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

High Frequency Oscillator Circuit Help?

  1. Dec 2, 2011 #1
    Hey guys,

    I'm designing a wireless charging system, and i've managed to take some measurements between two coils for the voltage transfer, but the signal generator i'm using doesn't seem to output any measurable current.

    What could I do to it, or what could I design from scratch that would enable me to select a frequency around 4MHz which could drive about 500mA at 20Vp-p?

    I've seen suggested added some sort of amplifier to the signal generator output so I can still experiment with frequency and power transfer without limiting myself to a set frequency at this early stage.

    (It might be possible that when I tune the coils I could make it transfer at around 200kHz meaning I could use a simple 555 timer? maybe? I'm not really sure :-S)

    Thanks in advance guys!!

    Anthony
     
  2. jcsd
  3. Dec 2, 2011 #2
    What are the diameters of your coils and how many turns do they have?
    What is the voltage and current capabilities of your signal generator?

    I think it would be easier to design one at 200 kHz than 4 MHz.

    You can probably increase the power transfer by making the coils series resonant at your frequency, preferably 200 kHz. I think it would be better to use a sinewave oscillator instead of a 555 timer.

    You may be further able to increase the power transfer by using a ferrite core to your coils and shield the outside of the coils with a magnetic material
     
  4. Dec 2, 2011 #3
    One coil is 5cm by 2.5cm with 32 turns of 0.711m diameter copper.
    The other is 2cm by 1cm with 16 turns.

    The voltage output is 20Vp-p with it failing to drive even 10mA out of it (although it doesn't specify anywhere it's actual limitation - not that I can see anyway).

    Thanks for the tips on the power transfer, it's definitely something I will have to investigate once I've achieved power transfer, rather than just a voltage.
     
  5. Dec 2, 2011 #4

    davenn

    User Avatar
    Science Advisor
    Gold Member

    how about drawing up your circuit. Annotate it with components used, voltages and currents at key points etc
    then we can all visualise what you are trying to achieve :)
    its going to be much easier that way, than playing 20 questions ;)

    Dave
     
  6. Dec 3, 2011 #5
    I've attached an image of the circuit (photo.jpg).
    When I place an Ammeter in series, even when putting a resistor (tried values from 10Ω to 1kΩ) I don't get any reading on the meter.

    I suppose what I think I'd like to come up with eventually is seen in the second attachment (photo2.jpg). Understandably I'll need to rectify the signal, but I've managed to build that circuit already (although I've been told I'll need shotkey diodes instead of 1N4001 to allow for the high frequency?)

    Thanks guys.
     

    Attached Files:

  7. Dec 3, 2011 #6

    gneill

    User Avatar

    Staff: Mentor

    You're going to want to make sure that you're not building a transmitter that's going to interfere with reserved frequency bands! Otherwise you may eventually get a knock on your door from the FCC (or whatever the local regulatory commission is where you live), or worse, an irate ham radio operator! 4MHz is at the top of the 80 meter band, bordering with a band used by maritime mobile services (at least that's the case in North America).

    A chart for the US spectrum allocation can be found here.

    Harmonics that result from rectification (rectification is a nonlinear process) may be nasty. Shielding and filtering will probably be necessary.
     
  8. Dec 3, 2011 #7

    vk6kro

    User Avatar
    Science Advisor

    Since you are using a signal generator as your signal source, you need to be driving a series tuned circuit, not a parallel one. Signal generators have low output impedances and this damps a parallel tuned circuit and stops resonant effects.
    Series tuned circuits work best with a low impedance drive.

    You could try 220 pF or anything like that that you have on hand. You could use a variable capacitor of about 400 pF maximum for this.

    Then tune with the signal generator for maximum voltage across the left coil.

    Make the second coil identical to the first.

    Now, since you want to drive low resistance loads, you should also use a series tuned circuit in the secondary.
    This time, though, the frequency is already fixed, so you need to adjust the capacitor in the secondary side until you ge maximum output. You could use a variable capacitor of about 400 pF maximum for this.

    Yes, you will need Schottky diodes for this. 1N4007s etc have a very poor reverse recovery time and don't work at HF as rectifiers. Observing with an oscilloscope is fine.
    BAT85 Schottky diodes work well.

    4 MHz crystal oscillator modules are readily available, so you could use one of those as a signal source although the frequency is fixed, of course.
     
  9. Dec 3, 2011 #8
    Thanks for the advice guys, I'm in Europe so I'll check the rules here.

    If I tune the circuit to around 500kHz I might not have a problem anyway.

    Is there any way of boosting the output power from the signal generator for the purpose of testing at least?
    I've been told power Mosfets might work? But I didn't really listen during the first year at uni so I'm not too sure.
     
  10. Dec 4, 2011 #9
    I understand that in the design phase you're using a signal generator as your source but in your final design what will be supplying the signal? Will it have the same output characteristics as the signal generator? If you will derive your power from a DC supply, it may make sense to incorporate the primary side of the transformer into a Hartley or Colpitts oscillator. That may solve the signal generator problem.
     
  11. Dec 13, 2011 #10
    Usually, there is alot of leakage inductance between the coils (depending on thier distance / geometery).
    Because of this, the current throught the primary must be much higher than that in the secondary. It's a terribly inefficeint process unless the primary winding is part of a tank circuit. Thus, most of the primary current is recirculating between the coil and capacitor.
    A fairly pain free way to start is to build a colpitts oscillator with a 2n2222A and include a small diode, such as a 1n914, in series with the collector. By choosing the correct set of components, it's fairly easy to get 2 amps p-p at 500kHz in a 25uH coil. You want the receive and transmit coils to be of at least 1-2 inches in diameter and aligned for good coupling.
    PS it really helps to have low loss capacitors in the oscillator. Some ceramics are okay, but silver mica are the best.
     
  12. Dec 13, 2011 #11

    vk6kro

    User Avatar
    Science Advisor

    Before you try getting more power, it would be good to get the circuits resonant. That is what this circuit is all about.

    It will still be inefficient because small coils make poor radiators of RF energy, but it will be a lot more inefficient if you can't get the tuned circuits optimized.

    You would probably need a circuit like this:

    http://dl.dropbox.com/u/4222062/RF%20transmission.PNG [Broken]

    You would make the coils as big as you can. Make them 6 to 8 inches in diameter, identical to each other and aligned parallel to each other.
    You should get sharp resonances indicated by sudden large peaks in output to the meter.

    The fixed capacitors in the voltage doubler should be about 100 pF at 4 MHz but this can be adjusted if you get output. These capacitors (particularly the left one) should be as small as you can use and still get good output.
     
    Last edited by a moderator: May 5, 2017
  13. Dec 14, 2011 #12

    NascentOxygen

    User Avatar

    Staff: Mentor

    Are these on an air core? Is so, then I think you have no hope of transferring much power. I think you'll have to use hundreds of MHz to get much power across.

    For lower frequencies, maybe it would be possible to wind the receiving coil on a ferrite core, then push this inside the transmitting coil when you want to transfer power, so one winding would totally surround the other, both on a common core. Would this still come within your concept of "wireless" transfer?

    Making the transmitting coil part of a resonant circuit would cause magnified currents to circulate in it, so the field around it would be stronger.

    Good luck!
     
  14. Dec 14, 2011 #13

    vk6kro

    User Avatar
    Science Advisor

    If you couple the two together, even with a common magnetic core, then it is not a wireless transmission any more.

    You could use a transmission line between the two and get excellent efficiency, but that is not the point of the exercise.

    With large diameter coils and short distances, the efficiency might approach 20 % at 4 MHz. It is just an inefficient transformer, not really a radio transmitter. It would be a fun thing to build anyway.

    Maybe AnthonyJames9 will let us know how he got on.

    We can look at MosFets if it doesn't work very well.
     
  15. Dec 15, 2011 #14

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    It may be better to use parallel tuned circuits and to couple into the 'primary' and out of the 'secondary' by tapping onto the bottom few turns of the coils (i.e.like two auto transformers) Once you have found your resonance, you can alter the tapping position to optimise the power delivered. Your coils will be single layer and fairly open spaced so re-soldering the taps should be easy. The advantage in this method would be that, although the two tank circuits may be identical, the optimum match will probably be with different tapping positions for primary and secondary and it should be easy to find that optimum.
     
  16. Dec 15, 2011 #15

    vk6kro

    User Avatar
    Science Advisor

    I think this will possibly end up as a pair of Pi networks. This will require two more variable capacitors on the left of the coils and to ground.

    The problem with tapping down on the coils is that these will not have a lot of turns on them because they need to be large diameter for coupling.

    So, the difference in impedance between adjacent tappings will be quite large even if adjacent turns are used as tappings. Certainly worth a try, though.

    The diagram shows many turns, but this will not be the case if a frequency of 4 MHz is used.

    A 10 turn coil 8 inches in diameter (and 1 inch long) will have an inductance of about 34 uH and resonate with 47 pF at 4 MHz.
    So the coil may end up with only 6 or 7 turns on it.

    Great project though, even if it will prove how inefficient such schemes are.
     
  17. Dec 15, 2011 #16

    sophiecentaur

    User Avatar
    Science Advisor
    Gold Member

    Yes. When people talk about 'wireless power transmission' they often forget all those -dB involved in the wireless transmission of information.

    I should have thought that hundreds of kHz would be a more suitable frequency range than several MHz for this. Circuits are much more likely to behave like the circuit diagram would suggest as the parasitics are less relevant.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: High Frequency Oscillator Circuit Help?
  1. Oscillating Circuit (Replies: 0)

  2. Oscillator frequency (Replies: 16)

Loading...