IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Improving Wireless Range, Getting the most out of your 743GE
Damion Milliken
post Jul 7 2003, 06:40 AM
Post #1


7500G Aficionado
******

Group: Members
Posts: 1,085
Joined: 28-March 03
From: Canberra
Member No.: 15



So, you're looking to get better wireless range from your 743GE? You've delved into a complicated topic that revolves heavily around the generation of electromagnetic signals and the interaction of matter with electromagnetic radiation. To be honest, there are no hard and fast rules, and nothing is guaranteed. You won't truely know the difference that you'll experience by trying any of these ideas until you actually test it out for real.

The first thing that you should do is align the antennas. The 743GE's antenna is vertically polarised. This means that it concentrates the signal in the left and right directions, rather than the up and down directions. Imagine a doughnut was placed around the antenna, and then pumped up to 100 times the size. This is approximately the shape of the signal that the 743GE puts out. (For a visual image, check out Interpreting Antenna Performance Parameters for EMC Applications, Radiation Patterns, or Development of vertical and horizontal patterns.)

Because of this, you'll generally want the 743GE's antenna aimed vertically, ie, directly up towards the sky. This allows the signal to spread left and right, which is typically towards your computer(s). If you happen to have your 743GE in your roof, or under your house, then you might wish to consider aligning the antenna differently, because directly above and below the antenna (near the tip and base) are the weakest signals. The same ideas apply to the antenna(s) on your computer(s). If you can move them, try to get them aligned so that their signal intersects (crosses) the 743GE's signal as much as possible, and preferrably so that the signals for each end overlap the actual device at the other end.

One additional point with respect to polarisation is that in general, you should align the antenna on the 743GE in the same plane as the antenna(s) of your computer(s). So if you have laptops with PCMCIA cards, which have horizontal antennas, then you'll want the antenna on your 743GE approximately horizontal. If you've laptop(s) with inbuilt wireless device(s), then find out where the antenna(s) lie(s) (often vertically alongside the screen), and align the 743GE's antenna appropriately. Likewise, if you've desktop computer(s) with external antenna(s), then they're probably vertical, so align the 743GE's antenna vertically. Cross polarisation of signals reduces the efficiency of communication significantly.

Then, you probably should start with the positioning and orientation of the 743GE and your computer(s). It's best to have the 743GE high, as that way the signal has to pass through less obstructions (most furniture, for example, is 1 m or so in height, most people less than 2 m). Place it on top of a bookshelf, desk hutch, cupboard, or other high object. However, avoid filing cabinets or metal shelving (for reasons of signal reflection, see below), and try to avoid placing the 743GE inside anything (for reasons of signal adsorption, again, see below).

The 2.4 GHz radio waves that the 743GE uses to communicate with your computer(s) are adsorbed fairly easily. It's best to try to get as clear a line of sight (LOS) between the 743GE and your computer(s) antenna(s) as possible. However, some things adsorb the radiation more strongly than others. In particular, try to avoid placing the router near large bodies of water (eg hot water tanks, fish tanks, and people). Likewise, try to avoid having such objects between the 743GE and wireless computer(s).

Concrete (believe it or not) contains quite a lot of water (chemically locked into the structure of the material). Concrete walls or building materials are good at blocking wi-fi signals. Concrete walls and other thick walls (such as double brick) also block the signal a fair bit just because they are thick and solid. It's better for the signal to go directly through the wall (at 90 degrees), rather than at a small angle. If you think of it from the signal's perspective, going through 30 cm of concrete at 90 degress is 30 cm of obstruction, but going at, say 15 degress (a low, glancing angle) makes it more like a meter of solid concrete.

Large metallic objects (eg filing cabinets, metal shelving, and even some window frames) are a source of reflection of the radio signal. The don't tend to adsorb the signal, but rather redirect it. Placing your 743GE on top of your metal filing cabinet can have wierd effects on the coverage of the wireless signal. Try to avoid placing the 743GE on or near any large metallic objects, and likewise try to avoid such objects being positioned between the 743GE and your wireless computer(s).

Also, watch out for interference from other devices, and try to avoid having these near the 743GE or computer(s) (eg microwaves, 2.4 GHz cordless phones, some wireless mice or keyboards, some wireless doorbells, video sender units, other wireless networks, etc). These sorts of devices operate on very similar (or the same) frequencies as your wireless network. They can also be a problem if they're between the 743GE and wireless computer(s), and even if they're nearby, such as in your neighbour's house. Interference, static, and radio noise will all degrade the performance of your wireless network.

