Hey everyone,

Im new to this site and am glad I found it! A few days ago something interesting had happened to the 400w GE Multi Vapor lamp I had been lighting my garage with, over the last few weeks it began to slowly shift from the typical 4000K (cool white) to a warmer color and just a couple days ago it went completely pink, it honestly looks just like a pink/magenta color saturated halide lamp now. I removed it from the fixture and inspected the arc tube, its slightly blackened on the ends but still relatively clear. I suspected a capacitor failing may have caused the severe color shift. Tried the lamp on another ballast and it gave off the same pink light. Lamp was burned base up, has approx. 1000 hours on it. does anybody have any idea what caused this? Or did I just get a lamp with poor gas/halide salt mixture?

welcome to lighting gallery!

The modern GE metal halide lamps are poor quality this is typical..usually lamps turn green (underpowered) when the capacitor fails

It look indeed like some contamination, electrode fault (cause accelerated blackening).
And how the lamp is operated, how long per start? Don't forget, than with less than 5hours/start the dominant limiting factor become the number of start cycles, HID lamps are rated for 10hours/start burning pattern, so 25khours mean only 2500 starts...

What I would expect close the EOL is the arctube blackening causing the tube to run hotter than designed, so it evaporate more sodium and probably other elements too, so the color shift (excess sodium cause the shift towards warmer color,...).

Thanks Alights,

Lamp is ran for only a few hours in the evening, I know GE halides arent exactly the best quality but ive never seen a color shift like this before, and so quick. Im going to keep burning it just to see what happens. I thought a different halide salt was used in the pink-color lamps not found it the standard 4000K lamps, which should be just sodium-scandium. I'll upload a pic soon. But it really looks no different from the Venture designer-color magenta lamp I have...

The 400 W metal halide lamps that used to light the high school gymnasium where I teach would turn pink very suddenly, and then after about two days of pink operation, they failed upon startup one morning. Similarly, the 1kW MH floodlight at church did the same thing last week. It produced, nevertheless, an interesting effect with the pink light on the church. It was relamped, though, as it failed completely. In both situations, one lamp usually lasts about 2-3 years (at least). At school, they were used about 10-14 hours a day from September to June (they were replaced with T8  ) and the one at church is on dusk-to-dawn service.

All american Metal Halide (MH) lamps are based on sodium and scandium iodides, which optimised for Constant Wattage Autotrasformer ballasts and have very high efficency and stable color with these ballasts.
In Europe and Israel (Where I live), we use tri-salt indium-thallium-sodium iodides lamps for the cool white light and thallium-dysprosium-casium for daylight light. These chemistries are better suited for our simple reactor systems.
European compact double ended linear and single ended pre-focussed and bi-pin MH lamps, have either dysprosium-holmium-thulium-thallium-sodium iodides for cool white light and tin bromide + indium-thallium-sodium iodides for warm white light.
Ceramic Metal Halide (CMH) lamps uses the same halides as the cool white quartz MH lamps. The Philips cosmopolis lamp halides are unknown to me except a sodium iodide.
Most MH lamps have argon as a starting gas (The role exception is the low wattage MH lamps and CMH lamps that had also krypton85 in the past, but it was banned. The Philips HPI MH lamps have a neon-argon penning mixture, and the cosmopolis have a xenon starting gas.