Page 1 of 2

Antenna differences

PostPosted: Sun Jul 28, 2002 4:48 pm
by Thorn
Q: What are the differences between an Omni antenna, Panel antenna, Parabolic antenna, Sectors antenna, and the Yagi antenna's?

A: Antenna differences: An Omni antenna is omnidirectional. In other words, the signal is transmitted and received equally in all directions. (Ignoring such influences as trees, terrain, large buildings, etc.)

Panel antennae, Parabolic antennae, Sectors antennae, and the Yagi antennae are all directional antennae. They all have different applications, which is highly dependant on the particular setup.

For example: A Sector antenna is usually design to cover a wide pie shaped area. The width of the sector they cover typically ranges from 60° to 180°. Usually they are used to provide specific regional coverage for broadcast areas of WISPs (Wireless Internet Service Providers) or similar applications.

Posted by tommy8834
Answered by Thorn

deciBels (dB)

PostPosted: Tue Jul 30, 2002 12:09 pm
by Thorn
Q: What is all this talk about dB and deciBels. I thought radio power is expressed in Watts?
Q: Why are deciBels expressed in negative numbers in NetStumbler?
Q: Why doesn't SNR equal Signal/Noise?

A: Decibels (or dB) is a REFERENCE value only. You have to reference it to a known value such as power in watts, milliwatts or voltage in volts, millivolts and microvolts. For power, dB
is expressed as:

dB = 10 * log10(p)

Now with that in mind:

Transmitters, such as those used in commercial microwave transmitters are referenced to 1 milliwatt. So

0 dBm = 10 log10(1 mW)

therefore a transmitter of 1 watt would be a +30 dbm transmitter, a 1/2 watt transmitter would be +27 dbm or a 0 dbm transmitter would be 1 milliwatt.

0 dbm= 1 milliwatt
10 dbm= 10 milliwatts
20 dbm= 100 milliwatts
30 dbm= 1000 milliwats or 1 watt

Received signal strength reads negative numbers because of the free space and path loss of the signal; a negative number simply means the power is less than 1mW. Thus, a negative dB value represents a loss, and a positive one represents a gain.

The noise floor is the signal strength of the background noise, which is caused by man made devices and nature, such as the sun.

Your signal to noise ratio (SNR) is how high your signal is above that noise floor. So if your noise floor is -92dB and your signal is -82dB, then your SNR is 10. Note that the SNR is obtained by subtracting the noise value from the signal instead of dividing because the signal and noise are expressed in dB. Thus:

SNR = S(mW)/N(mW) = S(dBm) - N(dBm)

Questions posted by many people. Answered by mshimek
Edited for the FAQ by Thorn

Omnidirectional Patterns and dB

PostPosted: Tue Aug 06, 2002 4:48 am
by Thorn
Q: I've been told that omnidirectional antennae will increase the range as the dB increases, but that the pattern gets flatter, and I may miss detecting APs. What does this mean?

A:For those of you who are scratching their heads over this, this may help visualize it better. Think of it this way: Omnidirectional antenna patterns are doughnut shaped. As you increase the dB rating, the doughnut's volume doesn't change, it gets flatter. It squishes flat. (yech!)

Here are two views of antenna patterns, a 5dB and a 10dB, using the same theoretical tx/rx point. ( BTW, not to belabor the obvious, but someone will probably point his out if I don't: The diagrams are not to scale with each orther.)

The top diagram shows the patterns from above, while the lower diagram shows the pattern from the side.

The black is the 5dB pattern, the blue is the 10dB pattern, and the red are APs that someone is trying to find.

AP1. Not detected. Within range of 10dB, but below signal area. Out of range for the 5dB.
AP2. Detected by 5db. Within range of of both; but only within the the signal area of 5dB. The 10dB would not have detected it.
AP3. Not detected. Within range of 10dB, but above signal area. Out of the range of the 5dB.
AP4. Detected by 10dB. Barely within range.
AP5. Detected by either 5dB or 10dB.

Re: Omnidirectional Patterns and dB

PostPosted: Tue Aug 06, 2002 12:01 pm
by bwsaloum
Q: Please excuse my ignorance, but this diagram (which is a pretty good explaination) makes the assumption that the antennae are stationary, correct?

