Murray would not divulge his friend’s name, but for practical purposes let us call him Don.
In his letter, Don said that Gunn diodes normally require less than a watt of power to operate, but that by working with the manufacturers they were able to engineer special high-power Gunn diodes suited to their project.
These were able to produce up to 10,000 watts of microwave power, and the various diodes that they fabricated functioned over a frequency range of 1 to 500 gigahertz.
Don did not specify whether this power rating referred to the average power of a single pulse or to the power that was put out when operating in a continuous mode.
The Gunn diode was first developed in 1963.
Later, much higher-power, more-efficient devices called impact ionization avalanche transit-time diodes (IMPATT diodes) were developed that were capable of higher microwave power outputs.
However, IMPATT diodes have the shortcoming that their signal has a much higher phase noise, meaning that their oscillation cycles may not be as precisely timed.
(A description of how these devices operate is given in the accompanying text box.) Such diodes have the advantage of being simpler to use than magnetrons, and they are more reliable in that they do not burn out as easily.
They are also able to produce much higher frequencies.
In the case of gallium nitride Gunn diodes, frequencies as high as 3,000 gigahertz have been achieved.
Commercially available Gunn diodes have an efficiency of only 2 to 5 percent, while IMPATT diodes have a somewhat higher efficiency of about 10 percent.
This is low in comparison with magnetrons, which are able to achieve efficiencies of 60 percent.
Gunn and IMPATT Diodes
The Gunn diode is named after J.
B.
Gunn, the physicist who in 1963 discovered that a crystal of gallium arsenide would spontaneously oscillate at microwave frequencies when a sufficiently high DC voltage was applied to either side of it.
This became known as the Gunn effect.
Gunn found that gallium arsenide exhibits a negative resistance when subjected to an electric field of greater than 3,000 volts per centimeter.
That is, below this critical threshold, the electric current passing through the crystal progressively increases with increasing voltage, as it does in most electrically resistive substances.
In this low-voltage region, the crystal is said to exhibit “positive resistance.”
However, at a critical threshold, the current-voltage curve plateaus and begins to bend downward such that the current now decreases with increasing voltage, a phenomenon called negative resistance.
Due to this peculiar characteristic, when a gallium arsenide Gunn diode is biased above its critical threshold at about 5 kilovolts per centimeter, it will spontaneously oscillate at a specific gigahertz frequency.
Some other substances found to exhibit this oscillatory effect are indium phosphide, cadmium telluride, zinc selenide, and wurtzite gallium nitride.
This latter material oscillates when biased at a higher potential of about 150 kilovolts per centimeter.
An IMPATT diode is a silicon p-n junction diode that normally operates in a reverse-biased mode.
Its principle of operation is different from a Gunn diode in that it involves impact ionization, which results in an electron avalanche electrical breakdown.
It is similar to the Gunn diode in that it is a negative-resistance device that begins to spontaneously oscillate when its bias voltage is increased past a certain point.
Commercially available Gunn and IMPATT diodes most commonly have power outputs in the milliwatt range, although it is possible for civilians to purchase Gunn diodes that produce up to 30 watts of power.
A thorough search for manufacturers of such solid-state oscillators carried out in the mid-1990s revealed that such diodes having power much higher than 30 watts are unavailable for public sale.
The story is quite different in the case of oscillators being used for military applications.
For example, in 2002 I learned through a personal contact that one defense aerospace avionics distributor had shipped out an experimental 1,040-gigahertz (300-micron) oscillator that reportedly had a rated output of 40 kilowatts!
This company routinely shipped items marked as “microwave oscillators” to defense aerospace corporations such as Northrop Grumman, Boeing, Lockheed Martin, and BAE Systems as well as to defense R&D contractors such as Raytheon and SAIC (Science Applications International Corporation).
A survey of cutting-edge developments in the field of solid-state microwave devices indicates that the high powers quoted for the modified Gunn diodes used in Project Skyvault are not all that out of line.
One literature review written in 1995 noted that gallium arsenide Gunn diodes were being combined to form units that could achieve the kilowatt level at frequencies above 30 gigahertz.
21
Also, in 2000 Purdue University researchers announced that they had developed a silicon carbide IMPATT diode that was able to achieve microwave power outputs four hundred times higher than silicon-based IMPATT diodes.
Their simulations projected the possibility of achieving power outputs as high as 4.2 kilowatts at 10 gigahertz.
22
A lightweight microwave-emitting tube called the Pasotron (for plasma-assisted slow-wave oscillator), developed in the early 1990s at Hughes Research Laboratories and Hughes Missile Systems Co., was able to achieve even higher outputs.
23
This uses an electron gun that generates high-energy electrons that emit a beam of microwaves as they pass through a low-pressure glow discharge.
The device produces microwave pulses lasting 100 microseconds with pulse voltages of 220 kilovolts, pulse power outputs of 1 to 5 megawatts, and effieciencies of between 20 and 50 percent.
More recently, Pasotrons have been reported to produce 100-nanosecond pulses with microwave powers of 7 gigawatts.
Although it is not a solid-state device, it has the advantages that it does not require a magnetic field for its operation, is much lighter in weight, and does not burn out easily.
It is not known whether the Project Skyvault team tested Pasotrons at some point in its research.
7.5 • THE BEAM AMPLIFIER
According to Murray’s friend Don, the high-power Gunn diode used in the Skyvault vehicle, like the magnetron, was mounted in a waveguide box.
This had an opening at one end and dimensions that matched the diode’s oscillatory characteristics.
In other words, the conduit’s length was made to equal some multiple of the wavelength of the microwaves emitted from the Gunn diode so that the waves would resonantly reinforce one another as they reflected back and forth along the length of the waveguide.
This resonance would increase the beam’s voltage.
Although Don did not mention their voltage requirements in his letter, Tom told me that he had learned that these special Gunn diodes were designed to operate in the range of a few hundred thousand volts to a million volts.
These voltages are unusually high in comparison with the voltages that commercially available Gunn diodes normally operate at, which is in the range of 5 to 100 volts DC.
One is left to wonder whether this voltage might refer to the voltage rating of the diodes, that is, the voltage they were designed to withstand that could be generated in the amplifier cavity.
The voltage of the amplified microwave beam, then, may have ranged up to several million volts.
As noted earlier, a simple waveguide cut to the proper dimensions would be able to increase the voltage of a microwave beam but not its total energy.
But in his letter, Don seems to be talking about a different sort of amplifier, one capable of increasing the total energy of the beam.
He said that this “amplifier” was needed to “extend the use” (i.e., the ability) of the Gunn diode so that it could “launch the .
.
.
vehicle” (see his letter in
appendix E
).
Although the modified Gunn diodes used in Project Skyvault had a power output far greater than those commercially available today, even a power output of 10 kilowatts would likely have fallen short of what was needed.
The magnetrons that the project had been using in their earlier work must have had power outputs several orders of magnitude higher than this.
So to match this, they would have had to boost the power of the Gunn diode beam in an “energy amplifier.”
Most likely the Skyvault project was doing this with a
parametric amplifier
, a device commonly used by microwave engineers to boost signal strength.
A parametric amplifier consists of a cavity containing a nonlinear medium such as a varactor capacitor diode.
The beam to be amplified is allowed to enter and exit the amplifier cavity through a port, and while in the cavity it passes through the diode where its energy is pumped up as the result of the action of a second microwave beam called the pump beam.
The pump beam typically has a frequency twice that of the main oscillator beam and is oriented at 90 degrees to the main beam so as not to directly interact with it (figure 7.9).
Figure 7.9.
Diagram of a microwave parametric amplifier.
(P.
LaViolette, © 2007)
The pump beam affects the main oscillator beam indirectly by varying the diode’s parameter, its dielectric constant, at just the right time in the main beam’s oscillation cycle.
*22
For example, by decreasing the varactor’s dielectric constant (K), the pump beam signal reduces the diode’s electric permittivity (ε); this, in turn, decreases the diode’s capacitance (C
∝
ε) and increases its voltage (V
∝
1/C).
By doing this at the phase of the cycle in which the main beam’s voltage is approaching either a positive or a negative voltage maximum, the amplitude and power of the main beam may be boosted.
In this way, the pump beam is able to progressively step up the power of the main beam.
Often, the parametric amplifier is interfaced with a circulator that allows the main beam to circulate in a loop with some of its energy being diverted into the parametric amplifier for amplification.
Parametric amplifiers are able to amplify a beam’s energy anywhere from one hundred to one thousand times.
Thus, a 10-kilowatt microwave oscillator signal could be boosted to create a 10-megawatt microwave beam.
One interesting thing about parametric amplifiers is that their energy output can greatly exceed their energy input.
The amount of energy that the pump beam requires to alter the permittivity or permeability of the amplifier’s nonlinear medium can be much less than the amount of energy that the amplified beam gains through the parametric excitation process.
The amount of this overunity output versus input depends on the type of nonlinear medium and its response in the frequency range used.
The magnetic resonance amplifier is an example of one such amplifier that operates in the audio frequency range rather than at microwave frequencies.
It is based on the work and theories of the ninteenth-century American inventor John Ernst Worrell Keely and has been extensively researched by hobbyists.
Circuit diagrams and research results on its operation are available on the Internet.
24,
25,
26,
27,
28
It uses a high-K dielectric such as a barium titanate capacitor hooked in series with a coil wound around a barium ferrite ceramic magnet core.
By exciting it at a frequency of around 20 to 40 kilohertz, this nonlinear tank circuit is made to oscillate at its resonant frequency of around 8,000 to 11,000 hertz.
Thus, the excitation frequency is chosen to be three times the resonance frequency, that is, three octaves (nine harmonics) above resonance.
Power is drawn from the oscillating ferrite core through a secondary winding that is connected to a bridge rectifier.
One such device built and tested by American researchers Joel McClain and Norman Wootan achieved a power output of 2.75 watts, for an input power of 0.7 watt or an overunity ratio of about 4.
29
At resonance, the voltage across the tank circuit ranged up to 1,000 volts when excited with a 30volt AC pump signal.
Even higher outputs than this have been reported for parametric amplifiers in the audio range.
For example, in 1949, Obolensky built a parametric amplifier that used Super Permalloy ferrite as its nonlinear medium and was able to achieve an overunity ratio of about a million to one when he pumped it at frequencies of 60 and 400 hertz.
30
Where does this excess energy come from?
Physicists aren’t really sure.
Obolensky suggests that the energy is cohered from noise present at the atomic level in the amplifier’s nonlinear medium and in the immediate space environment.
While it is possible to use a separate power source to generate the pump beam fed into the amplifier, it is also possible to draw off some of the energy surplus in the main beam that is being amplified and to recycle this to power the parametric excitation process.
This could be done by connecting the pump beam waveguide tube to a fourth port on the circulator cavity (figure 7.10).
The circulator would contain not only the fundamental frequency of the main beam but also its harmonics.
So by making the length of this connecting waveguide equal to an odd number of half wavelengths of the main beam’s second harmonic frequency, the fundamental frequency would be blocked and just the second harmonic (i.e., 2f
o
) would pass through to the parametric amplifier.
As the fundamental frequency becomes more intense, so would its harmonics, and a greater amount of power would become available in the second harmonic for parametric excitation.
As a result, the beam intensity would progressively increase.
Such a system, however, runs the risk of being unstable in that without proper regulation it could create an exponential buildup of energy that would ultimately result in an explosion.
That is, energy could be created in the amplifier faster than it could be removed.
Making such an amplifier work properly so that it is able to boost the wattage of the microwave beam without blowing the amplifier apart is quite tricky.
It requires ingenious engineering—such as incorporating a fast-acting servo control that automatically changes the phase of the pump beam frequency when the amplifier’s output power level gets too high.
This would squelch the amplification process and halt the exponential rise in energy production.
Figure 7.10.
A possible arrangement of components making up the microwave beam generator mounted on board the Skyvault craft.
(P.
LaViolette, © 2007)
Don did not mention the use of a sawtooth wave–shaping dielectric, so we do not know whether the wave-shaping dielectric was placed in the wave amplifier cavity or whether it was put in a separate wave-shaping resonator cavity.
We will suppose the latter was the case.
So once it had become amplified, the craft’s Gunn diode beam would have exited the circulator and entered a waveguide containing a polarized high-K dielectric that would have shaped the wave into a sawtooth waveform.
As in Brown’s vertical-lift electrokinetic apparatus, the incident microwave beam would have exerted a substantial electrogravitic thrust along the length of the dielectric that would have helped loft the craft.
Hence, this wave-shaping chamber was likely securely anchored so that its thrust would be transferred to the vehicle-support structure.
If the beam in the interior of the craft was directed upward through the wave-shaping dielectric, it may then have been made to pass into a convex slab of metamaterial having a negative index of refraction.
As mentioned earlier, a microwave beam tuned close to the material’s resonance would exert a strong repulsive force, so an upward-directed microwave beam would produce an upward propulsive force on the metamaterial slab as well.
Metamaterials have also been found to efficiently refract microwave beams through tight turns.
In fact, Pendry and Smith showed how a metamaterial slab having a convex lens shape refracts a beam through a 180-degree turn.
In a similar manner, the metamaterial thruster at the same time could have been used to redirect the beam downward through an adjacent waveguide, from which it would ultimately exit the saucer via a focusing lens and proceed toward the ground (see figure 7.10).
Chapter 8 further examines Don’s disclosure about Project Skyvault.
We will find that, besides pushing upward against the craft, it was necessary that the microwave propulsion beam also project downward and scatter back to the craft from a ground reference point.
In so doing, the beam could be made to resonantly store vast quantities of energy for supporting the craft, and the flight of the craft could be more precisely controlled.