Antenna for router

Just taken delivery of a "high power" Buffalo WHR-HP-G54 router and using it in Access point mode. Although it has increased the reach of my wireless network moderately, I think it could benefit from a better located antenna.

I am trying to get a decent signal in the kitchen from the router in a loft study two floors up and the other side of the house. Before I got the Buffalo, the signal was about 10-20% and 1-5mb/s. It is now 30-40% and 18-24mb/s. I would like to get that a touch higher, preferably 60-80% and 36mb/s+

Currently the Buffalo is located behind a desk and metal computer case with a metal desk and tons of copper wiring. Not ideal!

I have the ability to site the aerial clear of all this stuff in the neighbouring room.
So... I need a short cable that will attach to the Buffalo which uses a Reverse Polarity SMA (RP-SMA) connector. I can then either use the existing antenna on it or attach a new one.

What should I get? All advice gratefully recieved!

Have a look at Solwise.co.uk. for antennas and cables.

However be aware the loss at 2.4gig is horrendous on coax and ideally you would move the radio (i.e. router in your case) to the room too. We have wireless coverage in some areas and one in particular has around a 7metre drop from the sector antennas to the radio - it is on LMR400 coax (which is the thickness of your thumb) and looses 2.5db gain due to the cable and joins. A typical little antenna on a router is only 2db (or sometimes 5db) gain in the first instance, so you are going to need some darn chunky cable (and expensive).

I'd vote for an antenna with higher gain attached to the back of your router, or as short a cable as you can get away with (you also loose 0.5db per connection typically, as well as the actual wire). Ideally you should re-position the ensure router and run some cat-5 to it IMO.

Hope that helps, or at least aligns the train of thought!

my advice would be to put the Buffalo in the other room and run Cat5 to it, as the signal loss in a small pigtail cable is not insignificant.

Failng that a directional antenna pointing at the kitchen would be a useful gain over the curent omni. Have you tipped it on its side to help it go downstairs ?

I totally agree. The losses in the antenna cable would outweigh any advantages of moving it. Cat5 is much cheaper

Not often I see unanimity on these forums

Thanks for all the advice, folks. No problem moving the router / AP as it will just mean running the little power cable alongside the CAT5, which I already have.

I am always in favour of a free upgrade

On the technical points raised; yes the current antenna is tipped on its side (parallel to the floor) which gives the best reception throughout the house. I have spent hours (alright, minutes) fiddling with antenna orientation and channel choice and trotting up and down the two flights of stairs to check the signal on my laptop. I never thought setting up wi-fi would be such good exercise!

A directional antenna is not really suitable because the signal also needs to reach other areas of the house.

I have heard mixed views on whether a high gain antenna is much use in these situations. I have heard that the signal is more 2 dimensional, better if you are in the plane of transmission but if you are either above or below the antenna (or too far either side for a horizontal antenna) the signal can be worse. Is this true?

Regarding the cable and signal loss, I found this website:
http://www.wifi-antennas.co.uk/index.php
It seems to have an excellent choice of low cost antennae and low loss coax.
This cable would give only a 1.1db loss on the two metre run I would need. If I combined it with a new 7db antenna it might do the trick. Still, as you all say, probably better to site the whole AP in the next room and put a new antenna on it if necessary.

In reply to:

---

A directional antenna is not really suitable because the signal also needs to reach other areas of the house.

---

If you tipped the antenna on its side (parallel to the floor), you have already made it directional. The antenna is only omni-directional (in the horizontal plane) when it's vertical. Most wi-fi antennas claim to have "gain", and they do this by confining the radiation pattern to a horizontal plane - not too much radiated in the up or down direction. The radiation pattern is like a squashed sphere, rather like the shape of the Earth, with the antenna going through the "poles". That's why it works better upstairs when it's laid flat.

P.S. You can see what the radiation pattern looks like here

Edited by deleted (Thu 24-Jul-08 10:33:10)

In reply to:

---

If you tipped the antenna on its side (parallel to the floor), you have already made it directional.

---

I know, which is why I did it. Your post also confirms the view previously expressed to me that high gain antennae just flatten the dispersion pattern in the vertical plane. This one does not look too bad though.

When Yarwell referred to "directional" I assume he meant an antenna that "pointed" in one direction in the horizontal plane such as this . Actually the dispersion pattern of that antenna looks OK if I were to mount it at the apex of the roof and point it down, as less energy would be lost into the sky above my house, while still giving a reasonable amount of lateral coverage. The pattern is more like a doughnut with the hole off-centre rather than in the middle, which is the case with most omni-directional antenna.

Currently half the loft is a bedroom and a (tiny) study and the other half is rough storage full mainly of old books, clothes, suitcases etc. Currently the AP is in the study at the front of the house - behind the PC and metal desk as I said before. By moving it to the storage area I have the chance to mount the aerial clear of obstructions and/or mount an antenna under the apex of the roof. This will move it more into the centre of the house and the route to the kitchen will then be about 25 feet vertically through wooden floors rather than 30 feet diagonally through brick walls, as is the case at the moment.

I will try just moving the router for the moment, but may try the extension cable and antenna trick if that does not do the job.

Thanks for your interest.

In reply to:

---

A directional antenna is not really suitable because the signal also needs to reach other areas of the house.

---

but as the signal origin is at the top of the house an omnidirectional antenna is sending 50% of the signal out of the roof, so you might like to reconsider.

No directional is like a laser, they all focus the signal into one direction or plane but the pattern is still pretty wide.

In reply to:

---

high gain antennae just flatten the dispersion pattern in the vertical plane

---

Only true for high gain omnis mounted vertically - changes from a doughnut to a pizza sort of thing.

How about http://www.maplin.co.uk/Module.aspx?C=SO&U=strat320&T=sitenew&ModuleNo=38205

You might need to do what I did - have 2 wireless networks (both on the same channel, BTW). I live in a big house with thick walls, and this was the best way of doing it.