The experimental setup includes an electrical circuit for generating alternate currents of high voltage and high frequency.
There appears to be a short, pulse-like amplification of light intensity at the instant of switch closure. At voltages under 25 V (with 10 MΩ resistor), the circuit may have to be jump-started by short-circuiting the transistor's collector to its base/emitter with a quick tap.
At higher voltages, the intensity of light and lighting up distance increase, but at the same time the transistor gets very hot. The performance of the circuit seems to get better by adding a couple of switching diodes (for example 1N4001 or 1N4148).
After the first successful tuning (added a couple of diodes and changed the resistor to 10kΩ) there is no need to jump-start the circuit. The transistor still heats up at higher voltages, but within reasonable limits. The system is exciting and there is a lot to experiment and learn.
Thanks to Lidmotor and Michael Judd for the inspiring YouTube videos.
"I have made the discovery that an electrical current of an excessively small period and very high potential may be utilized economically and practicably to great advantage for the production of light."
Nikola Tesla 1891, System of Electric Lighting Patent.
According to George Trinkaus (1989) the primary and the secondary coils should be wound in the same direction. The primary coil has a role in the fine tuning the tesla coil. "One old rule of thumb says the primary and the secondary should have equal weights of copper." (1)
Although the experimental system is not exactly tesla coil, the same fine tuning methods seems to apply. The new primary coil has 8 loops of 2.15 mm diameter (~AWG 12) copper wire. The coil height and diameter is 80 mm. The fluorescent tubes generate soft light, and during low voltage operation, the tubes produce magical looking flows of varying light intensity.