“since lamps are running on AC, switching directions 50-60 times a second there for in fact dimming them 50-60 times a second does that cause them to wear out?” - propmonkey
Steve B is correct on the thermal inertia question. It extends their life or at least lets it attain it’s designed lamp life as opposed to full non Hz
voltage.
The
filament has under full
current sufficient resistance in the
wire that it incandesces - that
wire gets so hot it glows. Dependant upon how hot that
filament got, and it’s size it will take a while to loose enough heat so that any damage is done to it by instantly heating it up again and again because it’s not getting cold enough will not have sufficient
effect given the duration of this
power being shut off. The larger the
filament, the longer it takes to cool down. Instead as also thought of, it’s kind of like putting the lamp on a
dimmer when cycling on and off at 60Hz.
The lamp is designed for a specific
voltage every 60Hz. Should you speed this up or have 120v of non-cycling
current it would get much hotter and wear out faster given sufficient volume in
voltage. Not designed for this. The lamp would be as if at a higher
dimmer ratio, get hotter still and wear out faster. Should you slow down this speed or dim the lamp by way of
current chopping or
voltage lowering, the
filament is operating at a cooler temperature. Since the
filament is operating in a sustained one step below melt down when at peak
voltage, it does wear out and get thinner in some areas with time - even with the
halogen effect. Lowering how close to melt down the resister is operating at thus prolongs the life because it’s not getting so close to melt down and less of the
filament is burning off.
The
halogen effect is that of spent particles of the
filament in having burned off as if a candle, being re-deposited once it cools and is absorbed by the gas, back onto the hottest part of the
filament. A
halogen lamp however is not perfect in that while for the most part it re-deposits the
filament to continue burning or resisting, it’s not always placing the
filament particles back onto the points it burned off of. There are some parts of the
filament that are cooler than other parts by way of resistance thus those parts will still wear down in the extended life and higher operating temperature lamp still wearing out. Because the
filament can not burn hotter without wearing out as fast, it thus operates or incandesces under the
halogen effect at a higher temperature due to the heat and burning up being replentished but only for the most part. This also is reflected in a larger luminous output due to the increased resistance, and because it’s replentishing the
filament often an extended life. Just as with
incandescent lamps, you have certain checks and balances. You can have higher
voltage, life,
color temperature or luminous output
halogen or not. Change the balance of one, you change that of all. A
tungsten -
halogen or xenon repressed
filament or vacuum tube type will melt down once it recieves
voltage sufficient it gets somewhere beyond 3,500K, or with time wears out. Simple as that, in most cases, you balance lamp life with output and/or
color temperature, and price of materials used to suppress this burning up. This along with
voltage rating.
This Hz factor and materials used is unlike turning on and off the same lamp a few thousand times but waiting for it to cool down between turning on and off. The lamp that’s had time to cool down will suddenly get shocked into incandessing. This damages the
filament because of the
shock and is why most dimmers have a “warming
current” running
thru the lamps when at 0% on the
dimmer. A lamp with some
current running
thru it, not enough to incandess but enough to have some heat will both go to full faster and be less damaging by way of shocking the lamp’s
filament that otherwise at one moment is cold and in the next is burning somewhere between 2,800K and 3,500K. This is the main difference in wondering why a lamp that’s turned on and off will wear out fast, but in cycling
thru an
alternating current, it’s not really wearing out faster. It is in actuality but the AC lamp is designed around this 50 or 60 Hz cycling. A lamp rated for 60 cycle would wear out faster given 90 cycle because it’s not designed to get that hot, instead the lamp is designed to be on a
dimmer 60Hz, and it’s specified
voltage. Lower the
voltage to a say 120v lamp and it might be better able to withstand a faster cycling of
current peaks. Also conversely, put a 12v lamp on 120v and it will work, but wear out really really fast. I
call it going
super nova as that in many ways is what that
filament will do in being really really bright but short in life. Have a 110v lamp next to a 130v lamp and given all else is similar, the 110v lamp will wear out faster but in burning be much brighter. It also won’t be able to take
voltage spikes or turning on and off as it were as easily. Turning on the lamp when cold is a
voltage spike just as if a lightning
strike. What
effect turning on the lamp will have is in relation to the magnitude of the
current running
thru the
resistor and it’s rated resistance. Your 115v lamps will be very bright, but will be less sufficient in turning on and off in a cold start than that of a 120v lamp. This especially if the actual
voltage while at full to the lamp is at about 118v. Over
voltage for one in having that extreme output, under voltaging for the other. Yep, your
Par Can will stand up to a cold start better than S-4
Leko.
