Page 4 of 4

PostPosted: Mon Jan 30, 2006 12:53 pm
by Airstreamer
phlogiston wrote:Thanks. That makes sense.
I'm sure that this question could be answered on paper or with a simulation, but I wouldn't know where to begin. Perhaps I should have paid more attention in math class:) I guess I'll just have to make one of both with the same diameter, but very different shapes and see what happens. Next paycheck, that is.

Actually, I think you wind up with a sharper focal point, depending on the accuracy of the construction. (Spherical gives kind of a 'fuzzy ball of energy' type focus, and parabolic gives the sharpest.) I've also seen commercial microwave dishes with "blinders" along sections of the rim to limit interference, and I've also seen a short cylindrical tube section stuck on the front.

The best way to feed a dish is with a feed horn pointed back at the dish. With an antenna just stuck in front, part of the energy is radiated out directly from the antenna, and will be out of phase with the reflected signal, causing areas of lower signal levels due to cancellation zones.

PostPosted: Mon Jan 30, 2006 7:19 pm
by Scruge
Assuming the deep and shallow dishes are both parabolic, I would venture to guess that most designs lean towards the shallow depth because it uses less material and the angles involved are more forgiving.

As Airstreamer pointed out, one can provide less costly shielding techniques to the feed horn to eliminate or reduce signals approaching from acute angles.

PostPosted: Tue Jan 31, 2006 12:32 am
by phlogiston
Well, much to the dismay of my wife, I spend my entire day off building antennas instead of doing something useful. I built 3 antennas all exactly 40 inches across and 12 inches tall yielding an ideal 20 dbi of gain. The first was 25 inches deep, the second 10 inches and the third 5 inches. All were built from foamboard and Al tape (recycled as I went as I'm not made of money). I used the same Linksys wusb54g placed in each antenna's focal point.
The 25 inch one was a dismal failure, which was even more annoying as it took forever to tape down almost 6 square feet of that Al tape.
The 5 inch model worked very well, but the 10 inch antenna worked the best by a small margin. All were verified to be very close to a true parabola as I used a laser level to double-check the math. As the 10 inch deep antenna seemed to work the best, I finished it off. (almost)

http://img57.imageshack.us/img57/5962/test25eh.jpg

This weekend I plan on feeding the thing with a cantanna to see what that does, as suggested by Airstreamer. I can't yet say what the gain of this ... abomination is but I was able to connect to the target AP with no problem. (and can see almost 40 other AP's verses the 2 without the reflector)

PostPosted: Tue Jan 31, 2006 4:47 am
by Thorn
Nice work. Now go read the Rules, paying attention to the one about images, size and links. Both of the prior threads have been edited, but my patience wears thin after a couple of these.

PostPosted: Wed Feb 01, 2006 12:12 am
by phlogiston
Sorry about the links. I understand now.

So, tell me if this makes sense:
The calculated focal point from a parabola assume that the incoming waves arrive perpendicular to opening of the dish. That's middle school math. Well, it dawned on me that this is not the case when you are gathering radiated waves from an (effectively) single point such as an antenna. With my relatively deep reflector I noticed that the signal from near APs was less then that of far ones. That had me stumped until I tried moving the receiver back and forth and found that the focal point was a little different for APs at different distances. After some practice, I was able to mark focal points for different distances and either predict, very roughly, the distance of an unknown, or set it for best reception of a known. This effect will be the same for different shaped reflectors, but seems to be more pronounced when the shape of the parabola situates the focal point even with the opening of the dish. If my math is correct, then having a very shallow antenna (like almost all of those you find on the market) would be best because the focal point wouldn't "move" as much.

Which makes me think that this is already common knowledge. Man, I hate being the last guy in the room to get something.

I wonder if there is any use to being able to measure the distance of an unknown AP without GPS.

PostPosted: Wed Feb 01, 2006 7:29 am
by Thorn
phlogiston wrote:Sorry about the links. I understand now.
No problem.

phlogiston wrote:So, tell me if this makes sense:
The calculated focal point from a parabola assume that the incoming waves arrive perpendicular to opening of the dish. That's middle school math. Well, it dawned on me that this is not the case when you are gathering radiated waves from an (effectively) single point such as an antenna. With my relatively deep reflector I noticed that the signal from near APs was less then that of far ones. That had me stumped until I tried moving the receiver back and forth and found that the focal point was a little different for APs at different distances. After some practice, I was able to mark focal points for different distances and either predict, very roughly, the distance of an unknown, or set it for best reception of a known. This effect will be the same for different shaped reflectors, but seems to be more pronounced when the shape of the parabola situates the focal point even with the opening of the dish. If my math is correct, then having a very shallow antenna (like almost all of those you find on the market) would be best because the focal point wouldn't "move" as much.

Which makes me think that this is already common knowledge. Man, I hate being the last guy in the room to get something.
It makes sense, although I wouldn't say it is common knowledge within the WiFi community. It probably is common knowledge with designers of satellite communications, Point-to-Point microwave links, or similar communications gear.

As an example, look at satellite TV reception dishes: They are very shallow, and are designed to receive a signal anywhere on a given continent with no tuning, except for alignement. "More forgiving," as Scruge put it.

The opposite would be PtP microwave links on towers: Deep dishes, with a very tight beam. Some even have a short waveguide on the edge of the dish, as mentioned by Airstreamer. They tend to look like a hemisphere attached to a pipe of the same diameter, with the length of the pipe being about the depth of the hemisphere.

phlogiston wrote:I wonder if there is any use to being able to measure the distance of an unknown AP without GPS.
Distance measuring (with or without GPS) without triangulating is something that gets discussed ever couple of weeks.

You can't do it for two reasons: You don't know the power being transmitted, and you don't know what materials and objects are between you and the AP that are attenuting the signal.

In a purely theoretical situation, where you know the power level of the transmitter and where the signal is only lost in free space, then yes, you could measure the distance. It's actually pretty trivial, as it's the same formula used to compute the free space signal loss.

PostPosted: Wed Feb 01, 2006 12:41 pm
by phlogiston
Thorn wrote:
You can't do it for two reasons: You don't know the power being transmitted, and you don't know what materials and objects are between you and the AP that are attenuting the signal.

In a purely theoretical situation, where you know the power level of the transmitter and where the signal is only lost in free space, then yes, you could measure the distance. It's actually pretty trivial, as it's the same formula used to compute the free space signal loss.


I totally agree with you that there is no way to accurately calculate the distance of an AP in a real-life situation with so many unknown variables by measuring signal strength.
I wonder, however, if one could at least determine with reasonable accuracy weather the AP is "pretty close" or "really far" - independent of signal strength. With the simple rig based on a really wide parabola I've been playing with, I think that I could measure the distance of of an AP up to perhaps 50 meters in 10 meter units. Now, I don't yet understand many (most?) of the principles behind how this works, so it could just be a fluke.
In my understanding, when the radiation from a small antenna is focused by a parabolic reflector, the focal point would be slightly different depending on the distance of the antenna. In my case, an antenna 10 meters away focuses almost 50 cm closer to the ... apex of the reflector and an antenna 1000 meters away focuses very close to the calculated focal point. I imagine that this same principle allows the brain to determine depth - by analyzing the difference of angles of incoming light to each eye.
Now, this rig is definitely unwieldy and inconvenient, and I have yet to think of a situation in which this would be useful information. Also, we know the range of power most common APs transmit so there is an easier way to accomplish the same goal. Perhaps somebody else may be able to duplicate my results and even find a use for this.
I think this would make a totally awsome science fair experiment.

PostPosted: Wed Feb 01, 2006 12:54 pm
by saltman
Maybe you should look into a free to air dish instead of the dish 500. http://www.pansatusa.com/product/antenna.htm

PostPosted: Wed Feb 01, 2006 1:30 pm
by Scruge
phlogiston wrote: In my case, an antenna 10 meters away focuses almost 50 cm closer to the ... apex of the reflector and an antenna 1000 meters away focuses very close to the calculated focal point.


Unless I'm missing something I think what you are seeing isn't what one would expect from a perfect parabola. I think you are seeing hot zones brought about by imperfections in your construction.
In theory all source wave/signal should arrive at the focal point in phase having traveled the same distance, irregardless of distance from source.
Now on the other hand there may very well be some minimum distance between the source and dish that you begin to see gain diminish.

PostPosted: Wed Feb 01, 2006 1:43 pm
by phlogiston
Scruge wrote:Unless I'm missing something I think what you are seeing isn't what one would expect from a perfect parabola. I think you are seeing hot zones brought about by imperfections in your construction.
In theory all source wave/signal should arrive at the focal point in phase having traveled the same distance, irregardless of distance from source.
Now on the other hand there may very well be some minimum distance between the source and dish that you begin to see gain diminish.


Yes. That could very likely be the case. I should make another to see if I get the same result.
Thanks for the suggestion.

PostPosted: Fri Feb 10, 2006 9:03 am
by randomr8
Here is a link showing forumla calculation for using a wok scoop. This jpeg also mentions "lossless" long runs with USB cables.
http://www.usbwifi.orcon.net.nz/usbscoop.jpg

PostPosted: Fri Feb 10, 2006 9:06 am
by Mark57
randomr8 wrote:Here is a link showing forumla calculation for using a wok scoop. This jpeg also mentions "lossless" long runs with USB cables.


You might want to edit you post to include the link.

PostPosted: Fri Feb 10, 2006 10:23 am
by randomr8
http://www.usbwifi.orcon.net.nz/usbscoop.jpg

I apologize. Here is the link.

Re: Parabolic Dish Templates made easy

PostPosted: Mon May 07, 2012 1:15 am
by burning_wi_fi
You're a fun bunch. FWIW, just getting a Ku band satellite dish and mounting a small wi-fi antenna at the feed is about 500% +/- 2% easier than making a parabolic from templates. Want to get a distant access point with a big wi fi dish antenna? Ditch the template and get a cheap, accurate, mass produced dish and hack it for wi fi.