Let’s look at the homework. Once every 15 cycles, all four crystals line up. Once every 375 microseconds, all the crystals are pushing. So we could expect an audible tone at 2667 Hz. There is also a half-height reverse pulse in the middle of these, which combines to a second harmonic. In fact, a variety of pulses interact with the nonlinearities in the gadget itself to create all kinds of audible tones. Quoting CNN again,
This is about what these frequencies sound like. They are generated in the gadget itself, and by mixing of ultrasound inside the dwelling. Generation in the gadget is serious, because then we have a noisemaker in our hands, on the street, late at night.
We would like to get rid of the sounds in the gadget itself, but it is bound to be unstable, with constantly shifting frequencies and resonances, depending on internal tensions of the parts, ambient temperature, warming up, cooling down, and aging of the materials. We can put our brain fryer in a double or triple-wall case. But the radiator exposes the internal noise. Don’t underestimate Russian ingenuity. But an efficient acoustic mechanical high-pass filter may be impossible. A car muffler is a much easier low-pass filter.
Noise can be masked by spreading the spectrum. Some computer motherboards have an option to spread the clock frequency, to avoid interference with nearby devices. But since the power efficiency of this gadget derives from resonant principles, it would be complicated. If the amplifiers driving the crystals can recover reactive power, it might be possible to keep most of the power in the ultrasonic range, while spreading the audible sounds into Gaussian noise. With an operator switch to turn spread spectrum on and off, it would explain why some of the attack victims heard nothing at all.
That’s a tall order, so let’s consider if we could null out the unwanted sounds instead. How can we do this with a gadget that is constantly changing its notes? There are several possible approaches.
- In the coupling between the crystals and the radiator, include a 5th piezo to serve as both the input and output ports of an active, causal filter. Unlike a purely mechanical filter, it doesn’t soak up a lot of power.
- We can also do something with the other four crystals. Each crystal is driven by an individual power source. Instead of sine waves, we can use signals derived from a static model of the system, reducing output of the unwanted audible sounds.
- But we’re going to need a microprocessor anyway, so why not try to anticipate the unwanted sounds, and be more specific about reducing them electronically?
We can do this with a predictive algorithm that adapts in real time. Since we don’t want an earsplitting racket, even for an instant, this implies, on top of our pulse scheme, a second time variation, a ramp. On the scale of seconds, we ramp it up. This solves overheating too.
- As the gadget ramps up, starting from a whisper, the predictive algorithm takes note of audible sounds, and synthesizes an adjustment to the four crystal drive signals.
- With time allowed for the microprocessor to think, the correction is applied on the next cycle, or the one after.
- If the ramp up is slow enough, the error from this look-ahead is minimal.
- At maximum intensity, the gadget heats up. The algorithm continues to compensate until a limit is reached. Then the ramp drops to zero, the device cools a little, and the ramp repeats.
Everything about this problem is a little dirty, physically as well as motive. How do we know the window glass is vibrating to the max? Or if there is no glass, how do we get an idea of what is going on inside the room? Since the gadget has some frequency agility, we can tune it to the situation. But how do we know what to tune to?
If there is glass, traveling wave vibrations can be detected with a laser microphone, using an interferometer to measure the minute changes in path length. If there isn’t, then we listen, with a microphone for audible mixing products, which by signature are produced inside the dwelling.
This is an elegant yet simple approach. The ancestral gadget may have existed well back in the 20th century, held back by excessive audible noise. The American approach might be twice as good, and 10X the cost. My first thoughts were something akin to a traveling wave amplifier, or a photonic pumped system like a neodymium fiber amp. The pulse would be accurately synthesized, made to order with great flexibility, and amplified by a pumped system. But how much precision do you need to fry a brain?
Since that distant time of origin, numerous developments have disseminated to second-rate technical powers and even developing countries:
- Switching amplifiers.
- High frequency power converters.
- Supercapacitors and low ESR capacitors.
- Lithium batteries
- An incredibly wide variety of ceramics, and other materials with special properties.
This means that practically nothing electronic has to be the size of a car anymore. Current and power densities are immensely higher than just ten years back. The limit is now heat dissipation, not size. We may want to include a Peltier cooler.
Let’s party, comrades! I will get the Order of Lenin for this! What do you mean, “They don’t give it anymore”? Come and celebrate anyway. This evening we get drunk on Stoli instead of hooch. I don’t see any company, so I’ll just suck my pinkie like Dr. Evil. But how can I make some dough out of this? How about a low-fat fryer for late night TV ?
We’re done with the technical. If the device exists, this discussion might help, along with the Havana investigations, to discourage the use of it. It will be very discouraging if the F.B.I. determines that something like this was in fact used in Havana.
If an adversary concluded that frying the brains of Havana diplomats would be a profitable thing to do, we have only ourselves to blame. In what intelligent manner, devoid of emotional considerations, beneficial to us, our values, and our allies, should this influence foreign policy?
This will come in an article of the near future.