Take a piece of 12ga cable. It should be able to have 25 amps go
thru it, or a sustained 20 amps going
thru it for a period of time. This is not to say that it can’t have more, it’s just that due to size and resistance in the
wire it will heat up and wear out faster if not also incandess right before it melts down given high enough amperage or a lower amperage sustained amperage. Your 12ga cable given enough
current can right before it starts a fire or melts down be a
filament lamp. On the other
hand, a
strobe lamp that is at times very high in
current does not really wear out the
wire given time between that
current draw in the
wire and time to cool down.
Think also steel while in a molten state as similar to a
filament incandescing. That’s a lamp
filament. Note also the differences in tempering and quenching steel and it’s effects on the differences in strength of the metal. Tool steel or a knife
edge verses that of mild steel that can take an earth quake without breaking due to brittleness. Lots of things you can do to steel in making it or damaging it. Also take away the heat under the steel and it’s not going to get hot enough. Materials of the
filament, how they are treated or designed for use, and flame/incandescing supprescents also
play a factor in lamp design. Lamps lower than 40w are often just vacuum tubes without air. Others use argon or bromine gas under pressure to suppress the burning up of the
filament. - this even for
halogen lamps. Add a very expensive krypton or better still xenon gas, and it suppresses the
filament burning up further in making for a higher output lamp still. Add alloys to the
tungsten fiber and it has the same
effect in extending the life of the lamp
etc.
(For more information: Go to the Sylvania website, click onto photo optic lamps and do a browse/search for something like the EVC lamp. Click on one of them it pulls up. This will bring up that lamp’s specifications and at the bottom of the
page will be a further reading documents section. Click on it and you will have a PDF document called “Technology and Applications,
Tungsten Halogen Low
Voltage Lamps Photo Optics.” This free
book with another one you can find in searching for a arc lamp like the HSD 250/78 in the same photo optics section and looking for further documents on it are books that are rated amongst the best books on the subject and for free. Print them and read them. You will know all about
filament lamps in a language that’s both expert and easy enough to read and understand at 43 pages. Very highly recommended reading.)
“what if they were running on DC so they are not dimming 50-60 times a second, and therefore extending the life of the lamp?” - propmonkey
An AC lamp for the most part can be universally run off DC
voltage. This given there is not many 120v DC systems out there, so it’s primarily speaking as a general thing for low
voltage lamps.
Than on low
voltage and DC lamps we get from GE:
“The best
filament design for use in a
shock and vibration environment is a low
voltage, high amperage
filament. This represents a short, thick
filament. The higher you go in
voltage, the longer the
filament, which then necessitates the use of hangers. The hangers will increase the higher
voltage filament reliability, but the effects of notching due to the soret
effect are increased. Orientation of the
filament in relationship to the axis of vibration will also affect the lamp performance in a
shock and vibration environment. The most critical consideration in protecting a lamp against the effects of
shock and vibration is the packaging by the user.
Filament damage, due to
shock and vibration, can be reduces or amplified by the way the lamp has been assembled into the user’s product.”
Most
Stage and Studio type lamps are to some extent
shock resistant to some degree by way of
filament humm and vibration being designed against due to well supported filaments that don’t vibrate with the
current. At least when given the
alternating current a humm that’s detectable. The more supports or shorter the
wire, the less detectable humm and more rugged the
filament in support. On the other
hand, as above the more supports, especially once we add a DC
current to this
effect, the more a factor this support or even
grid will have on the lamp in how it cools down or on the molecular
level is effected by the
current passing
thru it.
From GE’s “Sealed Beam and Specialty Lamp Catalog” we get this:
“DC
voltage can reduce life by approximately 50% to 70% due to the “sawtooth” or notching
effect. Notching, also known as electromigration, is a change in the molecular structure of
tungsten due to the DC
voltage. The results are hot spots that accelerate the thermal conditions, the evaporation rate and embrittlement. DC
voltage operation is not recommended for
filament lamps.”
Think about it this way and in what’s not stated where the
halogen effect and filaments also burning up is involved. All the
current comes from one direction. Since it’s coming from one direction, one part of the
filament will have more resistance than another part of it. Dependant upon materials or instances of the lamp’s structure and manufacture, such parts of the lamp most effected by the magnetism might not be at the polar end of the lamp either. More heat from damaged parts of the
filament means more burning and heat. While by way of resistance, all the
filament gets hot enough to incandess, one part of it is going to be hotter than another part. This will cause it to be brittle, electrons to
line up due to the magnaitism and in the case of
halogen gas re-depositing spent
filament molecules back onto the hottest part of the
filament, one part of the
filament that does not get replenished as much. As shown in the catalog, under a microscope, you can see actual notches in the
wire that develop in the
filament thus the term “
filament notching.” These are most likely breaks in the
filament due to microscopic impurities in manufacture or material. It’s a bent
wire, urr yep, it can have problems in having been bent if not bent under the right conditions. Due to a specific
current flow,
filament notching and heating of one part of the
filament or certain parts of it over others can be much more of a problem on DC
current than on
AC current that balances this out.
