2008-03-08
Bicycle Rim Antenna for 20 Metres
Several months ago I was walking home from the post office, nearing my place I saw it was "clean-up week", the curb dotted with various piles of junk people had put out to be disposed off by the council pickup. The pile outside my block of units had mostly busted furniture, but one item caught my eye, an Aluminium bicycle rim. I dug it out and took a closer look. The spokes and hub were all rust-pitted chromed steel, and physically it was for a child's bike, only around 580 mm in diameter, but the Aluminium itself looked to be in good shape. Antenna was the immediate thought, so I carried it back to the shack.
The spokes and hub where removed and discarded. The rim had a join where steel pegs had been inserted into cylindrical openings in the extrusion and epoxy used to close and secure the join, forming the round shape. I used a cut-off wheel on the rotary tool to cut through the join, breaking the rim so I might measure and feed it.
Experiments
The rim is roughly 1.3 uH of inductance. This is a good fit with the "ring" inductor formula. I experimented with the loop of metal for some months before I finally settled on making it into an antenna for 20 metres. The efficiency is fairly poor on 40 metres and even somewhat marginal on 20, but on the higher bands it is an exceptional antenna. I even tuned it up on 11 metre CB and listened around, hearing not much but some Asian fishermen and the usual braindead 27.355 MHz crowd. It is self-resonant near 6 metres (distributed capacitance about 8 pF) and is therefore limited to upper-HF, roughly 30-10 metres.
For its life on 20 metres I chose the more electrically straight forward and tunable taped-capacitance feeding arrangement. The disadvantage of this arrangement is that it is a bit more difficult to tune, as the match and tune capacitances affect each other, but some iteration finds a good match quite quickly.
I did experiment with small coupling loop and gamma matching. Both work just fine but are a bit fiddly if you want to change bands a lot. Small driven loop has the advantage that you can twist the coupling loop with respect to the resonated loop to adjust the matching, and can pass the coax feed right through the main loop at its nodal-point where it has a voltage minima. The rim had a hole directly opposite the join where the tire valve likely was placed, this allowed hanging the loop by its coaxial feed. While interesting experiments and valuable lessons for future experiments, I like the tune/match capacitor coupling I ended up with.
Construction
The loop was screwed to a piece of timber (a poor insulator unfortunately) as a base. The tuning and matching capacitors were screwed to the timber base as well and the coaxial feed fed though some holes in the timber as strain relief. The tuning gang was used in "split-stator" mode to reduce losses, leaving it with a capacitance range of about 20 pF. Some binding posts were used to allow different silver-mica transmitting capacitors to be put across the loop to shift bands.
Two 160 pF capacitors are used for 20 metres, giving an equivalent capacitance of 80 pF. The loop tunes about 13.9 to 14.6 MHz, and the 2:1 VSWR bandwidth exceeds this.
The matching capacitor is a fixed 18 pF in parallel with a 10 pF trimmer. This value was determined experimentally with a signal generator, mini C-jig, and return loss bridge as the best arrangement for matching the loop over the frequency range of interest without excessive touchiness in the tuning.
Notes
Power handling is limited by the capacitors. The Silver-Mica caps used are rated to only 200 Volts, but for my current QRP use this isn't a problem. Efficiency if affected a little bit by using binding posts instead of soldered connections for the capacitors as well. I'll probably solder the caps in place permanently once I settle on them, but I am hoping to find higher voltage ones on eBay first.
The connection to the Aluminium of the rim is made with nickel plated hardware, thick copper wire and lots of star washers, etc. Ideally it should be spot-welded, as should all the rotor plates in the capacitor to the shaft and the stator to their connections. The stator plates are copper, I've never seen a capacitor quite like this one before, it would be possible to solder them together to reduce the losses. The rotor plates are Aluminium, the shaft brass, which presents more of a problem. The shaft has an inbuilt coaxial reduction drive with a 1/8" shaft that comes out of the middle of the 1/4" one. I'm having problems finding a suitable knob.
A ferrite suppression bead was slipped over the coax feedline near the feed-point to reduce a slight interaction with the coax position/body capacitance upon return-loss seen while experimenting.
9 comments.
Attachments
| title | type | size |
|---|---|---|
| Feed Network Diagram Source | application/postscript | 11.601 kbytes |
| Alternative Feeds Source | application/postscript | 10.002 kbytes |