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

Trying to wirelessly power an LED.

  1. Oct 21, 2009 #1
    I'm part of a Senior Design project involving Wireless Energy Transfer. We are trying to power a standard red LED as a demo, via resonant inductive coupling. Our setup is as follows:

    We have a 40.64 cm diameter transmitter coil (magnet wire) with 6 turns, hooked up to a signal generator. Our receiver coil has a 2 cm diameter, with 21 turns. We measured the respective resonant frequencies of these two coils, and used 522 pF of equivalent capacitance in parallel with our load (LED) to tune the receiver's resonant frequency to match the transmitter's.

    We placed the receiver coil inside the transmitter coil on a table with our signal generator at the resonant frequency. The result was a very very dimly lit LED, and it became brighter when we moved the receiver coil outwards, i.e closer to the transmitter coil. When we look at the output waveform on an oscilloscope, we get ~ 5 V peak-to-peak, and when we rotate the receiver coil so that it is perpendicular to the transmitter coil, the amplitude drops quite a bit, which seems to indicate that resonant inductive coupling is occurring.

    The thing is, we're not quite sure why the LED is barely lit. Adjusting the capacitance on our receiver coil does not seem to help, and changing the frequency only makes it dimmer. We are guessing that very very little current is going through our LED, which is why 5 V does not suffice to turn it on. We have tested the LED by hooking it up directly to the signal generator, and it lights up quite brightly.

    Any ideas why we are getting such poor efficiency? Even when the receiver coil is very close to the transmitter the LED barely lights up...

    Note: We got the LED to bright up very lightly when untwisted the twisted pair of alligator clips we used to connect the receiver coil to the LED. This would seem to indicate that the alligator clips were forming a loop and coupling with our other two loops, however, and that is not what we want.

    This is our setup. We've tried this with and without that resistor, we just did that because we thought it might increase the Q:


  2. jcsd
  3. Oct 21, 2009 #2
    Two possibilities.
    1) You have a voltage divider, because the small coil is intercepting only a small fraction of the magnetic flux from the large coil.
    2) You have a lot of leakage inductance from poor coupling.

    See Smythe, "Static and Dynamic Electricity", 3rd edition, page 335; coupling of circular loops.
    Bob S
  4. Oct 21, 2009 #3
    From what I understand, the main coil is connected directly to the signal generator via a twisted pair transmission line (crocodile clips).

    For max power transfer, you need to do some impedance matching - the impedance of the signal generator, to the transmission line, and then to the coil loop where the energy is to be dumped, otherwise some of the signal is reflected from the coil back to the generator and thus wasting power.

    Chances are that the pick up coil is highly untuned in your setup, the breadboard adds capacitance, and you have it going between the crocodile clips. Try putting the capacitor banks together with the coil, and add a small trimmer cap to tune with screw driver. Make sure that the tank is resonating at the desired frequency by checking it using a sweep generator or a network analyzer if there is one lying around somewhere.

    Also, air core coils don't have a high Q, increase the Q by adding a ferromagnetic core.
  5. Oct 21, 2009 #4
    Thanks for your responses. I will try your suggestions and see what I get.
  6. Oct 22, 2009 #5
    Try a SEC exiter.

    Last edited by a moderator: Sep 25, 2014
  7. Oct 24, 2009 #6
    In the photo, I see you have a single led and the impedance matching capacitors. The problem I see is that your transmitting and receiving coils are both AC devices and an led is a DC device. The led will present a load only to one half of the voltage swing, which may lead to saturation of your coil, I would try placing another led in anti-parallel with the other led.
  8. Oct 24, 2009 #7
    Watt, wouldn't the addition of core material also increase the coupling? Wireless energy transmission is beyond my kin so I can't comment on this problem.

    Would powdered iron core material also be an option?
  9. Oct 28, 2009 #8
    Hmmm.... more turns and a core were the first things which came to my mind.

    But.... if you're using a high frequency, that core can become a problem.
    As Phrak said, perhaps a powedered core will do the trick.
  10. Oct 28, 2009 #9
    two things:

    1)At higher frequencies, ferrite cores become lossy, and reduce the Q of the circuit. if you use a ferrite core, if possible choose a frequency under ~ 1 MHz.

    2) The impedance of the circuit at resonance is sqrt (L/C), so choose a lower inductance and higher capacitance. This will increase the resonant current in the coil. (Remember that the impedance of a simple antenna is near 72 ohms.) Using sqrt(L/C) over 300 ohms is too high.

    Bob S
    Last edited: Oct 28, 2009
  11. Oct 28, 2009 #10
    not much, the magnetic core would have to be closed like in a transformer.

    yes, but as stated, different magnetic cores respond differently to different frequency ranges. you have to choose a type that works at the frequency you want to use.
  12. Oct 28, 2009 #11
    Originally Posted by Phrak
    Watt, wouldn't the addition of core material also increase the coupling?
    Ferrite rod antennas are often used for AM reception 550-1600 kHz.
    Bob S
  13. Oct 28, 2009 #12
    Hi, I would just like to provide an update of our project.

    We haven't had a chance to look into all of your suggestions, but the impedance matching on our transmitter made a HUGE difference. The LED lit up at what looks like full brightness.

    We calculated our power transfer, and seems like we're supplying 92 mW to our transmitter and receiving 46 mW on our receiver at the closest range. The receiver can be a little more than a foot away from the transmitter before the LED turns off fully.

    We aren't positive about our power calculations, however... When we hook up an oscilloscope probe (with the ground clip too), the power over our load drops significantly, and our LED turns off. Is there a trick to measuring the voltage of an ungrounded circuit? The best we could do was hook up two probes without the ground clips across our load, and take the difference of the voltages we got.

    We are going to explore different antenna designs now; if you could maybe lend some insight regarding this oscilloscope probe, we'd be grateful. My best guess right now is that the probe's impedance is changing the resonant frequency of our receiver... which definitely makes it difficult to take accurate measurements.
  14. Oct 28, 2009 #13


    User Avatar
    Science Advisor

    A LED is a very non-linear load. That is its resistance varies a lot with the current that is flowing in it.
    For measurements, you could try to get a metal film (non inductive) 51 ohm resistor and take your readings with an oscilloscope across this. Not as spectacular as the LED but more useful.

    For low impedance drive and low impedance loads, you probably should be using series turned circuits, with the coil in series with the capacitor for the transmit and receive coils.

    You will get best coupling between the coils if they are of similar diameter. About a foot would be OK.

    What frequency are you using?
  15. Jan 18, 2010 #14
    Providing an update after one semester of work, and seeing if anyone can find a solution to our problem.

    Using equations for resonance and inductance, we have managed to light our LED up to a meter away using a big loop antenna transmitter and a spiral antenna receiver. We have also designed some external circuitry (i.e. amplifier, signal generator, etc.).

    Our goal now is to get some accurate measurements as far as efficiency. We measured the power supplied by our transmitter and theoretically verified it. The calculation verified our measurement.

    However, when we tried to measure the receiver power, we ended up getting more power out than we are putting in! We've tried measuring this various ways, using test loads (various resistors). Unfortunately, we don't know an easy way we can calculate this theoretically, because of factors such as mutual inductance.

    One guess we have is that the earth ground on our oscilloscope is messing up the measurement? We are fairly confident about our transmitter measurement, but we are pretty sure our receiver calculation is incorrect.

    Any ideas?
  16. Jan 18, 2010 #15
    Still recommend you look at the SEC exciter. As some claimed 100% power transfer using this method.

    Last edited by a moderator: Sep 25, 2014
  17. Jan 18, 2010 #16


    User Avatar
    Science Advisor

    You haven't said what frequency you are using, but there is no risk of you getting over 100 % efficiency. Maybe 5% if you are lucky.

    Ideally, I would like to see something like this:


    power coupling.PNG
  18. Jan 18, 2010 #17
    We are using 3.00 Mhz.

    The problem is our measurements are incorrect; we ARE getting over 100% efficiency with our measurements, which indicates something is wrong with our measurements. It is critical that we can record accurate measurements of the power transfer in our circuit. Here is the circuit we have set up:

    http://docs.google.com/Doc?docid=0AW92ycw2vmFnZDR0ajY1OV8xNWdkcHI5NWM5&hl=en [Broken]
    http://docs.google.com/Doc?docid=0AW92ycw2vmFnZDR0ajY1OV8xNWdkcHI5NWM5&hl=en" [Broken]

    R1 is the 50 ohm output impedance of our signal generator. We are using a 1k test load (R2), and using the Vrms value over that load to then calculate the current in our transmitter. We use that current and the voltage supplied to get the power supplied by our transmitter. We have verified this measurement theoretically.

    On our receiver, we are measuring the Vrms over our 1k load (indicated), squaring it, and dividing the resistance to get the power at the receiver.

    Doing these measurements, there is more power over the receiver than the transmitter is supplying, and we are trying to solve this problem.

    We kind of get the feeling that grounding isn't causing the measurement to be thrown off, that we are simply are missing something. Is there a radiative power equation we need to be using?
    Last edited by a moderator: May 4, 2017
  19. Jan 18, 2010 #18
    Is there any documentation this? As much documentation as possible would be nice since this is for a senior project, and we need to be using reliable sources.
    Last edited by a moderator: Sep 25, 2014
  20. Jan 18, 2010 #19


    User Avatar
    Science Advisor

    The power you are transmitting has nothing to do with R2. That is just in series with the transmitting coil.
    You would have to measure the current in R2 but the voltage across L1. This will be mostly reactive but a very small percentage of the volt-amps fed to L1 would be radiated.
    This is because the coil is too small to be an effective antenna at 3 MHz.
    I'm not even sure you could estimate the radiated power like this.

    To get more power radiated you could remove (or reduce) R2 as it is just limiting the power radiated.

    Also that 3 pF probably needs to be a variable C as stray capacitance would add up to more than that.
    The circuit may work better without it.
    L1 is very nearly self resonant so you may have to remove some turns to get a better tuning action.

    How did you measure L1? Do you know its dimensions and number of turns?
    I have a program somewhere that can estimate the inductance of such coils.
  21. Jan 18, 2010 #20
    I realize the power being transmitted has nothing to do with R2; R2 is ONLY USED to measure current. We're not actually measuring the power transmitted, we're measuring the power SUPPLIED by the signal generator. We get the voltage by measuring the voltage across L1, C2, and R2... which is the same voltage being supplied by the signal generator (or alternatively you could take a BNC cable and hook it up to an oscilloscope to see what the supply voltage is).

    Regardless, the power supplied by the signal generator should be greater than the power of the receiver load.

    I still have not gotten any helpful suggestions regarding why we are failing to measure the power on the receiver correctly.

    Isn't apparent power in = apparent power out? Or am I mistaken?

    We are simply comparing Vrms*Irms on the transmitter and Vrms*Irms on the receiver.

    We measured the value of L1 based upon its dimensions and number of turns. Using L1 and the resonance equation, we solved for the needed capacitance to resonate at 3 Mhz.

    Once again, at this point we don't care how well our circuit works, we just want to be able to measure it correctly.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook