RF SPST Switch Evaluation Board
The SKY13347-360LF 0.5 to 4.0 GHz RF SPST switch was initially chosen to control the radiation pattern of the antenna. Before the chip was implemented in the reconfigurable antenna, an evaluation board was designed and manufactured to test
and collect the S-parameter data of the RF switch. The evaluation board and the RF switch were designed using Keysight’s Advanced Design Systems (ADS). The evaluation board consisted of two co-planar waveguides with a characteristic impedance of 50 Ω,
two DC blocking 47 pF capacitors, and two DC bias lines. The layout of the RF switch can be seen below:
and the design of the evaluation board can be seen below:
Once the evaluation board was delivered, the RF switch and 47 pF capacitors were soldered on along with the wires to connect the DC power supply. The completed evaluation board can be seen below:
The motivation behind designing and testing the switch on the evaluation board was to find the insertion loss and isolation of the switch and compare it to the simulation results of a simple switch model and ideal capacitors. Doing this verifies that the RF
switch has the correct S-parameters. The data collected could then be used in HFSS and modeled using lumped RLC components that produce similar characteristics. The evaluation board was tested with a VNA to collect the S-parameter data. The results of the evaluation
board were then compared with two different simulations in ADS. The first simulation represented a simple switch model with an insertion loss of 0.55 dB with ideal 47 pF capacitors and two co-planar waveguides that had the same dimensions as the evaluation
board’s co-planar waveguides. The second simulation represented a simple switch model with an insertion loss of 0.55 dB and the manufacturer’s measurement data files of the 47 pF capacitors and two co-planar waveguides that were the same dimension as the evaluation
board’s co-planar waveguides. The schematic for the two simulations can be seen below:
The two simulations from ADS along with the evaluation board’s measured data was plotted and compared to one another. The results for the OFF and ON state can be seen below:
The OFF state in the first plot, shows the isolation for the measured results to be -29.5 dB at 2 GHz which is close to the simulated results using ideal capacitors and the simple model switch that shows the isolation around-35 dB at 2 GHz. The ON state in the second plot, shows the
insertion loss for the measured results to be -1.9 dB at 2 GHz. The ideal capacitor and simple model switch simulation show the insertion loss to be -0.7 dB at 2 GHz. The measured results have a higher insertion loss due to the evaluation board only being calibrated up to the SMA
connectors. The SMA connectors could account for an extra insertion loss of -0.3 dB on each connector. Furthermore, the dip around 2.7 GHz in Fig. 43 for both the measured and simulated results is due to the self resonant frequency (SRF) of the 47 pF capacitors. In the second plot, it can
be seen that only the measured evaluation board (blue), and the simulation using the manufacturer’s measurement data files for the 47 pF capacitors (black with dashed lines), have the dip at 2.7 GHz. The ideal capacitor simulation (green), does not have a self resonant frequency
due to having ideal 47 pF capacitors. Moreover, the measured results match up relatively closely with the simulated results of an ideal capacitor with a simple switch model.