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

My Design Project ( I NEED HELP)

  1. Aug 29, 2009 #1
    Hi Guys,

    I posted earlier two threads, one asking about how to design a 2.45 GHz oscillator,
    and the other asked from where to purchase one. To be honest, the oscillator is
    one part of the design project that I am actullay stuck with. The advising professors
    told us that they will help in a wireless power transmission project, which
    drived us to agree. However, at the end, they abandoned us. Now we are stuck. I would
    really appreciate your help and you will literlaly help me graduate.

    The project is a wireless power transmission system, divided into blocks

    1) A solar panel, which will be our supply.

    2) a regulator, since the solar panel output will be fluctuating.

    3) a 2.45 GHz oscillator, which will convert the DC to a high frequency AC

    4) A horn antenna for sending and receiving.

    5) down step the recived signal to a lower frequencey for an AC load, or convert it to
    dc for an ac load.

    This is the overall system, I need anything from links to helpful resources, general advices,
    proposed changes in the system it self, either removing blocks or adding them.

    Any help would be greatly appreciated...

    Thanks in advance
  2. jcsd
  3. Aug 29, 2009 #2


    User Avatar
    Science Advisor

    You need to get a new project.

    Sending power in the way you describe is potentially dangerous and unlikely to give useful results. It is also illegal.

    A big problem is safety. There is a good reason why microwave ovens turn off when you open the door. Microwaves cook meat, including human meat. People have been killed getting in front of radar dishes when they were operating.

    Another problem is that 2.45 GHz is widely used for short range radio communication. Some cordless phones use it. Wireless networks for computers use it.
    What would happen if you start transmitting 1000 watts or so on 2.45 GHz?
    Needless to say, it is also illegal to do this.

    But your main problem is that there is an inverse square law which means signals get a LOT weaker as they move away from the source. So, if you transmitted 1000 watts you are unlikely to capture 1 watt 100 metres away.

    Sorry to be so negative, but this is a bad idea and your teachers were right to discourage you.
  4. Aug 29, 2009 #3
    I concur, this is not a good idea. Besides breaking FCC rules, the power you can recover a foot away from the VCO wouldn't be enough to light up a single LED.

    Down converting the signal in order to receiver power wouldn't work because you need power to down convert the signal.

    Wireless transmission of power done at MIT was around 100 KHz, not GHz. It involved resonance between the transmitter and the receiver to suck out as much power as possible.

    If you are going to attempt this project see your prof ASAP.
  5. Aug 30, 2009 #4
    I don't perceive that there is a clear difficulty if:

    1. You have access to a screened room.
    2. You keep your personal exposure to a reasonable level (less than 5mw / cm^2 for 6min)

    I take it the principle interest is to validate or invalidate some assume regarding the validity of this project. I'm not at all convinced that you couldn't beat the efficiency of inductive systems using hundreds (or thousands) of amp turns to get a few feet.

    You may try minicircuits.com They take a black-box approach to oscillators, modulators, and amplifiers. They do a great job of keeping inventory.

    One warning about RF. RF of sufficient power can burn holes in your flesh, kill your bone marrow, or cook your eyes. The US Army has a fancy for turning high power RF on people for demonstrations without regard to the thousands of glass workers that suffered cataracts from similar heating of their cornea.

    If you lack a shielded room, signal source, amplifiers antennas, and field strength meter, you can contact any number of emc test facilities, which will have almost everything you need.
  6. Aug 30, 2009 #5
    Thanks Guys for your quick rich responses,


    You are right about sending alot of power, but ours will concentrate on small safe amounts only. Nevertheless, expermienting with such project might be dangourse. Can you propose something which is a bit different but havd, overall, the same idea. We already wrote a proposal about it and we are stuck right now. A slight change can be done, but a big one, I think is not.

    p.s. Our professors were the ones that propsed this idea, we wanted to use magnetic induction, which is much safer and more intersting, however, we couldnt disagree with our professors, we thought that we would get their help later on.


    You are correct. The losses is huge and wouldnt be able to light an LED. But in a sense, the porject is meant to immitate the projects NASA is working on, sending energy collected by solar panels floating in space to the earth using microwave. Of course the immitation would be so simplified. So sending alot of power is not our intention in this project. I know it does not seem logical but unfortunatlley this was our advising professors choice. That is why I am seeking your help here, WE ARE STUCK!!!


    I dont think we have a screened room, not to my knowledge. As I wrote earlier, we ar not intending to send alot of power.


    What do you propose? Do you think that with some modifications this project can see the light, or do we persue the magnetic induction instead of the microwave approach, that is if they allow us.

