DaveC wrote:
DirectTV 20" dish, offset fed, f/d~0.7, to be used for 2.4 GHz wlan link.

As I understand the issues, the best feed is one that -- for a given focal
length -- will "light" the dish with as spherical a wave as possible, that
has a broad enough dispersion angle to fully "paint" the dish without picking
up off-the-edge noise.

I've looked at all types of feeds, from cantennas to bi-quads to patch to
dual-dipole. All have their pros and cons.

I'd like to hear from anyone who has done this before about what feed they
chose and why.

Thanks,
Biquad , simple , safe ,cheap and you can build it with a minimm of tools.

In alt.internet.wireless DaveC wrote:
DirectTV 20" dish, offset fed, f/d~0.7, to be used for 2.4 GHz wlan link.

I've looked at all types of feeds, from cantennas to bi-quads to patch to
dual-dipole. All have their pros and cons.

The Trevor Marshall BiQuad is the most referenced feed for a dish.
http://martybugs.net/wireless/biquad/ refers to Trevor's site, and has
better instructions and pictures.

--
---
Clarence A Dold - Hidden Valley (Lake County) CA USA 38.8,-122.5

On Thu, 12 May 2005 22:28:26 -0700, DaveC wrote:

DirectTV 20" dish, offset fed, f/d~0.7, to be used for 2.4 GHz wlan link.

Offset feed dishes have different f/D ratios depending on
polarization. I just did a quicky measurement of an RCA (Thomson)
dish which yields 0.67. Close enough.
http://www.qsl.net/n1bwt/app-5a.pdf

As I understand the issues, the best feed is one that -- for a given focal
length -- will "light" the dish with as spherical a wave as possible, that
has a broad enough dispersion angle to fully "paint" the dish without picking
up off-the-edge noise.

Exactly. See:
http://www.qsl.net/n1bwt/contents.htm
for how it works. Be sure to read chaper 6 on feeds.

I've looked at all types of feeds, from cantennas to bi-quads to patch to
dual-dipole. All have their pros and cons.

Biquad should work.
http://www.trevormarshall.com/biquad.htm
http://martybugs.net/wireless/biquad/
-3dB beamwidth of the biquad is about 60 degrees. Using a protractor
to eyeball my pizza dish, it wants a 70 degree angle. Close enough.

I'd like to hear from anyone who has done this before about what feed they
chose and why.

See:
http://www.weijand.nl/wifi/
Have fun aiming the dish.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

On Fri, 13 May 2005 12:05:45 -0700, DaveC wrote:

On Fri, 13 May 2005 10:53:14 -0700, Jeff Liebermann wrote
(in article ):

-3dB beamwidth of the biquad is about 60 degrees. Using a protractor
to eyeball my pizza dish, it wants a 70 degree angle. Close enough.

Thanks for those numbers, Jeff. I hadn't intuited, nor could I find
confirmation of which feed would best illuminate my dish.

Bi-quad, here I come!

This might help:
http://802.11junk.com/jeffl/pics/antennas/Biquad/
It's the 4NEC2 output for Trevor Marshall's biquad design. I'll tweak
it to include a VSWR over frequency graph later.




--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

On Wed, 18 May 2005 00:47:01 -0700, DaveC wrote:

On Sat, 14 May 2005 11:56:51 -0700, Jeff Liebermann wrote
(in article ):

This might help:
http://802.11junk.com/jeffl/pics/antennas/Biquad/
It's the 4NEC2 output for Trevor Marshall's biquad design. I'll tweak
it to include a VSWR over frequency graph later.

Not to look a gift horse in the mouth, but how about a plot of VSWR vs.
groundplane-to-element spacing?

If I did everything I promised to do, I'd never get anything done.

See:
http://802.11junk.com/jeffl/pics/ant...uad/index.html
I've added a VSWR and reflection coefficient graph, main page, and
NEC2 file.
http://802.11junk.com/jeffl/pics/ant.../biqdfeed2.nec

The 300MHz frequency used is handy because everything is in
wavelengths. In this case, to get 2400MHz, just multiply everything
by 8. For example, the 288-312MHz frequency plot is actually
2304-2496MHz. If you extract dimensions from the model, it's the same
story. Multiple all dimensions that are in wavelengths by 8 to get
something useable at 2400Mhz. If you want to use the design for a
cell phone antenna or other frequency, just use a different scaling
factor.

I think the NEC2 deck I posted is the one that came with the samples
in the 4NEC2 program. (Actually, I'm not sure where it came from.)
You can download the program and sample antennas from:
http://www.si-list.org/swindex2.html#4nec2_

The frequency plot was done by changing the FR card to:
FR 0 7 0 0 288 4
This results in the plots being slightly off the ideal 300MHz
frequency. Therefore, I posted the frequency plot and main page that
was generated with the above frequency scan. There are probably
better ways to do this, but with 9 minutes per compile, I just used
what popped into my foggy brain first.

For construction details and dimensions, see:
http://martybugs.net/wireless/biquad/
The antenna to reflector spacing isn't horribly critical. The above
construction article uses 15mm. I've been adjusting the height during
construction for maximum gain and/or best VSWR and ending up with
about 16mm (measured to center line of copper wire).



--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558

Dave

Have you considered using a USB wireless adapter as the feed so the line
loss might be lower?
I have built some illuminators for the off set feed satellite TV dish
using a USB wireles adapter. But, I dont yet have any definitive results.

Jerry



"DaveC" wrote in message
news.net...
DirectTV 20" dish, offset fed, f/d~0.7, to be used for 2.4 GHz wlan link.

As I understand the issues, the best feed is one that -- for a given focal
length -- will "light" the dish with as spherical a wave as possible, that
has a broad enough dispersion angle to fully "paint" the dish without
picking
up off-the-edge noise.

I've looked at all types of feeds, from cantennas to bi-quads to patch to
dual-dipole. All have their pros and cons.

I'd like to hear from anyone who has done this before about what feed they
chose and why.

Thanks,
--
Please, no "Go Google this" replies. I wouldn't
ask a question here if I hadn't done that already.

DaveC

This is an invalid return address
Please reply in the news group

On Wed, 18 May 2005 10:48:29 -0700, Jeff Liebermann
wrote:

If I did everything I promised to do, I'd never get anything done.

There's always time to do it over again after a rush job.

I found a better biquad model at:
http://pow.za.net
that compiles quickly, is much simpler, doesn't require wavelength
scaling, is vertically polarized, and is much simpler than what I
previously posted. I re-ran the 4NEC2 model and dumped the output at:
http://802.11junk.com/jeffl/pics/antennas/Biquad/
The NEC2 file is at:
http://802.11junk.com/jeffl/pics/ant...biquad.nec.txt
Gain is about 11dBi. -3dB beamwidth is 60 degrees horizontal and 50
degrees vertical.

You can download the 4NEC2 program and sample antennas from:
http://www.si-list.org/swindex2.html#4nec2_

For construction details and dimensions, see:
http://martybugs.net/wireless/biquad/


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 AE6KS 831-336-2558