Filament noise in question might also be interesting with DC
current.
Consider this
halogen effect, or in general the cooling of the lamp as you get further away from the
filament as per a wind that’s generated by the thermal effects of the heat. Touch the glass of a
halogen lamp and either that heat is reflected or attracted to that spot touched. This as opposed to normal cooling cycle by lamp verses
filament shape engineering than effects it’s heating and cooling cycle as if a wind
blocking wall or tunnel
effect. One can only imagine the amount of cooling cycle thermal wind inside of a lamp - sufficient to make it explode at times. Often the
filament will be either attracted or blown towards or away from where it’s touched - depending upon say skin oil reflects or dirt attracts and retains heat left on the surface. Normal cold lamp +1 AU (Atmospheric
Unit) pressures inside a lamp once heated at the
filament are higher than those areas further away from the
filament no matter the +AU filler gas or -AU vacuum in use, pressure both increases heat, but also suppresses it once it cannot expand further. Used to be you could take a bunch of lamps and
throw them at a brick wall and get a big pop from them as great fun. Not as much fun in taking a bunch of
HPL lamps and throwing up against a wall. This noise was from the vacuum in the lamp. Most normal lamps don’t pop unless super heated in very high AU pressure.
(A xenon follow spot lamp given it’s pressure when hot will explode with the concussion of a
hand grenade due to AU pressure. Break it while at room temperature and it’s a major mess but not nearly as loud or dangerous.)
Instead the use of noble gasses (and we all hate chemistry class) does not require the vacuum in suppressing the fire as it were in the
filament. No high or low pressure, no pop. On the other
hand, you
drop a even high pressure
mercury vapor lamp and it’s the same pop and explosion in this case as opposed to implosion but similar noise. (Doot’ there goes $250.00 in lamp and an hour to clean up.) Noble gas fillers are safer to be around at room temperature and much more efficient in suppressing the burn up.
It’s the same with any gas or vacuum within the lamp in that this wind goes from hot to cool, but has to
return in a sort of weather
pattern back to the heat source/
filament due to vacuum
effect.
In getting back to your flash light or car lamp, given such an above
effect on the lamp is known, by design, while a AC lamp is just a
wire and it’s resistance incandesses, for DC situations you can design a lamp that when
polarized as to which direction the
current comes from design a lamp stronger
in one area than another, much more heavy duty or just down rate the lamp for DC usage over what it could be otherwise. Also given it’s lower
current, such effects of notching
etc, will be to a much less extent. The DC specific lamp once designed for such effects DC
current will have on it would not react well to
AC current. At times you will find some DC specific lamps if not a lot of arc lamps that require a specific direction of
current to
strike and maintain the arc.
On the other
hand, (I trust no offense on his part)
Steve B was correct but incorrect in the case of this “
halogen effect” as it relates to dimming of the lamp. The actual statement was from something brought up in a I think 1998 Philips catalog:
“Low
Voltage halogen lamps should not be dimmed by more than 10% of their rated
voltage since this will result in a reduction in life. Standard
tungsten filament lamps (with no
halogen filling), can be dimmed to zero volts, resulting in virtually endless life. However if low
voltage tungsten halogen lamps are dimmed by more than 10%, the lamp will be operating at too low a temperature and the free halogens in the gas fill, will attack the cooler parts of the
tungsten filament i.e. where enters the quartz or glass
envelope. The
wire at that
point will then be eroded and eventually will fail. So if dimmed by 10% or more low
voltage tungsten halogen lamps will not have an extended life but are unlikely even to reach their rated life.” - Philips Website, Optical p-1
Higher 120v+
voltage lamps have sufficient volume of
current going
thru them that such an
effect of dimming is not much a factor here. A 120v lamp even at 20%
power and no longer incandescing - realistically
dimmer curve level of 0% if well maintained at about 20%. That’s sufficient in
power to heat the iodine in the gas sufficient that it’s still functioning to catch the molecules of the
filament and re-depositing it on the hot spot of the lamp’s
filament. It’s also by volume of
voltage still much more than the common 12 and 6v lamps stated above as a problem. On the other
hand, once at 0% by dimming curve (80% range not 0-100% in reality), once you get to 0% if your dimmers are left on day and night, just as a fully loaded 12ga cable under full load will wear out and burn up with time, this lamp in getting constant heat and burning even if not incandescing will still wear out. As also stated above, the
