DIY WiFi antenna (to increase reception)

Discussion in 'Linux Networking' started by gaikokujinkyofusho, Jun 9, 2005.

  1. Hi, I kinda new to wireless networking but wanted to find out if it was
    possible to make an antenna that increases the range of signals my
    notebook can receive. I have looked around and most antennas that I
    have seen are for increasing the range of a network (the sending of the
    signals?) not necessarily for reception... or if it works for
    transmission would it also work for reception? Is it possible to build
    (or buy on the cheap) some sort of omni-directional antenna that will
    allow me to receive signals that are far away? Any help would be
    greatly appreciated!

    Cheers

    -Gaiko
     
    gaikokujinkyofusho, Jun 9, 2005
    #1
    1. Advertisements

  2. gaikokujinkyofusho

    Postmaster Guest

    Construct a metal parabola with a focal point that is
    1/2 * (C/2.4 Ghz) ...( about 2.27 inches) from the
    surface of the parabola. Then place your antenna at the
    focal point, and point the parabola at the transmitter :)

    You can also construct similar antennas for cell phones....

    Enjoy,
    Postmaster
     
    Postmaster, Jun 9, 2005
    #2
    1. Advertisements

  3. Yes, it is reciprocal, and any improvement in transmission will be
    identical to the improvement for reception.

    (Note that increasing the transmit power of an Access Point might
    be possible too, but has exactly the effect you are concerned with,
    because it has value in one direction only.)
    Definitely! Just be warned that high gain omni-directional
    antennas are not simple, and not inexpensive.

    Let me provide a few technical details that you need in order to
    understand what you'll be looking at.

    First, antennas are measured in numbers called "dBi", which is a
    logarithmic ratio. The 'd' for deci, or 1/10th. the 'B' is for
    Bell, the name of the guy who came up with it. You'll see all
    kinds of dB ratios used in radio work. As just dB, it's a
    ratio. With a letter after it, it's a ratio compared to
    something specific. The 'i' stands for a point source antenna
    in outer space, called an "isotropic radiator". It's a truly
    omni-directional antenna, and no such thing exists in reality.
    This (imaginary) isotropic radiator as 0 dBi gain.

    What you'll see called "omni" antennas are omni-directional
    around the points of a compass (azimuth), but not vertically.
    That's good, because you don't want to waste your radio's power
    by transmitting it directly up, or directly down. Instead, that
    power is redirected perpendicular to the antenna, which gives it
    gain in those directions.

    Because of that, a plain old every day omni antenna has a gain
    of about 2 dBi. And everything else you see, you will want to
    compare to that. The comparison is that a 6 dB increase in
    signal strength (3 dB is twice as much power, because dB's are
    logarithmic, so 6 dB is 4 times as much power) will just about
    double the range of a radio link. Hence if you can walk down
    the road 300 feet before you lose your connection, if you switch
    from a 2 dBi antenna to an 8 dBi antenna you would probably be
    able to go about 600 feet before losing the connection.

    You can buy 7 dBi (almost a 6 dB increase) antennas for something
    like $40 or so. More gain than that in an omni requires some
    significant hardware, and will cost relatively more.

    "Directional antennas", where that means directional in azimuth
    instead of only in elevation, are much easier to get a lot of
    gain from, and hence are less expensive (but don't do you much
    good either).

    There are other considerations too. The higher the antenna, the
    farther it's view will be. Also the fewer obstructions, the
    better. It happens that 802.11b/g wireless uses 2400MHz, and
    at that frequency almost *anything* absorbs radio signals. Trees,
    walls, and basically anything you can't see through. So locate
    your Access Point accordingly!
     
    Floyd L. Davidson, Jun 9, 2005
    #3
  4. gaikokujinkyofusho

    Postmaster Guest

    A few more details:

    You'll need a 9 inch piece of cardboard,
    and some aluminum foil. Then create a parabola
    See:
    http://www.sciences.univ-nantes.fr/physique/perso/gtulloue/conics/drawing/para_string.html

    Be sure that the focal point is ~ 2.267 inches from the nearest
    point on the parabola. Now cover your cardboard with the foil,
    and bend into the parabola that you just drew. Put the wireless
    antenna at 'F' ( the focal point ) and point this at the nearest
    access point. You got it...

    Note: You can create two parabola shapes out of a piece
    of cardboard, and bend the aluminum foil/cardboard around
    these two pieces. Then glue, to hold it in place.

    Nothing like a bit of math and a homebrew project to add
    fun to computing :)


    Enjoy,
    Postmaster
     
    Postmaster, Jun 9, 2005
    #4
  5. gaikokujinkyofusho

    Unruh Guest

    This however also means that you cannot have it on a laptop which you then
    operate on your lap--ie with the perpendicular of the antenna pointing
    everywhich way. The antenna must be oriented so that the perp (axis) really
    is perpendicular, or you will get less, rather than more gain. The higher
    the gain the worst this is.
     
    Unruh, Jun 9, 2005
    #5
  6. gaikokujinkyofusho

    ynotssor Guest

    You should look up the definition of "omni-directional".
     
    ynotssor, Jun 9, 2005
    #6
  7. True, but even *worse* than you are saying!

    2.4GHz RF will bounce off any metal object larger than about 3"
    square, hence there are bound to be lots of reflected signals in
    almost any environment. In some cases one of those reflected
    signals is likely to be the strongest signal. Because it is
    reflected, it no longer necessarily has vertical polarization,
    and might just be almost anything.

    If you put the laptop in your lap, and tilt one way... you may
    get an increase in signal, but tilting in some other direction
    may get an even deeper fade than just having it sitting straight
    a level. And of source in many laptops the antenna is in the
    lid, which isn't necessarily straight up and down either!

    What it all boils down to is that having a good line of sight
    location only a short distance from the Access Point is best,
    and for anything else a little experimentation with orientation
    might do wonders if there is a marginal connection.
     
    Floyd L. Davidson, Jun 10, 2005
    #7
  8. gaikokujinkyofusho

    Postmaster Guest

    One may find that a directional antenna, pointed at the
    access point, can greatly improve range. If the antenna
    is lite, and easily turned, one can quickly find the access
    point and be a happy camper :) The method above
    increases the range by a rather significant distance....
    In some cases, around 1/2 mile !

    Enjoy,
    Postmaster.
     
    Postmaster, Jun 10, 2005
    #8
  9. gaikokujinkyofusho

    James Knott Guest

    With very few exceptions, antennas work equally well for transmission and
    reception. Now what you're looking for is gain, which involves tradeoffs.
    You get more signal in one direction, only by reducing it in others. With
    a standard vertical antenna, you'll have a donut shaped pattern. With an
    omnidirectional gain antenna, you'll have a squashed donut, where range is
    extended at the expense of signal at higher angles. You can also get
    directional antennas, with more gain. However, the more gain, the more
    complex (read expensive) the antenna becomes. There are commercial antenna
    available and you can roll your own as well. But no matter how you obtain
    an external antenna, you'll need some method to connect it. Does your wifi
    adapter have a connector, that can be used for an external antenna?
     
    James Knott, Jun 10, 2005
    #9
  10. The OP has a laptop, with a builtin wifi client. He could go to
    a lot of trouble to figure out how to get an external antenna to
    work, and it would no doubt cost a wee bit too.

    But tell me, is he really going to pack a parabolic around and
    set it up everywhere he wants to use the lap top from? I doubt
    it.

    Perhaps, the appropriate discussion is about what he can do with
    high gain omni-directional antennas for the AP.
    And indeed, there is some remote chance that he would be
    interested in doing it that way. But you need to point out the
    detractions too, such as needing to buy another wifi client and
    get one with a connector for an external antenna.
     
    Floyd L. Davidson, Jun 10, 2005
    #10
  11. gaikokujinkyofusho

    Stan Goodman Guest

    It is in the nature of antennas that changes for the better or worse produce
    the same effects on reception and transmission.
     
    Stan Goodman, Jun 10, 2005
    #11
  12. gaikokujinkyofusho

    Unruh Guest

    On a laptop, a high gain omni would not be terribly useful. The antenna
    would have to be very accurately mounted (re the vertical direction) and
    laptops never are. It would in fact be easier for him to pack around a
    parabolic and aim it when he wants to use it.


    Very true.
     
    Unruh, Jun 10, 2005
    #12
  13. Nobody suggested using a high gain omni on a laptop. But what you
    say is certainly the reason nobody mentioned it.
     
    Floyd L. Davidson, Jun 10, 2005
    #13
  14. This is absurdly pessimistic, in my experience.
    I don't believe there should be any problem in using WiFi
    in 99% or houses or apartments.
    If there is, I would suggest trying a different AP or PCMCIA card.

    I have a PCI-to-PCMCIA adaptor on my desktop
    with a Lucent/Orinoco Gold PCMCIA card in it.
    This has a separate antenna,
    but I haven't found the orientation of the antenna makes much difference.
    I can access the desktop from my laptop (also with an Orinoco gold card)
    throughout a very large (4-storey) building
    with some extremely thick (2 feet) walls.
     
    Timothy Murphy, Jun 10, 2005
    #14
  15. Lets look at it one part at a time, and follow the logic.

    1) good line of sight at short distance is best

    That's an *obvious* truth!

    2) for anything else

    Means that what follows does not apply for such wonderful
    situations described in 1). It doesn't apply because if there is
    only a short distance and there is perfect line of sight, it is
    very unlikely that a reflection provides the major signal
    component. Therefore the polarization is unlikely to be
    changed. And even if it is, a 10-15 dB drop in signal probably
    won't hurt, since the signal would have a 20-30 dB or more fade
    margin anyway.

    3) if there is a marginal connection

    In other words, this part makes no difference if you have
    a good connection.

    4) experimenting with orientation might do wonders.

    That is the only part left for your claim that it is absurd.
    And of course 4) is not at all absurd, but is absolutely true.

    Can you follow the logic when it is laid out one step at
    a time?
    That is a meaningless statement. No problem doing *what*?
    Connecting across the length of the living room? I agree.
    Connecting through the exterior wall to the patio? I doubt
    it comes close to 99%. Maybe 50-50. It depends on the position
    of the windows and the composition of the walls. And flat
    statement just are not useful.
    Which of course has *nothing* to do with whether signals
    propagate or not, so that is not likely to be the correct
    solution.
    If you make a valid test, you'll find that the null can be as
    deep a 30+ dB. (Some applications *require* a null better than
    30 dB!)

    The reason you haven't found much difference is probably because
    you are experimenting in an environment that has hugely
    significant amounts of multi-path signals. That means no matter
    how you orient the antenna, there is a signal relatively close
    to that polarization. You won't see more than a few dB change.

    A typical example of that would be a pci bus card in a tower
    case with the antenna on the back. You simply can't escape all
    of the reflections off the case itself. With your separate
    antenna it depends on where you put it. If you happen to be in
    a building that has aluminum foil backed insulation in the walls
    (which is exceedingly common), you'll have a hard time finding
    any polarization that is unworkable.

    Regardless of that, if you want to see an example of just
    exactly how significant cross-polarization nulls are... look at
    virtually every microwave used by the telecommunications
    industry and virtually every C band satellite system, and most X
    band satellite systems other than those providing TV to
    consumers. All of those cited use both horizontal and vertical
    signals at the exact same frequency, with the antennas very
    carefully set at 180 degree angles. Starband satellite Internet
    service is a good example.
    So? As with another recent thread, you seem to think that a
    single experience is indicative of what all experience will be.

    However... in this case I have no doubt you are exaggerating
    greatly. A large 4 story building with 2 foot thick walls?
    2400MHz signals don't go through concrete, don't go through foil
    back insulation, and don't go through plaster walls that use
    wire mesh to support the plaster.

    Your not being honest about this 4 story building.
     
    Floyd L. Davidson, Jun 10, 2005
    #15
  16. My appologies to Timothy Murphy, I confused him with someone
    else, and he doesn't deserve the implication of that statement.
    My mistake, not his.
     
