nschutten
Member level 3
I wanted to determine the antenna efficiency of the printed antennas that I designed. I constructed a Wheeler cap for this purpose and carried out the respective measurements.
The Wheeler cap dimensions are said to be not critical; in my setup, at 2.4GHz, the radian sphere should be ~2cm in radius. So what I did is ensure that the box is never closer than 2cm to the antenna under test.
In my measurements I distinctly see that wire antennas have higher efficiency than printed antennas and that the electrically smaller printed antennas have lower efficiency than the larger ones. So this is all according to expectations.
However, I am struggling with the following: the literature dictates that the S11 measurements with and without Wheeler cap must be "all real" measurements. But when I take a reference point on the Smith chart of the antenna impedance on the real axis and then place the Wheeler cap on, there's always a non-negligible imaginary part of the impedance at this reference frequency point (and this is true for any antenna). So it seems to me that "all real" measurements cannot be taken with the Wheeler cap method (at least not at one frequency point for both with and without Wheeler cap). Should I just take the modules of S11, or use a different Wheeler cap size, or ...?
Any help or user experience with actual Wheeler cap measurements is highly appreciated!
The Wheeler cap dimensions are said to be not critical; in my setup, at 2.4GHz, the radian sphere should be ~2cm in radius. So what I did is ensure that the box is never closer than 2cm to the antenna under test.
In my measurements I distinctly see that wire antennas have higher efficiency than printed antennas and that the electrically smaller printed antennas have lower efficiency than the larger ones. So this is all according to expectations.
However, I am struggling with the following: the literature dictates that the S11 measurements with and without Wheeler cap must be "all real" measurements. But when I take a reference point on the Smith chart of the antenna impedance on the real axis and then place the Wheeler cap on, there's always a non-negligible imaginary part of the impedance at this reference frequency point (and this is true for any antenna). So it seems to me that "all real" measurements cannot be taken with the Wheeler cap method (at least not at one frequency point for both with and without Wheeler cap). Should I just take the modules of S11, or use a different Wheeler cap size, or ...?
Any help or user experience with actual Wheeler cap measurements is highly appreciated!