Jeff Liebermann wrote:
> On Wed, 03 Dec 2008 01:48:00 GMT, "me here"
> <(E-Mail Removed)> wrote:
>
> > I recently did a comparo between a corner reflector antenna based on
> > the EZ-10 (of quoted gain = 10 dbi) and a 30 cm diameter stainless
> > salad dish /usb adapter reflector. Both 54G.
>
> Oh-oh. I think I know where you found the salad bowl.
> <http://802.11junk.com/jeffl/antennas/Salad-Dish/index.html>
>
> > The weak reference signal was 100 metres away.
>
> Problem.... The earth is inside the Fresnel Zone which will cause some
> weirdness in trying to measure signal strength. See:
> <http://www.terabeam.com/support/calculations/fresnel-zone.php>
> For 2.4GHz at 100 meters range, the end points have to be at least 1.4
> meters above ground level. Get two ladders.
>
> > The corner reflector hanging off a Belkin PCI card and very short
> > pigtail pulled in the chosen signal at 18 - 20%.
> >
> > The dish in the same position using a USB Belkin adapter pulled it
> > in at a constant 40%.
> >
> > The internal antenna on the laptop used for the USB didn't pick up
> > anything.
>
> It's going to be rather difficult trying to guess the gain using
> percentages. Percentage of what? The numbers don't mean much except
> for relative improvements.
>
> > So can I make an assumption on the gain of the USB dish, based on
> > the supposed gain of the EZ-10, given the different transmitters?
>
> Nope. If you had numbers in dBm or milliwatts, it might be possible.
> Can you run Netstumbler?
>
> > The EZ-10 clone is actually my EZY-F freestanding unit - but the
> > specs are identical.
>
> Incidentally, this floated to the surface of PC World Magazine
> recently:
> <http://www.pcworld.com/article/15095...range_with_a_p
> arabolic_reflector.html>
>
> > The salad bowl is a perfect half sphere except for the 10 cm flat
> > base, so the beamwidth should equate closely to a half sphere.
>
> Find a dark room and a light bulb (LED with wire leads works) and see
> the pattern on the wall. Instead of a concentrated spot, as you would
> expect from a parabola, you get an annualar ring. It's still useful,
> but there's a substantial dip in the signal level at dead center. I've
> never bothered to measure the salad bowl gain.
>
> However, it is possible to guesstimate the maximum gain for a given
> diameter parabolic dish antenna. See:
> <http://groups.google.com/group/alt.i...g/3a58d0c911f6
> e30c>
>
> > I had a browse on the net but it wasn't very helpfull.
> >
> > Can I assume the dish is greater than 10 dbi?
>
> Nope. Besides the hemispherical shape and weird pattern, the USB
> dongle is not exactly an ideal feed element. Much of the RF radiates
> in directions that do NOT hit the salad bowl. My guess is about half,
> for a net loss of -3dB. In addition, the tiny ceramic antenna inside
> the USB dongle has a gain of perhaps -4dBi, which is why it didn't
> hear anything. It's difficult to guess the gain in absolute terms.
> Best to have a test antenna, with a known gain, and do a comparison.
Hi Jeff,
Thanks for the indepth response.
Some interesting stuff in those links.
My salad bowl and your salad bowl must be related :-) Yep, they're
identical. Hadn't seen that photo before.
I mounted the dongle differently, I pop rivetted an L shaped tab at the
back of the flat section (underside) of the bowl and held the back end
of the adapter against it and in the focal position with two stretchy
"O" rings. Works well.
I can run Netstumbler on both the PCI based PC and also the USB based
laptop.
However, I get strange happenings with Netstumbler and BeLkin wifi.
Often Netstumbler will not pick up all the networks that the Belkin
software shows.
I am well aware that Belkin and windows wifi management do not get on
well together and you must tick box to turn it off :-).
Netstumbler suspends the Belkin process when it runs, and I suspect
there's a conflict there somewhere. It seem to turn Windows management
on again, so I doubt thats the problem.
The end result of the comparo is that the dish easily outperforms the
corner reflector, even if the USB signal pattern is dodgy.
Incidentally, before I rigged up the bowl, I tried to replicate my
EZY-F corner reflector setup with the USB adapter instead of the dipole.
Absolute failure. I suspect that as you suggest, the USB dongle just
sprayed signal everywhere and the reflector was useless - couldn't
focus it.
With the bowl, the USB adapter works better with one side against the
locating tab than the other. Could be some interference on one side of
the adapter from the tab.
The bowl however definitely gets stuff the corner reflector can't.
So is it worth a try of the USB adapter with the PC World parabola? Or
is the USB going to get more lucky spraying signals within a dish.
Cheers
Rob
Similar Threads
Linear Mode
