Categories
Electrics

Lamp life, color temperature

In response to a question to me off Pro-Sound Web's Lighting Forum. 

 "If I went with the identical lamp that's in this fixture, are you saying that it would be a more "brown appearing lamp"? You made that reference but am wondering if there is a lamp that is a more white appearing lamp as well. In the application that I will be using these lights, the distance to the stage is approximately 20 feet. I personally think that 575w is a bunch, although I am planning on hooking it up to a 600w per channel dim pack. Is it better to have the wattage there if needed, and just dim it when not? That sounds logical to me" …


A specific answer to your question would be no, installing an identical lamp to one that’s in the fixture will not appear more brown, it will appear the same.  Answering what I think is your real question, installing the same HPL 575w/115vX lamp in the fixture and dimming it down will make that lamp appear more brown than installing a lesser wattage lamp and not dimming it or only dimming it slightly.  Amber Shift.  Get out a stage lighting book, most will cover the subject.  A lamp operating at or over it’s rated voltage will appear more white than one that is dimmed.  With color temperature, distance has little effect on the appearance of the light’s color.  The color of a beam at 20' is the same as the color of a 100' beam.  It’s output that drops the further away you get.  Output of a dimmed lamp verses that of a lesser wattage lamp is effected the same by distance.  Light is light no matter if it comes from a dimmer or not.

 
The major choices you need to figure is lamp life, maximum output desired and cost effectiveness.  If your fixtures need to give out as much light as possible, and you can budget for shorter life lamps, those are the ones to choose.  If lamp life is your major consideration over output, than longer life lamps should be chosen.  If you need long life lamps but at a certain amount of intensity, than you would need to use a higher wattage lamp to achieve the same intensity.  If you need less intensity, and only have so much wattage available, than you will have to sacrifice life for output.  If you plan to leave your lights dimmed, than they will be further extended in life but will loose the color temperature that’s the major selling point of the lamp.  In that case, installing a lower wattage less life lamp in your fixture will be better to preserve the color temperature.  Such lamps will cost more money in the long run to keep replacing, but will save money in size of dimmers needed to operate them and energy costs.  Lots of things to consider. 

Here is a much more detailed description of what’s going on.  If you can  follow it, you will learn a lot about lamps and the factors that go into design of them.  Not all about them, I’m even still learning, but a good part specifically about color temperature and life.  Many more details yet.

 Lamp color temperatures, wattages and life or at least small tidbits of the equation. 
A lamp that appears more brown is an observation of it having a lower color temperature than your mental reference color appearance of what a light should look like by memory or visually in comparison to other beams of light near it.  It’s subjective unless verified by a light meter or individual lamp specification test data.  Color appearance is dependant upon many things such as the angle you view it at especially in reference angle to other beams of light, surface reflection and coating the beam of light is bouncing off of, differences between beams in similar areas, operating voltage and dimmer intensity – amber shift, fixture efficiency and lens characteristics, age of lamp with some lamps, and design values of it, etc.

In other words, operating a 115v lamp at 120v will be causing the filament to heat up more and thus give off a slight increase in color temperature which is directly related to the temperature the filament is operating at up to the filament’s maximum usable temperature.  Beyond that, you can also use color boosting filters at a slight loss of light to boost the color temperature of the lamp.  However, since lenses and reflectors kind of filter a beam of light while it bounces off or passes thru them, this will also effect the beam’s color and output dependant upon their efficiency or purity.  Color temperature unlike output is not effected by distance as long as there is not atmospheric filters involved in the light. 

Color temperature is also related to light output in it’s spectral graph of the emissions from the burning source.  The spectral graph as opposed to the spectral curve is a slightly more accurate telling of lamp specific output in that it shows spikes of light output at certain nanometers of wavelength as opposed to rounding them out into a more general curve of output with the average output being what color temperature the lamp is rated for burning at. 

Burn salt, and it gives off a certain color temperature in general – sodium vapor lamps, but more specifically, it gives off a wide range of colors both visible and not, corresponding to spikes in output in certain areas of color temperature as plotted on a spectral graph.  (The same type of thing astronomers use to get data on distant stars.)  The pressure, dichroic coatings and gas fillings of a lamp have a large factor in an incandescent lamp on what color temperature they burn at, or what spikes that lamp’s spectral graph has it’s spikes at.  A incandescent vacuum lamp is going to have a lower color temperature than a Nitrogen filled lamp, and that’s less than a Xenon lamp because with the pressure, those chemicals allow the filament to burn hotter without burning up.  They also effect certain parts of the light as plotted by adding their own composition when burning up to that of the spectral graph for a normal filament.  Krypton for instance would have more spikes in the green area than Xenon.  But this part is my assumption because the gas is not really burning.  It is however to some degree incandessing and filtering the light providing and blocking certain wavelengths of otherwise normal tungsten incandescence. 

Other factors such as halogen gas or dichroic coatings will also effect the operating temperature of the filament in allowing it to safely operate hotter. Halogen because it is replenishing the filament by re-depositing what burns up and falls off back on the filament so it can burn again once cool.  The lamp is able to operate at the higher temperature in burning itself up but being re-supplied to a point as long as that re-depositing of the filament is even and not just in certain areas of it.  It’s not perfect and the lamp will eventually have a part of the filament not having enough mass to resist breaking, but it in general extends lamp life given it’s operating at the right temperature.  A dichroic lamp coating such as on a HPL/HX-600 lamp, takes the IR heat out of the beam and reflects it back to the filament letting it operate at a higher temperature by convection than applied voltage.  In other words, it is getting it’s source of heat not only from the voltage applied to it, but it is also heating up by getting heat reflected from the light it is putting out, back on the part that’s generating the heat making it heat up more yet above the voltage applied. Since this will cause the filament to deteriorate faster, there had to be improvements in the filament design and halogen cycle to implement this. 

A 500w Halogen lamp in general is as bright as a 750w incandescent lamp, as is bright as a 375w Dichroic Halogen lamp in the most broad sense.  It’s also going to have a higher color temperature due to the improvements because the filament is allowed to burn brighter – at least in parts of it’s spectral spikes such as on the higher wavelengths.  A Halogen lamp and a Dichroic Halogen lamp might be rated for the same color temperature, but because of the heat applied to the filament, a larger portion of it’s average spikes will be in the higher wavelengths.  Both lamps have the same average color temperature in reference to the range they burn at, but the Dichroic lamp is going to have a larger percentage of spikes in the higher spectrum.  Thus lamps rated for a color temperature based upon a spectral curve is misleading in actual color of light given out especially when not corrected for in the case of a 115v lamp it operated over voltage. 

In specific reference to your lamps HPL 575w/115v Extended life lamps, the individual color temperature or color appearance of a HPL lamp in a S-4 fixture (also dependant upon it’s type of reflector because if I remember right, ETC makes two types of reflector at least for the PAR cans,) is very much dependant upon the brand of the lamp and what mix or chemicals are use in it’s makeup.  This data is published in the lamp specifications for each individual lamp along with life and luminous output.  These published specifications change from year to year because how a lamp is made or what percentages of gas or types of materials used for it change year to year and lot number to lot number.  One brand to another, due to differing manufacturing processes, materials and mixtures, output will be different in many cases be very noticeable such as the case between brands of HMI1200w /GSlamps and refrenced in the manufacturer data or at least in the spectral spikes that are a little harder to see but are still there and present many times as you gel or dim the lamps. 

