Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

How rapidly can an electromagnet get magnetised and de-magnetised?

  1. Aug 13, 2012 #1
    Hello all!

    How fast can an electromagnet get magnetised and de-magnetised, by DC-pulsing it and without having any hysteresis & retentivity effects in the coil itself? Obviously its not going to be a ferrite core coil(because of retentivity,hysteresis). How fast can a low power air-core electromagnet be switched on and off in such a way that at the very instant the electromagnet is turned off, the magnetic field is also completely gone at the same instant without any delay?. Can we get in the 10-100GHz range? or is it basically limitless from the
    "coil´s physical point of view"?

  2. jcsd
  3. Aug 13, 2012 #2

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    are properties of ferromagnetic material and you later specified air core.

    try google on inductors, both air and iron core.

    You'll find that in iron, eddy currents place a very modest frequency limit and that's why there's powdered ferrite cores for high frequency

    "retardation of magnetization due to eddy currents" should produce rich search results for the case of soft iron cores.

    A fellow named Bozorth published a book "Ferromagnetism" which i think still stands as the ultimate reference.

    Happy hunting !
  4. Aug 13, 2012 #3
    Jim, yes I mentioned hysteresis & retentivity that are related to ferromagnetic cores trying to make my point clearer. "in other words, I want to avoid any similar effects in air core copper electromagnet maybe I should I ask if diamagnetic copper have super tiny hysteresis & retentivity effects?"I am therefore intending for using only air-core electromagnet in my project. EXTREMELY high Frequencies are greatly limited by ferromagnetic cores or any magnetic cores. Maybe my question should be formulated this way:

    If I DC pulse a small copper air-core electromagnet with 10ghz, will the coil retain some magnetic field (for a very short time) after every "off"? or will the magnetic field PRECISELY follow the extremely high Ghz DC pulses??

    My entire wondering is really because of the very high frequencies involved?

  5. Aug 13, 2012 #4

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    now you're into microwave realm, completely beyond my experience.

    I'll hope somebody experienced in high fvrequency RF can help.

    Best i can offer is that microwave oven magnetrons have air core inductors in series with their heater leads - so they must still work at ~2 ghz.

    I do point out that flux and current are related by proportional constants not time dependent functions
    so , unless your excitation current induces secondary currents nearby, i would look for flux and current to be similar functions of time.
    Voltage relation to flux of course is a different story

    old jim
  6. Aug 13, 2012 #5
    "Best i can offer is that microwave oven magnetrons have air core inductors in series with their heater leads - so they must still work at ~2 ghz."

    Thanx for the tip Jim..., Im going to be waiting for some experienced in the microwave area. BTW, are these air core inductors in the oven fed DC pulsed or AC?

  7. Aug 13, 2012 #6


    User Avatar
    Science Advisor
    Homework Helper

    At 10 GHz, the wavelength of the EM radiation is only 30mm. If that length is comparable with the size of the device you are thinking about (i.e. the size of the complete circuit, not just the coil), you would have to take the finite speed of light into account to answer the question. A simple idea like "a DC pulse" doesn't mean much at those frequencies.
  8. Aug 13, 2012 #7


    User Avatar

    Staff: Mentor

    Welcome to the PF, Peter.

    Can you tell us about your project? What are you wanting to do? There are lots of considerations at those frequencies, and hysteresis in an air core is not one of them AFAIK.
  9. Aug 13, 2012 #8

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    they are in series with a low voltage winding (5volts i think) of the 60 hz power transformer
    and their purpose is to pass the 60 hz into the magnetron's heater but block the 2ghz keeping it from travelling back up the wires and past the shielding thence into the kitchen.

    So they pass 60 hz AC and block 2ghz AC, both continuous for all practical purposes.

  10. Aug 13, 2012 #9
    AlephZero, berkeman

    Hey and thank you for the info and welcoming.
    As far as I am concerned there will be no other part in the circuit except the air core electro magnet(well it is going to be hooked to a some kind of signal/function generator), so thats all there is going to be, a coil hooked to a function generator, unless I find out I have to install some other electronic parts into circuit to help me reach my goal which is extremely fast & clean DC square-wave pulsing without having any delay between the current and flux field, the coils magnetic field is precisely off at an off pulse, nothing else matters..(I believe so far?)
  11. Aug 13, 2012 #10


    User Avatar

    Staff: Mentor

    So you want to create an oscillating RF magnetic field in the 100GHz range inside a small air-core coil? What signal level are you going to drive through how many turns for what size volume?
  12. Aug 13, 2012 #11
    Im sorry but Im not quite exactly sure what you mean by "RF" but yes I want to create a rapid DC pulsing in an air- core coil in the 100ghz i.e. an oscillating magnetic field that goes: on off, on off. 10ghz may be enough. But it must be 10ghz minimum

    Since this is an experiment, none of the coil specs really matters as long as I have some weak magnetic field in the coil. All I need is to get Clean DC high pulses. The air core coil itself could be as small or as big as required all to help reaching my main goal, also voltage & current could be as small or as big all to help reaching my main goal which is: DC pulsed air core in the microwave band.
    So, I thought that things like the power, volts, amps, turns, inductance etc would be decided accordingly to and towards what is ultimately is needed to achieve 100ghz pulses.
    Last edited: Aug 13, 2012
  13. Aug 13, 2012 #12


    User Avatar

    Staff: Mentor

    I think your concept of the signal that you will drive is a bit off. Don't use the term "DC" for something that happens at 10GHz. If you are going to drive a signal at 10GHz, it will be a sine wave, or perhaps an isolated pulse with subsequent ringing.

