Copper Toilet Float Antenna

The main points of this was

• to have some fun
• while trying to use only off-the-shelf hardware
• to make a Wide-bandwidth antenna.

I succeeded on all three points! ** Well except for 'off-the-shelf'. I had to cut some holes and there is one step that is tricky involving drilling a hole lengthwise down a brass screw... but other than that.

Below I give some rough ideas how to make one of your own but I don't give step-by-step construction details.

Note that this thing as currently built isn't rugged enough for permanent outdoor mounting.

Content
• Overview Photo
• VSWR Plots
• Summary Construction Instructions
• Slightly More Detailed Steps
• Detail Photos
• Where Did the Idea Come From

A Picture of the overall antenna mounted on an angle bracket.

  VSWR data:

These VSWR measurements were made with a HP 8753E Network Analyzer. I took the antenna into work and grabbed the data off the analyzer then plotted it using MS Excel® . I hypothesize that the main reason the VSWR is not so pretty above 2 GHZ is that the cable and BNC connectors are not really good enough. It would be interesting to see what would happen if good quality microwave connectors were used.

Wait a minute ! What about Gain !

Sorry, I didn't have the facilities available to make gain measurements. Based on basic antenna principles it has to be at least as good as a dipole antenna. If it weren't for the loss from using off-the-shelf BNC hardware it might even be better at the higher frequencies due to it simulating a flared horn rotated about the vertical axis.

Maybe I can convince someone at work or somewhere into doing gain measurements for me later.

Even though I don't have gain measurements, I know it does work in real life on a real radio system. Right after I built it, for fun I took it out on a business trip to a military flight-test range where our company was testing one of our new data-link radios with military jets and other aircraft. After the official testing for the day was done at our main ground site, we took the rather sizable and (supposedly) optimized omni-directional antenna we use for testing down off of the mast and replaced it with this thing. One of our other ground network stations on a mountain some miles away was still operating and we compared notes. It worked great! It came out only a few points lower on the linear correlation score. It also survived the ~ 100 watts of power at L-band that the transmitter put into it.  

Construction

I am not going to give detailed instructions but here are the basics.

On the left you see a photo of the assembled antenna without the mounting bracket. The ruler gives a size reference. In what follows I will refer to the two copper floats that make up the antenna as the "top float" and the "bottom float" which correspond to the top and bottom floats as shown in this photo. Only the bottom float has to be modified; the top float is used as-is right from the hardware store bin. The bottom float is only 'modified' by drilling some holes. The hardware is then attached. The idea was to do as little work as possible and use as much off-the-shelf stuff as possible so I just used prefabbed BNC connectorized cables and didn't roll my own. One end of the BNC cable gets connected to whatever electronic or radio project you want to use with the antenna; the other end gets connected to the BNC panel mount connector that is mounted to the bottom float. This technique eliminates a lot of guesswork about how to best make a feedpoint that keeps the impedance match correct. The use of the off-the-shelf BNC cable and BNC panel mount is probably the main reason the antenna doesn't work better above 2 GHz.

 

OK, slightly more detailed steps (more detailed pictures below):

• Go to the hardware store, find plumbing, find the toilet floats, find the copper version. Buy 3. Why 3? because if you are like me you will almost certainly mess up the first one. They are made of very thin metal and because of their shape are hard to hold in place while drilling.
• With floats in hand, wander over to the screw, washer, and nut, department and locate at least 2 brass screws that fit the screw receptacle on the floats. One of them will be used to mount the bottom float to the bracket. The other screw will be modified by drilling a hole length-wise down the end, cut off and soldered to the center pin of the BNC panel mount connector.
• Go over to the shelving department and get an angle bracket that looks about the right size and with a hole that will accept the brass screw (or resign yourself to drilling out a hole).
• Go back to the screw, washer, and nut, department and get at least one nut and two washers that will fit on either side of the bracket.
• Go to Radio Shack® or you favorite electronics store and by a short length of 50 ohm coax cable with BNC connectors at each end (or maybe your hardware store is well stocked in this; beware: most of the cables sold in hardware stores are for "video" use and are 75 ohm cables). Also obtain a 50 ohm BNC panel mount connector. You might need two just to get the mounting hardware; you need at least two of the mounting nuts for the panel mount BNC. These might also be found at the better equipped hardware stores.
• Chose one of the brass screws to modify. A hole slightly larger than the center pin of the BNC panel mount will be needed in the end. At home or your favorite machine shop find a drill press and some wood and some clamps and some drill bits and carefully drill lengthwise down the center of the brass screw. The depth of the hole only has to be as deep as the center pin of the BNC panel mount is long.
  • My method was to drill a hole in some wood blocks big enough to accept the brass screw, fasten the screw to the wood and then fasten the wood block with the screw in it to the drill press.
• After the hole is drilled, thread the screw into one of the floats to see how far it 'goes' mark the threads just outside the float, take the screw out then use a hacksaw to cut the end of the screw off.
• place the short section of thread with the hole in it on the BNC panel mount and use a hot soldering iron to solder the center pin to the BNC connector. Don't get any solder on the threads of the screw or you won't be able to screw the top float on later.
• Chose one of the floats to be the "bottom" float.
• Carefully drill a hole in the end opposite of where the threaded end is that just big enough to accept the main body of the BNC panel mount connector (but small enough so that the panel mount's flange, nut, and washer can't go through). This looks easier than it is.
• Drill a larger hole an inch or so away from the threaded end. This hole has to be large enough to accept the whole of the BNC panel mount connector with the BNC cable attached to it.
• Connect one end of the BNC cable to the BNC panel mount.
• Pass the assembled cable and panel mount through the larger hole and wiggle it around to get the brass screw end out the smaller hole. One of the panel mount nuts should be on the panel mount to provide a way of adjusting the amount of protrusion of the panel mount connector to the outside.
• Put the panel mount washer and the other nut on the panel mount BNC connector and tighten. Remember, if the panel mount does not come with two nuts, you may have to find a second one somewhere. One of the two should be on the inside of the float and will need to be adjusted to get the right amount of protrusion of the part on the outside.
• To help keep the sharp metal edges of the larger hole from digging into the insulation of the BNC cable, I cut a grommet and formed it around the edges of the hole then used a hot-glue gun to hold the pieces in place.
• I also used a hot-glue gun to put a small bead of glue around the bottom of the junction of the brass thread and the BNC panel mount insulation. This is mainly to aid in mechanical support. Solder is not known for its mechanical properties.
• Attach the mounting bracket.

Here is a photo of the bottom and top separated. The one on the left is the "bottom" piece and is the only one that is modified. The top piece (on the right) is used "as is" from the hardware store.

A close-up photo of what the brass thread soldered to the BNC panel mount looks like with the panel-mount properly installed into the float and with a small bead of 'hot-glue' strengthing the junction between the thread and the panel mount.

Close-up of what the feed point should look like with the whole thing put together.

The BNC coax cable, the panel mount connector and the washer and nut that is used outside the float to hold it in place. The brass screw shown is actually the remnant of the one used to create the short piece with the hole in it. (I neglected to take a photo of that piece before it got soldered to the {other} BNC panel mount -- sorry). The BNC coax cable is connected to the panel mount and then the panel mount (along with the one nut that is show on it) is put inside the float.

The "bigger" hole near the bottom of the bottom float through which the whole BNC panel mount connector attached to the BNC cable is passed. A grommet was cut and 'hot-glued' into place to avoid having the sharp edges of the hole chafe the cable insulation.

The bracket hardware to attach to the bottom float (fairly obvious).

So why did I think this idea might work before I started?

Well, antennas have been a sort of interest of mine ever since I was in high school and became a ham radio operator. I have read a lot of articles on antennas over the years and knew that biconical antennas were wideband and also that tapered horn antennas were wideband. More recently there have been papers collecting wideband antenna information for the purpose of UWB (ultra-wide-band) radio systems that show the use of tapered and elliptical planar elements. Since every trip to the hardware store (for whatever purpose) is also for me an exercise in thinking about what weird use I could put all the various hardware and plumbing supplies for radio and antenna purposes, it was only natural that one day the idea of trying copper toilet floats for something would come around. When it did I held one above the other and realized that they looked just about right (from memories of figures in antenna books) to be approximately correct angles for a 50 ohm biconical antenna or the right shape to form a type of tapered horn / slot of revolution about the vertical axis. Both of this properties speak immediately of broadband potential. Also, I thought it might be possible that a slight gain over a dipole might be possible at higher frequencies where the wavelength is small enough so that the gap between the two floats simulates a tapered horn in cross-section.

I must admit that I am not the first person to think of using copper floats for electronic purposes. Once, in a former job, I traveled to a lightning testing laboratory to have some aircraft skin samples tested. Now this was many years ago so things may have changed but, I was highly impressed that they had constructed most of the test apparatus from locally available hardware-store and electric utility parts. This included the 1,000,000 Volt + voltage multiplier tower that used... yep! copper toilet floats at the elements of the spark gap switches. These switches are used to switch the tower from charge to strike through a physical movement of one set of float towards another. There was no need for those highly polished silver colored spheres that you see in the high energy physics labs. So that planted a seed for their use also I suspect.

Comments? Then email me (but you have to type in the email address yourself): Ray L. Cross, , WK0O,
BSEE, MSEE,
and sometimes Webpage Curator

Back to Sponsor's Homepage

Last modified on: 3 November 2007