Reason I ask, if I were to mount a big (and presumably flat) omni on my jeep, I'd be able to catch AP's further out, on either side of my vehicle.

A: The antenna doesn't have to be stationary. It applies to mobiles as well. (The physics stays the same.) So yes, adding the higher gain omni will cut a wider swath as you wardrive. Just remember, you may see less APs that are above (or below) street level, because of the flatter pattern.

Asked by bwsaloum
Answered by Thorn

Multiple Antennae

PostPosted: Tue Aug 06, 2002 12:20 pm
by bwsaloum
Q: What about adding multiple antennae? An you run two omni's in tandem, or would the additional impedance screw up both?

A: Directly connecting two antennae into one lead will cause impedience problems. You'd need a splitter/combiner or switch for it to work right. That adds loss to the system, so you'd probably not gain too much.

NOTE: Since this was first answered, some reasonable cost and low loss splitter/combiners have appeared on the 2.4GHz consumer market. Links will be added here as time permints.

Asked by bwsaloum
Answered by Thorn

PostPosted: Tue Aug 06, 2002 12:29 pm
by lincomatic
one idea that i thought of but haven't tried is to take a high gain omni and tilt it at an angle (forwards or backwards in relation to the direction of travel). this way the off polarized AP's don't get penalized as much (but the correctly polarized ones do), and you don't have the problem w/ the flat pattern. but i guess you would need to make NS scan faster and not drive too fast....it's like an MRI...taking slices as you drive.

PostPosted: Tue Aug 06, 2002 4:02 pm
by lincomatic
Q: What does polarization mean?
Q: Is the radiation pattern of the horizontal simply flipped 90 degress as opposed to the vertical?
Q: Very few antennea seem to be horizontal polarized and tend to cost quite a bit more. Why are horizontal polarized antennae costlier?
Q: I built a homebrew antenna. How do I find its polarization?

A: Here is an answer from:
http://www.larsenantennas.com/technical/basics.html
-----------
Antenna Polarization
Polarization is defined as the orientation of the electric field of an electromagnetic wave. Polarization is in general described by an ellipse. Two often used special cases of elliptical polarization are linear polarization and circular polarization. The initial polarization of a radio wave is determined by the antenna that launches the waves into space. The environment through which the radio wave passes on its way from the transmit antenna to the receive antenna may cause a change in polarization.

With linear polarization the electric field vector stays in the same plane. In circular polarization the electric field vector appears to be rotating with circular motion about the direction of propagation, making one full turn for each RF cycle. The rotation may be right-hand or left-hand.

Choice of polarization is one of the design choices available to the RF system designer. For example, low frequency (< 1 MHz) vertically polarized radio waves propagate much more successfully near the earth than horizontally polarized radio waves, because horizontally polarized waves will be cancelled out by reflections from the earth. Mobile radio systems waves generally are vertically polarized. TV broadcasting has adopted horizontal polarization as a standard. This choice was made to maximize signal-to-noise ratios. At frequencies above 1 GHz, there is little basis for a choice of horizontal or vertical polarization, although in specific applications, there may be some possible advantage in one or the other. Circular polarization has also been found to be of advantage in some microwave radar applications to minimize the "clutter" echoes received from raindrops, in relation to the echoes from larger targets such as aircraft. Circular polarization can also be used to reduce multipath.
--------------

Basically, the radiation pattern is independent of the polarization. If you take 2 vertically polarized antennae and turn one on its side, you'll receive almost no signal between the two - this is called cross-polarization. If you're having trouble visualizing polarization, think of it in terms of light; it works the same way w/ other electromagnetic radiation such as WiFi. LCD panels work on the principle of polarization, by the way. the dark bits of the screen have the polarization 90 deg off.

The reason the vertically polarized omni antennae tend to be cheaper is that they are usually collinears, which are cheap and easy to make because basically, they are just plastic tubes with some wire inside. A slotted waveguide is an example of a horizontally polarized omni - they are big metal tubes which are machined at close tolerances, so you can see why the materials and labor cost more.

Asked by TheSovereign and rerunn
Answered by lincomatic

Antenna 101

PostPosted: Thu Aug 08, 2002 4:44 am
by Thorn
Q: Where can I learn more about antennae?

