3
\$\begingroup\$

I was playing around with two EM coils where they have a steady flow of DC current through them.

Theoretically, both of them should be magnetised to have a stable north and south pole, a typical dipole field.

When I make both their like poles face one another, they are able to repel but when they get close enough, they somehow "merged" and attracted one another.

If I put them at a close distance but not too close together(~0.8cm,) they will repel as expected.

Why is this happening? I am using an Arduino with some drivers to drive the EM coils.

I have read this but it seems that for this case,the two magnets of interest have different magnetic fields. The magnetic field and current through the coils should be approximately the same but it is still happening.

EDIT

Added photo of the two coils where the like poles are facing one another. enter image description here

\$\endgroup\$
10
  • \$\begingroup\$ Can this happen due to mutual inductance. Are the coils of the same number of turns or different number of turns? \$\endgroup\$ Sep 21 at 12:11
  • \$\begingroup\$ @Jonathan_the_seagull i would say they are slightly different. The thing is they are supposed to be identical coils which i bought online. They have slight resistance differences, around 0.1-0.2 ohms difference \$\endgroup\$
    – Iberico
    Sep 21 at 13:19
  • \$\begingroup\$ Steady DC driven with Ardiuno? Are you sure you are not PWMing them? Have you tried just using the power supply directly? Maybe with a series resistor to limit the current. \$\endgroup\$
    – Aaron
    Sep 21 at 13:51
  • \$\begingroup\$ Are there ferrous armatures inside the coils? If so, the core's reluctance may overcome the coil's repulsion. \$\endgroup\$ Sep 21 at 13:52
  • 1
    \$\begingroup\$ If the Ardiuno is using an H-Bridge to drive the coil (even from separate power supply), there is a good possibility that they are being PWMed. Most H-bridges made with all NFETs can not do 100% duty cycle. So it'll have some amount of field fluctuation. \$\endgroup\$
    – Aaron
    Sep 21 at 14:15
2
\$\begingroup\$

Look at systems where you can get force in both directions.

Dismantle a loudspeaker. There is one permanently activated magnet (in the 1930s this was usually a DC driven electromagnet) and a voice coil which can move in the gap in both directions.

Note also the design that concentrates the magnetic field in that gap where the voice coil acts on it directly, and never completely leaves the gap. This is quite an efficient magnetic circuit unlike the long air paths between your two solenoids.

There is no ferrous component in the voice coil. If there was, it would just glom onto the magnet regardless of whatever the current was.

Imagine one of your solenoids is DC driven and call it your magnet. It attracts the core in your other solenoid, regardless of the current direction in the magnet. Clunk! The magnetic field stores energy in the air gap between them (creating a force pulling them together). As they move together the energy is released and the magnetic reluctance decreases.

Current in the other winding creates flux which will either add or subtract to this attraction. The core reduces reluctance increasing the flux over an air core inductor, but you'll still need quite some current to cancel it out altogether and start seeing repulsion.

(For another style of voice coil actuator, dismantle a hard disk drive, and you'll see a rotary one that moves the heads in an arc across the drive surface)

\$\endgroup\$
2
  • \$\begingroup\$ Does PWM-ing also affect the reluctance to a great extent? Or do i not need to be that bothered with it \$\endgroup\$
    – Iberico
    Sep 22 at 2:23
  • \$\begingroup\$ The reluctance is a property of the magnetic circuit which controls the path of the flux, not the current which creates (or is created by) the flux. PWM controls the current, so, no. \$\endgroup\$ Sep 22 at 12:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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