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

Eddy current losses in core of a electromagnetic motor

  1. Apr 6, 2013 #1
    Hi here,

    I am trying to make an efficient ribbon microphone motor.
    My target is to get a highest possible output voltage so it can be plugged into a regular mic pre-amp.
    It should have possibly flat output on frequences 20Hz to 20KHz.
    I know that electric output depends on magnetic field B and velocity of a wire in it. (and velocity is proportional to frequency and amplitude.)
    So target is to achieve highest possible B in air gap where ribbon oscillates.

    I've made a model of motor using Gmsh and GetDP. It shows more than 1 Tesla in the gap, and as high as 2.5 T in core tip. I plan to use cobalt-iron core because of its high saturation capabilities. But I also read that it is not very good for high frequences like 20KHz because of Eddie currents.

    My question is, should I consider to suppress Eddie currents that should appear in the core as a reaction on movind a wire in B field?
    Is my understanding correct that Eddie currents should appear in this case?
    Will be Eddie currents significant enough at high frequences (5KHz-20KHz) that I should take precausitions to suppress them?

    I have this concern because all microphone/speaker motors I saw were made of just solid soft steel, not of electrical steel laminates like transformers are.

    Is my concern about Eddie curents valid?
    I am not sure how to calculate Eddie curents/losses in core for my case where a wire moves in mag field.
    What portion of it will produce a voltage in the wire and what portion will impact mag field B so it will create Eddie currents in core? If such proportion exists, what it depends on? Does it depend on frequency?
    Could you please advise?

  2. jcsd
  3. Apr 6, 2013 #2
    This is the first time I ever heard of a ribbon microphone motor.
    However after some Googling, it is my opinion that;
    As long as there is no current in the ribbon, there will be no eddie current in the steel.
    As long as there is no eddie current, solid steel will work as well as laminated steel.
  4. Apr 6, 2013 #3
    My understanding is that current in ribbon should appear as a result of moving it in magnetic field and electrons should go back and forth in ribbon/wire accordingly. Main part of these electron movement should go to preamp. But some part of this energy, I assume, should create eddy currents in core.

    I've googled this article http://www.tymphany.com/files/resources/papers/AES122nd-Impedance.pdf [Broken]
    that I not 100% understood, but it looks it confirms my assumption.

    though the article says that eddy currents in speaker core actually helpful, it helps to extend diapasone higher.
    Some speakers even have copper rings in core to help eddy currents.
    I completely puzzled now. Should I fight with these currents or should I help them?

    I found also another, more simplistic article http://www.diy-audio.narod.ru/litr/FaradayRingsVoiceCoilImpedance.pdf
    that I am reading now....
    Last edited by a moderator: May 6, 2017
  5. Apr 8, 2013 #4
    well, from articles I read I understood that B will not alter if flux is saturated.
    Then Eddy currents should not appear? Is it correct?
    Typically all ribbon mikes have flux saturated in core and especially in poles where core is narrowing to concentrate the flux to a maximum before it goes to air gap.
    Does it mean that movements of ribbon will not modulate flux behind saturated areas?


    Here on this picture B is saturated in inside tips of pole pieces, but usually saturation disappears in wider outer volume of pole pieces.
    Will that not saturated field that is located in outside volume of pole be modulated by ribbon movements while the field in insider narrow pole tips is saturated?
    Other words can modulations of B be transmitted via saturated field areas to farther not saturated areas?

    Another thing, according to my GetDP model, B decreases below saturation point in very internal volumes of pole tips, that touch air gap, on surface and a little bit deeper. That means that very tips of poles are not saturated and flux can be modulated in them. Volume of these areas is not very deep, and area is not wide, but still it may impact system on high frequencies, I assume.
    So, even despite of saturation in major part of pole, Eddy currents may appear on pole tips internal surfaces and a little bit deeper.

    Question is what should I do about this?
    a. suppress them with using thin isolated laminates that poles should be made of?
    b. place copper ring(s) around poles to create short-circuit secondary for transformer-like system - ribbon-pole-ring? (similarly to what they do in loudspeakers)
    c. both
  6. Apr 8, 2013 #5
    This is not my area of expertise, however these are my beliefs:
    The magnetic field produced by the permanent magnets is maybe 1000 to 1,000,000 times larger than the magnetic field produced by the ribbon.
    The magnetic field produced by the ribbon is so small that it may be ignored.
    Therefore the magnetic field is does not change when the ribbon is vibrated by an audible signal.

    To answer your question:
    a. Will not make any difference.
    b. Will not make any difference.

    It would appear that the secret to making the ribbon microphone motor is:
    Type of material used for the ribbon,
    Number and shape of bends in ribbon,
    Thickness of ribbon,
    How ribbon is supported,
    The design of the magnetic field is not important as long as it's uniform and strong.
  7. Apr 8, 2013 #6
    Carl, thanks for response.

    Yes, magnetic field produced by moving ribbon is extremely low, but the EMF produced by the ribbon is also very low.
    Yes, ribbon's magnetic field is incomparable by level with field from magnets. But magnet's field is constant. ribbon's field is changing. This modulation still can create Eddy currents. Yes, these currents are extremely low too, but they may be same magnitude with "good" current inside ribbon. Let say 25%. I do not know. And they also tend to grow with frequency.

    Other words, in mike acoustic energy is converted from mechanical to electrical. Some portion of electrical energy creates current in ribbon, it is "good" energy that is picked up and sent to preamp, but some portion creates Eddy currents in core poles, I assume it is as "bad", as losses.
    I am trying to avoid these losses.

    What is puzzling me is that loudspeaker's cores are not designed to suppress Eddy currents in pole and core. These are not laminated - solid steel - some even have conductive ring/sleeve/cup. So these intentionally maintain currents in that.
    My understanding was it would be better to suppress Eddy currents in core, so more electric energy will go to mechanical. But they made opposite way. I do not understand this for all 100%.
    But, by analogy, I am looking, what if to introduce this idea to mic, to add rings around poles there? Will it make sense?
  8. Apr 8, 2013 #7
    Sometimes it's easier to build something and try it. This may be one of these times.

    However from a design viewpoint, the current produced by the ribbon is insignificant. There will be no eddy current in the magnets or pole pieces.
    There may be some eddy currents in the ribbon caused by the magnetic field. This eddy current will tend to dampen the ribbon and prevent it from vibrating.
  9. Apr 8, 2013 #8
    May be. But how can I find from experiment are there Eddy currents or there are no. How could I measure them?
    What makes you be so sure? Could you prove your point?
    These are not Eddy currents. This is good EMF.
  10. Apr 9, 2013 #9
    Eddy currents are caused by changing magnetic fields.
    Insulated wire wound around the poles will detect this changing magnetic field (this voltage can be displayed on an oscilloscope). This changing magnetic field is so low that most likely the only thing detected will be 60 hertz from the power lines.

    The output voltage from the ribbon is so low and the load impedance is so high that there will be practically no current. Since there is no current there will not be any magnetic field. Since there is no magnetic field there will not be any eddy current in the magnets or pole pieces.

    The ribbon will have a voltage induced in it. Most or all of the voltage will appear on the ends of the ribbon. It is possible for a small portion of the voltage to be shorted internally by the ribbon. This internal voltage that is shorted is usually called eddy current. This shorted voltage will have very little or no effect and may usually be ignored.
  11. Apr 9, 2013 #10
    This method is possible, but it is indirect. Though nature of current in this wire will be same with nature of eddy currents, value will be different, and it will be a portion of value of eddy currents.
    Some portion of energy will create current in the wire and some still create eddy currents in core.

