The design called for a voltage source of 20kV. I obtained an old CRT television and drilled holes in its back, through which I strung wires soldered to its flyback transformer and to a ground plate inside it. This provided me with 30kV to work with. Finally, about two years ago, I built the first of many prototypes.
Early designs were yielding no movement. I experimented with different dielectric materials, and with different plate sizes, shapes and arrangements. I finally realized that in my calculations, I had not accounted for friction. The rotors I was using up to that point were all heavy enough that any force they might have generated could not overcome the friction in their bearings. Thus, I redesigned and rebuilt the engine with lighter materials, eliminated as much weight as possible, and experimented with various lubricants.
Finally, in December of 2016, a prototype began spinning. This convinced me that the principle behind the design itself was sound. Unfortunately, it did not spin nearly as fast as it should have. By this point, I had measured the friction and factored it in to my calculations. Thus, after several tests, I decided the problem must be the amount of force being generated.
I have since built a test apparatus to measure the generated force directly. I attached rotor plates to a wooden beam, aligned the stator as necessary, and placed the beam on a digital scale. When I power it up, the force pulls the beam downward and the scale measures the force as an apparent increase in weight. Thirty tests showed me that I was producing an average of about 26% of the force I had calculated.
I had few ideas for adjustments that might increase this. I made one such adjustment, ran 15 more tests, and measured an increase in force of about 3%. However, it was around this time that I also realized my source voltage was fairly unstable.
The tests were yielding slightly different results, so I had been averaging them together, but I decided to measure the source voltage without the test apparatus attached to see if it might be part of the problem. I had done this a few time before, but on this occasion, I decided to do a more thorough test. I turned the TV on and off several times (making sure to completely discharge it each time), and I left it on for periods ranging from 10 seconds to over a minute. In each test, the voltage was somewhat different, and it changed over time somewhat differently with each test. I have therefore decided that I need a new power supply—one that is more stable and reliable—to continue my work.
The power supply I have in mind is a 25kV supply from Analog Technologies Inc. The picture below shows the model and cost.
