modify if someone wants to make one a different size.
What I’m not clear on is where the official documentation say interference should be under 30% at 50-75% throttle.
Is this before or after motor compensation?
@amilcarlucas I do kinda use the methodic configurator in that I follow the steps but I prefer to set my params manually. Just reviewed it now, it skips over the Compass/Motor Interference check, just says to calibrate the compass by moving it around and then use MagFit.
You do know that due to a bug in the parameter protocol used by ArduPilot, MP and others sometimes fail silently to set a parameter?
AMC validates each parameter set operation and is therefore safer.
it is your vehicle. You decide.
You need to calibrate the compass while having a GPS lock. Did you do that?