    Another way would be to measure rise of temperature in core.
    But both methods require extremely precise instrumentation that I do not have.
    For start I'd prefer theoretical, not empirical, analysis.
    Something like - current in ribbon Ir will cause field change delta B, this delta B will cause Eddy current Ie in core of width W at distance L for frequency F ...
    something like that.
    what load impedance do you refer to? My understanding is that impedance of ribbon will be just portion of Ohm. However it will be multiplied by transformer (1:37) to increase voltage that will go to preamp. If the load is preamp, then yes, it is going to have at least 5 times higher impedance. 5 times does not mean that current in ribbon will disappear. It should still exist. Am I right?
    Also, if there is no current in ribbon, there will be no current in primary of mic transformer, then how a signal will be passed via transformer? Transformer requires some current to create flux in its core. And this current will go through ribbon too. And will/should impact flux in mic core.
    Value of this impact I'd like to find.
    I agree this should be ignored. Because these currents will be situated perpendicularly to ribbon surface (coaxial with flux in air gap). So diameter of these eddy circles will be equal or less than thickness of the ribbon, which is ~2 micron. Such small circles may have some impact only on very high frequencies.
  12. Apr 10, 2013 #11
    "For start I'd prefer theoretical, not empirical, analysis.
    Something like - current in ribbon Ir will cause field change delta B, this delta B will cause Eddy current Ie in core of width W at distance L for frequency F ... "

    Sounds like a good approach.
    Assume that the sound moves the ribbon a certain distance at a certain frequency. You have calculated the flux density previously, so voltage = change of flux/change of time
    Next calculate or measure the impedance of the transformer with the operational amplifier connected.
    Now calculate the current through the ribbon.
    Now calculate the change in magnetic flux. Most of the opposition to the flux will be in the air gap between the ribbon and the pole pieces, so the pole pieces and the magnets can be ignored.

    This is getting so complicated that I wouldn't trust the calculations and would do one of the following;
    1.Get the proper equipment and measure the change in flux density using a coil having insulated wire.
    2.Build two electromagnetic motors, one with laminations and one with solid steel.
    3:Just ignore the eddy currents.
  13. Apr 11, 2013 #12
    Hmm. That would be very nice.

    #1. What would be equipment that can measure so weak fluctuations of flux in cores?
    I do not believe that can be available for me from budget perspective.
    #2. It is possible, but also requires couple of hundreds for machining solid cores and thousands to make custom stamp for stamping laminates.
    #3. This is practical! :) However I'd like to know _what_ I've ignored. :)

    What I am thinking is to analyze all three cases (solid, laminated, laminated+ring) with SPICE method to get FRCs.
    From them I hope to see the impact.
  14. Apr 11, 2013 #13


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I can't find anywhere in this thread that explains the use of the term "motor", here. A ribbon microphone is a transducer that produces an emf from air motion. A motor produces motion, using supplied electrical power. The two terms seems to be self contradictory. Is this just a language thing and are we really just talking about ribbon mike design?
  15. Apr 11, 2013 #14
    Yes, it is a language thing. it is about ribbon mike design. "ribbon mic motor" is wide used term, though, I agree, it may be not correct from terminology point of view.

    Any dynamic/ribbon mic can be used as a loudspeaker and vice versa, dynamic/ribbon loudspeaker can be used as a mic. I believe that physical principles of one are applicable to another. So if some solution decreases impedance for a speaker it can decrease impedance for mic. If "flux stabilizing Faraday ring" helps in loudspeaker it can help in mic. Am I right?
    Or is impact opposite?
  16. Apr 11, 2013 #15

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Consider relative currents in the ribbon for motor vs microphone.

    Current that doesn't flow can't cause eddy currents elsewhere.

    Hence the advice to present a ribbon mike with high impedance preamp.
    courtesy this free book preview:

    sound logical?
  17. Apr 11, 2013 #16
    preamp is going to be tube preamp with very high input impedance, dozens of K Ohms, according to http://www.aikenamps.com/CommonCathode.htm [Broken]

    Does it mean there will be practically no current in ribbon, thus no currents in core?
    Last edited by a moderator: May 6, 2017
  18. Apr 11, 2013 #17

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    That is what I believe.
    Disclaimer - I am not a microphone designer.

    if I treat your ribbon microphone as an electric machine ,

    without armature current there's no "armature reaction" to perturb the flux.

