Solar power Satellite

I think I posted a variation of this previously.
We have a sizeable group of alarmists who will oppose even a 1/4 sunlight energy beam, whether microwaves, monochromatic light or sunlight reflected from a mirror. Similar groups oppose most any change advocated by anyone. Abiding by the wishes of minorities has some advantages, but we need to realize that expediency is occasionally essential to provide even marginally for the soon to be 7 billion humans on this planet.
With cheap access to space we can perhaps build a square 4 kilometers by 4 kilometers = 16 square kilometers covered with solar cells or a mirror surface orbiting at an altitude of about 20,000 kilometers. This is tethered to a transmitting assembly. The tether typically has some slack so the two assemblies can be positioned separately. In the case of the photo voltaic cells, the giant panel is positioned to receive the sun's rays most efficiently. This may allow 20% 24/7 except for twice per year when the earth eclipses the sun briefly on about five consectutive days. The moon will eclipse the sun less often, I think. There will be small losses in collecting the electricity from the solar cells and the power line in the tether. With premium solar cells we can get 274 megawatts per square kilometer = 4.38 gigawatts for 16 square kilometers. If the cells are connected in series parallel, we can put one million volts dc at 4380 amps into the tether power line and receive perhaps 980,000 volts at 4370 amps at the transmitter end. I suggest ten traveling wave tubes in series (many series strings will be needed to use the 4370 amps) producing about 2.22 gigawatts of microwave RF. This will be fed through wave guides to three feed horns on a giant parabolic dish perhaps a kilometer in diameter. The feed horns allow three separate beams to three separate rectennas up to perhaps 1000 kilometers apart on Earth's surface. The beams will deliver about 0.6 gigawatts = 600 megawatts to each rectenna during the peak demand period early each evening. The wave guides have low loss, the transmitting feed horns lose 10 %, the beam scatters 10% of the energy and the rectenna converts 80% to low voltage dc. The rectenna dipole diode outputs are connected in series to produce about 1000 volts dc input to each inverter. The inverter outputs are in series to produce about 400,000 volts 60 hertz single phase ac or whatever the grid voltage is near the rectenna. A multiple of three series strings are needed to to produce the three phase power for the grid. If the rectenna puts 400 megawatts on the grid, several thousand inverters are needed in many series strings.
You noticed that some power is lost at each step, likely more than I optimistically suggested. If we can't run the inverters in series, we need three huge transformers with thousands of low voltage primaries /one for each inverter output. With so many primaries and a very high step up ratio, and very high output voltage, we will be lucky to get 80% efficiency from these transformers.
400 megawatts put on the grid may be close to optimum at most rectenna locations as more means some big power plants on the surface will produce less than full output at peak demand. The system can be scaled up by adding more area to the giant solar panel in the sky, by increasing the dc power line voltage to more than one million volts, and by adding more klystrons in each series string. The rectennas can likely handle a much stronger beam, but the risk to people will be increased slightly and we will need more inverters.
Addendum
Magnetrons are more efficient than klystrons, but klystrons can be modulated with broadband data, which can be received over about 1/2 of the solar system, plus about 1/2 of Earth's surface with a modest antenna due to the 180 megawatts plus that is scattered by the three beams, Earth's atmosphere and reflected by the three rectennas. This communication bonus may be worth more than the electricity delivered.
Instead of solar cells a mirror the same size can send a beam of concentrated sunlight to the transmitter assembly, where a heat exchanger makes steam which turns a turbine which turns a dc generator which powers the klystrons that produce the three beams etc. This system may be slightly more efficient (and lower initial cost) than the solar cells, but it will require human oversight at the satellite, whereas the solar cell system can possibly be serviced by robots.
The three beams can possibly power three spacecraft going in approximately the same direction, be used to deflect slightly an asteroid or comet threatening Earth, power a colony at L1, the moon or other uses, while the satellite is over the Atlantic or Pacific ocean. The cities depending heavily on the satellite during peak demand may experience rolling blackouts, if their beam is pre-empted for other uses. Neil