A: Tim Pozar, one of the founders of the Bay Area User Group has a good primer on antennae at: http://www.lns.com/papers/BAWUG-antenna101/ . It is in the form of a slide show, as it was originally a PowerPoint presentation.

PostPosted: Sat Oct 05, 2002 7:02 pm
by Thorn
Q:I'm trying to build high-gain yagi/directional antenna. I've looked at a lot of plans (including pringles), but I'm not certain of what to build. What would you experienced stumblers recommend?

A: You may want to check here:
NetStumbler.com Forums > FAQ's > Antennas > Directional Antenna Links
http://forums.netstumbler.com/showthread.php?s=&threadid=2750

as well as:

NetStumbler.com Forums > Wireless Forums > Hardware > Post Your Homebrew Antennas Here
http://forums.netstumbler.com/showthread.php?s=&threadid=2034

Member gerlag also recommends building a circular polarized antenna. Such antennas have high gain and are less sentitive to reflections from buildings etc. Thus reduces delay spread, the link error rate will be better.

Posted by ziral
Answered by Blackwave and gerlag

*With additional thanks to Blackwave for catching the cut and paste problem on the original.*

PostPosted: Fri Dec 13, 2002 6:13 am
by Chris
Q: Does anyone know of a good reference for calculating the antenna strength for a point to point setup between two line of sight distances?

A: YDI has some easy to use web-based calculators.

Asked by Chris.
Answered by Thorn.

Re: Omnidirectional Patterns and dB

PostPosted: Sun Dec 15, 2002 8:19 am
by dax
Q: How do I determine what the thickness of the volume is? Is that a pre-determined measure based on db rating? It seems that a omni would be at least like a dome from left extreme to right extreme, but how tall?

A: You don't measure this in physical units (like feet, meters or kilometers), it is measured in dB. Specifically, the -3db measurement is important, as that is the halfway point in measuring the signal strength.

In order to "see" the pattern, you should look at the manufacturer's diagrams "e" and "h" plane diagrams. This will show the most information.

The signal from an omni doesn't look like a dome, as the signal comes from along the length of the wire. As you move to a point in space where the wire appears head on (a little round dot) the signal drops off to zero, as little or no signal comes off the end of the wire. Hence, you get a donut shape.

A sufficiently sensitive receiver, moving closing around a powerful transmitter's omni, might show a dome pattern, but for most real world RF applications the doughnut pattern is a good working model.

Asked by dax
Answered by Thorn

Antenna Tower / Airport Questions

PostPosted: Wed Jan 22, 2003 8:51 am
by paran0id
Q:My shop is on a rather high hill in my town with a decent line of sight to a lot of homes. I would like to start providing 802.11 access to as many people as I can, however and I know that the higher I can mount an antenna(s), the more potential people I can service. When I asked my local building inspector about how high I could mount a tower on the roof of our building the answer was, more or less, show us your plans and well tell you if we approve it. My question is what can I expect to get away with? What is the average overall hight of an antenna in a buisness district? (considering how tall the building is allready).
Also, another factor for me is that there is an airport runway about 300 yards up the hill from me. Does anyone have any experience installing antennas near an airport? Will I have to have it approved by the FAA, or am I ok because it is in the public ISM band?
Also can anyone reccomend a company that makes antenna towers suitable for mounting on a flat rooftop?


A: Usually, tower heights are specified someplace in the local zoning ordinances. That kind of "show us first" line is BS. Hell, if that's they make a decision individually, start off at 1000' and whittle them down to 500'! :)

They have to have some guidelines someplace. I'd suggest talking to Planning/Zoning Dept.

Towers are available from a variety of sources. Electrocomm and YDI both carry a number of lines. For a small one, I like the Rohn 25 and 45 models.

If you're that close to an airport, most likely the FAA will be involved. Band and licensing has nothing to do with it. Normally, unless you go over 200' there's no requirement to file with the FCC and FAA. However, there is a requirement to file with the FAA is you are within the confines of the controlled space, which varies according to airport size and air traffic, if I recall correctly. There is also a glide path ratio (7:1, I think.) Anyway here's some links for the FCC and FAA regs:

http://www2.faa.gov/ats/ata/ata400/oeaaa.html

http://wireless.fcc.gov/antenna/documentation/

There should also be an FAA office at the airport, and they can advise you. I haven't filed one in years, but hear the FAA currently has a three to four week turn around on approvals/denials.