    Since per your sketch the ribbon moves in a uniform field, there's noplace for current to flow except out the ends (QVcrossB) and you're connecting the ends to a high impedance tube amp.. So,, Circular eddy currents in the ribbon look to me impossible.
    That's unlike the magnetic braking of a disk by a small magnet where current can circulate around the magnet pole.

    Corrections welcome.....

    Here's an appnote for your audio circuit collection:

    Last edited by a moderator: May 6, 2017
  19. Apr 11, 2013 #18


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    I find it hard to believe that a ribbon mic (without a matching transformer - which they were always fitted with afaik) could have as high an impedance as that graph suggests. It only consists of a single turn of wire, effectively. Where could 250Ohms come from, let alone nearly 2k at low frequency. The magnets they used were old fashioned low flux type. Were they really like that in the early days?
  20. Apr 11, 2013 #19

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    Hmmm looks like you're right on as usual - the old timers used transformers..

    So - there's a transformer in the picture too ? the plot thickens.....

    It's still an electrical machine driving almost an open circuit.
    Last edited: Apr 11, 2013
  21. Apr 11, 2013 #20
    It is still unclear to me how mic transformer would work if there is no current? How it passess and increases voltage?

    Thaks for the link. I am thinking which amp to make. I'll take a look. Currently I am thinking about RCA OP-6 because of its huge 90dB gain.
  22. Apr 11, 2013 #21
    250Ohms is after transformer. 2-3K is input impedance of preamp, I believe.
    Old magnets were huge, so they may produce flux equivalent to today's small neodims.
  23. Apr 11, 2013 #22
    Today's ribbons also use transformers. It is the only way to increase such low voltage without adding noise.

    I have another question. There is one transformer in mic ~ 1:35, and another one in preamp.
    Is it really required to have two? Can they be replaced by one with "summarized" ratio?
    In this case it should be less distortions, right?
  24. Apr 12, 2013 #23


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Unless you are intending to use a valve amplifier in the pre-amp, it shouldn't be necessary to have one there. There are many more options for pre-amp designs now and those would not usually involve a transformer. The reason for the transformer in the microphone is to make sure that the impedance it presents to the line is more suitable and reduces the effects of interference and hum. Nowadays, I would imagine that a suitable pre-amp could easily be placed inside the microphone body and so you'd need no transformers at all. But it depends whether you want optimum performance or authentic 'retro' equipment.
  25. Apr 12, 2013 #24

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    A good transformer requires only small magnetizing current.
    Load current is in addition to that.

    So the old advice about connecting the ribbon microphone to a high Z preamp is to reduce the current demanded of it.

    Doubtless those old-timers calculated the damping provided by electrical loading.... and designed partly 'by ear'.
    This could become a VERY interesting project.

    Interestingly your turns ratio of 35^2 is 1225 which X ~0.2 ohm Zribbbon is your 245 ohm Zout...

    In your pre-amp schematic in last post, that input transformer I think provides noise rejection in addition to stepup. Pins 2-3 are differential input for the long microphone cord..
    Here's all I could find on that transformer part number:

    With today's opamps I would think you can get by with with one less transformer - but which one to skip?

    As I said this is out of my field. My freshman physics is not too bad but you need a genuine audio guy.
    To me this is just a small dynamo.

    thanks - i'm learning new stuff.

    old jim
  26. Apr 12, 2013 #25


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Providing the optimum amplifier is not simply a matter of considering the volts. It's the Power that counts and how to match the source resistance (0.5Ω) best to the most appropriate amplifying device in the best configuration. The last thing you'd choose to do, I reckon, would be to use a bog standard Common Emitter, voltage amplifier, configuration. Common Base mode would probably be most suitable - or the equivalent thing with an OP amp. You want Volts out for Current Into a low input impedance. But I'm no expert in the details of low noise system design. It's never straightforward - I do know that! I'm pretty sure that you can do well enough without a transformer with the available devices.
    But we're still not sure what the OP actually wants to make - optimum or retro.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook