Hawking 8db gain Dish woes.

Discussion in 'Wireless Internet' started by Don W. McCollough, Jan 13, 2006.

  1. Ok. I purchased one of these silly 50s sci-fi looking
    adapters to see if I could lock onto more wi-fi routers
    in my neighborhood; I was going to go with a cantenna
    but I didn't want to buy a new PCI card and pigtail.

    This is the dish


    So, I hooked it up to my computer which has
    a cheap Wireless B PCI card. Normally, I get about 7
    or 8 neighborhood wireless networks and can connect
    to them sometimes with "Very Good" or "Excellent" signal readings.
    My own Wireless B router usually comes in "Excellent."

    I was expecting to see more networks and pull better
    signals from the preexisting networks (routers) with this
    dish, but after installing the driver and using its networking
    configuration software, I found that all I could get is
    "Low" "Very Low" and rarely "Good" signals.

    The dish should have a 8db gain. I tried pointing it in all
    directions and angles...outside my windows...but this thing
    would not get any additional router signals and when it
    found the older signals it didn't connect very well. The cheap
    wireless B PCI card did much better without having to aim it.

    I've contacted their tech support...and the thing might be
    defective, but am I missing something here?? Do these
    "gain" antennas only connect to networks that are normally
    out of the distance of typical wi-fi adapters and do little to
    lock onto to closer signals? One would expect that these
    kind of dishes would be useful in determining from which direction
    a signal is coming from...especially with built in signal locator LEDs.

    This Hawking dish is a big let down. Its supposed to be Plug and Play.
    Any input would be much appreciated.

    Don W. McCollough, Jan 13, 2006
    1. Advertisements

  2. Chuckle.

    I can't really tell what's inside or how it works, but let's pretend
    it's a properly designed parabolic dish. Could you measure the dish
    diameter? My guess(tm) is about 15cm.

    The theoretical power gain of a dish is roughly:
    (circumference / wavelength) ^2
    (Pi * 15cm / 12.5cm ) ^2 = (3.14 * 15 / 12.5)^2 = 14.2
    Converting to decibels:
    Gain = 10 * log(14.2) = 11.5 dBi
    However, that's assuming perfect illumination and efficiency. Well
    designed dish antennas have about 50% efficient or a -3dB loss. Do
    the maximum gain of this contrivance is:
    11.5 - 3 = 8.5dBi
    Yeah, 8dBi is theoretically possible.

    The beam width would be:
    1.2 * wavelength / dish diameter) in radians
    (where 2 * 3.14 radians equals 180 degrees).

    wavelength = 3x10^8 meters/sec / 2400x10^6 Hz = 0.125 meters
    Dish diameter = 15 cm (my guess)
    1.2 * 0.125 / 0.15 = 1.0 radians
    1.0 radians * 180 / Pi = 57 degrees.
    With 57 degree beam width, this is not a very directional antenna. Of
    course this assume that's its really a parabolic reflector. My guess
    is that the feed is anything but ideal for illuminating such a small
    dish. Methinks that the USB dongle at the center has a PIFA or
    meandering 1/4 wave element for an antenna. These radiate roughly in
    a spherical pattern, where the energy goes in all directions roughly
    equally. That means a large part of the signal goes in directions
    other than hitting the reflector and going in the direction of the
    access point. I previous ground through the numbers at:
    | http://groups.google.com/group/alt.internet.wireless/msg/af7ddbbf5aa01dd7
    Jeff Liebermann, Jan 13, 2006
    1. Advertisements

  3. Thanks for the calculations. The dish is actually (more like) 12cm. And
    57 degrees isn't that directional, which makes me wonder why this thing
    won't pull in a better signal than my standard PCI card. It should, if
    at a nearby broadcasting router give *some* gain in signal...but nada.
    I've contacted
    Hawking Tech tech support, lets see what they say.
    Don W. McCollough, Jan 13, 2006
  4. Well, the numbers I calculated are for an idea dish antenna. My
    guess(tm) is that this thing is anything but idea. Did you read my
    Google Groups reference to what happens when you use a USB dongle as a
    feed for a parabolic 18" satellite dish? The illumination pattern of
    the feed (i.e. the dongle) is VERY lossy, with much of the signal
    going off in useless directions. My guess(tm) is only about 15% of
    the RF that leaves the USB dongle, hits the reflector and goes in the
    general direction of the access point. 0.15 is about:
    10 log (0.15) = -8dB
    of loss. That makes the antenna gain about:
    8dBi (theory max gain) - 8dB (illum loss) = 0dB
    In other words, this contrivance is slightly better than the USB
    dongle by itself.
    Jeff Liebermann, Jan 13, 2006

  5. That's dissapointing to hear. Hawking tech support claims that the thing
    defective...send in for a replacement etc.

    Out of curiosity would the output of a PCI card be significantly better
    than a USB dongle?

    Don W. McCollough, Jan 14, 2006
  6. Jeff,

    Something like this would probably be a better bet eh? 24db gain.


    But could I mount this thing indoors?

    Don W. McCollough, Jan 14, 2006
  7. Your dime. I would ask for a refund and buy something that works.
    No, not really. The typical USB dongle and PCI cards vary between
    +13dBm and +17dBm, mostly towards the lower end. However, there is a
    difference in antenna gain. The typical USB device, with a PIFA or
    meandering monopole antenna, has a gain of about -2dBi. The PCI
    antenna is much better at about 2 dBi. The 4dB difference should
    theoretically yield about 70% more range.
    Jeff Liebermann, Jan 14, 2006
  8. Yep. Much better. That's the old PacificWireless antenna. I have a
    pile and use them for point to point links. It has a rather narrow
    beamwidth (about 5 to 7 degrees) and may be difficult to aim and keep
    Are you married? Are you planning to stay married?
    Technically, you can mount it in a window and use it indoors. If the
    walls are like paper, you might be able to shoot through them.
    Jeff Liebermann, Jan 14, 2006
  9. Don W. McCollough

    Bob Alston Guest

    Of course, unless your PCI card has a cable allowing repositioning of
    the antenna, the dongle has the advantage of being able to position it
    for best reception.

    Bob Alston, Jan 14, 2006
  10. Don't know the experiences others have had with Hawking Tech .
    They seem to be somewhat slow on the draw - service and even rebates.
    The last simple question I had took weeks to answer with the tech
    apologizing that " we have worked off our feet and yes it should work".
    Anyone else have similar or diffirent experiences ?
    frankdowling1, Jan 15, 2006
  11. Don W. McCollough

    dold Guest

    I have used the "Hawking HAI6SDA Directional 6dBi 2.4GHz Antenna" with good
    success on a Netgear WG311 PCI card.

    I wouldn't expect their 8db USB widget to be junk. Jeff's calculations
    indicate that their claim could be correct, but then he wanders off
    claiming that it is no better than the New Zealand dishes, which are rather
    artsy, but lack the real design that Hawking might have done.

    If you just bought it, why can't you return it and get a replacement. If
    the replacement is the same, then you should return that for a refund.

    Did you remove your original PCI card? Are you sure that you are looking
    at the signal from the new gadget? Did it come with a USB cable? Could
    you try another cable?
    dold, Jan 15, 2006
  12. Don W. McCollough

    sded Guest

    I just tried and am returning an HWU8DD. Seemed well built, software was very
    good, no installation problems, worked fine. BUT the sensitivity was not nearly
    as good as my old D-Link DWL120+ with its small dipole antenna. And I pointed
    the HWU8DD at known targets and rotated it through 360 degrees. The DWL120+ saw
    and associated with signals the HWU8DD never saw, and provided a much better
    (10dB or more) signal for those both could see. This is my second bad
    experience with a similar Hawking product (wasted time a bit ago with an
    HWU36D); never again! It is awkward to store and use and does not provide
    higher performance than a simple dipole unit from other vendors.
    sded, Jan 15, 2006
  13. hath wroth:
    Yeah, something like that. Some more guesswork. What I found
    interesting is that the "dish" was 12cm in diameter. That's almost
    12.5cm which is 1.0 wavelengths at 2.4GHz. If the dish were measured
    along the curved diameter, I suspect it would be exactly 12.5cm.

    Well, that's a simple 2 element "yagi like" reflector. You can take
    any dipole antenna and fasten a reflector at a distance of
    approximately 1/4 wavelength and get some gain. Instead of a dish, it
    could have been a flat plate and achieved the same results. Any
    multiple of 1/2 wavelength diameter will work. In other words, this
    is not a "well designed" reflective dish antenna. I may be a "well
    designed" flat plate reflector antenna where the aesthetics department
    decided would look better if it were a dish instead of a flat disk.

    I'm not sure how much gain such a flat plate reflector will produce.
    I'll try to shove it into 4NEC2 and see what happens. It will
    probably be over 3dB but surely less than the promised 8dB. I'll
    assume a common USB dongle with a ceramic backed antenna feed (-2dBi

    This the 2nd time that Hawking played games with the specs. Last
    time, it was specifying the gain of their external patch antenna at
    6dBi without the slightest consideration for the 3dB loss in the 3
    meters of tiny RG-316/u coax cable.

    Jeff Liebermann, Jan 15, 2006
  14. Don W. McCollough

    dold Guest

    Reading you draw conclusions is like reading Sherlock Holmes.

    Did you look at the FCC sheet for it? That would show whether there's an
    active USB element in the "feedhorn", or if it's just an aesthetically
    pleasing dipole.

    Hawking didn't seem to have a problem with the flat plate on the 6dBi that
    I bought. Marketing could have appended an empty "pointer" to that one
    dold, Jan 16, 2006
  15. hath wroth:
    Like Sherlock Holmes, I don't know all the answers until the
    investigation is done.
    Yep. I used it to extract some dimensions from the photos.
    The test report shows that it was made by Edimax in Taiwan.
    The antenna gain is claimed at 8.00dBi with +17dBm output for 802.11b
    modes, and +15dBm output for 802.11g (OFDM) modes.

    The construction was not what I expected. I previously guessed that
    there would be a USB dongle hanging in the center feed area. Instead,
    the USB radio is all in the base of the unit, with a coax cable going
    to the antenna feed in the center cylinder. There are no details or
    photos that show the type of feed. It could easily be a ceramic
    backed 1/4 monopole, PIFA antenna, or some manner of loop or squashed
    dipole. I can't tell without breaking the center cylinder open.

    I have an NEC2 model constructed that seems to show that it is
    possible to get 8dBi of gain out this 12cm dia reflector using a
    simple dipole feed. I get slightly less gain with a flat plate
    (circular) reflector. The gain will be less when I try it with a
    smaller (and lower gain) center feed antenna. I'm having some
    problems getting the VSWR down to below 2:1 so I'm not quite ready to
    post quite yet.
    Except for not specifying the coax loss, I don't have any objections
    to that antenna.
    Jeff Liebermann, Jan 16, 2006
  16. http://802.11junk.com/jeffl/antennas/

    I gave up on getting the VSWR down and simply cheated by using a 75
    ohm feed instead of 50 ohms. I'll play with the optimizer later. I
    also couldn't figure out what was inside the Hawking antenna feed, so
    I just left it a simple dipole. It's probably a circuit board with
    some manner of loop, PIFA, meandering 1/4 wave line, or something
    similarly miniature that will fit. That will reduce the gain

    The 1.0 wave diameter dish antenna yields a gain of about 7.7dBi.
    The 1.0 wave flat plate reflector yields a gain of about 8.3dBi.
    However, that's well within the tolerances of my guesswork as to the
    location of the feed in the flat plate reflector and f/D ratio on the
    dish. However, if I use a shortened feed, the gain will drop about
    2-4dB, putting the antenna well below the advertised 8.0dBi gain. I'll
    do a shortened feed model later.

    FCD ID web page:

    Hawking HWU8DD web page:

    NEC2 data files (not optimized):

    4NEC2 antenna modeling program:
    Jeff Liebermann, Jan 16, 2006
  17. Jeff Liebermann, Jan 16, 2006
  18. Don W. McCollough

    dold Guest

    I expected Hawking to be better than the gang in New Zealand.
    I am a bit surprised by that. I guess we don't typically "see" what's
    inside the rubber duckie on a WAP either, but this seems different. I
    expected to see an inside photo.
    Sherlock on the trail again. Maybe it's a perfectly optimized feed. After
    all, these guys do this for a living. ;-)

    The HWU54D is my flat panel HAI6SDP + HWU54G USB in one unit. A little
    less Sci-Fi looking than the HWU8DD mini-dish, at 6dBi, and $10 cheaper.

    But the OP never said what kind of PCI card he had. Maybe he could just
    add the external antenna.

    And then we could talk about which of his neighbors' connections he intends
    to poach.
    dold, Jan 16, 2006
  19. hath wroth:
    When you have eliminated all which is impossible, beaten to death that
    which is improbable, pounded into the ground that which is
    unproveable, contemplated the rediculous, denounced the political,
    dubunked the miracles, removed the personalities, and read the FAQ,
    then whatever remains, however absurd, must be the truth.
    Not me. I just answer the techy questions and avoid the politics.
    You're welcome to provide moral support or legal direction if
    required. Incidentally, I read somewhere that the term is now
    Jeff Liebermann, Jan 16, 2006
  20. Don W. McCollough, Jan 17, 2006
    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.