Here's what I have - I made it as to be as classical of a electromagnet as possible:

a) Soft iron "U" shaped round bar-stock, ends machined to be flat & parallel. Spool of copper wire slid over one leg of this "U" shape iron.

b) Soft iron flat "keeper" bar to go across the ends of part "a" above

Momentarily connect the wire across a battery: the flat bar jumps to, and becomes firmly attached to the electromagnet.

Remove battery.

Flat bar stays firmly attached...

as long as I don't pull it off...

...for months without losing strength.


Yet, when I do remove it, there is no remaining "pull" that attracts the flat bar back to the U shape rod... so it's not becoming "permanently magnetized"... which is why soft iron was used.

Also, if a flashlight bulb is connected across the ends of the wire before removing the "keeper" bar, the bulb flashes. (in this case, the "keeper" is quickly smacked with a screwdriver handle in order to get high detachment speed.

This happens no matter how long it's been sitting on my shelf... or even hanging upside-down by the "keeper" bar... with the weight of the electromagnet hanging by the keeper bar.

  • 2
    \$\begingroup\$ YOu mean this... youtube.com/watch?v=r9Kg69cQteg \$\endgroup\$
    – Trevor_G
    Commented Dec 28, 2017 at 13:50
  • \$\begingroup\$ Did you use DC power or AC? \$\endgroup\$ Commented Dec 28, 2017 at 21:33
  • \$\begingroup\$ Yes... the PMH, with DC. Lots of responses, but I guess what I was looking for was what is it that keeps ion flowing - in order to "keep the keeper" strongly attached. I've always heard that in an electromagnet, the magnetic flux ceases to flow once power is removed. If the iron got magnetized, then it should retain that same pull even if I remove the flat bar/keeper... and it does not. I'm not aware of any classical explanation that defines perpetual magnetic flux in a magnetic closed loop... if there's no external excitation. \$\endgroup\$
    – ronbot
    Commented Dec 28, 2017 at 23:16
  • \$\begingroup\$ @ronbot Here's the proper terminology for it. \$\endgroup\$ Commented Jan 2, 2018 at 10:09
  • \$\begingroup\$ Yes, that's the static remanence- "the magnetization remaining in zero field after a large magnetic field is applied". The iron has a tiny amount of remanence, but I'm speaking of the effect while it remains in a closed magnetic circuit. The difference between the static remanence and when while the magnetic circuit remains closed is VERY significant. There is a strong magnetic flux "flow/storage" that continues after the electromagnet is turned off - in the exact way that electron flow continues in a superconductor after the energizing source is removed. It's that flux flow I'm interested in. \$\endgroup\$
    – ronbot
    Commented Jan 3, 2018 at 13:16

3 Answers 3


The soft iron indeed got magnetized. Your closed soft iron loop has a flux running in it as soon you magnetized it and connected with that is a mechanical force to minimize the magnetic resistance in the loop. That's why the bar is stuck to the U.

Removing the bar from the U inserts an air gap, which has a large magnetic resistance and will dissolve all the flux previously running in the core. That's why you cannot get the bar stuck again.


enter image description here

Figure 1. Magnetic field (green) of a typical electromagnet, with the iron core C forming a closed loop with two air gaps G in it.
B – magnetic field in the core
BF – "fringing fields". In the gaps G the magnetic field lines "bulge" out, so the field strength is less than in the core: BF < B
BL – leakage flux; magnetic field lines which don't follow complete magnetic circuit
L – average length of the magnetic circuit used in eq. 1 below. It is the sum of the length Lcore in the iron core pieces and the length Lgap in the air gaps G. Both the leakage flux and the fringing fields get larger as the gaps are increased, reducing the force exerted by the magnet.
Source: Wikipedia Electromagnet.

  • When you switch off a DC electromagnet the domains tend to remain aligned somewhat. This is greatly aided by the presence of the flat-bar "keeper" - so called because it "keeps" the magnetic circuit closed.
  • Forcing the keeper bar to quickly break contact with the magnet causes a sudden change in the magnetic field strength in the core. As Michael Faraday discovered, a changing magnetic field induces a current in a coil. This is what lights your lamp.

enter image description here

Figure 2. A simple generator schematic with automatic voltage regulator (AVR). Source: Generator Guide.

Remnant magnetism is very useful. Many generators rely on having some remnant magnetism left on the rotor to excite the stator on startup. As the rotor spins up the resultant magnetism generates a very weak current in the excitation winding. This feeds back through the rectifier, through the slip rings and onto the rotor, reinforcing the rotating magnetic field. This in turn increases the excitation current and the generator quickly "boots" itself up.

Meanwhile Vout starts to rise and when it gets to the AVR setting the voltage detection circuit starts to turn off Q to limit the excitation current to the value that maintains the desired output voltage.

I mention this because generators can lose their remnant magnetism if left idle for a long period. The fix is to inject a little DC onto the slip-rings while the generator is spinning. If you get the polarity right the output voltage will rise, regulation will be achieved and the rotor become magnetised again very quickly.

  • \$\begingroup\$ +1 Hey.. If you watched that video I linked to the second part of it is far more interesting. \$\endgroup\$
    – Trevor_G
    Commented Dec 28, 2017 at 16:17
  • \$\begingroup\$ Yes, I've had to do that to generators. Big ones. Important to get the polarity right. \$\endgroup\$ Commented Dec 28, 2017 at 21:34

As I visualize the scheme, I assume that the total airgap is constant.

I think the culprit is the inducted eddy current on the free bar by the motion. This ceases the flux (which Janka described) as it ceases along and so borrowed energy is transferred out.

edit: (expressing my thoughts a bit further)

If I understood correctly, so if the total airgap is constant despite the moving parts, then it has to be something other than mechanically affecting the flux. It has to be induction of eddy currents on the keeper bar.

An objection to that would be the idea of the magnetic field is parallel with the movement of bar. But, it won't be that much parallel, because of the vast difference of µ of mediums.

enter image description here

My claim is, when it is moved, inducted currents will effect on the skewed pattern of flux, oppositing it, making it flux pattern smoother, this is where the flux is ceasing.


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.