Compact rf polarizer and its application to pulse compression systems

We present a novel method of reducing the footprint and increasing the efficiency of the modern multi-MW rf pulse compressor. This system utilizes a high power rf polarizer to couple two circular waveguide modes in quadrature to a single resonant cavity in order to replicate the response of a traditional two cavity configuration using a 4-port hybrid. The 11.424 GHz, high-Q, spherical cavity has a 5.875 cm radius and is fed by the circularly polarized signal to simultaneously excite the degenerate $T{E}_{114}$ modes. The overcoupled spherical cavity has a $Q_0$ of $9.4\times {10}^4$ and coupling factor ($\beta$) of 7.69 thus providing a loaded quality factor $Q_L$ of $1.06\times {10}^4$ with a fill time of 150 ns. Cold tests of the polarizer demonstrated good agreement with the numerical design, showing transmission of $-0.05\mathrm{dB}$ and reflection back to the input rectangular WR 90 waveguide less than $-40\mathrm{dB}$ over a 100 MHz bandwidth. This novel rf pulse compressor was tested at SLAC using XL-4 Klystron that provided rf power up to 32 MW and generated peak output power of 205 MW and an average of 135 MW over the discharged signal. A general network analysis of the polarizer is discussed as well as the design and high power test of the rf pulse compressor.

Figure 1

3D rendering of an 11.424 GHz, x-band, rf polarizer with physical ports and waveguide types labeled accordingly.

Figure 2

The rf polarizer divided into each submatrix for cascade analysis.

Figure 3

Subnetworks A and BC used for the second cascade of the polarizer model with labeled port assignments.

Figure 4

A cross section view of network-B with the boundary conditions used to isolate the even mode (top) and odd mode (bottom) for this analysis.

Figure 5

A range of valid phase delays $\varphi$ (black-line) based on the magnitude of the reflection $\theta$ at the cylindrical waveguide for network-BC with annotations denoting operating points of the polarizer for mode-1 (blue-X) and mode-2 (red-O) as measured in HFSS.

Figure 6

Subnetworks A and BC used for the second cascade of the polarizer model with labeled port assignments.

Figure 7

S-parameters of the ideal circuit model with $S_{1(3:1)}$ (black-o), $S_{1(3:2)}$ (gray-line), $S_{21}$ (dashed-red), and the relative phase between each $T{E}_{11}^O$ mode (blue-squares), $S_{11}$ (not shown) is zero by design for this analytic model.

Figure 8

Numerically derived output power (dashed:blue) using the convolution of an RLC resonator model and standard input pulse (black) for passive pulse compression cavities.

Figure 9

(a) Cutaway CAD drawing of the spherical cavity used for pulse compression with RF polarizer, and (b) contour plot $T{E}_{141}$ mode in the spherical cavity calculated using HFSS.

Figure 10

(a) Photograph of the rf polarizer with no attachments, and (b) photograph of two rf polarizers connected via the cylindrical waveguide.

Figure 11

Cold test results of configuration:A, measured with an Agilent N5242A PNA-x showing (a) the reflection characteristics S11 and S22, and (b) the transmission properties S21 and S12.

Figure 12

Cold test results of configuration-C measured with an Agilent N5242A PNA-x showing the transmission properties S21 of the network.

Figure 13

(a) Photograph of the XL-4 klystron used to test the SLED cavity and (b) a circuit diagram of the experimental setup: (1) local oscillator, (2) arbitrary function generator (AFG), (3) 20db amplifier, (4) traveling wave tube, (5) SLAC XL4 klystron, (6a-6b) peak power meters before SLED, (7) polarizer, (8) SLED, (9a–9b) peak power meters after SLED, and (10) high power rf load.

Figure 14

A chart showing the measured input pulse envelope (dotted:Red), the measured output pulse (black), and the predicted output pulse (dashed:Blue) based on the cold test measurement.