You can sometimes avoid the problems associated with interference by changing channels. Many wireless networking devices operate on channel 1 or 6 by default. So you'd be advised to go to a high channel such as 11, 12, or 13 to avoid potential interference from other people's wi-fi gear. Supposedly, channels overlap, so if your neighbour uses channel 3, then using channel 4 is not sufficient to avoid interference. Again, supposedly, channels 1, 6, and 11 are the only non-overlapping channels. The 743GE can use channels up to 13, which many other wireless devices are not capable of, so make good use of the higher numbered channels. The flip side of this, of course, is if someone else nearby has a 743GE and is following this same advice tongue.gif (watch out if your wireless network cards are unable to use these high channels as well). Also, with interference from non-wi-fi equipment, there's no certainty as to what the frequency range that the equipment uses will be. Try every and all channel if you've got problems with your neighbours 2.4 GHz cordless phone or something else interfering.

You can also add home made accessories to your omni-directional antennas to make them more directional. This idea exploits the "reflection by metallic objects" that I suggested above could be a problem. Take a look at Poorman solution improving WLAN performance, Simple Method for Wireless LAN extension, How To Build A Tin Can Waveguide Antenna, WWW.FreeAntennas.Com, Parabolic Template, and Deep Dish Cylindrical Parabolic Template for ideas.

Finally, if you're looking to purchase an extra gadget to extend the range of your network, make sure it can do the job. Many routers and access points can only operate in "master" mode. You'll need one that can be a "relay", "slave", or "repeater".

For further reading, I suggest the Wireless LAN Performance Improvement NTK, which has much of the information that I've posted above, as well as some great ideas on improving signal from the client (ie, computer) side of things, and a bit of a guide to changing antennas. The 743GE has a removable antenna of the "Reverse SMA Bulkhead" type (reverse polarity/pin, not reverse thread). Additionally, the D-Link DI-614+ Manual has some good information about placement of your wireless router, in Chapter 3.

My final words of advice are to EXPERIMENT. Put your wireless computer(s) where you want to use it/them, and move the 743GE around, wiggle the antenna on the 743GE and the antenna on your computer(s) (if possible), and see what sort of signal you get. Likewise, you could try shifting a wireless computer a bit to the side, or rotate it a little, and see if the signal gets any better. The Network Stumbler utility can be helpful for experimenting, as it gives precise information about the signal strength and quality. You pretty much want to maximise SNR.

cool.gif
Go to the top of the page
 
+Quote Post
_Raz_
post Apr 6 2004, 05:02 AM
Post #2


Advanced Member
******

Group: Admin
Posts: 2,748
Joined: 8-September 03
From: Melbourne
Member No.: 701



Excellent guide Damion, but may I suggest one small correction?

QUOTE
The 743GE's antenna is horizontally polarised.


The antenna on the 743 (in its default position) is vertically polarised. The field from an antenna contains both electrical and magnetic components, the polarisation of an antenna is generally taken from the polarisation of the electrical component, which (in this case) is vertical.

You are spot on with the description of the radiation pattern, vertically polarised antennas radiate most energy in a plane parallel to the earth.

Another thing worth mentioning; If you adjust the orientation of the 743 antenna for horizontal polarisation, it is a good idea to also adjust the client antennas, as there is an approx 20dB penalty for cross polarisation.
Go to the top of the page
 
+Quote Post
Damion Milliken
post Apr 7 2004, 02:19 AM
Post #3


7500G Aficionado
******

Group: Members
Posts: 1,085
Joined: 28-March 03
From: Canberra
Member No.: 15



Good points, _Raz_. I mixed up the polarisations biggrin.gif. I used to have a bit in there about matching polarisations, but I deleted it because it got people confused. However, I wasn't aware of the severity of the losses for cross polarisation (20 dB is a lot!). I'll see if I can put it back in with simpler wording.
Go to the top of the page
 
+Quote Post
_Raz_
post Apr 7 2004, 05:40 AM
Post #4


Advanced Member
******

Group: Admin
Posts: 2,748
Joined: 8-September 03
From: Melbourne
Member No.: 701



QUOTE
I wasn't aware of the severity of the losses for cross polarisation (20 dB is a lot!).

It is probably not as bad as it seems, the 20dB attenuation is true if the antennas are modelled in free space, but in the real world (especially at 2.4Ghz), we have all kinds of reflections and refractions etc. which tend to skew the polarity of the signal at the receiving antenna, so real world attenuation because of cross polarisation would probably be less than 20dB, but still significant.

Years ago, when I was playing around with receiving signals from satellites, one of the largest problems to overcome was cross polarisation. Most sats tumble in their orbits, so their signal polarity tends to rotate with them. The solution was to have both a horizontal and vertically polarised array, and feed the two arrays 90° out of phase, by mounting them 1/4 wave apart on the antenna boom, or inserting a 1/4 wave delay line on one of the arrays. This resulted in circular polarisation.

Sorry for going off topic and waffling on here, but radio and antennas has been a hobby of mine for a long time. That is why I like playing with wireless networking, it allows me to combine my two favourite hobbies, computers and radio.
Go to the top of the page
 
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



Lo-Fi Version Time is now: 4th September 2010 - 05:04 AM