A cantenna is a nickname for a type of Do-It-Yourself directional waveguide antenna. It is so named because it is made out of a metal can of some kind–coffee cans and large soup cans are popular. These antennas allow you to focus your WiFi signal for higher gain and greater range, and generally cost a lot less to make than it would be to buy a comparable directional antenna.
I built the main part of this cantenna a couple years ago, but did not have the card and dongle part to use it effectively. The can was a pineapple juice can that was about the right size, and which I already had around. Into the can, you mount an N-Type Panel connector. This requires one large center hole and four mounting holes, but the can isn’t difficult to drill through.
There are two important calculations to make. I simply used an online calculator to find the ideal specifications for my can. The one at http://www.turnpoint.net/wireless/cantennahowto.html seems to do the trick, although I’m not sure if this is the one I used.
The can I’m using is 4.125″ in diameter and 7″ long. Using that diameter, the above site tells me:
Cutoff Frequency in MHz for TE11 mode: 1676.91
Cutoff Frequency in MHz for TM01 mode: 2190.27
Guide Wavelength in Inches: 6.67
1/4 Guide Wavelength: 1.67
3/4 Guide Wavelength: 5
This seems to indicate that my cantenna is a bit too wide to be in the proper frequency range, but they recommend experimenting, so experiment I did. Another crucial part of building this is to solder a length of 12-gauge copper wire into the top of the N-Type connector so that from the base (which you solder into) to the top of the wire is as closed to 1.21″ as possible. Once everything is set with the cantenna, there’s another element of the equation to work on.
The PC Card that goes into the laptop is the next part that needs to be prepared for this project. Most typical laptop PC Cards don’t have an external connector to plug in an auxiliary antenna, which is kind of unfortunate. This means you either have to buy a special card for the purpose, or possibly modify one you already have. I opted to attempt the latter.
I had an Orinoco 802.11a/b card from several years ago. It supports 2.4GHz 802.11b and 5GHz 802.11a. However, when attached to the cantenna, 5GHz wouldn’t be received, but that’s a minor issue since 2.4GHz is the most widely used standard. To prepare the card, I had to open it up and see what I was working with. After a bit of difficulty, I managed to do this, and found what seemed to be a conveniently located U.FL connector on the PCB. This seems to be one of many standards that exist for wireless communications connectivity. The next step was to order a dongle that would allow me to connect the U.FL connector on the card to the N-Type connector on the cantenna.
Upon receiving the dongle, I knew I would have a little bit of modification to do. At the very least, I would have to drill a hole in the casing for the wire to fit through. I found I also had to make another modification, as the case would not fully shut with the added height of the connector. Another hole above it solved that issue.
Once everything was assembled, it was time to test it. This is where things did not turn out as I had hoped. It seemed the cantenna made no difference in the wireless signal. I directed it at an access point that was weak without an antenna, and it remained weak (or worsened) with the cantenna. This is a bit of a perplexing problem, and a disappointing end to this project, but hopefully in the future I’ll revive it and get the cantenna working after all.
1. The can is the wrong size. This is evidenced by the calculations I ran. It’s possible the cantenna is only picking up frequencies in the 1.6GHz to 2.1GHz range, which is not at all useful to me. I could easily and cheaply pick up another can and quickly modify it to mount the components I have into it.
2. The Orinoco card is not working properly. I don’t know if the U.FL connector I found works, or if connecting into it works properly. I may need to get a card with a real external connector, or try some other means of doing this.
With some luck, and persistence, this project may prove to be useful after all. One thing I do know is I’m already through one big hurdle: I have most of the parts I need to pull it off.
All content ?2006 Tony Magri