    Floyd L. Davidson, Jun 10, 2005
    #16
  17. This is like saying that phones are best
    if the person you are talking to is in the same room.
    True but ridiculous.
    I'm sure you are right -
    you obviously know much more about the subject than me.
    But in my opinion, anyone who has serious WiFi problems in a normal house
    (or apartment) would do better to try different WiFi cards,
    rather than trying to make antennae out of cardboard and tin foil.
    I've used WiFi on my laptop in lots of places eg around my college.
    I just took my house because it seemed to offer the greatest difficulty.
    I'm in it now,
    and it is exactly as I described.
    It is an 1860 orphanage.
    The reason why it has some very thick walls
    is because the building was extended almost as soon as it was built,
    and what were exterior walls became interior walls.
    [I just checked, and the wall is more than 26 inches thick.]
    These walls are granite, and the signal has to pass through this wall
    (and 1 ceiling and 1 other thinner wall)
    when I am in the bedroom, as I was when I wrote before.
    Presently it only has to go through one (very substantial) ceiling.

    I've seen colleagues with WiFi problems,
    and it seems to me that these are nearly always due to the WiFi devices,
    and not to the environment.

    Incidentally, even Bluetooth is almost usable throughout this house,
    though it won't go through the thick wall reliably.
     
    Timothy Murphy, Jun 10, 2005
    #17
  18. gaikokujinkyofusho

    Stan Goodman Guest

    Not so. What he is trying hard to tell you is that your best chance of
    getting a good and reliable signal with a known polarization is to have a
    line-of-sight between the two antennas. In the absense of such a geometry,
    you may have multiple paths from reflections, which may produce interference
    patterns so that shifting one station by centimeters may make the received
    amplitude go to an unusably low level; or a 90 degree rotated polarization
    plane that does about the same thing; or high attenuation by the object that
    breaks the line of sight. The analogy with two telephones in the same room
    is a false one, based on a deceptively oversimplified understanding of what
    happens between the two antennas. The fact that you can listen to a
    broadcast station that is e.g. over the horizon does not imply that the same
    thing can happen at these wavelengths.

    He is trying very patiently to talk sense to you in non-technical language
    that a non-engineer can understand if he wants to do so. Your reaction, on
    the other hand, relying as it does on an ignorance of electromagnetic
    propagation, illustrates the saying that "a little knowledge is a dangerous
    thing".

     
    Stan Goodman, Jun 11, 2005
    #18
  19. Your statement explains why you can't seem to understand exactly
    how it does work. Fundamentals are *not* ridiculous, they are
    *required*.

    I assume you have read Stan Goodman's article, and I'm not going
    to repeat what he has stated so ably.
    If you don't understand what I'm saying, feel free to point out
    that my words are confusing, and ask for a different
    explanation. Tell me what you do understand, and how you
    understand it, and I'll try to go from there and provide as good
    a tutorial as I can. Skip the adversarial part of it, because
    it isn't a contest, and if you make it one you can't win.
    That is untrue. What value is there in spending money to buy
    another equal product? All of these units have about the same
    capabilities.
    If I had suggested that, you might have a point. I didn't, and
    you don't. (And I might note that what you are suggesting is
    far less likely to be effective than antennas from cardboard and
    tin foil, which do happen to work.)
    Then you are not describing it well. That would be for the
    simple reason that you don't know what makes a difference to
    2400 MHz radio signals, and simply can't tell me what is
    happening. Most people wouldn't, and have no reason to!
    The signal does not necessarily have to pass through the wall.
    It is probably reflecting off something above it and coming back
    down on the other side.
    A ceiling made from wood that has been in place for 140 years, getting
    drier all the time, would not be much of a problem. That is assuming
    nobody has stretched wire mesh across it to hold plaster, which would
    be a serious block.
    Clearly *every* such setup has environmental problems. We all
    expect there to be a limit to how far away we can go and still
    connect, and you are just ignoring that because you know about
    it. Someone who has a good background in radio signal
    propagation will see significant things most people wouldn't
    notice or suspect.
    I doubt that the 2400 MHz wifi signal is going through your wall
    either, though I don't really know for sure, as granite walls
    aren't something I have handy to test with. I'm sure that
    concrete is bad, but that does have a lot of moisture in it that
    granite would not have. Hence it is quite possible.

    Regardless, just be glad you get the signal coverage you want
    without having to do much to accomplish it. Often enough that
    is not the case.
     
    Floyd L. Davidson, Jun 11, 2005
    #19
  20. gaikokujinkyofusho

    James Knott Guest

    Try holding a conversation with someone in the same room, via a pair of
    digital cell phones. ;-)
     
    James Knott, Jun 11, 2005
    #20
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.