Cantenna Mark 2
My interest raised, I then set about building Cantenna, the sequel, this time using some of the more tried techniques. The spreadsheet for optimal can design here http://www.concentrate.com.au/misc/circular_waveguide_optimise.xls, suggests that the right length is about 2 tins instead of just the one. So I decided to try to make a much longer can and this time build it with rain in mind!
(Click on any of the images to enlarge)
Start off with two suitably sized dog tins, although I haven't done conclusive testing, I do not believe that it makes any difference if the dog food is with gravy or jelly, mine was with gravy as my dogs prefer that variety. Nor do I believe it makes any difference whether the choice is chicken or beef.
The bottom of the tins are the same size as the overall can diameter, the top of the tins is a couple of millimetres larger. I removed the top of both tins, including the ridge by using a tin opener horizontally and the bottom of one of the tins was removed by using the tin opener vertically, thus removing just the inner disc, leaving the seam intact.
This allowed one tin to be pushed into the other ready for soldering rather than trying to butt solder rough edges.
When the tins were soldered at about one inch intervals, some self adhesive, conductive copper tape was used to ensure continuity all around the tins.
A piece of round kitchen extractor ducting provides a perfect fitting shroud for the tins
I have two plans for this antenna, the first is to construct the de-facto N-type socket and the second is to attach the USB NIC once more but this time with a weatherproof shroud. For this, I needed to bring out an external antenna cable, terminated in an N-type connector. The next picture shows the inside of the tins with the antenna feed soldered to the N-type chassis socket.
Weatherproofing equipment! :-
The finished antenna, attached to the flexible suction cup (with locking clamp):
Netstumbler output of the effect of adding the antenna, the NIC in use for this was a Compaq WL110, on the left is the basic NIC and then the larger signal strength is shown quite clearly when the antenna is plugged in. The left hand part of the output is without the antenna, the right is the signal with it connected.
The USB NIC that that featured in mark 1 of the antenna, showing the built in antenna on a small piece of coaxial cable. I needed to remove this to take off an external antenna feed.
The next picture shows the socket which is pre-drilled for an SMA connector. I'll add one later so for now, it's a case of soldering the mini-coax straight to the board.
Originally I had planned to stick the NIC inside the plastic casing but the tins take up all the space so simplicy rules and a couple of bits of velcro and a cable tie results in a quick fix (Present waterproofing for the NIC being provided by a polythene bag - not shown for clarity):-
Finally, I was interested in the difference in using two tins compared with just the one, Netstumbler output yields the following result, the first part of the signal test is with the two tins (Cantenna mk 2) and the right hand side is just using one tin:-
I have now had an opportunity to compare the mk 1 milk tub against the mk 2 twin tin design. In summary, the mk 1 performed just about as well. One problem would appear to be that when using an Orinoco NIC, the extra gain that the mk 2 produces is lost in the couplings and the cable run to the PC which is a shame. Even when using the USB NIC, the signal seen by the NetGear utility is not really much better. Again, this might be due to my external flylead and the extra losses incurred with the N-type connectors whereas in the original mk 1, I simply pushed the NetGear antenna through the tub.
(Further testing shows that the signal loss might also be largely due to the extra foliage on the trees since the first test 3 weeks ago)