# Directional vs Omni question

• posted

In THEORY would a 6 dBi directional antenna send a signal at the same strength in a given direction as a 6 dBi omni directional antenna? If dBi is the same would the only advantage to a directional be to remove noise for receiving?

Pardon my question into basic theory but I have done lots of reading just because I find it entertaining and still don't know understand this area well.

• posted

"ahh" hath wroth:

Yes. 6dB of gain is exactly the same no mattery what is used to create it. However, that's only at the maximum gain point. As soon as you get off axis or away from the maximum gain point, the omni and the directional antennas work very differently.

That's really the big advantage. Sources of interference that are NOT along the maximum gain axis of a directional antenna rapidly disappear with a directional antenna. However, it's not perfect. If you're unlucky enough to have a source of interference that is along the line of sight axis, the directional gain antenna will actually make the interference worse. It's not unusual for someone to install a big

24dBi dish antenna, and suddenly discover that another wireless network that's located miles away, along the line of sight, is now a source of interference.

Other benifits of directional antennas are:

1. Larger vertical radiation angle for directional versus omni (for the same gain) which makes keeping the antenna vertical less of a headache.
2. Directional antennas also reduce the effects of reflections, which can be a serious problem in some highly reflective locations.
3. Directional antennas are much easier to mount.

There's also a big problem with end fed vertical collinear omni antennas, which are what most omnis are inside. The patterns shown in the data sheets look like a squashed donut with the major axis exactly

90 degrees (perpendicular) to the antenna. If such an antenna actually worked like that, it would be idea with the bulk of the signal pointed at the horizon. However, that's not reality. Those patterns are the free space antenna pattern and totally ignore the effects of the mounting structure or any metal directly under the antenna. Both of these tend to make the squashed donut move upward causing what's called uptilt. Typical is about 3-5 degrees. That may not seem like much but when combined with an antenna that perhaps has only about 11 degrees of beamwidth, it means that at least half the signal is going towards the sky instead of downward towards the users. It's especially a problem with rooftop and mountain top systems where talking to airplanes is not the intent. There are high gain omni's that don't have intentional downltilt that are a considerable help.

Directional antennas generally don't have this problem with one exception. Sector antennas are designed to have a very broad horizontal pattern, and a narrow vertical pattern. The idea is to cover as many square miles of area as possible. This is done by trading vertical coverage for horizontal. The result is that the vertical radiation pattern of a sector antenna is quite narrow. Pointed exactly horizontally, it will send the signal over everyone's head, which is not exactly desireable. So, sector antenna always have a mounting contrivance that allows for downtilt.

When I was much younger, I always wanted a magic wand. Antennas are magic and will suffice.

• posted

I'm still trying to figure out how to point/tilt/aim/angle the two big honkin 9dB booster antennas I put on the back of the WRT44G router, or the standard antennas, for that matter. Do they point vertically, horizontally, or somewhere in between? I

s the max signal boost at right angles to the two antennas (back of router box pointing towards maximum signal lobe), or parallel to the plane of the antennas(side of router box pointing towards maximum signal lobe)?

I can't seem to find any literature on antenna positioining - it's not mentioned anywhere in the Linksys router manual.

Sheesh, they give you all these rotation options, but not a clue as to where to point them. Are they omnis, or dipoles? If omni, why two of them?

• posted

On Fri, 07 Jul 2006 15:18:27 -0500, Harvey Gerst wrote in :

Max signal from a "stick" antenna is at right angles to the axis of the stick.

See

They expect you to make them vertical, like in the pictures, which optimizes horizontal coverage, the usually desired result.

Omni (horizontal).

Diversity (space). See

• posted

I couldn't find the "Big Honkin" company or a WRT44G. I think you might mean one of the aftermarket Linksys omni antennas. Perhaps this one? |

Well, here a differ from orthodoxy. Somewhere in the docs, it says to point both antennas straight up. That gives the maximum gain horizontally (perpendicular to the antennas). Only one antenna is "on" at a time, so it really doesn't matter (much) which way the WRT54G is pointed. I suppose there might be a tiny bit more signal towards the back of the router because the router itself is not blocking the signal.

However, most houses are not built flat. They have upstairs, downstairs, and basements. If you need to cover these areas, what I recommend is to point both antennas to form a "V" or about 45 degrees from vertical. The idea is point the antennas up and down somewhat and also make it somewhat polarization insensitive. Unfortunately, there's no free lunch, so you're going to lose some gain in the horizontal plane.

I think there was some documentation on the back of the antenna box.

They're omnis. You want two of them for diversity reception. The problem is called frequency selective fadeing, which I don't wanna expound on right now. Basically, having two antennas improves the probability that at least one of them is going to receive the signal. The router picks the antenna with the best signal.

• posted

Are you familiar with a common balloon? Think of one that is so strong, that it won't break no matter how to squeeze or pinch it.

Fill that balloon with water so that is two units in diameter (inches, cm, feet, meters, miles, whatever) - nice and round, and every point on the surface is one unit from the exact center. That is an isotopic balloon - the 'i' in dBi. Now, squeeze the balloon from the top and bottom. Notice that as you squeeze in there, the surface of the balloon pushes _out_ all around the sides. This is the same as an "omni" antenna with gain. It gets that gain by reducing the signal out the top and bottom, and adding that signal to go further out the sides.

Now, squeeze the top and bottom, and both sides. Notice how the balloon extends further in two directions (call it "front" and "back") as you do so. That's a "bi-directional antenna. Now, while you continue to squeeze the top and bottom, and both sides, squeeze the back as well. Notice how the front extends even more - a uni-directional antenna.

Some people also use the idea of a ball of "Play-Doh" (a clay like soft plastic sold as a toy). You can squeeze that material into a ball of two units in diameter. You can also take a board, and flatten that ball into a disk - the disk will be thin, but much larger in diameter. Same idea.

An antenna produces the same gain in transmit and receive. It does this by not transmitting (or receiving) in "that" direction over there, but increases the signal in "this" direction. Think about a flashlight bulb - the bulb in a 2xD cell (NEDA 13F) is rated at about 0.8 candela (close enough to the light output of a candle) or 1.2 Watts. Compared to a common "night light" which is rated at 4 or 7 Watts, it shouldn't be very bright. In fact, the night light produces about 3.3 or 6 times as much light. Compare that to the light of a single birthday candle. But you notice that the flashlight is "directional". Off to the side or behind, there isn't that much light. But when the flashlight points directly at you, it's often as bright as a much larger lamp. The flashlight concentrates the light in one direction (directional) while the night light spreads the light in all directions (nearly isotopic).

Old guy

• posted

Good analogy. However, the last time I used it in front of a knowledgeable audience, some clown ask "Does it extrude linearly or logarithmically like in the antenna plot"? Oops. I instantly lost the audience when I stated explaining why antenna plots use logarithmic scales.

• posted

Probably not completey relevant, but I've been studying the demod output of trunk radio systems and noticed that the signal quality with directional antennas is better, even at similar gain. You can see multipath in the digital signal. If there isn't too much, the effects of multipath will get removed in the slicing, but for noisy signals, directional antennas do have an advantage.

With wifi, the higher frequency signals should have less multipath than a UHF trunk system, but still, I think there would be advantages with a directional antenna.

I've been thinking about this in relationship to war driving, where a figure 8 pattern with the gain lobes firing out the sides of the car and the nulls front and back. Should be possible with two verticals spaced a half wave apart according to the ARRL antenna book.

• posted

You notice I'm using a water filled balloon - that was after someone came back when things didn't work as well as they thought using an air filled one.

With a water filled one, it's linear. If your want the logarithmic expansion, you have to add a tablet of metallic sodium to the water - the tablet size relates to the size of the balloon of course. Do be careful.