    Waiting for your responses...
  7. Aug 30, 2009 #6
    If I understand your situation, you are students outside the united states? If so, the local laws regarding power limits / band allocation are likely different, but the 5 mw/cm^2 exposure level is something good and safe to abide by. After all, if you stay a few wavelengths away from a 20cm x 20cm horn, your talking about 2 watts. That's a lot more than you should need to tune up your system.

    Since you're uncertain, and definitely need local resources, you may look to the following:
    1. rfcafe.com - They have an RF forum with contacts
    2. Amateur radio operators. These guys have a great deal of amassed knowledge and contacts. You can reach some via rfcafe.com
    3. Companies via internships. If any of you work for a large company as an intern, you may be surprised how far reaching your company's resources run. Most large companies have their own test RF facilities.

    Remember, you don't want your project to fail because it turned into a mega-project:
    1. RF stuff is difficult for a novice to design, so borrow it or buy it.
    2. Before there was an internet, there were people and knowing people is the best way to get things done. So, call, visit, discuss. This is how you get solid resources.
    3. Perfection is the enemy of the good. Identify what seems reasonable and abandon what is interesting but difficult.
    4. Make mistakes quickly. It's better to make mistakes, learn and change than it is to suffer from "analysis paralysis."
    5. Recycle as much as you can. If I part is assembled, use it. If somebody else has a good design, steal it (if you can..), if your old work had something useful, stick with it. Successful design is often about knowing what stride to take - how much to invent.

    If I were in your shoes, I'd aim for at least six conversations a day with area technical people. Try to find:
    1. A test facility
    2. An old lab
    3. A signal generator
    4. A screen room
    5. A pair of antenna's

    I'd purchase:
    1. An amplifier that runs off the solar panel
    2. A project box with a connector for the receiving antenna

    I'd design:
    1. A microwave matching circuit (unless you can buy one)
    2. A microwave rectifier circuit
    3. A load that clearly indicates the received power
    4. A means of reading the load (i.e. a shielded meter or old fashioned meter)

    In the end, you may have nothing more than an RF Schottky rectifier bridge being loaded by a high efficiency LED and series resistor. Then tap off of the load resistor and go to an old, old panel meter - one that doesn't have tubes or transistors. There you go, Viola! Vary your distance, tune your matching circuit and record the input and output readings.

    But, first things first. Find someone with a test facility, and get your materials together. The facility will undoubtedly have most everything you need (except for an amplifier that runs on low voltage).

    Good Luck,

  8. Sep 11, 2009 #7
    Thanks Mike your suggestions and advice were really helpful and it shows that you want to help

    I will show your reply to my professors and collegues and we will see what happens.

    By the way, is there any good resources that would show how to simulate antennas and allow me to measure their features, gain, directivity, etc.. in matlab, or if there is any other software please mention it as well.

  9. Sep 16, 2009 #8
    I know of segment based software. You can probably simulate a Yagi or helix with it, but not a horn or parabololic reflector. Try searching for "Antenna W7EL"
    But remember, time is ticking on your due date. Rather than design an antenna, it's better to buy one, and if you have to make one, it's better to steal the design.
    You're operating near the wifi frequency, so there's any number of off-the-shelf Yagi's that are fairly inexpensive. Also, you can visit this site for various wifi designs:
    A good many people use discarded satellite television dishes for high gain wifi antennas. In the USA, these are very common at apartment complexes.
    Another fairly simple antenna is the helix. These are simple to construct and have fairly high gain. The following web site shows how to construct a simple one:
    By the auther's guess, this antenna has about a +15dB gain - comparable to a good Yagi. With a quad version of this antenna, the gain will likely be closer to +22dB. The following site exhibits a quad, but since he's targeting a different frequency, you'll have to adjust the dimensions accordingly:
    If you do choose to build an antenna, it's best to have a spectrum analyzer with built-in tracking generator for trimming to the correct frequency. This arrangement is also good for characterizing the antennas. To do this, perform the following operation:
    1. Acquire two coaxial cables and build / tune three antennas (Antenna A,B,C).
    2. Use the spectrum analyzer to measure the loss through each cable.
    3. Transmit through antenna A and receive through Antenna B at a given distance (i.e 3meters). Record the value and subtract out the loss that is due to the two cables.
    4. Repeat procedure 3 with antennas A and C.
    5. Repeat procedure 3 with antennas B and C.
    6. Each antenna has an associated loss at 3 meters (i.e. dB_A for antenna A). You've measured (dB_A + dB_B) (dB_B + dB_C) and (dB_A + dB_C).
    7. To ascertain the antenna factor (or loss for each antenna at 3m), solve for unkowns, dB_A, dB_B, and dB_C, using the values you've measured.
    . Best of Luck,
    . Mike
    Last edited by a moderator: Apr 24, 2017
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook