Nowadays, compact fluorescent lamps are quite common in replacing incandescent lamps. The advantage of compact fluorescent lamps as compared to an incandescent lamp resides in their high efficiency and lifetime. They are nice and small and actually fit in the same fixture as most incandescent lamps.
Compact fluorescent lamps (cfl) are the small and `rolled up' version of their large and straight TL counterpart. These fairly long lamps are placed in a fixture which includes some additional electronic components. These components serve two purposes: they provide the high voltage which is necessary to start the lamp and they limit the current through the lamp when it is operating. In order to be able to operate cfl directly on the power grid (like it is possible for incandescent lamps) they need similar electronics, but they need to have it included within the lamp, and not in the fixture. In the figure below an early type cfl is shown: inside you can see the coil which serves to reduce the current through the discharge tube.
Modern cfl do not include such a large chunk of electronics as shown in the picture: instead they use smaller electronic components mounted in the base of the lamp. These components operate the tube at a higher frequency than the 50 Hz at which the tube in the picture operates.
As mentioned above, a part of the job of the circuitry in the lamp is to provide the high voltage which is necessary to start the lamp, i.e. to change the gas in the tube from being an isolator to being a conductor. The process is called breakdown and is relatively poorly understood. A good understanding of the breakdown process might make it possible to ignite a cfl at lower voltages, which would make the electronics in the lamp cheaper.
The simulations and the related experimental work on clf ignition are therefore aimed at understanding the breakdown process. The alternating voltage and the curved geometry of the tube make it difficult to interpret the interplay of the different processes taking place. In order to be able to separate the various mechanisms that could play a role in a real cfl, the system is simplified to a straight glass tube, operated on a DC voltage.
On the left one can see a `film' of iCCD camera pictures taken of such a lamp by Maxime Gendre. The electrode at 0 cm is the cathode, the one at 14 cm is the anode. At t = 4 microseconds after the voltage has been applied, an ionisation front starts to move from cathode to anode. When it reaches the anode (at about 12 microseconds) a `return strike' fills the tube with a fairly homogeneous emission. From then on it will take some time (several tens of microseconds) before the current rises and the lamp can be said to have ignited. Depending on the conditions striations can be observed.
By modelling the system (see the false-colour map of the calculated emission on the right) we think we gained a fair understanding of what happens in the phase in which the ionisation front moves from cathode to anode. For more information on this one can refer to W.J.M. Brok et al. J. Phys. D: Appl. Phys., 36(16):1967-1979, 2003.
The results for subsequent phases are less consistent with the experimental results and are under investigation at present. When these have been understood to some degree, the step can be made to e.g. high frequency operation of the lamp. In such operation the electrodes reverse polarity at a timescale which is smaller or comparable to the timescale of the ionisation wave's motion through the tube.