I'm putting IMU sensor to read earth-magnetic field. Does metal housing affect it?

It might be basic question, but I lack experience in this exact scenario. (My device needs a bit of a reliability, so I'm asking this)

Currently, metal case is aluminium; about 0.8~0.9mm thick.

  • 1
    \$\begingroup\$ Depends on the metal. Aluminum, not too much. Iron, lots. \$\endgroup\$
    – Hearth
    Commented Jul 6, 2021 at 5:55
  • \$\begingroup\$ Any ferrous fastenings in the alli case? \$\endgroup\$
    – Neil_UK
    Commented Jul 6, 2021 at 6:06
  • \$\begingroup\$ Connector shields of things like Sub-D, USB etc. are often magnetic steel, which will affect your accuracy. \$\endgroup\$
    – tobalt
    Commented Jul 6, 2021 at 12:55

3 Answers 3


Aluminium doesn't affect magnetic field as long as the magnetic field is constant. Others have warned you that you should prevent all current in the aluminium housing because current make electric magnet which disturbs your measurements.

If the magnetic field is not constant but changes there's some current in aluminium because changing magnetic field generates a voltage (=induction, the basic principle behind generators) and voltage causes current in metals. The total effect is that induced current cancels partially the changing magnetic field.

If the magnetic field stays a long time constant but then changes to a new value and stays at that value again a long time the sensor will see the change slowed, but finally the new value is seen right.

If the magnetic field changes slowly enough the effect can be neglible, but it exists and must be examined carefully. We should know the spectral content of the change of the magnetic field. The skin depth at the highest remarkable frequency would give some idea of the magnitude of the effect when compared to the thickness of the housing.

See this: https://en.wikipedia.org/wiki/Skin_effect

If your magnetic field hasn't any remarkable spectral components (due the change) say at 100 Hz and higher frequencies, the skin depth is about 10mm or more ==> Less than 1mm thick housing causes well below 1% cancellation.

If you plan to some control system where the magnetic field is controlled by feedback you have an extra RC lowpass filtering like effect in series with the sensor and that must be taken into the account in controller design.

Testing with and without the aluminium housing would be an excellent way to see the effect if the skin effect analysis suggests that the effect cannot be skipped. I guess you must build a test lab because the real conditions need the housing to protect the sensor.


Aluminium does not have magnetic properties. Therefore magnetic sensor will not be affected. Please make sure that no current is flowing through the housing, even tough aluminium is not magnetic itself it can have a magnetic field if current is flowing trough it. Copper is also ok, magnesium, zinc, even some types of austenitic stainless steel alloys like type 316L which have FCC face centered cubic crystal lattice.


In addition to the other answers, I would like to point out that the "reach" of magnetic disturbances is quite far. A magnetic sensor can be disturbed by metal parts or strong currents even meters away. I once got big measurement errors because I attached an IMU with steel screws to the underlying fiberglass instead of the included ones made of brass.

You will need to calibrate the sensor in the exact situation where you use it, together with whatever system you install it into (i.e. the whole car or boat around it). Read up about magnetic deviation. Basically, you need a table that gives, for every possible (real) direction, the compass measurement and vice versa.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.