Short of using a calibrated light meter on each brand and type of lamp, the best way to tell what the color temperature is going to be of specific brands of lamps is by using the published data on them.  Remember how many factors go into what a color temperature appears to be thus how in-accurate it depending upon what you perceive should be in the color, such as efficiency of the fixture and even where you stand. Differences in brand to your eyes in comparing lamps unless drastic in difference such as say over 1,000°K is hard to tell.  That said, as I inferred, the Ushio lamps by specification have better color temperature in general than Osram HPL or GLA lamps.  That’s based upon the data each company has provided at least this year and it’s probably going to change.  Will you be able to tell the difference between even a GE and Philips lamp with only a few hundred degrees difference in color temperature (taken as a example and not specific lamps) given the data provided is accurate to the lot number of lamp you use?  Probably not with your eyes.  However, once such lamps are filtered with the same color, since individual brands of lamp have a different actual color temperature and thus different spectral graph spikes, they will effect the gel or even paint on a stage differently.  A lamp with a red gel such as say a RX27 will react differently with a 2,950°K lamp than with a 3,050°K lamp in general and specifically with spikes of light on the lower end of the spectrum. 

Wiko/Eiko/SoLux, Philips, Osram/Sylvania, GE/Thorn/Koto, Ushio/Reflekto as the major brands of bulb many times if ANSI coded will have similar outputs, life and color temperature on paper, or if accurate for the current catalog, slightly different outputs, it all depends upon the brand and lamps change year to year or lot to lot. 

How they rate their bulbs can also be wishful thinking, inaccuracies in the test data, different mixtures or materials year to year, even hour to hour, or even atmospheric or locution differences on the test facilities.  Than there is what is being tested such as initial verses mean output or in life what they call the average life of a bulb, be it 40% burning out after a period of time, 10% burning out, 50%, etc and how large that sample was, across how many lot numbers of the test sample and how controlled the experiment is, or how many they thru out.  It can also be rated by how many lamps in that sample blew out or what percentage of the expected life the lamps blew out at.  For instance, in  Osram – Technology and Application, Tungsten Halogen Low Voltage Lamps Photo Optics p.32 “The lamp specified for tungsten halogen Low Voltage lamps is based upon defined “average lamp life.â€?  This is the time after which, on a statistical average, half of a not too small number of lamps fail.  “Failâ€? means that the filament burns out.  To be on the safe side, lamp manufacturers as a rule set the design value slightly above the promised “average lamp life.â€?

This modifies the above definition to the time after which, on statistical average, half the lamps may fail.  The lamp life distribution of individual lamps in a group approximately follows a Gaussian bell-shaped curvve.  Lamp manufacturers have the following to say about the width of this curve: individual lamp life is at least 70% of average lamp life.  If for example the average lamp life is 100 hours, every lamp will last for at least 70 hours, except for premature failures – the black sheep of mass production which can never be entirely avoided.  A mandatary percentage limit laid down internationally – the AQL – is specified for these premature failures (AQL stands for “Accepted Quality Levelâ€? and is part of a comprehensive statistical quality system in common use internationally, see DIN 40080) the AQL value varies for different groups of lamps (general lighting service, photo-optic applications, etc.)  The tungsten halogen LV lamps under consideration here normally have an AQL of 6.5 which means in practical terms that 6.5% of the lamps in a sufficiently large random sample do not have to achieve the individual lamp life.  In accordance with the lamp life definition, they may fail shortly after being switched on for the first time or, as in the above example, after 69 hours.â€? 

Lots of differences between brands in addition to differing materials and quality of workmanship going into individual lamps that would be factors both in specified data and spectral graph output.  Lots of quality control or AQL levels that can be used.  In general, once you get a brand of lamp, stick with it for similar fixtures doing the same work.  Differing materials making up the lamp will even react to voltage applied to it differently. Granted most of what I am writing is in the most finite of measurements on the data.  Differences between HPL lamps can be large by using the specifie  numbers even if the actual visual differences are possibly too small to be seen.  Differences between say FLK lamps in general on paper are not noticeable and only the spectral curve and materials and quality of the lamp have effects that can be judged but almost certainly not noticed unless you are dimming them or filtering them. 

By the way, a Osram HPL 575w/C lamp has a very slightly larger color  temperature than a Ushio lamp by the specifications, but the same is not the case in the HPL 375w/C lamp.  For me at least, the lamp and it’s heat sink on the Osram lamp don’t have the bond of a Ushio lamp to it’s heat sink and the Osram lamp frequently pulls out of the heat sink.  That’s why I don’t buy them. even at a lower cost, I don’t even consider Wiko lamps for S-4 fixtures.  Sometimes, it’s not lamp data that is a factor in buying lamps, in the case of a 2Kw CYX lamp, the shipping boxes that package GE and Philips lamps doesn’t support the bulb well enough for it to survive being bounced around in the back of a truck as a spare lamp well enough for me to buy them even if more in output.  Ushio and Osram CYX lamps hold up better to transport and thus I buy them.  For me, the Osram lamp is cheaper than  the Ushio lamp so it’s my primary lamp in spite of any loss in output.  Is the packaging of a Ushio HPL lamp better than that of a GE or Osram lamp, good debate, but not much different in quality once it does some travel or gets wet. 

Try lighting a bloody scene on stage with a incandescent plano-convex fixture such as a Bantam Super Spot, than with a S-4 fixture.  You can even use a radial mounted Altman #360 for this.  Use the same voltage, percentage of dimmer, and say a 750w lamp in the Plano Convex verses a 375w lamp in the ETC fixture.  Not only especially with gel will each beam of light appear much different, but the color of the blood, and it’s sparkle or pop will be totally different.  Now start to dim them.  As you dim a lamp, you get “amber shiftâ€? going on.  That’s the result of the lamp’s filament burning cooler and not putting out as much light, but also the filament’s temperature not burning at the same color temperature or heat from the voltage, thus it drops as you dim the lamp.  There will be a different dimming curve between types of lamps that can be noticeable.  In general, when you dim a lamp however, it will be effected by amber shift.  That’s why it is better to put a 375w S-4 lamp into a fixture as opposed to leaving it on a dimmer with a 575w lamp to provide the same intensity while dimmed.Lamp might last longer, the intensity might be the same, but the output in color is going to be crap – like lighting the stage with candles.  Since different lamps have different places they spike in color – or groupings of color’s the filament is burning at, a lamp when dimmed will drop in output and color temperature following that graph with the spikes that are largest lingering the longest in the light beam still present in the dimmed beam of light.  That’s dependant upon the chemical fillers making up the lamp and what color temperature or heat it’s burning at.  After a certain point, all filament lamps will no longer have the benefits of the filler boosting color temperature and will burn similar.  A HPL lamp with a dichroic coating reflecting heat back to the filament, and having a halogen (Bromine or Iodine) and Krypton or Xenon filler will have a different normal operating color temperature than a lamp having a nitrogen/argon filler because it cannot burn as hot in suppressing the rate of vaporization, given the same wattage or resistance present in the filament. It’s spikes thus as you drop the power into the filament will be highly different with the HPL lamp lingering longer in a brighter/more white output than with a normal halogen or incandescent lamp, though both at some point will have similar outputs at lower dimmer ratios.  Thus, in at least my theory, a HPL/HX-600 lamp will have less problems with amber shift up to a point when those advantages will rapidly drop off. 

(Osram – Technology and Application, Tungsten Halogen Low Voltage Lamps Photo Optics.)  “The reduced rate of vaporization of the tungsten can either be used to increase lamp life or – if the life remains the same – to increase the luminous efficacy and the color temperature by raising the temperature of the tungsten.  In both cases, using the standard krypton lamp as a starting point, the filament dimensions have to be recalculated and the lamp filling modified. Luminous efficacy can be increased by about 5-10% with the “Xenon Effectâ€?, which corresponds to a color temperature increase of about 100K.  Xenophot technology can only be used for low-voltage lamps.  In high-voltage lamps the lower ionizing energy of Xenon would lead to electrical discharge in the lamp bulb.â€? 

That resistance in the filament is the wattage of the filament as modified by the voltage it is designed to operate at.  The larger the voltage, the larger the filament needs to be to carry the current safely where life and cold starting is concerned amongst other factors.  The larger the filament, the longer it’s going when dimmed to retain it’s heat and thus color temperature for the initial dimming up or down.  In many cases, that’s coming close to the rate your eyes  adjust for the drop in color temperature or output without you noticing it.  The larger the filament, the less resistant the materials comparatively will be to the flow of electricity due to the mass of the wire radiating the same amount of heat.  It’s still giving off the same amount of heat, just doing less work to do so and thus burning up less.

Another way of controlling resistance in the wire is by changing the percentage of tungsten to other materials in it.  A long life lamp can have the same size of filament wire, but have longer lasting – more resistant to heat materials making it up that incandess a little less or even a higher percentage of halogen in the gas or be operating at a higher temperature allowing the halogen cycle to operate more effectively.  Differences in how the bulb is designed or the gas flows within the lamp will also effect this. With any of these methods the long life lamp in general will have less output, but the same color temperature in most instances, but you can retain the same output and life by adjusting the color temperature the filament burns at.  There are three primary factors life, output and color temperature to a lamp given it’s resistance and voltage by design are the same.  Adjusting any of them is a question of fillers, coatings, voltage, filament composition and winding of that filament.  A filament designed for a high color temperature, and high voltage such as 125v will when at a lesser voltage have a similar color temperature to a lamp designed for 115v operation but more life when operated under voltage given the same life rating at the start.  The only thing that will drop is luminous output.  On the other hand, when you operate a 120v lamp rated exactly the same as a 115v lamp at 115v, it’s going to have a longer rated life, but less color temperature and output.  The 120v lamp will appear less bright in both color and intensity.  The main difference between 130v and 120v incandescent lamps in a household fixture.  The larger filament will also be more resistant to voltage spikes and cold starting in-rush currents effecting the filament by making it operate at a higher voltage and temperature if only for a few moments. 

Since filaments have different compositions, in addition to the fillers, closeness of filament wires to each other having a thermal effect on them, and coatings on the lamps, they on a spectral graph will have differing spikes on the chart brand to brand and type to type.  A long life lamp will have differing dimming characteristics than a high output lamp due to what’s burning inside of it and what spikes they have.  Also if the lamp is say already a higher voltage lamp that’s operating on a lesser voltage, than it will tend to more rapidly be effected by amber shift than one that is operating at it’s peak output because it’s already not at it’s peak values and some parts of the range of light are already not there. 

All of that said, when you operate a 115v lamp over it’s rated voltage, such as on a HPL lamp at 120 or more realistically 117v, than it’s going to have a higher color temperature than it’s rated and published color temperature. HPL/HX-600 lamps appear more blue than other lamps in older stage lighting fixtures in part due to fixture efficiency.  The design color temperature is usually about the same as with 120v lamps, (the color temperature difference between a EHD lamp and a HPL lamp is 250°K and that’s not noticeable in theory,)  but the voltage is boosting the color temperature to make it look different in a factor of 2% color for 5% in volts (making it seem as if the lamp had a 120v. 3,770°K color temperature instead of a 115v, 3,250°K color temperature, or 2,950°K color temperature of a EHD lamp) in addition to it’s differing spectral spikes from operating at a higher filament temperature, while sacrificing lamp life at operating over voltage.  (That’s 50% less life when using a 115v HPL high output lamp on a 120v circuit or 150 hours without dimming.  Don’t believe me, check the math, for every 1% of difference in supply voltage, life is effected by 12%.  Large increase in color temperature not to mention actual output.  Remember also that the actual amount of time such lamps are on is not much especially when dimmed down to voltages below 115v which go back to extending their life, plus line voltage after voltage drop is usually much less than the calculated 120v.) 

HPL/HX-600 lamps operated over voltage and with their various improvements are kind of similar to car engines with a nitro boost.  It’s the same basic engine though probably improved for the best output, but that nitro boost makes it go faster and burn out the engine faster as a secondary result.  A HPL lamp appears brighter in color temperature and has more output much due to the voltage.  A HPL 575w lamp operated over voltage, and with it’s improved dichroic coating and gas mixtures, puts out as much light (17,208.333 Lum/120v out of a 16,520 Lum lamp) as a average between a EHG and EHF 750w Quartz lamp.  More than the EHG (usually 15,400 Lum)  with it’s longer life, and less than a 750w EHF (Usually 20,400 Lum) with it’s similar life to that of a HPL lamp when at differing design voltages.  Thus, a HPL 575w lamp, in a higher and more efficient fixture puts out about as much light as a 750w lamp, but in the higher efficiency fixture might even put out slightly more say 800w worth of halogen light because the light is collected and focused more efficiently.  That 800w figure is also based upon how the light appears.  Since as you raise the voltage that 4%, your color temperature also goes up, the light is going to appear more blue especially with better lenses on a ETC fixture in addition to differences in the lamp itself.  A lamp operating at a higher color temperature seems to be brighter even if the same or less in actual lumens coming out of the fixture.  It appears to be brighter and we perceive it to have more luminous output because of it.  However actual output in many cases can be less such as on a multi-vapor lamp.  It’s usually the case that a lamp having a larger color temperature will have less of a CRI rating.  That’s the case even if the actual lamp has the same luminous output on paper.  It’s a question of how natural that light looks in being useful verses just plain how bright it appears. 

The maximum burning temperature of a average filament is about 3,550°K (3383°C) when operated at it’s rated voltage.  There are some incandescent lamps out there that burn at about that color temperature without using any filters to boost it.  However any time you put a filament at it’s maximum burning temperature, or the closer you get to it, the faster it will burn up or larger chance it will be adversely effected by variations in voltage applied to it.  Normal maximum color temperature of stage and studio bulbs is between 2,800°K and 3,200°K which leaves somewhere around 20 Volts (my figure) of margin of error before the filament burns itself up too rapidly for it to be used.  A better figure would be using a 10% maximum variation in over-Voltage.   For a HPL lamp designed for a 115v lamp, you don’t want to operate it at over 126.5v for semi extended use or 131.43v (14%) for a voltage spike.  Osram says in their below book, start up lamp filament resistance can be as much as 20 times less than operating resistance, and most lamps are designed for a start up voltage of 108%.   With every 3 lumens per watt applied to the lamp, color temperature changes by 100K. That’s a base way of determining color temperature when not given.  Remember this figure for special effects and low voltage lights. 

(Osram – Tungsten Halogen Low Voltage Lamps Photo Optics p.21 as referenced from IES Lighting Handbook & The Science of Color as a refrence) “The following variables can be related in a fixed formula for incandescent lamps.

  1. Luminous flux
  2. Luminous efficacy
  3. Color temperature
  4. Electrical voltage
  5. Electrical current
  6. Electrical power consumption

In non-tungsten halogen lamps, lamp life can also be added to this list as it is only determined by the physically measurable evaporation rate of the tungsten filament.  In tungsten-halogen lamps, lamp life is also affected by the chemistry of the tungsten halogen cycle.  A fixed mathematical relationship with the above variables therefore only exists in a small, well-defined range. The mutual dependence of these variables can be shown very clearly in a diagram id the deviation from the rated lamp voltage us used as the abscissa. The following rule of thumb can be derived:

A 5% change in voltage applied to the lamp results in

  1. halving or doubling the lamp life
  2. a 15% change in luminous flux
  3. an 8% change in power
  4. a 3% change in current
  5. a 2% change in color temperature

The limitation described above applies to lamp life.  It must also be noted that increasing the voltage may in some circumstances not be permissible, depending on the design of the lamp; if it causes the tungsten filament to reach its melting point the lamp will burn out.â€? 

Review of this only small portion of the subject as I understand it: A long life lamp will last longer than a high output lamp in exchange for output or real light coming out of it, or exchange color temperature for life and it has to be one of the two if you don’t change the voltage or wattage given the same fillers. The long life lamp should react just slightly different under a dimmer or over voltage than a high output lamp also due to the differing materials making it up as plotted on a spectral graph. 

Such lamps as a HPL lamp are more efficient by design and fixtures they are used in than halogen lamps used in older fixtures, just as halogen fixtures were a vast improvement over incandescent sources. 

Any filament lamp is limited in it’s maximum color temperature by the filament itself and what pressure or gasses surround it preventing it from evaporating or burning up too rapidly which is also effected by voltage applied to it in addition to other things such as frequency. 

When you operate a lamp at too high a voltage, it gets really bright but goes super nova just as fast.   Otherwise in the case of a HPL lamp, it has more color temperature and output but less life.  A HPL 575w/115v lamp will look very different than a HPL 575w/120v lamp when operated at the same voltage no matter what it is.  Those differences are enough to notice even though there is only a 4% change in voltage applied to it and that on a dimmer usually is not enough to notice in difference between the same lamps. 

A lamp when dimmed is going to have amber shift effecting it and will provide light corresponding to the spikes on the output graph up to a point when special gasses, proximity to other parts of the filament or dichroic coatings stop effecting the output and it will than return to normal incandescent output.  Those spikes on a dimmed lamp will make it linger in certain ranges of spectral color and appear different,  making say a HPL lamp look different on a dimmer look different in color temperature than a lesser wattage lamp not dimmed.  It is going to have amber shift and loose much of the usable light in it’s full range of colors, but it will linger at certain points differently. 

A lamp with differing compositions of the filament, or what is “dopedâ€? into it’s make up will also have slightly differing spectral spikes as would a larger filament lamp when dimmed to a point that it is operating at the same temperature.  Tin will have a different burning spike pattern than that of a copper doping given that’s what’s used. 

A dimmed lamp in comparison to a lamp operating at it’s rated voltage but at a smaller wattage will have about the same luminous output at some point in dimming no matter what the color temperature, and both will be effected exactly the same by the law of squares or law of inverse squares which ever applies the further away from the fixture you get. 

The color temperature and life of that dimmed lamp will be inversely effected by dimming to life but less so effected than Lumionous Output will be in going down as the lamp is dimmed.  This is also effected by the types of chemicals, proximity of the filament wires to each other or thickness of the wire or other factors such as pressure, chemicals used and dichroic coatings as they relate to filament heat at voltage to the lack of benefits such things offer.  At some point, a lamp given current is just heating the wire and not incandessing, at some point before that, no matter what chemical or pressure you are using to allow for a higher burning temperature of the filament, the lamp is acting as if a normal incandescent lamp in life and output in a broad sense even with spikes in spectral output considered. 

When you have a need for a lower intensity on a source and don’t need to go above it, it is better to perhaps dim it very slightly to extend life, but always go for the lower wattage lamp that is operating at peak color temperature because the actual radiation of the lamp in the visable spectrum of light will be more full in all areas of light and operating at it’s design peak. If you only need 10,500 Lumens out of a fixture, rather than dimming a 575w lamp to about 66%, you are better off putting a 375w lamp in the fixture, it ’s giving design color temperature with all light present in it’s spectral graph.

Note: HPL lamps and FLK/HX-600 series lamps are for all intensive purposes the same lamp see the GLA series of lamp that used to be able to be used for either type of fixture.  You can get a HX-400 lamp that’s going to be about the same as a HPL375, just as you can get a HX-754 or HX-800 lamp that’s going to be the same as a HPL 750.  Just a question of what fixture it’s in and your need for output.  All styles have long life variants.  A Shakespeare and a ETC S-4 fixture use those different lamps but can be expected to have similar outputs coming out of them. 

Notes: (Anything without a source following it probably comes from a GE catalog especially the GE-Spectrum Catalog.) Cand. = Candlepower, Candlepower is the normal rating method of the total light output of miniature lamps.  To convert this rating to lumens multiply it by 12.57 (4 pi).

 Mean spherical candlepower MSCP is the initial mean candlepower at the design voltage.   It is subject to manufacturing tolerances.

 Mean spherical candlepower is the generally accepted method of rating the total light output of miniature lamps.

cd  = Candela.  The international unit (SI) of luminous intensity.  The term has been retained from the early days of lighting when a standard candle of a fixed size and composition was used as a basis for evaluating the intensity of other light sources.

Chromacity = See Color Temperature

Color Rendering = As a rule, artificial light should enable the human eye to perceive colors correctly, as it would in natural daylight.  Obviously, this depends to some extent on the location and purpose for which light is required.  The criterion here is the color rendering property of a light source.  This is expressed as a “general color rendering indexâ€? (CRI).  The color rendering index is a measure of the correspondence between the color of an object (its “selfluminous colorâ€?) and its appearance under a reference light source.  To determine the CRI values, eight test colors defined in accordance with DIN 6169 are illuminated with the reference light source and the light source under test.  The smaller the discrepance, the better the color rendering property of the lamp tested.  A light source with a CRI value of 100 displays all colors exactly as they appear under the reference light source.  The lower the CRI value, the poorer the colors are  ndered.   – Osram Photo-Optic Lighting Products, 1999

Color Temperature = Originally, a term used to describe the “whitenessâ€? of ncandescent lamp light.   Color temperature is directly related to the physical temperature of the filament in incandescent lamps so the Kelvin (absolute) temperature is used to describe color temperature.   For discharge lamps where no hot filament is involved, the term “correlated color temperatureâ€? is used to indicate that the light appears “as ifâ€? the discharge lamp is operating at a giving color temperature.   More recently, the term “chromaticityâ€? has been used in place of color temperature.

Chromacity is expressed either in Kelvins (K) or as “X� and “Y� coordinated on the CIE Standard Chrom-aticity Diagram.

 Although it may not seem sensible, a high color temperature (K) describes a visually cooler, bluer light source.

 Typical color temperatures are 2,800°K (incandescent), 3,000°K (halogen), 4,100°K (cool white or sp41 fluorescent), and 5,000°K (daylight-simulating fluorescent colors such as Chroma 50 and SPX 50.

 Unit of measurement: Kelvin (K) the color temperature os a light source is defined in comparison with a “black body radiatorâ€? and plotted on what is known as the “Planckian curve.â€?  The higher the temperature of this “black body radiatorâ€? the greater the blue component in the spectrum and the smaller the red component.  An incandescent lamp with a warm white light, for example, has a color temperature of 2,700°K, whereas a daylight has a color temperature of 6,000°K.  – Osram Photo-Optic Lighting Products, 1999    Light color = The light color of a lamp can be neatly defined in terms of color temperature.  There are three main categories here: warm<3,300°K, intermediate 3,300 to 5,000°K, and daylight > 5,000°K.  Despite havig the same light color, lamps may have very different color rendering properties owing to the spectral composition of the light.  – Osram Photo-Optic Lighting Products, 1999

Hal = Halogen Lamp – A short name for the tungsten-halogen lamp.  Halogen lamps are high pressure incandescent lamps containing halogen gasses such as iodine or bromine which allow the filaments to be operated at higher temperatures and higher efficacies.   A high-temperature chemical reaction involving tungsten and the halogen gas recycles evaporated particles of tungsten back onto the filament surface. Also called a Quartz lamp, though this is a term for the higher melting temperature glass enclosure used on halogen lamp

HIR = Halogen – IR Lamp. Dichroic Lamp Coatings. G.E. designation for a new form of high-efficiency tungsten halogen lamp.  HIR lamps utilize shaped filament tubes coated with numerous layers of materials which selectively reflect and transmit infrared energy and light.   Reflecting the infrared back onto the filament reduces the power needed to keep the filament hot.

Illuminance = The “densityâ€? of light (lumens/area) incident on a surface. Illuminance is measured in footcandles or lux. – GE Spectrum Catalog Illuminance = The “densityâ€? of light (lumens/area) incident on a surface.

Illuminance is measured in footcandles or lux.

 A unit of measurement: lux (lx) illuminance E is the ratio between the luminous flux and the area to be illuminated.  An illuminance of 1 lx occurs when a luminous flux of 1lumen is evenly distributed over an area of one square meter.

 – Osram Photo-Optic Lighting Products, 1999

Lamps with Blue Dichroic Reflectors:  Lamps with Semi-Clear Blue Reflectors reflect less unwanted visible light above the 70nm range.

Lum. = Lumen – The international (SI) unit of luminous flux or quantity of light.  For example, a dinner candle provides about 12 lumens.  A 60-watt Soft White incandescent lamp provides 840 lumens.  (Lumens = Mean Spherical Candlepower x 12.57)

 Luminance L = A unit of measurement: candelas per square metre (cd/m²) The luminance L of a light source or an illuminated area is a measure of how great an impression of brightness is created in the brain. – Osram Photo-Optic Lighting Products, 1999

 Luminous efficacy ɳ = Unit of measurement: lumens per watt (lm/W). Luminous efficacy indicates the efficiency with which the electrical power consumed is converted into light.         – Osram Photo-Optic Lighting Products, 1999

Luminous Flux Ф = a unit of measurement: Lumen (lm).  All the radiated power emitted by a light source and perceived by the eye is called luminous ux.   – Osram Photo-Optic Lighting Products, 1999

 Luminous Intensity I = Unit of measurement: candela (cd).  Generally speaking, a light source emits its luminous flux in different directions and a different intensities.  The visible radiant intensity in a particular direction is called luminous intensity. – Osram Photo-Optic Lighting Products, 1999

Lumen Maintenance = A measure of how a lamp maintains its light output over time.  It may be expressed as a graph of light output verses time or numerically.

 All metal halide lamps experience a reduction in light output and a very slight increase in power consumption through life.   Consequently there is an economic life when the efficacy of the lamp falls to a level at which is better to replace the lamp and restore the illumination.   Where a number of lamps are used within the same area it may be well worth considering a group lamp replacement programme to ensure uniform output from all the lamp.

Luminarie Efficiency = The ratio of total lumens emitted by a luminary to those emitted by the lamp or lamps used.

 Luminarie efficiency (also known as light output ratio) is an important criterion in gauging the energy efficiency of a luminarie.  This is the ratio between the luminous flux emitted by the luminarie and the luminous flux of the lamp (or lamps) installed in the luminarie.  For detailed information on indoor lighting with artificial light, see DIN 5035.  – Osram Photo-Optic Lighting Products, 1999

 Luminance = Formerly, a measure of photometric brightness.   Luminance has a rather complicated mathematical definition involving the intensity and direction of light.   It should be expressed in candelas per square inch or candelas per square meter although an older unit, the “footlambertâ€? , is still sometimes used.   Luminance is a measurable quantity whereas brightness is subjective sensation.

Luminous Efficacy = The light output of a high source divided by the total power input to that source.   It is expressed in lumens per watt.

Lux (lx) = The SI (International) unit of illuminance.   One lux is equal to 1 lumen per square meter.  See also footcandle.

MSCP = Mean Spherical Candlepower, this value is the initial mean spherical candlepower at design voltage, subject to manufacturer tolerances, generally the accepted method of rating the total light output of miniature lamps.

See Candle Power above.

Mean Lumens = The average light output of a lamp over its rated life.   For fluorescent and metal halide lamps, mean lumen ratings are measured at 40% of rated lamp life.   For mercury, high pressure sodium and incandescent lamps, mean lumen ratings are measured at 50% of rated lamp life.

Neodymium Coating, a Dichroic Coating on the lamp which reduces the yellow content of light, enhancing whites, reds, blues & Greens.  These lamps are useful for merchandise displays, or on dimmed circuits to correct for amber shift.

Nitrogen = Common inert gas filling other than halogen for inside incandescent lamps, This is usually a mixture of nitrogen and argon used in lamps 40watts and over to retard evaporation of the filament.

Smaller bulbs usually do not require gas and therefore are vacuum bulbs Krypton is limited in output and Nitrogen/Argon gasses Tungsten = Tungsten filaments change electrical energy to radiant energy. The light generated results from the filament being resistance heated to a temperature high enough to produce visible light.

Filaments can not be operated in air see seal and vacuum.  Tungsten is used for the filaments because of its low rate of evaporation at temperatures of incandescence and its high melting point 3,655°K.

There are grades of tungsten purity and different grain structures.

Only the highest grade of an elongated grain structure guarantees maximum life and reliability during shock and vibration.

Heat treatment of the tungsten filaments is one of the most critical factors in lamp manufacturing..

Proper heat treatment prevents  filament sag, abnormal coil shorting or premature breakage.

Tungsten Halogen Lamps = Halogen Lamps are tungsten fliament incandescent lamps filled with an inert gas (usually krypton or xenon to insulate the filament and decrease heat losses) to which a trace of halogen vapor (bromine) has been added.

Tungsten vaporized from the filament wire is intercepted by the halogen gas before it reaches the wall of the bulb, and is returned to the filament.

Therefore, the glass bulb stays clean and the light output remains constant over the entire life of the lamp. (p33, Sylvania Lamp & Ballast Product Catalog 2002)

Halogen lamps are high pressure incandescent lamps containing halogen gasses such as iodine or bromine which allow the filaments to be operated at higher temperatures and higher efficacies.

A high-temperature chemical reaction involving tungsten and the halogen gas recycles evaporated particles of tungsten back onto the filament surface. Also called a Quartz lamp, though this is a term for the higher melting temperature glass enclosure used on halogen lamp

v = Volts – A measurement of the electromotive force in an electrical circuit or device expressed in volts.

Voltage can be thought of as being analogous to the pressure in a waterline.

The effect of voltage on a lamp will cause a significant change in lamp performance.

For any particular lamp, light output varies by a factor of 3.6 times and life varies inversely by a factor of 12 times any percentage variation in supply.

For every 1% change in supply voltage light output will rise by 3.6% and lamp life will be reduced by 12%.  This applies to both DC and AC current.  Most standard line voltage lamps are offered at 130v.   Since most line voltage power is applied at 120volts, the result is a slight under voltaging of the filament. The effect of this is substantially enhanced lifehours, protection from voltage spikes and energy cost savings.

 Voltage and Light Output: The effect of voltage on the light output of a lamp is ±1% voltage over the rated amount stamped on the lamp, gives 3.1/2% more light or Lumens output but decreases the life by 73% and vise a versa.

Do not operate quartz Projection lamps at over 110% of their design voltage as rupture might occur.  GE Projection, Ibid p.13

Xenon (High output halogen lamps using Xenon filler instead of krypton producing a luminous flux up to 10% higher; with otherwise identical lamp data 

Quartz Lamp “QIâ€?, or Quartz-Iodine Lamp.  Introduced in 1959, this small, compact, long-life lamp consisted of a tungsten filament enclosed in a transparent quartz envelope partially filled with vaporized iodine.

When an ordinary lamp burns, tiny particles of tungsten are released from the filament and are deposited on the glass envelope as a black film, gradually reducing the intensity of the light.

During the burning process of the quartz-iodine lamp,  released particles of tungsten reacted chemically with vaporized iodine and returned to the filament.  Not only was the life of the lamp improved by this, but the black deposits on the inside of the envelope were eliminated.

The ideal lamp had been created except for one small detail: as iodine sublimes, it turns a purple-violet color in both the warming (dim-up) and cooling cycles.

Clearly, the untenable situation for theater lighting.

Further experiments substituted a related element, halogen, for iodine and heat resistant quartz glass for the quartz envelope, producing a lamp that retained the favorable characteristics of the quartz-iodine lamp and eliminated the purple discoloration.

The new lamp was redesigned and introduced to the market as a tungsten-halogen (TH) lamp.

The term “Quartz� carried over.

Tungsten Halogen (TH, quartz iodine, QI). A lamp using a halogen gas around a compact filament.

Used in instruments designed specifically for this type of lamp, the TH lamp can also be retrofitted for older instruments.

It should be noted that the terms “QIâ€?, “Quartzâ€?, and “quartz iodineâ€? are “misnomersâ€? in common usage. (Theatre Lighting from A to Z by Norman C Boulanger and  Warren C Lounsbury, University of Washington Press, Seattle 1992)  The Halogen lamp was invented by G.E. Lighting in 1957. (G.E Spectrum, Ibid p.2-1)

Tungsten-Halogen Lamp (TH, quartz, QI).  The tungsten-halogen lamp is made with a heat-resistant synthetic quartz envelope, filled with halogen gas. Under the intense heat of the burning process, bits of tungsten released from the filament react chemically with the halogen to return to the filament.  The process not only improves the life of the lamp but eliminates black deposits on the inside of the envelope that with standard tungsten lamps filled with inert gases.  Another favorable feature of TH lamps is that they burn equally well in any position and therefore have made possible improvements in the design of instruments, including the axial mount ellipsoidal reflector spot light such as the Altman 360 being made into the 360Q.

Because TH lamps offer higher intensity, longer life, and soot-free envelopes, they are obviously the favored lamps for stage-lighting instruments.

Warning however, do not touch the synthetic quartz envelope of the lamp with bare fingers; skin oil deposited on the envelope will cause hot spots to develop when the light is turned on, shortening the life of the lamp. (Theatre Lighting from A to Z)  Normal lamp globe temperature is 482°F minimum, hot spots on the bulb wall itself can go as high as 1,230°F. in normal operation.   Use the paper or plastic wrap which comes with the lamp to shield it while handling.  Clean dirty or touched lamps only with alcohol or grease free solvent.  Keep sealed fixture temperatures below 350°C. Bulbs on the other hand must maintain (482°F) 250°C for operation of the halogen cycle..   To avoid shock when on, do not operate them beyond 8-10% of their total rated voltage (by the safety specs), 3,400K Quartz lamps should not be operated above 105% of their voltage or life will be seriously effected, under voltage operation under 90% of their rated voltage gives longer but un-predictable length extended life, however transformer type  dimmers adjusting the voltage of a quartz lamp will preserve more lamp life than semi-conductor dimmers due to the type of dimming work actually done. (G.E, Ibid p58)  Quartz lamps may begin to devitrify at temperatures above 1,832°F.  The best operating range for a halogen lamp is 482-1,472°F. Oxidation on the sealing foil carrying current from the base to the filament however begins to oxidize at temperatures above 662°F.  Lamp life may be shortened by premature seal failure if this temperature is exceeded. (G.E. 99, Ibid p.6-5)  Contact pins are plated to ensure good electrical connection with the lampholder.  However, at temperatures above 662°F. the plating may loose adhesion, leading to deterioration in contact and possibly local hot spots, arcing and consequent irreparable damage to both lamp and holder.   Note that if there is evidence that this has occurred, the lampholder should be replaced before the next lamp is fitted, otherwise it is likely to fail prematurely for the same reason.  Lamps normally fail by fusing of the filament.  This is often followed by arcing, leading to very high currents which can cause the envelope and seals to fail and the lamp to shatter.  A quick-acting, high breaking capacity fuse should therefore be connected to the supply line in all applications suitable types are given is IEC 127, 241, and 269.    Because of the heat involved with line voltage halogen lamps, do not use them in fixtures not rated for their use, or at least 660V constant operation high temperature plastic or porcelain, or in fixtures with cooling fins on their base, reflectors or anything else needed for extra cooling of the equipment. (G.E Spectrum, Ibid p.2-17)  Normal operating temperatures of a halogen lamp are above the flash point and kindling temperatures of many materials and lamp bases, care should be taken when using them.  Temperatures above 350°C should be avoided when using a halogen lamp as they might deteriorate the lead wires and basing cement can loosen causing lamp failure. (GE Miniature & Sealed Beam Lamp Catalog, G.E.

Lighting # 208-21121 (9/92) p. 23)

Halogen lamps operate at near 100% efficiency throughout their life, and generates 1/3 more light per watt than conventional incandescent lamps (Philips, Ibid p.111) 68% more energy cost savings over Incandescent and 50% more life. (G.E. 99, Ibid p. I-5)  Substantial heat is generated in all halogen lamps (90% of their light is infrared and a small amount is UV which can be protected against by almost any screen or lens) (G.E Spectrum, Ibid p.17), so equipment design should make allowance for the dissipation of excessive heat.   Certain lamps and extremely confined fixtures may require additional ventilation or heat sinking to ensure proper operation of the halogen cycle and to prevent damage to the fixture.  It is a good practice to test the lamp in the operating environment early in the design cycle to ensure adequate performance.  Precautions must be taken in the selection of materials for lamp holders, reflectors, and lamp housings because the 1230°F. bulb wall temperature is greater than the kindling temperature of many materials.   Lamp base temperatures should not exceed 662°F. because above that point, lead wires may deteriorate and the basing cement loosen, causing premature lamp failure (G.E.99, Ibid p.2-15) Avoid lamp use on dimmers which can deliver voltage over their rated voltage, do not allow one lamp to directly touch another lamp, and do not allow particles to fall on the lamp they can cause hot spots on the lamp. (Ushio, Ibid p.28)  Extended exposure to un-jacketed lamps rated at 3,200K and above, or to any un-jacketed quartz lamps operated above rated voltage, may lead to ultraviolet irritation of skin and eyes.  Passing the light through ordinary glass or plastic provides adequate protection.  Such protection is automatically provided by the glass of outer bulbs of quartz Par and R-lamps.  (G.E, Ibid p.54)  Noise – all Quartz stage and studio type lamps except Par types have special “low noizeâ€? construction to minimize generation of audible noise when operatid on A.C. circuits.  In addition all Quartz RSC lamps have such construction. (G.E, Ibid p.57)  The most powerful Quartz lamp is 20,000 KW.

