Ballast Design for Low-Power DC Arc Lamps?

[Wirth's Law]

Recently, I acquired a low-voltage direct current germicidal lamp. The lamp uses elemental mercury to generate the germicidal 254 nm ultraviolet C (UVC) wavelength. I don't have an exact part number, however it is functionally comparable to an UShio GTL3 (a 3-watt Germicidal UVC lamp with an E17 Base).

Unlike larger mercury vapor lamps that are used for germicidal or lighting purposes, this unit is very small: close to 1 cm x 3 cm. It also operates off close to 10.5 volts DC, but you need to briefly apply about 15 volts DC to start it. It is apparently a type of arc lamp, which means it has negative resistance characteristics that need to be ballasted.

Question: Does anyone know of any guidelines for how to design a ballast for DC light sources?

Note: I did confirm via dosimeter that the lamp does emit real ultraviolet-C at germicidal levels. Presumably, it also has ozone-generating emissions near 185 nm. To avoid injury from the UVC light or ozone, the lamp is remotely operated inside a shielded area. The estimated UVC dose from 3 cm away is between 50 mJ/cm2 and 100 mJ/cm2 in 30 seconds.

 

[unclejed613]

this is a gas tube. if it's srike voltage is 15V, and it's run voltage is 10V, and it's a 3W tube, it's simple to figure out, you need a 15V supply capable of delivering 300mA. your ballast resistor needs to drop 5V @ 300mA so it will be (5V/.3A)= 16 ohms, and since the dissipation is (5V*.3A)= 1.5W, use a 3 watt resistor for safety factor... the closest higher standard value is 18 ohms, so use an 18 ohm 3 watt resistor as the ballast. with the tube not conducting, the full 15V will appear across the tube and cause it to go into conduction, after that, there is current flowing and the resistor drops the excess 5V, so you have the tube with 10V across it.