We have built and tested electrically small (∼γ/10) resonant patch antennas as proposed in recent literature [1, 2]. The metamaterial array loading the antennas formed a rough cylinder axially enclosed by a patch antenna and a ground plane. The fill ratio, or ratio of the metamaterial array's radius to the patch radius, was less than one. Given a particular negative permeability metamaterial (copper spiral rings printed on circuit board in this case), the fill ratio dictates the lower of two resonant frequencies of the antenna. The higher frequency resonance is characteristic of the patch.

We observed that each of the antennas radiated at two resonant frequencies, as predicted. The lower frequency resonance disappeared when the metamaterial was removed. We built two versions of this antenna, one (Design I) with a lower resonant frequency of 756 MHz and higher resonant frequency of 3.3 GHz, and a second antenna (Design II) with a lower resonant frequency of 385 MHz and higher resonant frequency of 1.8 GHz. Because we were interested in reducing the size of patch antennas, we focused on the lower frequency resonances in this work. The antennas' return loss was measured at -23 dB and -28 dB, the gains were -11 dBi and -13 dBi, and the return loss was less than -10 dB over bandwidths of 4.7% and 1.8% for the lower frequency resonances of Design I and Design II, respectively.

We also predicted the trend of increasing resonant frequency with decreased metamaterial fill ratio. We varied the fill ratio was by changing the patch size while maintaining the same metamaterial array. As predicted, resonant frequency increased with increasing patch size, an opposite trend to what one would expect without the loading metamaterial. Altering the patch size allows simple tuning during the assembly and test process.