I'm trying to think of the easiest/best way to steer a compass needle using a PCB. Specifically, placing a PCB of some kind under a compass (assume no shielding between the PCB and the compass) and electrically steering the needle in the direction I choose using a circuit on the board.

I feel like this is best achieved simply by conducting current on some sort of pattern-like trace on the board, but for some reason I'm having trouble visualizing the kind of pattern this would take, and how to calculate the current needed in a given pattern to produce a sufficient field to drive the needle.

Can anyone point me to the right idea with the kind of pattern/approach that would realize this sort of actuation and how I can calculate requisite current for the given solution/pattern?


  • \$\begingroup\$ Can you change PCB orientation? Or is the PCB fixed and un-moveable? \$\endgroup\$
    – glen_geek
    Commented Sep 23, 2019 at 0:55
  • \$\begingroup\$ @glen_geek PCB is fixed. So assume that you have a compass, and immediately under it is a PCB of the same diameter, rigidly mounted to it, and the goal is to control the needle in the compass. \$\endgroup\$ Commented Sep 23, 2019 at 1:12
  • 1
    \$\begingroup\$ Sin cos oriented current (90 deg) traces with bipolar control may be controlled as a vector angle to force the magnet like a weak motor coil into a static position \$\endgroup\$ Commented Sep 23, 2019 at 1:18
  • \$\begingroup\$ Think...motor stator on a PCB. The compass is the rotor. \$\endgroup\$
    – DKNguyen
    Commented Sep 23, 2019 at 1:23

1 Answer 1


Make four printed coils in a square, arrange for current to flow opposite directions in each of each pair of diagonally opposite coils, place the compass in the centre. Driving one pair will align the compass with that pair, reverse polarity will reverse the direction.

enter image description here Using PWM, or analogue proportional control of the currents you can point the needle any direction, but there will be some deflection due to environmental masgnetic fields (such as the Earth's magnetic field).

  • \$\begingroup\$ So this makes sense, but as I understand it, you don't have continuous control, yes? You only have four states total? \$\endgroup\$ Commented Sep 23, 2019 at 3:58
  • \$\begingroup\$ using about 100Hz Pulse Width Modulation ( a lot faster than the compass needle can respond), the "width" parameter (aka Duty Cycle) provides a continuously-variable magnetic field. \$\endgroup\$ Commented Sep 23, 2019 at 4:00
  • 1
    \$\begingroup\$ using pwm, or analogue proportional control of the currents you can point the neede any direction. \$\endgroup\$ Commented Sep 23, 2019 at 4:14
  • \$\begingroup\$ OK, this all makes sense and I'm on-board -- do you guys have an idea of how to figure the necessary currents necessary for steering the needle? \$\endgroup\$ Commented Sep 23, 2019 at 5:44
  • \$\begingroup\$ it's probably easier to experiment with sticky tape and magnet wire and measure how much current it takes. I show 5 turns stacked 2 deep, more turns would be better. \$\endgroup\$ Commented Sep 23, 2019 at 9:59

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.