1996 Philips TL-D 18W 521-840

The lamp featured here has a design which solves a limitations inherent to all fluorescent tubes: their fixed light color temperature. This experimental tube was made for a study at the Technische Universiteit Eindhoven (the Netherlands) and it is built exactly like a standard 18 W tri-band /840 lamp, except for two things: the buffer gas fill consists of pure neon instead of the usual argon-krypton mix, and a pair of external electrodes run the whole length of the lamp. These aluminum electrodes are connected to the shell of the lamp’s end caps which are used as a terminals, besides the standard G13 pins connected to the internal hot cathodes.

The color temperature of the light emitted by this lamp can be tuned from 2100 to 4000 K depending on the relative power fraction coupled to the hot internal electrodes and the cold external ones. When current is fed at HF frequency (28 kHz) via the normal internal electrodes, the resulting discharge is dominated by mercury vapor and the normal excitation of the phosphor ensues, resulting in the emission of 4000 K light typical of standard tri-band fluorescent lamps. However, the situation changes considerably when the discharge is driven at radio frequency (13.56 MHz) via the external electrodes: the electron energy in the capacitively-coupled plasma increases so much that neon becomes the dominant emitter of optical energy, which consists of deep UV radiation and of orange-red light. The fluorescent layer is still excited to some extent and the resulting light output has a very low color temperature of 2100 K.

This unusual lamp design was first developed in Japan during the early 1990s and Philips became interested in it as the company was keen to improve the versatility of its fluorescent lighting systems. During the 1990s the Dutch were particularly active developing electronic control gears for fluorescent lamps with the intention to add new functionalities, such as dimming and presence detection. Color variation was an extra degree of freedom that Engineering and Marketting were interested on implementing. Various lamp designs were investigated to that end, including this HF/RF-driven neon-filled fluorescent lamp option. Although the light color can be varied over a wide range using this particular approach, the lamp efficacy drops considerably when neon dominates the discharge’s radiative output, going from 80 lm/W to only 20 as the light color temperature decreases. Neon does not convert electrical energy into light nearly as efficiently as mercury vapor. This, the high cost of RF electronics, and practical considerations related to an operation at radio frequency (i.e., EMI emission and parasitic capacitive coupling to the fixture) effectively prevented the technology from leaving the laboratory stage. Philips abandoned this approach completely after the study at the TU Eindhoven ended in 2000 since the problem behind the efficacy drop at low color temperature was found to be a fundamental one and could not be solved.