Halogen Lamps: to clean touched lamps use alcohol and a clean cloth if touched or dirty, better yet do not touch a halogen lamp as the oils from ones fingers will stay on the glass and cause heat to not dissipate as well. Sometimes these areas can burst or swell up in time.  They can also reflect heat and cause the filament to become misshaped even to the point of it touching the opposite side of the lamp and melting its way thru the glass. In this case, even if the filament does not break, the focus point of light will be out of focus.  Also always allow a lamp or fuse to cool before touching it even with gloved hands, as the glass might explode. ANSI lamps and generalized data do not necessarily mean every lamp brand producing the same lamp will have the exact performance data.  Materials which make up the lamp play a large part in the lumen output and life of a lamp.  Factors affecting this are: the grade of quartz (it purity its preparation and transparency) (Ushio, All Lamps Are Not Created Equal, Ushio Pamphlet), the cement and ceramic materials strength and durability, the gas selection – mixture and fill pressure.   (The choice of gas is critical as well as its pressures and organic carriers: see chart below.) The tungsten filament ([K2O-SiO2-Al2O3 family] having a low rate of evaporation at high temperatures, and is easily formed into complex shapes necessary for the filament.  Different treatments during the production of the tungsten wire affect the filament’s properties.   For example, the introduction of re-crystallized particles along the length of the wire makes it possible to produce filaments which remain distortion free.   Such non-sagging filaments are critical in many applications.) ( Ushio All Lamps, Ibid)  The filament must be formed and coiled to the right specifications, and assembly must be done in a clean environment. (the sealing must withstand an increase in Temperature from ambient to 250°C. and still keep its seal.  Forming the seal is critical to making a good lamp, molybenum foil is used since it expands at almost the same rate as quartz when it is heated.  Since the rates do not match perfectly, the stress on the seal area must still be minimized by chemically milling the edges of the foil of the thinnest feasible cross-section, it is possible to improve the seal performance further.  Such proprietary techniques differ from one lamp maker to another and serve as examples of the differences in manufacturing technique which impact on lamp performance and consistency.) (Ushio All Lamps, Ibid)  Any scaling down of these features will probably be reflected in the price and quality of a lamp. (Ushio Lamp Promotion, Special Promotional Pricing for Distributors, Ushio#P004/0500  c5/1/2000   p.5)  There are more than twenty companies which manufacture lamps today.   There are also a number of companies selling lamps that are private labeled for them.  The manufacturers are generally divided into two groups: companies products primarily for general lighting and those producing lamps for special applications.   The requirements for success are different.  Products for general lighting are typically manufactured in high volumes.   Being able to design, build and operate high speed production lines is critical. Specialty product manufacturers usually concentrate on producing small quantities often with more specific design goals and tighter tolerances. Their challenge is to maintain consistency since unexpected lamp failures can result in down time costing many thousands of dollars per hour. (Ushio All Lamps, Ibid) 

Most typically today, bromine or iodine are used as the active halogen components.   Nitrogen, argon and sometimes krypton gases from the atmosphere.   The choice considers thermal losses, arcing voltage, molecular mass and cost among other factors. ( Ushio All lamps, Ibid)

Heat Impact Resistance – The quartz glass envelope signifies that halogen lamps are much more resistant to heat impact than ordinary incandescent lamps.   There is almost no danger that a lit halogen lamp will break even if it should come into contact with cold water.    Halogen Cycle – When the filament is heated to a high degree, the tungsten evaporates and reacts chemically with tie iodine gas (halogen gas) inside the bulb to produce tungsten iodide near the bulb wall.   The tungsten iodide particles are moved by convection within the bulb and, when they approach the highly heated filament, they are decomposed once again into iodine and tungsten.   The tungsten returns to the filament once more and the same cycle is then repeated.   The process, called “Halogen Cycle,â€? effectively prevents blackening of the bulb wall and thinning of the filament’s tungsten, thus resulting in longer lamp life. (Ushio Halogen Lamps, Ushio Pamphlet #94-3-1000 YO(24) Japan    pp.1-2)

Interference Filters: These filters are sometimes called “Dichroicâ€?and provide selective transmission of radiant energy.  They are generally used to transmit light and reflect the invisible radiation. (1) Infra-red in the beam is minimized (up to 85% reduction) with no significant loss of light. Re-directed radiant energy is deflected to a heat Absorbing collecting surface which must be cooled by more conventional air or water techniques.

Note (1): Interference filters are also available as “cold mirrorsâ€? to reflect light and transmit infrared.  These are useful for reflecting contours.  “Dichroic Beam Splittersâ€? act down range of the lamp, and act as a lens transmitting light while reflecting radiant heat.

Transmission:  Light Approx 92%,     IR Approx 15%

Heat Absorbing Glass: These materials tend to absorb some energy in the visible spectrum as well as infra-red.  However, some types are relatively effective—absorbing as much as 80% of the infra-red while transmitting approximately 75% of the light.   Because heat is principally absorbed (rather than reflected) a temperature rise occurs n the glass it-self.  This surface tends to become a radiant heating panel unless effective air circulation is provided to minimize the build-up of heat. Transmission:  Light Approx 75%,  IR Approx 20% Water Filters: Many liquids will absorb large portions of the infra-red energy while transmitting most of the visible wavelengths.   A one-inch thickness of water, for example, will absorb approximately two-thirds of the invisible energy.  While such a circulating water system is not a normal procedure, it may be useful in limited situations, particularly in conjunction with other water-cooled panels.

Transmission:  Light Approx: 85%  I.R. Approx 30%

Incandescent Lamps: The efficiency and operation of a filament lamp is relatively unaffected by temperature.  However, the effect of heat on lamp and fixture materials may be a critical design consideration.  (Also see “Lamp Heat Emissionâ€?)

Ambient Temperature: The filament itself operates at a very high temperature (E.G. 4,000-5,000°F.),  so any normal change in air surrounding a bulb is relatively insignificant and will not affect filament temperature.  Since filament temperature is neither increased nor decreased, there is no adverse effect on lamp life or light output.

Bulb Temperature: If a region on the bulb is heated to the softening point of glass, a blister or bubble will develop due to the pressure of the gas inside.   Most general-purpose lamps produce maximum bulb temperatures below 500°F (and often below 300°.)  With higher wattage lamps and with compact special-purpose sources, however, the glass temperatures may be a design consideration.