    And by RF I mean radio frequency. A 10-100GHz signal is radio frequency signal.

    If you want to make a gated pulse of some number of 10GHz sinewaves, then you can do that.

    If you want to make a 10GHz sine wave with a DC offset of half the amplitude of the sine wave, then you can do that. That is the closest you will get to an "ON/OFF" 10GHz magnetic field.
  14. Aug 13, 2012 #13
    You really don't get into GHz with any solid core, even the powder stuff goes to a few hundred MHz. For more info, go to Mini-Circuit site and look at the RF transformers and see what is the upper limit. For air core, you can go a little higher, but as Alphazero said, when you approach 1GHz, even a one turn coil is quite high inductance and the wave length come into play. In air, speed of light is 3EE8m/s, so for 1GHz, λ=10cm. At 10GHz, λ=1cm. As you can see, you really cannot get many turns to be longer than λ. Any length of wire longer than 1/20λ need to be treated as transmission line and has to use phasor notation for calculation. It is no longer a wire. So your assertion of a "coil" is out the window quite fast when frequency goes up.

    In fact we often use a section of wire( trace on pcb) shorter than 1/8λ as inductor with value calculated by the length. There nothing coil about it.
  15. Aug 13, 2012 #14
    I wrote dc to not confuse with ac, big difference. Though everything will tend to have a round triangular waveform at such high frequencies, my goal is to have square waves.

    Anyway I found this:

    http://apl.aip.org/resource/1/applab/v58/i12/p1253_s1?isAuthorized=no [Broken]

    This GaAs direction seems to have pretty clean square wave pulses at fairly high frequencies. Its seems to be laser pulses though..
    Last edited by a moderator: May 6, 2017
  16. Aug 13, 2012 #15


    User Avatar

    Staff: Mentor

    What would you like to do with the 100GHz square waves if you had them?
    At 100GHz you are in the realm of microwaves, but getting close to light. Infra-red light is from 300Ghz up.

    Are you wanting to construct a maser cannon? :wink:
  17. Aug 13, 2012 #16
    Naah, Im trying to reconstruct Teslas Death ray... :wink:

    Yes, Im aware that Im closing to light, that is sort like the idea. Honestly I have some kind of a little "theory" in my head(for some years now) that I really want to confirm by an experiment. I dont see really why it shouldn't work, I have went/revised through it countless times. That is, if I can oscillate a magnetic field at a frequency (which happens to be above 1ghz and simply can never work at any lower frequency), with actually a somewhat clean square wave pulse-train, the theory should really work. I can in fact say that I am confident that it will work if I reach the correct high frequency pulse-train/oscillating magnetic field in the electromagnet.

    I have read some threads considering signal generation and if we can output a nice high speed square wave pulse train(1ghz+) and it seems that despite being difficult its manageable (?) I will probably though face a problem to actually output such a high speed clean square wave pulse train, but according to my (unfortunately and recently deceased) friend, he never mentioned that to be a problem.
    So therefore I believe as he did that the main problem seems to be into "constructing such a physical coil" that can take extremely high pulses.
    Last edited by a moderator: Aug 14, 2012
  18. Aug 14, 2012 #17

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    i'd think the trouble would be in generating picosecond risetimes.

    that's a lot of di/dt.... extreme voltage rquired ?
  19. Aug 14, 2012 #18


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    1GHz = 30 cm; 10 GHz = 3 cm wavelengths ;)

  20. Sep 17, 2012 #19
    Hi again, I didnt want to start a new thread so here I continue. I decided to let you know what Im working/researching on. If you find that this concept is totally impossible it is fine, but please give it a thought and if it cant work, please specify the the reason why.

    Anyway the whole idea could be summerized in that all I want to do is to create a short moment (nanosecond ? maybe less..?) where a delay exist between a magnetic field and a coils activation. in other words: a very short moment where the electromagnet is off while the magnetic field is still there at the same time. if you understood me correctly you will realise that there is a great potential for creating a net force, i.e. propulsion.

    Take a look at Sun to Earth light propagation, it takes 8 minutes. magnetic fields also propagate at speed of light.

    Please ask if you didnt get me, Im not grasping this completely really either as its get real hard imagining the whole procedure

  21. Sep 17, 2012 #20


    User Avatar

    Staff: Mentor

    What makes you think you will get any measurable propulsion out of this?
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook