In art. said:
As I've mentioned here several times: Take the wheels out from under a
lighting installation and I understand a great deal less of the science
involved. So, what causes each phase of the cycling observed at and beyond
end of life in HID lamps including (but not limited to) HPS and MH? I may
not even be using the correct term. The phenomenon manifests in phases
thus:
(1) The lamp begins to emit a red-orange glow, which gradually increases
in intensity.
(2) The arc starts, sometimes after several failed attempts. Arc colour
changes fairly rapidly (apparent CCT drops in HPS and rises in MH, for
instance) until the lamp is producing light of approximately the
normal/expected color
(3) The arc stops, and the capsule cools down as evidenced by the gradual
dimming of the red-orange glow within the lamp.
(4) GOTO (1)
Curious, DS
I can certainly say how this works with HPS:
HPS arc tubes are not perfect at containing sodium. Sodium is extremely
chemically rective, and under normal conditions some will react with the
arc tube (and get "lost" as a result). Some will alloy into metal parts
of the arc tube, and some will permeate out of the arc tube. All of the
"alkali metals" (metals of the first column of the periodic table) are
chemically very nasty. They even have a tendency to spontaneously combust
if dropped into water, corrode in seconds if exposed to air, and are
typically stored submerged in kerosene or some other petroleum oil (I
don't know how they do that with lithium, which floats on anything that is
liquid at room temperature and atmospheric pressure).
So HPS arc tubes have a surplus of sodium. That means that the "amalgam
reservoirs" at the ends of the arc tube have to be at or near some proper
temperature in order for the sodium vapor pressure to be what it should
be.
But when the arc tube ages, it is "blackened" at the ends, and these
dark regions absorb light and get hotter and make the amalgam reservoirs
hotter. That increases the sodium vapor pressure. If this happens
enough, then the arc's electrical characteristics change so badly that the
ballast cannot maintain the arc.
As for metal halide: I suspect two things:
1. The arc tube gradually blackens and gets hotter, vaporizing more of
the halides (which there is normally a surplus of).
2. Some halogen from the halides corrodes the electrodes, making them
less efficient and that increases the arc voltage, as well as arc tube
heating which increases vaporization of halides and you get more arc
voltage and more extreme arc "negative dynamic resistance".
2a. Corrosion chops an electrode from its lead, and the arc has increased
voltage drop along with adding extra heat to one end of the arc tube. I
suspect this is how some aging metal halides that are supposed to have
color temperature near 4000K or in the 4,000's K turn pink in color -
anyone please correct me here? I do see some metal halides of color
temperature in the low (maybe mid) 4,000's K somewhat suddenly have a
color change from white to pink, and they usually completely conk out a
couple weeks to a few weeks afterwards. I don't see them cycling, but
maybe with a different ballast they may cycle.
I see cycling mainly with HPS, hardly ever with other HID lamps. Most
cycling that I see with non-HPS HID lamps involves either lamps that are
failed and cycling before full warmup, or having clues that indicate
excessive voltage drop in the wires feeding them or something else wrong
such as lamp (bulb) / ballast mismatch or damaged ballasts or
(less frequently) ballasts incorrectly wired (wrong voltage tap used or
improper use/nonuse of a capacitor that is intended to be in series with
the lamp socket).
- Don Klipstein (
[email protected])
