Wifi Aerials: Which direction should they aim?

[Zak33]

Lo dudes

I've been wondering: what's the best way to "aim" wifi aerials on my router?

I googled "Draytek Wifi Aerial" and got this
http://www.draytek.co.uk/products/aerials.html

and I quote

One, Two or Three ?

Vigor routers have one, two or three aerial connectors. Having more than one gives the advantage of 'diversity' which means (put simply) that a signal subject to some local interference in one particular plane, may be better on the other aerial. On a Vigor router with two aerials, the second aerial is for receive only. The main aerial both receives and transmits. Therefore, if you are fitting a replacement aerial, it it not necessary to replace both or have two fitted. If you do have two fitted, however, they should be the same. On 802.11n 'MIMO' Vigor routers, there are three aerials. The outer two are Transmit/Receive and the centre one is Receive only.

so.. my aerials do different jobs (I've got 3) but I still don't know how to aim them.

I've got OMNI directional....

so reading further on....

An Omnidirectional angle transmits its beam in all horizontal or vertical directionals equally (360°).

but their chart shows

HPBW 360° Horizontal 40° Vertical

so I googled HPBW and got a Beamwidth website.. and gave up on paragraph one.

So... my question: DO I AIM THEM? in my case the router is upstairs and I need coverage on both floors... so do I point one of the other ones DOWN and one out sideways?

Ignorin the "try it to see what's best" option because it seems to fluctuate.... what is the THEORY to begin with?

[dave87]

Theory? If it works, then leave it be?


Actually, where would the fun be in that. In my case all three aerials are exactly as your picture, vertical, and I get pretty good coverage across the entirety of the house (weather hasn't been warm enough to try the garden ).

[Zak33]

so yours point UP.... and is the router flat?

Mine is hung on a wall..... sideways... so the aerials point down in a line. Might need to move the router position.

[dave87]

Yup, router is flat on the side with the aerials pointing vertically.

I did ponder the benefit of having the outer two antennas at an angle, but never tried it.

[MaddAussie]

Zak

Omni-directional aerials (as I seem to remember) direct their signal in all planes apart from directly above/below the aerial think of it as a large doughnut shape.

This may help as well
http://www.ekahau.com/products/heatmapper/overview.html

[Zak33]

Zak

Omni-directional aerials (as I seem to remember) direct their signal in all planes apart from directly above/below the aerial think of it as a large doughnut shape.

This may help as well
http://www.ekahau.com/products/heatmapper/overview.html

that looks ACE

thanks chap

[Saracen]

The best advice I can give on this question, Zak, is to directly quote Draytek ....

You can experiment for best performance, particularly with regard to aerial position and angle.

That, exactly.

Or to put it another way, what works, works.

Every given location is different, which is why we do site surveys before installing radio gear for commercial use. Even in a home, you can find that a relatively small change of position of the router can make a big difference .... especially if you currently have a dense object (radio-dense, I mean, like a metal metal slab of a boiler, freezer, etc) in-between where you are and the router you're talking to.

I remember talking to a friend about 6 or 7 miles away, on a radio, while parked on another friend's drive. Signal was perfect. I moved about 6 feet forward (pulling off the drive) and lost the signal entirely. Move another 3 or 5 feet and the signal was back to perfect. And that's with a FAR higher power level, over several miles. So .... a small but perfectly placed obstruction can be a major headache.

Therefore, experiment. Try the aerials one way, transfer a large file and monitor. Then adjust aerials, and repeat.

You'll soon find out if the aerial positioning makes any difference to you, in your circumstances, and if it does, how much. Forget the theory and rely on .... whatever works, works.

[Mama Sumae]

Well imagine a dipole antenna (most common/basic example of a omnidirectional antenna) is vertically polarized (perpendicular to the ground) the radiation pattern (both for transmission and reception) will form a doughnut shaped ring around it similar to what you see below:



The weakest point of propagation will be directly above and below the antenna (in the doughnut hole area), given that if one of two antennas was directly above the other with no obstructions then it would make sense to keep them parallel to each other.

But that's not purely the case here, you have to account reflections and attenuation caused by nearby objects and structures (floors and so on), in a clustered environment such as a house you might find that the shortest most logical path/polarization isn't the most efficient.

The best method really is to "try it to see what's best" and tbh I doubt you would see much difference regardless.

[watercooled]

As MaddAussie said an omnidirectional antenna will transmit in a torus (doughnut) shape, higher gain antennas producing a flatter radiation pattern. Aside from the radiation pattern, moving the antenna also affects the polarisation of the signal i.e. if the transmitting antenna is vertically polarised, you'll theoretically get better reception with a vertically polarised receiving antenna but in practice I find it doesn't make a huge amount of difference with WiFi. Internal laptop antennas are generally vertically polarised, running up the side of the monitor, so in theory you'd want your router antenna positioned the same way. Also, MIMO actually relies on simultaneous transmission through multiple antennas, so having them each positioned differently may not be the best solution for best throughput, but then you may create weak spots directly above/below the router.

Inside an average house though you'll have large metal objects, sources of interference, etc so it's not as simple as that so trial-and-error is often the best solution, but theory is nice eh?

Edit: Beaten to it by several more posts by the looks of it.

[Mama Sumae]

Btw



Purely for scientific purposes ofc.

[Zak33]

I'm pointedly and deliberately ignoring the "try it and see what works" camp cos I said I would.

I wanted (and have learned) to know the why's and wherefore's

thanks to TEAM WIFI DONUT!

[JamieH]

I have a sudden urge to grab a coffee and go to the bakery now. Hmmm I wonder why...