That's only a mental convenience to allow you to see more details. In these days of pocket calcula^Wcomputers, people forget that adding and subtracting is easier to do in your head than multiplying/dividing. The HP 8552B IF section on the spectrum analyzer offered both log and linear presentations. I don't think I ever used linear even though the display graticule had 70.7%, 50% and 10% markings in addition to 2 or 10 dB per cm.

Smoke and mirrors? Wazzat? "Ignore the man behind the curtain..."

Old guy

• posted

Using water also adds considerable suspense to public demonstrations. I've noticed that the front rows of the audience tends to empty as I squeeze the balloon.

Metallic sodium in water creates lots of hydrogen gas and sufficient heat to explode the hydrogen in air. I have some of the stuff if you want to Learn By Destroying(tm).

I know the feeling. In 1970, I bought a TI SR-10 calculator in college because I couldn't afford an HP-35. Anyway, my checkbook and bank statements haven't balance since then. Even with the help of various arithmetic accelerators, I still can't add and subtract.

Incidentally, I bought two K&E slide rules last weekend for \$10 in perfect condition. Don't ask me why.

I have one of those in an HP140T main framis. It really impresses the visitors. Ummm... my graticule isn't stock so it's missing the NTSC references and linear confusion. I have used the linear scale to do IF and xtal filter tuning. It's way more sensitive and I needed to see ripple down to fractions of a decibel.

Another great antenna demonstration is to rotate the antenna while measuring the detected signal on a polar display. The display shows the antenna pattern (when run through a log amp). However, what was normally suppose to be an outdoor demonstration turned into an indoor demonstration because the parking lot had just been re-paved. I also managed to mismount the 5 element VHF yagi off the center of gravity. When the rotator motor got up to speed, the assembly took off in the general direction of the audience. It was almost as good at clearing the front rows as the water balloon antenna pattern demo.

• posted

How big are these balloons you're squeezing?

Where do you think that logarithmic expansion comes from ;-)

The SR-10 rings a bell, I think I had one as well. I also bought a TI Programmer Calculator, which was somewhat similar to the TI-30 except that it did octal and hex. I've NO idea where it disappeared to, but the battery pack had finally died and I was running it off a NEDA-216 9 volt taped into the battery compartment. The TI's I have handy are the later version of the TI-30 with LCD display rather than the LEDs.

Looked through some interesting boxes, but didn't find the Programmer - did find a brand new L2B flight jacket still in the factory box, and a flight suit in real need of the washing machine though.

Crude order of magnitude stuff - accurate to 1.5 digits - I still do in the head because it's faster than finding where the calculator is hiding. Of course having access to a computer

[compton ~]\$ whatis bc bc (1) - An arbitrary precision calculator language [compton ~]\$ echo "2^156/47" | bc -l 1943486219854924093355963873292929547414803913.53191489361702127659 [compton ~]\$

Try that on your SR10 ;-)

WANT!!!

Most of the stuff I worked with wasn't stagger-tuned, and I tended to be more interested in the spurs and harmonics. The few projects I worked on where ripple was a factor were normally tuned using fix frequencies - set the source to X, tune L3 for a peak, set the source to Y, tune L5 to peak, and so on.

I co-authored a paper on the modulation effects observed at an antenna behind an aircraft propeller, and it was decided by someone high in the chain of command that a movie version would be beneficial. Being rather brain-dead, we set up a transmitter on the top of a hanger, then put a receiving setup (and the camera man) a few feet behind the wing of a Convair CV-340 parked on the ramp (R2800-CB17, spinning a three bladed prop 13 foot 6 inch in diameter - the blade chord varied from 10 to 18 inches). Of course, we need to vary the speed of the prop from idle (about 350 RPM) up to nearly full speed (about 1500 RPM at the prop, 2700 RPM on the engine). We're giving hand signals to the crew chief, who is standing 50 feet in front of aircraft where the pilot can see his hand signals. We had our gear tied down. The camera man? What's that song from the 50's Broadway show "Peter Pan"? I'm Flying!

Old guy

• posted

snipped-for-privacy@painkiller.example.tld (Moe Trin) hath wroth:

Not too huge. About 12" in diameter which should be about a quart of water.

It had a square root key, which I thought was really cool. I think I paid about \$120 for it. The HP-35 was about \$350, which was more than I could afford.

I have one of those. TI SR-22 circa 1973. Does Hex, Octal, Binary, etc:

Yuck. Calculator collectors (like me) prefer LED's over LCD's. Part of my HP collection. I'll update the photos one of these daze:

have about twice as many now.

Wash it. That will let you claim that it shrank when you discover you no longer fit into it.

No way. I actually use a slide rule sometimes. Usually to irritate someone while I claim to check their calculations.

you can use the online simulation:
the other hand, if you want to support my decadent and lavish lifestyle with a substantial "contribution", I could probably be convinced to part with one of them. (Not the 6" K&E. That one sqeezed me through kollege).

No problem. Just turn the input attenuator to 0dB and the front end will create all the spurs and harmonics you've ever wanted.

We just stuck a clear acetate overlay on the screen and scribbled the correct response on the overlay. The tuning procedure was something like "twiddle the coils and caps to make it look like this...". It worked.

You're lucky you didn't kill him with the debris from the hanger floor.

Reminds me of the idiot that taxied to the run up area but rotated the plane in the wrong direction. Instead of kicking all the rocks into the wall, he managed to launch them into the parking lot. I think he totaled about 5 car windows and dinged another 10 paint jobs.

• posted

hmmm... 4 pi r cubed / 3 231 cubes to the gallon...

Where can I buy gasoline (or better, Scotch) in those quarts?

I don't remember, but it _wasn't_ cheap. The HP-35 sounds about right. My boss had one, and it was secured to his desk in a locked cradle with a security cable.

The advantage of the readability is offset by the battery life. Last year, I got a bee in the bonnet, and replaced the batteries in all of the calculators. Several had notations that they had last been replaced in the late 1990s. And I also just found the manual to my (1974) SR-50 "Slide Rule Calculator" - which was LEDs. No idea where that one is either.

Nomex doesn't shrink (or at least not significantly). I did wash it, but decided it needed more help. Dry cleaner wants \$29 plus tax. Sigh... My cardiologist has me on a regular exercise program riding a bike for 20 minutes a day. This is really fun in the summer here (ride early or very late - carry large water container).

Why not. I occasionally do it to baffle the young'uns

I think that's how you toasted that 435 a while ago. No, I usually had requirements to have the junk down by a minimum 30 dB - which is fun when you're at C or X band and the source is being multiplied up with a SRD.

I presume using a sweep-gen. I was dealing with "factory" procedures, and setting up with CW sources at discrete frequencies seemed to be quickest.

No, we were out on the ramp, with the bird facing the hanger which was at least a 100 yards away. The camera got dinged, but surprisingly I don't think we damaged the lenses.

Ouch! I'm trying to think of an airport I've been in were cars were allowed near the bird paths (other than parking ramp/hangers). But then I also stay on paved surfaces, and complain loudly if there is debris on the ramp. Approved mechanics cost money to file out nicks in a prop. While GenAv birds typically don't have that much of a wake (heck, we used to stand 50 feet behind a T-33 to warm up on cold mornings), larger birds - well, if you are familiar with KSFO, you're supposed to use minimum power to taxi the first 600 feet of runway 1R before applying full power and that's even with the blast fence protecting the frontage road (and vehicles on the Bayshore) nearly 1000 feet away. One of my (many) FAA guides talks about the breeze behind an early 747 (with something like 45k of thrust per engine - small compared to a 777-300ER with 115k per engine). You small guys better stay back a bit, ya hear.

Old guy

Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.