Metal Halide Flicker

[pSlawinski]

Ok, I have a 1000 watt Metal Halide lamp from Eye Lighting International.? I have it running a lot because I enjoy very bright lights.? I have noticed when I run this lamp there seems to be some kind of color shift occuring.? The light gets slightly red then it gets slightly blue.? This color shift occurs in a sinusoidal pattern.? Does anyone know what might cause this?

[Alan Franzman]

While I haven't heard of this exact symptom before, metal halides are notorious for color inconsistency. Several years ago when I was still an apprentice electrician, I worked on an installation in an old building somewhere in the L.A. area - don't remember exactly where - the ground floor was a very tall atrium-like space, some 30-50 feet high with interesting architecture, all painted white, and the new tenant had a large square "island" of cubicles in the middle.  The cubicles were not the typical modular "furniture" that is used now, they were bult in place by carpenters. We electricians installed several very bright MH fixtures atop the cubicle wall intersections to illuminate the ceiling architecture, and I noticed immediately that they changed color during warm-up (no real surprise), and that after warm-up it looked like they were a mix of pinkish and greenish shades (slight surprise to me but not my journeyman). The architect of the remodel was apparently aware of the situation, so the customer was not dissatisfied - I guess it added some amount of interest to all the white space.

The only way I can think of this having any bearing on your situation is that the temperature might be fluctuating - if the fixture is not in a draft, perhaps the bulb is going bad (or maybe even the ballast, but I doubt this).

So to sum up, if you want a specific color or several lights to all be the same color, (or apparently even one light to stay a constant color...) don't use metal halide.

[pSlawinski]

Well I know this color change is not due to age, the lamp is brand new.  I do not run this in a fixture designed for a metal halide lamp, rather an old mogul base floor lamp.  I have the lamp plugged into my 1000 watt metal halide ballast.  I do realize the dangers of running a metal halide lamp in an unenclosed fixture, but I do not run this for very long, perhaps a few hours at the most, and as I stated earlier this lamp is brand new.
http://www.metal-halide.net/index.php?id=gallery&cat=HID&scat=MH&bulb=M1000BX-U

[Max]

Hi,

To me, the only reason why a MH lamps would experience a repetitive variation of color would be an unstable arc. Have you tried to observe the lamp arc during operation? If not try with an arc welder protective goggle, and see if the arc is stable or moving continuously...

Max

[pSlawinski]

I checked the lamp out today and it does wobble a little, but not continuously.  Most of the tube was covered in the halide salts (I operated base up yesterday, today I ran it base down), so I couldn’t see much.

[Alan Franzman]

I do realize the dangers of running a metal halide lamp in an unenclosed fixture...

That's probably the source of the problem - bulb is getting too much air circulation, keeping its envelope too cool. There's a slim chance that the lamp wiring could be part of the problem too but I doubt it - I've worked on street lamps with ballasts remoted all the way to the bottom wiring access compartment, and they didn't seem to mind the extra wire length, though I think they were high pressure sodium, not metal halide.

[pSlawinski]

I think it might be the envlope.   I have run this off a socket with 3" leads before and noticed the same problem.  I also noticed that the envlope  has  a spot where it looks like there was a bubble or something when they made it.

[Max]

Most of the tube was covered in the halide salts (I operated base up yesterday, today I ran it base down), so I couldn’t see much.

That's strange! Normally, as you operate the lamp base up or down, most of the salts should form a pool in the coldest region of the arc tube, that is in its lowest part. Are you sure you operate the lamp in the optimal regime? Have you checked that the lamp is fully compatible with the ballast?
Arc helical instabilities originates either from a bad arc tube design (which can be accidental), or from an improper ballast. Some high-wattage mercury lamps show this substantial instability when ran at lower power.

A for the outer envelope, this is not likely the cause of the problem. The arc tube is thermally protected from the outside world by an atomosphere of nitrogen, so the cooling of the outer bulb by natural convection in a room does not present any problems. Having run some 2kW MH lamps, I never encountered this problem.
Another possibility is that there is? maybe a leak in the outer bulb that fills it with air. This would increase the thermal losses of the arc tube and decrease its operating temperature and internal pressure. A possible candidate for the onset of arc instabilities.

Max

[pSlawinski]

I do know that the ballast is compatible with the lamp.  It was designed specifically to run a 1000W metal halide lamp.  I suppose there could be  a leak, however the arc tube supports show absolutely no sign of oxidization, nor does anything else.  I also imagine that if there was a leak the outer envelope would run much hotter than normal.  I have measured the temperature of the outer envelope during operation and the maximum temperature is usually around 300F and 400F.  As for the salts, I ran the lamp base up the night before, allowing the salts to condense in the top of the arc tube (it was the bottom at that time).  The day I took the pictures the lamp was operating in a base down position, so there was a migration of the halide salts.  During that time the entire tube was covered with the salts.

[Max]

I do know that the ballast is compatible with the lamp.? It was designed specifically to run a 1000W metal halide lamp.?

There a different ballasts that can different 1kW metal halide lamps. Iwasaki's specs for your lamp states that you must use a MH CWA ballast without ignitor, or a mercury reactor ballast with ignitor. Is that the precise case for you?

? As for the salts, I ran the lamp base up the night before, allowing the salts to condense in the top of the arc tube (it was the bottom at that time).? The day I took the pictures the lamp was operating in a base down position, so there was a migration of the halide salts.? During that time the entire tube was covered with the salts.

So when the salts have settled in the coldest region, and the arc tube is clear, do you observe some subtantial wobbling of the arc. If so, does it coincide with the change of color.
Have you tried to operate the lamp in a non-vertical position? If so, are there still some cyclic variation of lamp color?

Max

[pSlawinski]

Yes the ballast is a CWA Metal Halide ballast.  Most of the salts have settled, but I'm going to have it run for a while and make sure they have all settled.  I will post pictures of the whole setup later this evening.  Unfortunately, I have no means to run this in a horizontal position, aside from holding the hot ceramic base in my hand.  There is no way i could hold it long enough for the salts to settle.

[pSlawinski]

I have it running behind me, and I can see the arc wobbling again.  It seems that the arc is slightly bowed at the bottom towards the tube wall.  It rotates at random intervals to point in different directions.  The best way to demonstrate this would be to take a string and hold it above a flat surface.  Then lower the string slightly so that the bottom touches the flat surface and the end is slightly bent.  Finally, twist the string so that it rotates slightly.

[pSlawinski]

I have had the lamp running for about an hour now, and the color shifts seem to have ceased.  I checked and sure enough the arc has become stable.  I stood next to the lamp looking through the welding goggles for a good 5 minutes and the arc did not move at all.  I doubt this is because of the halide salts settling I have run the lamp with the salts all condensed in the bottom of the arc tube before and noticed the color change before.  Perhaps the lamp take a long time to fully warm up.  I don’t know why it would take 30 minutes or more to warm up though.

[Max]

This instability may originate either from impurities in the arc tube that are cleaned off after a sufficient runtime. Or else, the arc tube geometry and fill is such as the lamp is highly sensitive to the migration the halide salts, and hence stabilises in operation when the salt pool has completely setteled in the coldest regions...

Max

[pSlawinski]

Could arc instability be a characteristic of this lamp?  I have run three lamps of the same type since posting this topic.  Each lamp showed some arc wobbling during warm-up and sometime after warming up.  I ran one brand new lamp, one slightly used lamp, and one heavily used lamp.  The new lamp and the slightly used lamp were almost the same as my original lamp.  The heavily used lamp was by far the most unstable of the group of lamps I ran.  The arc wobbled all around the arc tube during warm-up and after the lamp finished warming up it ran along the side of the arc tube.  I think it is worth mentioning that I ran all of these lamps were run in the base-down position.  The two used lamps were run horizontal before I got them.