RADIOFREQUENCY WAVES, HEATING AND CURRENT DRIVE IN MAGNETICALLY CONFINED PLASMAS
6.4. GYROTRONS FOR ECR HEATING AND CURRENT DRIVE
6.4.4. Engineering features of the gyrotron
As can be seen from Fig. 6.69, the gyrotron has several major assemblies.
The electron beam is created by the electron gun, transported through the vacuum tube by a magnetic field created by a superconducting magnet and deposited at the collector. Microwave radiation created in the cavity exits through an uptaper into a mode converting launcher. The microwave beam is radiated to a set of mirrors and exits the gyrotron through a dielectric window. A vacuum pump is also necessary in the gyrotron to help assure ultra high vacuum conditions.
6.4.4.1. Electron gun and beam tunnel
The electron gun must supply the electron beam power for the gyrotron. If the overall output power is of the order of 1 MW and the efficiency of order 50%, the electron beam power must be in the 2–3 MW range. Typical voltages range from 70 to 100 kV with associated currents of 35–60 A to assure adequate electron beam power while maintaining space charge effects at a modest level. Electron gun designs are carried out with a variety of electrostatic codes. An example of an electron gun design for a 1 MW, 110 GHz gyrotron is shown in Fig. 6.77.
The goal of the design is to produce an electron beam with the proper values of beam radius and beam helicity factor ( v / v ) z at the cavity. The design must also minimize the spread in the velocity of the electron beam. The electron beam tunnel, the region between the electron gun and the cavity, is shown in the simulation in Fig. 6.78 as a smooth wall. In fact, the beam tunnel cannot be a smooth wall since that would allow the onset of unwanted oscillations. The beam tunnel is often built of alternating rings of metal and dielectric to suppress instabilities.
FIG. 6.77. Design of the electron gun for a 1 MW, 110 GHz gyrotron operating at a cathode voltage Va of 96 kV and an electron beam current of 40 A [6.263]. Reprinted from Ref. [6.263].
Copyright (2011) by the IEEE.
6.4.4.2. Gyrotron cavity
The gyrotron cavity must be an open resonator to permit the transit of the electron beam through the cavity. The simplest version of such a cavity is just a straight cylindrical pipe with a downtaper at the electron beam entrance end, to prevent power from passing out towards the electron gun, and an uptaper at the other end to permit passage of the microwaves out of the cavity. Figure 6.78 illustrates a simple cavity design for a 1 MW, 110 GHz gyrotron operating in the TE22,6,1 mode. Modern cavity designs are optimized to achieve high efficiency while limiting mode competition. For power levels above 1 MW, it may be useful to introduce a coaxial insert at the centre of the cavity to reduce space charge forces and possibly to reduce mode competition.
FIG. 6.78. Design of a tapered gyrotron cavity for a 1 MW, 110 GHz gyrotron. The mode pattern of the TE22,6,1 mode is shown to the right [6.256]. Reprinted from Ref. [6.256].
Copyright (2011), American Institute of Physics.
6.4.4.3. Internal mode converter (IMC)
A major challenge of gyrotron design is to transform the cavity mode into a free space, Gaussian-like beam. Gyrotron modes such as the TE22,6,1 mode cannot be transported over long distances in waveguide because of losses due to both ohmic loss and mode conversion loss at bends. The free space Gaussian TEM00 mode (the fundamental transverse electromagnetic mode in free space) is generated inside the gyrotron by an internal mode converter (IMC). This mode can be coupled to a corrugated metallic transmission line with very high efficiency.
The first approach to the IMC was suggested by Vlasov and coworkers in the form of a helically cut length of waveguide feeding power to a nearly parabolic collecting mirror. This concept is relatively easy to implement but is limited to an efficiency of about 80%. A major breakthrough was the invention of a dimpled wall launcher, which can achieve a mode conversion efficiency approaching 100% [6.264]. Further optimization of the mode launcher has been made possible
by improved optimization methods and improved codes [6.265]. An optimized launcher design with dimpled walls is shown in Fig. 6.79.
FIG. 6.79. Design of the internal mode converter for a 1 MW, 110 GHz gyrotron. The waveguide launcher wall ripples are shown at the left and the resulting field pattern on the waveguide wall is shown at the right. At the edge of the launcher, a nearly Gaussian beam has been created and is subsequently launched towards the first mirror [6.258].
6.4.4.4. Phase correcting mirrors and output window
The microwave beam from the launcher is transported from the launcher to the output window by a set of mirrors. Figure 6.69 shows an example using four mirrors, but a smaller or larger number can be used. The mirror set can be used to correct phase and amplitude errors in the microwave beam from the launcher. In modern designs, the launcher output beam is of sufficient quality to be transmitted to the window by mirrors with smooth surfaces. Since the output window must transmit high power, a CVD diamond window is often used due to its very low microwave loss (tan ~ 2 10 5, where tan is defined as the imaginary part of the dielectric constant divided by the real part of the dielectric constant), high strength and excellent thermal conductivity. The diamond window can be edge cooled with water without a large temperature rise or stress in the window material.
6.4.4.5. Depressed collector
The spent electron beam is guided to a collector which must be able to operate under conditions of very high average power. The tendency of the beam to strike a small area of the collector must be counteracted by use of small sweep coils that produce a local magnetic field in the collector region that oscillates in time at several hertz. The beam is then swept up and down the collector and also, in some cases, around azimuthally. A major advance in the development of gyrotrons was the successful demonstration of operation of the gyrotron with a depressed collector [6.266]. A gyrotron with a depressed collector is illustrated in Fig. 6.80.
FIG. 6.80. Schematic of a gyrotron with a depressed collector. Reprinted from Ref. [6.267].
Copyright (2011), American Nuclear Society.
Megawatt power level gyrotrons typically have an efficiency of about 30 to 35% before application of the depression voltage. After application of the depression voltage, the efficiency can rise to well above 50%. The depression voltage is limited by the onset of beam interception on the body of the gyrotron.
Experimental results on depressed collector operation of a short pulse gyrotron are illustrated in Fig. 6.81.
FIG. 6.81. The power and efficiency of the MIT 1.5 MW, 110 GHz gyrotron, operating at 96 kV and 42 A, are shown as a function of the depression voltage. The efficiency reaches 50% at a voltage depression of 25 kV. Ibody is the current intercepted by the gyrotron internal parts; this current does not reach the collector. Reprinted from Ref. [6.254]. Copyright (2011), American Institute of Physics.
6.4.4.6. Auxiliary components
The gyrotron output power must be coupled into a transmission line.
Very low loss is achieved in evacuated metallic corrugated waveguide systems.
Transmission by a set of mirrors is an alternative approach. A matching optics unit is used between the gyrotron and the transmission line to assure that the gyrotron beam couples efficiently to the waveguide and to remove stray radiation.
The transmission line for the ECH system at DIII-D is illustrated in Fig. 6.82 [6.268]. A comprehensive review of the passive components needed for efficient operation of the transmission line has been given in Ref. [6.269].
FIG. 6.82. Components of an ECH system, where MOU is a matching optics unit; DL is a dummy load; CDL is a compact DL; SDL is a small DL. Reprinted from Ref. [6.267]. Copyright (2011), American Nuclear Society.