Maximum Safe Operating Temperature for Bulb Glass:    (Approximate)  Soft, Lime Glass   700°F.  Hard, Heat-Resistant Glass 855°F.  Molded, Heat-Resistant Glass  975°F.  Quartz Tubing   3,000°F. Bulb Position: Because Convection Heat Rises, location of the “Hot Spotâ€? will vary with the bulb position.  Some lamp types are limited to certain burning positions to insure that glass temperature limits are not exceeded. Base Up lamps have convection of heat flowing upwards from the filament along the lead-in and support wires (at the center) to the base of the lamp. From there, it is turned around (in a high pressure exchange due to the amount of heat convection verses the size of the stem,) and flows along the outside of the bulb until it hits the top of the envelope which is in a down position, than back into the filament to be re-circulated.  How the filament is supported, especially on C type single filament lamps is also a major factor in burning position, horizontal/base up or down. are all factored into the design and layout of the hangers/supports, and how tight they keep the filament or how much sag/stretch and eventual breakage is countered by the supports fixed in a ceratin position. Internal Convection: Base Down lamps flow in the opposite direction (filament to top of envelope, around the bulb to the stem/base, than back up the center to the filament.) except not all of the circulating air reaches the base of the lamp. The lamp base on these lamps is slightly cooler than on base up lamps because less convection heat is directed or forced into the smaller turbulent area of the lamp base.  The heated air/gas flowing in this area (having already circulated over ½ way around the bulb does not have the pressure to force its way into the turbulence of the lamp base/stem, thus leaving the base cooler because it does not contact as much heat.  The overall globe temperature and amplitude of heat circulating is more however through the filament because of the shorter path of circulation of the heat. These differences in circulation of heat within the lamp are important factors when things like porcelain verses plastic lamp bases are in question (See “Chimney Effectâ€? Below,) or in the composition of the materials making the lamp and its efficiency verses wattage are involved.

Reflector Focus of Energy: When circular or spherical reflectors are used to re-focus light, the physical position of vulnerable lamp parts becomes a design consideration — to prevent a focus of radiant energy on the bulb filament.  Such concentrations of heat, whether caused by faulty design or maladjustment of the unit, can cause glass failure. Exposure to Water: Gas-Filled lamps must be protected from localized cooling (thermal shock) due to rain, snow, or even large bugs.  This causes bulb breakage.  Glass cover plate (or screens) are used for protection (given proper ventilation or high temperature lamps to counteract the increased heat) or hard glass bulbs may be used.

Contact with Metal: Thermal cracks may result from metal fixture parts touching the bulb.  Localized cooling causes internal stress and can cause glass failure.  Note if the lamp is rated higher than the reflector or fixture, lamps which are out of focal adjustment and too close in proximity to the fixture, can also cause burning, rusting, or other fatigue on the fixture in addition to the lamp – especially with adjustable focus bases on quartz fixtures.

Lamp Base Deterioration: Lamp base temperatures are a basic consideration in fixture design.  While most fixtures are properly designed to dissipate the heat, excessive temperature can be caused by over-voltage operation or by the use of lamps of higher wattage than recommended.  This can adversely affect the bulb seal and cause failure.  In extreme cases, heat can also damage the socket and adjacent wiring. Maximum Safe Operating Temp. for Bulb Bases:    (Approximate)

 Regular Basing Cement  345°F.

 High Temperature Basing Cement 500-600°F.

 Mechanical Base   450°F.

Ventilated Fixtures: Vent Slots must be located below the lamp base to minimize the “Chimney Effectâ€? of hot air rising past the base itself.  Heat baffles are also useful for controlling convection currents — to reduce pockets of hot air near vulnerable parts of the assembly such as the areas where color is used, where the ballast is, where the fixture comes into contact with wood framing materials, where the fixture might be adjusted or handled by operators or service personal or for the purposes of heating and cooling in a space.

Housing Materials: Thermo-plastics (I.E. Acrylic, Styrene, Vinyl) are generally acceptable as components in fluorescent fixtures or systems, but their low resistance to heat makes them unsatisfactory for murcury and incandescent units.  With these sources, metal, glass, or thermo-setting plastics (I.E. Polyester) are required.

Lamp Heat: a 300 watt halogen lamp burns at 1,000 degrees, (Home Depot 1999 Calender Sept. 28.)  The temperature of a 1,000 watt Par can is 180 degrees, a Source Four heats up to 240 degrees. (Upstaging co. 1999 shop temperature test)

Fixture Efficiency: (Lighting Dimensions, April/May 1983    “….Worldâ€? p.?) (c.1983) “Unlike the late 1970s, few wholly new systems are being built today. Therefore for most shops any “third generationâ€? solution is going to have to be so spectacularly good or spectacularly cheap that it’s worth replacing existing equipment to get.â€?

(Four years later, ETC and Altman came out with their new fixtures and opened the floodgates.)

 â€œImproving fixture efficiency means increasing the amount of light a fixture of a given size and wattage produces or decreasing the size of the fixture required to produce a given amount of light.   Miniaturizing fixtures isn’t a new idea; theatrical designers have asked for decades why a smaller leko can’t be built so more fixtures can be crammed into positions with limited capacity. (See the MR-16 Par Can)   The biggest problem (given a compact enough light source) has always been heat.  Most of the electrical energy pumped into a tungsten-halogen bulb is wasted as heat and the size of the fixture cannot be reduced beyond the point at which its internal temperature climbs beyond the limits of the materials in the fixture or bulb. (eg. the FEL and TP22)

 One fix, of course, is to reduce temperatures by increasing the rate at which heat is transferred to the outside world.  Performance lighting is not a stranger to the technique.  Fifty years ago some carbon-arsenic projectors were circulating water through their condenser lenses to protect delicate slides from heat.   Today there are a variety of materials, components and techniques for heat control (many spin-offs of military electronics packaging and the space program.)  A miniaturized fixture built with them would have the advantage of small size, comparable operating cost and allow the use of current dimmer equipment.  The question is whether anyone particularly outside the tour market, is willing to pay the premium prices required for a fixture that is “onlyâ€? smaller than its predecessor – or even the investment of the funds it would take to figure out just how much more it would cost.

Another method of increasing efficiency is to use some new-fangled light source that produces more light (and less heat ) from the same amount of power: a high lumen to watt efficiency.

(See HPL, HX600, MSR and MR-16 technology as compared to standard quartz lamps.)

There are many other light sources with far higher lumen/watt efficiency than the quartz-halogen bulb.   But if efficiency were the only important criteria, we would have fluorescent tubes in our fixtures.  In fact light sources for performance lighting have to satisfy some very demanding criteria and no commercially available source yet satisfies them all at a total cost comparable to the tungsten-halogen. Sources for fixtures with controlled beamspreads require a luminous area small enough for a reflector of reasonable size to collect.   They require a relatively continuous spectral output if we are to filter out a wide range of color using current techniques.   And they require a close color and intensity match from lamp to lamp across the life of the lamp despite aging, input power variation, and operating temperature swings.   Measured against these criteria, the field narrows before you factor in three more problems:

1) Operating cost.   At rated life, a PAR64 has a life cycle cost of about $0.10 (1983) per hour.  The sources touted as its replacement have much higher operating costs – and higher fixture costs.  

2) Suitable Higher Efficiency (discharge sources need high-voltage ignitors the start) and some form of power conditioner which varies from type to type to run.   Therefore the system user gets a choice between a simple magnetic ballast relatively cheap but heavy and large) and an electronic ballast which generally trades weight for cost and complexity. 

3) Discharge sources are not electronically “dimmableâ€? in the sense that we use it, instead it is much like a follow spot, it can be dimmed only by mechanical gating means such as the shutter/iris dimming technique. 

The ceramic arc tube resists this material loss, can be manufactured to tighter tolerances and withstans a higher temperature to provide a more constant colour.

Filament lamps also have a major advantage over diode or cathode type fixtures, in that they are flicker free, instead of a using a pulsed arc of light to illuminate surfaces, incandescent types gain light by resistance to the filament which shows less variation from pulses in current than the arcs of light in other fixtures.  This creates a more natural mood (GE Halogen Performance Plus Bulbs, G.E. Lighting #202-81341 p.2)   Ceramic burner tubes will reduce the flicker