|
|
Under-driving HPI: an alternative to clear mercury
|
The HPI Plus lamps from Philips are designed to operate either at low current on a mercury lamp ballast or at high current on the sodium lamp one (both of the same wattage rating as that of the lamp of course), thus making them quite flexible. What is little known, however, is that it's also possible to run them reliably on the control gear of next lower wattage/current rating. If we take the HPI-T Plus 250W shown here for instance, it is possible to run it on a 150 W sodium lamp ballast without any problem. The lamp voltage remains around 120 V, which keeps the drawn current below the rated one for that ballast (i.e., 1.60 vs. 1.80 A), thus resulting in reduced ballast losses (16 vs. 20 W). What is even more interesting is that the arc tube temperature drops low enough to cause the light color to shift towards a cool daylight tone, a tint very much similar to that of clear mercury lamps but with a better color quality, while the salt vapor pressure remains high enough for an efficient arc operation.
With a power dissipation of 185 W, this HPI-T Plus 250W has a greater light output than that of a 250 W clear mercury lamp (14.9 vs. 11.8 klm). So, that's a useful trick if you're looking to reproduce a typical pre-1950s mercury lighting atmosphere but can't find a clear mercury lamp (those have become quite rare). This is also very useful for garden/landscape lighting in order to enhance the green of plants and grass, an area where uncorrected mercury lighting used to shine. Note that this is possible only because of HPI's NTI-type of fill chemistry, whose blend of sodium, thallium and indium iodides results in a spectral emission which is similar to that of mercury, only with small shifts of the main line wavelengths. The reduced operating temperature balances the optical output of these additives in a way that results in a mercury-like light emission, and a more efficient one too. Moreover, this also extends the lamp service life considerably given the reduced rates of chemical reactions and of tungsten transport in the burner.
|
|