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

Voltage spike in alternator when phases are shorted?

  1. May 12, 2012 #1
    inductive spiking as a viable electricity generation method?

    i have made a multi phase alternator to be used in a permenant magnet rotor setup , where the rectified DC output is 3V on average at the operating rate of rotation, as you can see in this video
    however, the AC wires leading to the rectifier had exposed solder joints, and while holding the positive and negative DC output, i received an electric shock to my fingers ...

    this is a picture of the alternator https://fbcdn-sphotos-a.akamaihd.net/hphotos-ak-ash3/523686_10150938016016197_502061196_11628304_387425437_n.jpg [Broken]
    as you can see, it is a 5 phase alternator (notice the 5 brown AC output wires) , with 3 coils in each phase for a total of 15 coils (most obvious by the T-shaped structure potruding from each coil) ... the coils have 400 turns each, and they have basically an iron core... i am estimating their inductance values to be in the millihenry range

    i also have a total of 4.7 ╬╝farad capacitance in parallel with the DC output

    after abit of testing, i have found that if i short out certain pairs of AC phases , the DC output would jump to ~12V before falling down back to slightly below 3V , and by releasing the short, the DC voltage jumps again before falling... and by continuously shorting and releasing the phases, the high voltage can be mantained...
    this suggests that the sudden change in current resulting from the short was enough to cause a flyback voltage spike, and fill the capacitors to 12V

    bear in mind that the only thing i have ever connected to the DC output was this multimeter that i have used in the video...
    i am thinking of using diodes to 'switch' the shorting between the phases, hopefully this will sustain the high voltage when a heavier load is applied across the DC ouput?
    Last edited by a moderator: May 6, 2017
  2. jcsd
  3. May 13, 2012 #2
    it seems that the time for editing my post has elapsed

    but if flyback voltage works the same way in both AC motors and alternators... only that in motors it is something which is unwanted , and hence you have the flyback diode to relieve the charge buildup when the terminal is disconnected... in an alternator, would any voltage generated be welcome, even if it was flyback voltage?

    like, could the coils act as part of their own flyback voltage converter if i leave out the capacitors from the DC output, and connected it to something which would switch the current on and off periodically? this would cause current fluctuations in the entire alternator, and trigger the inductive voltage spikes that way
  4. May 13, 2012 #3

    jim hardy

    User Avatar
    Science Advisor
    Gold Member

    i've no idea how that thing is wound.

    But since you have it in front of you, try this thought experiment:

    A shorted coil allows high current, and causes significant magnetomotive force that opposes your field (permanent magnet?) in the vicinity of that coil.
    So where does the flux go ?
    Maybe into adjacent coils.
    Trace it out.

    It appears that by shorting coils you are pushing flux around more quickly than the rotating permanent magnets do.
  5. May 14, 2012 #4
    yeah that is exactly what i thought of ... by the way, the coils are all wound in the same direction
    in fact, i am now planning to exploit the flyback spike by using transistors

    suppose that for each phase, the peak voltage between the center tap and the AC output is V volts
    since 360 degrees divided by 5 = 72 degrees, all the phases are offset by 72 degrees from each other
    so at the instant where one phase is at its peak voltage, 2 phases would be:
    (cos(72 degrees))V = 0.309V ,
    while the 2 other phases would be:
    (cos(144 degrees))V = -0.809V

    hence, there are two different inter-phase voltages, V - 0.309V = 0.691V and V - (-0.809V) = 1.809V

    from my testing, the voltage spike happens when i short a pair of high difference phases... it would make sense, since a high voltage would induce a high current, which in an inductive environment, would go on to induce an even greater voltage

    i think that if i were to bridge between 2 high-difference phases with a transistor, i could exploit the voltage spiking in a managable way... lets assign these phases, in an alphabetical circular nomenclature, as phase A and phase C...
    there should be a high-voltage tolerant transistor, whose collector and emitter are connected to the phases, and whose base is connected to a phase which has a low voltage difference to one of the phases connected to the transistor, and a high voltage difference to the other , which would mean either phase D or E , depending on the transistor being NPN or PNP , and the direction of rotation

    the idea being that the base of the transistor would be exposed to a high voltage difference with respect to the phase connected to the collector or emitter, so it would saturate the transistor and short the phases in time with the rotation of the magnetic field...the direction of rotation will matter, since the inductive effect means the current will lag behind the voltage ... hence, when the voltage of one phase is at maximum, the maximum current is probably in the phase which had the maximum voltage before...
    so it is possible for the transistor's base to be exposed to the phase containing the maximum voltage, at the same time that the collector or emitter is connected to the phase containing maximum current ...
    since whatever change in the current which happens during a short is a function of the current itself, having the short during maximum current would also cause maximum change in current...
    this would then cause the maximum amount of inductive voltage spike

    also, i have rearranged the AC output leads at the star-shaped rectifier so that the high-difference phases are adjacent to each other, and each rectifier would have the 1.809V voltage across its AC side, leaving the phase with 0.691V voltage connected indirectly and so acting as the voltage differential between the AC and DC sides... before, it was arranged randomly, i was assuming that the rectifiers were simply 4 diodes in a housing , but maybe they dont work so well when there is a voltage difference between their AC and DC sides, and i would get more voltage out of the rectifier if i minimize this voltage bias, and maximize the voltage across the AC
    Last edited: May 14, 2012
  6. May 15, 2012 #5
    the gist of it is that, in other alternator designs, the output voltage is low, but the resistance is also low, so you could exploit that by pulling alot of current into a switching power supply or voltage booster or somesuch

    but in my alternator, the output voltage is also low, but the conductivity is low too ... however, the inductance is quite high, so i think that it would be best to exploit that
  7. May 17, 2012 #6
    harnessing the inductive spike may allow more energy to be harnessed from the kinetic energy of the windmill, as the

    after some testing, i would have to say that if the inductive spike can be harnessed, it would have to be using methods other than MOSFETs to short out the phases, because they require a high gate voltage to conduct appreciable current, and this high voltage would have to be applied suddenly to maximize the current delta , and maximize the inductive effect
    the way to get the sudden, high voltage at the MOSFET gates is to use Zener diodes, they would be connected to a common voltage multiplier, with the voltage differences in the phases adding to it to bring it over the zener threshold voltage, switching on the MOSFETs one by one... the problem is, at the high voltages involved, in the diodes i could buy around here, their 10% tolerance means the variations in the Zener threshold voltages is as great as the AC voltage coming out of the phases
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook