An AM radio transmitter requires an antenna to radiate its signal to the receiver. An ideal antenna would be very, very long (usually coiled for most of this length to keep it small), but we can get a decent signal by using a 40" wire that is connected to the output signal pad A0. (You can effectively double the range by using an 80" length of wire, and gain a bit more at 120", however beyond that the ambient noise will increase a lot.)
To make this simple antenna, measure and cut a 40" length of enamel coated copper wire (called "magnet" wire because it's often wound into coils to create electromagnets.)
The wire is coated in enamel, so it is electrically insulated. We need to expose a bit of the copper at onw to connect it to the Gemma M0. You can use a flame to burn away the enamel, or scrape it away carefully with a knife blade as shown here.
Use alligator clip leads to connect one of the antenna wire to A0 on the Gemma M0.
Special thanks to Jan Goolsbey for his great insights on antenna design and testing for this guide. Here are his notes on the antenna experiments he conducted:
A 40- to 80-inch long wire antenna is the better solution. After looking at the carrier and modulation waveform on the 'scope, I began to test it without an antenna, playing with loading on the analog output pin. As the load dropped below 5k ohms, I noticed an increase in radiated energy and signal distortion. When shorted to ground, the circuit created a distorted tone with lots of carrier sidebands, but could only be picked up by the radio within about 10 inches. Adding a 40-inch antenna (shorted to ground) seemed to increase the range, but only along the length of the antenna wire -- the signal died out quickly if the radio was more than about 10 inches away from the antenna wire. Testing an open-ended 40-inch long wire was the next test. Without a ground connection, the range was about 8 to 10 feet. Adding a 40-inch ground plane wire (dipole configuration) to the ground had no noticeable effect.
Increasing the antenna length to 80 inches almost doubled the range to 18 feet. 120 inches increased it to about 25 feet. The range stayed at 25 feet when the antenna length was increased to 160 inches. The ambient noise at the low end of the band was just too much for the Gemma to overcome even with longer antenna lengths.
I also created a helical antenna using the 160-inch antenna wire wound around a paper towel tube. The range dropped to about a foot. In theory, 600 feet of wire wound around a tube could work if you need a more compact antenna than a 586-foot quarter-wave wire length. I'll leave that idea there for the discussion...
The basis of it is understanding how a wave travels over time. In this case, it's a 450kHz sine wave traveling through air at 186,272 mi/sec. Note that synchronized peak energy (the absolute value of the area under the curve) happens at a full wavelength, 3/4, 1/2, and 1/4. Anything between those values creates a moving wave across the length of the antenna, lowering the antenna's ability to transmit the power of the wave. Our 40-inch antenna is very inefficient since it represents only 0.0015 of the 450kHz wavelength.
Tuning an antenna is the process of reducing the destructive energy caused by wave movement across the antenna. It's accomplished by trimming the length of the antenna to match the 0.25, 0.50, 0.75, or 1.0 wavelength nodes of the carrier wave.
If you'd like to try some of these same experiments, here's an example of a dipole antenna, which would have equal lengths of wire connected to A0 and GND.
To make the dipole antenna, you'll use two wires instead. Prepare them both as before by scraping off the insulation from one end.
Use alligator clip leads to connect one of the antenna wires to A0 on the Gemma M0, and the other to GND.
You can try out other arrangements and lengths of antennas to see how the signal strength and noise levels change. What if you orient the antenna differently? What happens if you become part of the antenna by touching the wire?
Next, we'll send some secret messages...