Compass calibration - bad fit, using previous

[seandepagnier]

You need to calibrate on the boat because magnetic objects on the boat itself interfere with the field. The field in a house is also very distorted as most houses have steel in the walls, and small movements in position will greatly affect the field, so if you calibrate in a house be careful to rotate the sensors but not move them to different locations in the house.. You will have to recalibrate anyway on the boat to fix the bias.

[jamos.tan@gmail.com]

You need to calibrate on the boat because magnetic objects on the boat itself interfere with the field.   The field in a house is also very distorted as most houses have steel in the walls, and small movements in position will greatly affect the field, so if you calibrate in a house be careful to rotate the sensors but not move them to different locations in the house..    You will have to recalibrate anyway on the boat to fix the bias.

Thanks, our house is mostly concrete / brick, I was calibrating on a desk chair that turns around its own axis. But, better to calibrate on the boat I guess.

[Joel]

Hi Guys, 

Im also getting a 'bad fit' compass calibration result.... and "using 1d fit"

....Although Im also still indoors (for now)
This thread suggested re-trying the calibration on the boat. Although I have a wood framed house, and imagine it shouldnt interfere. 

Heres my calibration view....
I see a circle of dots.....
but I dont yet inderstand the dots, dot colors, or pattern.
[/img]


... and this is peculiar.... 
.... the headings response to a quick step (a step re-orientation of the IMU along the boat's 'flat' axis).....
The heading overshoots, then creeps slowly back down towards the original heading (resulting in a heading delta much less than the actual physical rotation applied).
It doesnt look good. 

(IMU is the IvenSense MPU-9250)

[/img]


thanks,

[seandepagnier]

it appears you have significant magnetic distortion. Any metal that a magnet attracts near the sensors?

[Joel]

it appears you have significant magnetic distortion.  Any metal that a magnet attracts near the sensors?

Hi Sean, 

Yes, the metal of the RPi itself.... I must have the IMU too close to the RPi (on top of plastic case).... I'll move it 6 inches away and re-try

Joel

[seandepagnier]

hmm maybe the pi4 is a problem?? the pi zero isn't a problem for this.

[rastam4n]

it appears you have significant magnetic distortion.  Any metal that a magnet attracts near the sensors?

Hi Sean, 

Yes, the metal of the RPi itself.... I must have the IMU too close to the RPi (on top of plastic case).... I'll move it 6 inches away and re-try

Joel

This is what I did, I mounted the Pi4 and cooling fan etc and used jumper wires to connect to my Moiteissier Hat. which is mounted about 6 inches away, there is still a small amount of magnetic interference but it calibrates fine, Just my offset is not aligned as well going south

[DanielGoedhuis]

I had much the same issues as Joel, when trying to calibrate behind my desk at home. The garden was a much better option, leading to reasonable calibration and readings that are fairly accurate. Will have to see how this works out on the boat, but that will take a while, due to travel restrictions due to Corona (boat is in Spain, I'm in The Netherlands)