HPL Lamp question

[firewater88]

I have attached a pic of a lamp out of our lobby lights, I have 37 ETC S4 50 degree fixtures as my main down light in my lobby. I use 575X lamps and get pretty good life out of them. My question is: can dust make the lamp burn out like this? The fixtures are usually covered in dust when I change the bulbs. I have just noticed that several of the reflectors are flaking off on the inside. All the ones I found had a lamp that looked like this, this being the worst. It does have a slight bulge on the side. This one and the others were installed about 18 months ago (I date them).
Wondering if it is getting super hot right before the lamp goes and that is what is causing the reflectors to flake. These are not on during the day, my 200' x 60' glass wall lets in enough light.
Any thoughts? or is it just age? These lamps? My theatre lights don't suffer this much damage and they are much older than these units.

 

[Les]

It looks to me like the lamp lived out its normal life. Look around the site and you'll see much more dramatic failures. The flaking reflectors need to be replaced (they begin to flake in their old age -- again, a normal end-of-life effect). Dust buildup on the outside of the instruments probably didn't directly cause either issue, however it will cause the fixture to retain more heat.

It looks like you have three separate issues that while in need of address, are more related to age/duty cycle than dust.

As for your theatre lights outlasting these; take in to account the fact that these instruments may be used at full intensity for longer amounts of time. Burn position (fixtures pointing down) could also have something to do with it.

 

[derekleffew]

Sorry to say, I have more questions than answers. More a collection of random thoughts than anything.

Do you only use Ushio lamps? I would guess that the ceramic lamp base would be less efficient as a heat sink than the metal bases of other brands, but since you're not experiencing pinch seal failures, that may not matter. When the lamps are on, are they at Full? If yes, what is the voltage at the socket? Do you use HPL120X or 115X? When you change burned out lamps, is the filament broken in only one spot? Have you ever experienced an envelope rupture? Do you also inspect pins and sockets? Do you bench focus the fixture with every lamp change?

As for flaking reflectors, from KnowledgeBase: Flaking Source Four Reflectors - Electronic Theatre Controls :

Source Four reflectors will flake if the reflector has been cleaned with an abrasive cleaning material such as bleach or Ajax. This will damage the reflector and cause it to flake when heat, such as an activated lamp, is applied.

See also http://www.controlbooth.com/forums/lighting-electrics/8981-reflector-removal.html . Reflector cleaning abuse isn't the only reason for a reflector to lose its dichroic coating. Age and heat have a lot to do with it. Eighteen months seems like good lamp life in a theatrical fixture used architecturally. Maybe your lamps should fail more often so you'd clean the luminaire s more often?

While dust will act as a thermal insulator, causing the fixture to run slightly hotter, I don't think it's a significant contributing factor. The gate temperature of a SourceFour is 600-900°F; skin temp. of rear body is 423° max. (source); I'm guessing the envelope's surface temperature would be around 1200°F. Your theatre fixtures don't have the same duty cycle and lead a much less stressful life than the lobby fixtures. Have you had any rear body paint failure (a big problem with early units used architecturally)? How old exactly are the lobby instrument s? Are they the 750W style?

In summary, minor lamp envelope distortion is nothing to worry about in a fixture that burns for long periods of time and never moves. Reflector flaking happens, even to good people. Try to dust your units more often, but more to remove a potential fire hazard than any other reason.

 

[JD]

Normal burnout and/or the bases are too cool causing the deposit. (Near air vents?) If these fixtures are not on a dimmer and are on for extended periods of time, then the HID version of the S4 would be more appropriate. The HID lamped fixture and ballast are more expensive, but are designed for architectural applications as HID lamps have many times the lifespan. As for the reflectors: Again, if the are on long enough then this is what happens. Again, the HID fixtures run cooler and should solve this problem.

 

[firewater88]

I do run Ushio lamps in the lobby, HPL115X is what I run (I run Sylvania HPL in the theatre). The units are 7 years old and yes they are the 750 variety. They are controlled through an ETC Unison dimmer setup. I do NOT run them at full, ever. All the presets run them around 90 %. They are 30' off the floor, with another 30' or so above them to the ceiling. There is plenty of air space around the fixtures and no direct air blowing on them. The building is just inherently dusty. I have never used any chemical cleaner on the unit or reflector- only a micro fiber cloth.
Typically when I change lamps, I inspect the fixture, clean off any dust on and clean cobwebs out of the barrel. None of the fixtures have experienced the paint flaking on the can. I have had a few bases burn out and replaced those out. Normally the bulb is broken in one spot, but these (I had three of them like this) were the only ones this change round out of 7 that looked like this. All the ones that looked like this I noticed flaking reflectors.
Unfortunately, after a few cuts when building was opened, these are my primary lobby light source. I have some generic fixtures that point up for accent, but these were supposed to be accent/effect lighting, but wound up being primary lights. I would love HID lights, but I do need them to dim- so they are out.

Thanks for the questions/thoughts- no matter how random you think they are!

Sorry to say, I have more questions than answers. More a collection of random thoughts than anything.

Do you only use Ushio lamps? I would guess that the ceramic lamp base would be less efficient as a heat sink than the metal bases of other brands, but since you're not experiencing pinch seal failures, that may not matter. When the lamps are on, are they at Full? If yes, what is the voltage at the socket? Do you use HPL120X or 115X? When you change burned out lamps, is the filament broken in only one spot? Have you ever experienced an envelope rupture? Do you also inspect pins and sockets? Do you bench focus the fixture with every lamp change?

As for flaking reflectors, from KnowledgeBase: Flaking Source Four Reflectors - Electronic Theatre Controls :
See also http://www.controlbooth.com/forums/lighting-electrics/8981-reflector-removal.html . Reflector cleaning abuse isn't the only reason for a reflector to lose its dichroic coating. Age and heat have a lot to do with it. Eighteen months seems like good lamp life in a theatrical fixture used architecturally. Maybe your lamps should fail more often so you'd clean the luminaire s more often?

While dust will act as a thermal insulator, causing the fixture to run slightly hotter, I don't think it's a significant contributing factor. The gate temperature of a SourceFour is 600-900°F; skin temp. of rear body is 423° max. (source); I'm guessing the envelope's surface temperature would be around 1200°F. Your theatre fixtures don't have the same duty cycle and lead a much less stressful life than the lobby fixtures. Have you had any rear body paint failure (a big problem with early units used architecturally)? How old exactly are the lobby instrument s? Are they the 750W style?

In summary, minor lamp envelope distortion is nothing to worry about in a fixture that burns for long periods of time and never moves. Reflector flaking happens, even to good people. Try to dust your units more often, but more to remove a potential fire hazard than any other reason.

 

[JD]

They are controlled through an ETC Unison dimmer setup. I do NOT run them at full, ever. All the presets run them around 90 %.

BINGO!
Although 90% sounds like it should not be a problem, dimmers do not use a liner curve. Halogen lamps require a certain envelope temperature to be reached before the halogen cycle works. As the quartz envelope toward the rear of the lamp was not reaching that temperature, there is where the tungsten got deposited! (Thus, the odd look to the failed lamps.)

 

[firewater88]

BINGO!
Although 90% sounds like it should not be a problem, dimmers do not use a liner curve. Halogen lamps require a certain envelope temperature to be reached before the halogen cycle works. As the quartz envelope toward the rear of the lamp was not reaching that temperature, there is where the tungsten got deposited! (Thus, the odd look to the failed lamps.)

So, even though I get a good life out of the bulbs, should I change my percentages to reach the envelope temp? If so, what would you recommend? Or just keep everything at par for now and keep an eye on the fixtures.

 

[derekleffew]

BINGO!
Although 90% sounds like it should not be a problem, dimmers do not use a linear curve. Halogen lamps require a certain envelope temperature to be reached before the halogen cycle works. As the quartz envelope toward the rear of the lamp was not reaching that temperature, there is where the tungsten got deposited! (Thus, the odd look to the failed lamps.)

JD, while I, kind of, agree with your hypothesis, a couple of points defy logic:

1. These lamps are burning base up, fixtures likely straight down.
2. Is the temperature of the envelope nearest the base really that much cooler than nearer the filament, considering that heat rises?
3. Interesting that the envelope toward the tip from the filament is perfectly clear. If this is the hotter portion, wouldn't the bulge be there?
4. Distance from center of filament to top of base in an HPL is ~34mm.
5. We've had this discussion before, can't remember where, but the voltage needed to activate the halogen cycle is significantly lower than 90%, perhaps as low as 30% of rated voltage?
6. The Source Four's dichroic reflector removes most, but not all, of the infrared energy (heat) emitted from the lamp.

 

[DELO72]

Hi everyone,

Derek gave me a poke to get me involved and put in some input. I happen to be sitting next to my quality manager and head of engineering at the factory where are HPLs are made at the moment. I showed them the photos and their take on this is that those lamps are showing a complete lack of Halogen/Bromine. The belief is that they are failing early because the halogen cycle is not happening at all. Usually this is due to a clogged or not-working exhaust machine, and as a result the lamps were made and no fill gas was added to the capsule. This results in the lamps operating in a vacuum instead of a pressurized atmosphere (with those HBr gas mixture to allow the halogen cycle to occur).

As for the halogen cycle, I think the lamps need to run at full for the halogen cycle to work-- that said, it doesn't need to work ALL of the time. I always recommend doing a warm-up cue that at SOME point during the night brings the lamps to full for a minute. This allows the Halogen cycle to completely operate, and can "clean" the lamp. Then, if you are only running the lamp at 20% during the show, it's still fine as the halogen cycle at least gets to operate fully at some point each night.

But- getting back to the original issue- the theory based on the images are that you simply got a bad batch and can probably return them and get replacements. It can happen to any one. We make hundreds of thousands of HPLs each year, and I hate to admit it, but even some of ours have problems from time to time-- sssh! but don't tell anyone.

Cheers,

Mark

 

[derekleffew]

Thanks Mark, but I think you missed a key point in the original post:

...All the ones I found had a lamp that looked like this, this being the worst. It does have a slight bulge on the side. This one and the others were installed about 18 months ago (I date them). ...

(Emphasis added.) Surely improper fill during manufacture wouldn't allow a lamp to last 18 months? So could a pinch seal failure have allowed the halogen to escape? Once the halogen is gone and oxygen enters, wouldn't the filament break dramatically?

 

[JD]

I actually stand by my statement. When you get a shiny tungsten deposit, the halogen cycle is not working. Yes, less than an inch can make all the difference in the world. As for the lamps not getting a gas charge, could be! This would be the point of negotiation if you are trying to return them.

Now, as for lifespan, you may actually have gotten a real good lifespan out of them. Don't know. The cooler the filament runs, the less tungsten "boils" off, which increases life. As the lamp reaches full intensity, the "boil off" is compensated for by the halogen cycle. There is a danger spot where the filament is hot enough to be evaporating a lot of tungsten, but not hot enough for the cycle to be re-depositing it. The photo provided is almost textbook material. Close to the filament, the cycle is working, as you move closer to the base, it is not. (Otherwise the whole thing would be coated.) Near the base and pinch seal, the lamp runs the coolest.

If you know for a fact that the lifespan was poor, switch to a lower wattage and run them at 100%.

 

[firewater88]

Hi everyone,

Derek gave me a poke to get me involved and put in some input. I happen to be sitting next to my quality manager and head of engineering at the factory where are HPLs are made at the moment. I showed them the photos and their take on this is that those lamps are showing a complete lack of Halogen/Bromine. The belief is that they are failing early because the halogen cycle is not happening at all. Usually this is due to a clogged or not-working exhaust machine, and as a result the lamps were made and no fill gas was added to the capsule. This results in the lamps operating in a vacuum instead of a pressurized atmosphere (with those HBr gas mixture to allow the halogen cycle to occur).

As for the halogen cycle, I think the lamps need to run at full for the halogen cycle to work-- that said, it doesn't need to work ALL of the time. I always recommend doing a warm-up cue that at SOME point during the night brings the lamps to full for a minute. This allows the Halogen cycle to completely operate, and can "clean" the lamp. Then, if you are only running the lamp at 20% during the show, it's still fine as the halogen cycle at least gets to operate fully at some point each night.

But- getting back to the original issue- the theory based on the images are that you simply got a bad batch and can probably return them and get replacements. It can happen to any one. We make hundreds of thousands of HPLs each year, and I hate to admit it, but even some of ours have problems from time to time-- sssh! but don't tell anyone.

Cheers,

Mark

Thanks guys. I would tend to think I had a bad batch. I think I will plan on keeping my current levels programed into the wall stations and creating a 100% at the wall touch screen in the main office. Nobody else here knows how to use it so i am safe there! I will then apply this 100% preset once in a while to make sure the cycle is run. I do still get pretty decent lifespan out of them.
Mark, I don't plan on sending these back (these are USHIO ) and besides I already trashed them.
Now I need to work on replacing the reflectors in the few that I found that were flaking, or just find a different fixture all together to replace these with.
Thanks