2.4 GHz Omnidirectional Antenna Build (802.11b/g)

[systems_glitch 98]

I'm going to be doing a few 802.11b/g experiments soon and decided to build another "spider" omnidirectional antenna. They are cheap, easy to build, and pretty rugged. I made my first years ago to get 802.11b through the brick firewall separating two halves of a house. The site explaining the design can be found here:

 

http://articles.adelayde.net/antenna/omni/quarter/

 

The components required are cheap and easy to acquire. You need a panel-mount female N connector with 4-hole flange, some wire and a bit of solder. I used #10 house wire (30 amp, orange jacket) for the ground plane and #12 house wire (20 amp, yellow jacket) for the driven element. The N connector was picked up for $1 at a local hamfest, but they're available online.

 

Tools you'll need:

 

- Big soldering device (big iron, soldering gun, or small torch)

- Pliers to bend the wire

- Cutters to cut the wire

- "Helping hands," small vise, or lock pliers (the N connector will be hot enough to melt plastic and ruin table finish)

- Round file

- Flat file

- Metric ruler/scale or calipers

 

What I used:

 

- Hexacon 100 Watt iron

- Klein/Bell System lineman's pliers (cutter combo)

- Radio Shack "Helping Hands"

- Home Depot hobby files

- $9 Harbor Freight calipers

 

 

The N connector needs cleaned to take solder. If it's a really nice silver-plated one, it should be ready to go. If it's a more common chrome-plated kind, you need to remove the chrome or solder won't stick. Use the round file to remove the chrome from the flange holes and the flat file to remove it around the holes. You should see yellow/gold brass under the chrome.

 

 

Once it's clean, cut your #10 wire into four lengths. They should be at least 40mm long. I left mine much longer to make handling the hot N connector easier. Bend the ends at a right angle with your pliers, so that they can be inserted into the mounting holes on the connector flange.

 

 

Use the holding device to hold the N connector while you solder. DO NOT TOUCH ANY OF THE METAL! Even the ends of the long #10 wires were far too hot to touch.

 

 

Cut a piece of #12 wire for the driven element. It needs to be at least 32mm long, from the end of the N connector flange to the tip of the wire. I cut mine long for easy handling and cut to length later. Make sure it's as square as possible to the N connector.

 

 

Soldering is now finished, and you're ready to trim down the leads to their required lengths. Let the antenna cool down to room temperature first! We're going to cut them to 1/4 wavelength dimensions. Use the handy diagram in the original linked article.

 

 

Cut the driven element a bit long and then file it flat. This will make the cut more precise. You want the end of the element to be as close to flat as possible.

 

 

Bend the ground plane elements to 45 degrees. The original article suggests 30 but reports that people had better results with 45. I used 30 with my first build and then corrected to 45 later on. Once they're bent, measure the leads and cut them. These don't have to be perfectly square, but I filed down the sharp edges.

[tekio 104]

Thanks for posting! 

 

I want to learn to solder. Been waiting for the right project. Think I'll give this one a try.

[systems_glitch 98]

This is a good one to learn on...no tiny joints to make! Make sure your soldering device (iron, soldering gun, torch) is capable of heating the large N connector and #10 wire. You probably want to practice on scraps of #10 wire if this is your first time ever soldering. Check if any local hackerspaces are doing soldering workshops near you.

[tekio 104]

Thanks for the tips, Glitch!

 

Just so I'll know what to expect, what kind of gain are you getting with the antenna?

[systems_glitch 98]

I haven't measured with this one yet, and I don't know that I ever wrote down what I got with the last one. My antenna pigtail should be in today, so stand by!

[systems_glitch 98]

Antenna pigtail arrived today! I tested with an old PCMCIA Senao 200mW NL-2511 card, which is 802.11b only (thus, 2.4 GHz spectrum only). I have a "rubber duck" MMCX antenna for the Senao card, it's an inch or two long. I'm not sure of its effective gain, but when using the spider omni, access points that were reachable with the rubber duck antenna were 3-4 dBi stronger, which matches with the claimed 3 dBi center frequency performance of this antenna.

 

I now consistently see traffic from other access points that before were only identified by the occasional single packet captured by Kismet. There are some interesting SSIDs out there, like "WerewolfBarMitzvah" and "DEA_surv"!

[tekio 104]

Those Senaoa's were nice back in the days. I think they only see 802.11B, since they have the Prism2 chipset. 

[systems_glitch 98]

Yeah, they're b-only cards, but that's fine for testing. The antenna is designed for 2.4 GHz so any card used with it should be b/g or non-5 GHz N. I'm looking at a few high-power b/g cards at the moment.

[Ticom 7]

1/4 wave ground plane antennas work really well for their intended frequency of use, and they're cheap to build - just some wire and a connector.  You can also improve the performance of a lot of antennas by putting a 1/4" ground plane under them. Those of you who remember the old G-Net web site from the l0pht, their colinear performance was greatly enhanced with a ground plane under it.