Get it ALL in writing.

BTW, "higher is better" isn't always true. With these freqs, you can actually get an umbrella effect where you get under the antenna and cannot get the signal. If you are going to do this as a WISP, then you might want to get some proper help in planning it.
Good luck.

Q&A How do I switch off antenna diversity

PostPosted: Wed Jan 22, 2003 1:42 pm
by MarcusJClifford
Question: Several access points come with 2 antennae, I want to replace one of them with a higher gain directional / omni / patch antenna. How do I do this?

Answer:
Most access points that come with 2 antennae are setup to use a system called diversity. If you are going to use only one antenna then you must switch off diversity and specify which antenna to use.

For access points based on the Amtel Chipset (Linksys 1.1 / Belkin / D-Link) then you must use the Linksys USB Utility to set the setting on the advanced tab as in the screen below.

Other Access points may have this setting elsewhere, please post below for specific access points

Thorn Question on Omnidirectional Patterns and dB

PostPosted: Mon Jan 27, 2003 4:21 pm
by L8Nighter
Thorn thanks for the image in the FAQ to visualize how you increase db the pattern gets flatter.

Now my question: Say one has a 5db mag mount on the roof of their car. In the manual of most cards they will get Up to 1500 ft outside. How high in elevation does that go, im confused on height vs distance. Don't worry about stuff getting in the way I just want to get the basic logic down. Does that mean 1500 ft distance and 1500 ft up into the air? And if you put a 10db antenna on your car that looks like the elevation of the signal would go down but your distance you cover would go up?

Is there any formula to calcuate height vs distance?

So with your 5db antenna driving around in the city, are you going to be able to reach access points that are 50 some stories in the air or would you have to tilt your antenna a certain way.

Thanks if you can clear any of that up for me.


A:Thanks, I'm glad the diagrams helped. I can't take too much credit for them though; they're hardly unique. Most books on rudimentary RF stuff have better ones.

Under perfect (theoretical) conditions, the signal would originate from a single point in space, and radiate outward in all directions, giving a sphere of signal. So -in theory- the 1500' would reach in all directions (BTW, take that 1500' with a grain of salt. Hell, with most manufacturers, take a whole salt shaker...)

In reality, the signal orginates along the length of the antenna, hence the doughnut shape, and that's for an omni. Cards with a built in antenna, would probably have a pattern closer to a wedge.

As to formulas to calculate height vs. distance, I'm sure are some, someplace, but I don't have them. They would most likely be highly theoretical. Check the ARRL antenna book. In reality, the problem with what you are asking is it is highly dependant on the strength of the transmitted signal, which is an unknown. And remember, we are talking 2 two-way radios here: NS transmits a beacon request, which the AP must receive. The AP transmits the response, which your stumbling rig must receive. Two transmitters, two receivers. In addition to the signals, noise levels between them will also affect the whole equation.

Since dB is logarithmic, replacing a 5dB antenna with a 10dB antenna should give a distance almost 4 times the diameter of the 5dB unit, at a corresponding decrease in the height. But it is still dependant on the signal transmitted to you.

In actual wardrivng, an AP 50 stories up is probably effectively shielded by the building. Using a directional, you should be able to scan a building window by window that high up, but just driving by would not pick it up in all likelihood.

PostPosted: Fri Jan 31, 2003 5:34 pm
by Thorn
Q: How do you measure the gain? I would like to visually see the difference of my antenna vs. plain card.

A: Below is the quick and dirty way. Be advised this is very rough and very prone to error.
1) Get the best (strongest) dB level reading from a known AP using the card. This X dB. (eg -58dB)

2) Don't change the location or orientation. Plug in the antenna, get a new reading. This is Y dB. (eg -50dB)

Y + X = Antenna gain in dB. (Addition because these are negative values.)

-58dB + -50dB = 8dB.

In this example the gain would be 8dB.

Asked by hevnsnt and several others.
Answered by Thorn