halogen effect might be too cold to continue working. While it’s doubtful that the iodine in the gas will be sufficient to attack the
pinch seal if 120v as a
base, this lamp under warming
current will still for all intensive purposes be an
incandescent lamp thus if left on 24/7 have the same effects on lamp life as that of a lamp on a
dimmer in extension of life but it still burning up. In this case, at some
point the
halogen effect will stop for all intensive purposes working and it will only be a
incandescent lamp in life. The iodine in getting board might not attack the
pinch seal or cooler parts of the
filament, but it’s also not interested enough to replenish the lamp’s
filament. By assumption or theory, you will probably see more blackened
halogen lamps similar to that of
incandescent on dimmers left on day and night (unless given loss of signal they do go to a actual 0% in
voltage I don’t think they do.) This would be from a now non-halogen
effect lamp burning up it’s
filament over time. Might only be a warming
current, but it’s still heat burning up the
filament. Once the
filament is expended and not replenished, it won’t be ever again to any significant extent heated up again sufficiently to go back to the
filament. Once burned off in other than
halogen effect conditions, that part of the lamp life and
filament is instead spent permanently.
Low
voltage by code is what... about ?64 volts and it’s also supported in this less than range for consideration of what’s a low
voltage lamp by lamp specification. Just as in design, you tend for highest
color temperature and output want to install lamps in the fixtures most close to the expected use, on a low
voltage halogen lamp,
power is not always good. Yea, you can install a FEL in a
Altman 360Q and it will be more
power. But on the other
hand, if normally at 60%, you loose both
color temperature and output over that of a 575w lamp or even 400w lamp given improved efficiency and it at operating capacity. That 400w lamp at full next to a 1Kw lamp
bit dimmed to 60% might just have more output and a better less amber
shift color temperature.
For this reason, low
voltage lamps are possibly best specified to be not dimmed. Instead just change the wattage of the lamp to fit the need. Cam you dim them for a show, sure, should you use them under a
dimmer at home, perhaps not as smart. There is also some especially domestic as with DC lamp corrections possible for this stated
effect. Often the premium brand especially
stage and studio
line of low
voltage lamp with improved pinch technology will suffer less from these problems than other brands and types. You get what you pay for.
While iodine - the primary
halogen effect chemical is not an acid, it has the same
effect on the
pinch seal otherwise by way of escaping
thru the glass sealed metal
envelope or attacking the metal in this coolest area. But this
effect is only where the
voltage is too low to sufficiently heat the gas but hot enough to incandess the small
filament. It should not be the same with
line voltage lamps under a
dimmer - this unless their
voltage while powered up at some
point matches that of the 24v or lower common low
voltage lamps above stated. For 24v as say a normal for low
voltage, 10% of this is 21.6v. That’s right about at the warming
current for a 120v lamp thus the effects of leaving your dimmers on 24/7 is of concern but not yet studied by the lamp companies. I don’t think the
dimmer companies on the other
hand note a serious problem with this so the actual problem
voltage might be more like 11v.
Might not have as much a
halogen gas eating away at the
pinch seal, but it’s still a lamp operating that is expected not to
gain the extended life of the
halogen effect. Instead it’s only a
incandescent lamp while under a
dimmer. Lamp life will still be good, just not improved by the
halogen effect in being great. Should your
trim setting on the
dimmer be off either in the plus or minus way, this can also have an
effect on lamp life. Get your dimmers maintinenced professionally yearly. This will save money and optimize your performance. A
dimmer that’s actually set for a real 26% when at 0% on the
dimmer, will clip off at 96% instead of 100%. The other discussion of a dim glow to one’s lamp thus is say at 30% and his maximum output of the lamps given 80% ratio but 120v is only at 90% of the maximum output of the lamp by way of potential but nothing there to maximize it. This as opposed to a
dimmer trimmed to 10% that also will only get to 90 % because that’s all that is available in a actual 80% dimming range. 5% less - 15% warmer
current can extend expected lamp life to a degree that probably won’t be noticed in maxing out at 95%, but more than this you potentially damage your lamps. Even with this you might be.
There is some science that is probably easy but at this
point is not studied or known for sure at this
point. That Euro people and those using low
voltage cyc lights working in series find as exception to these rules. The low
voltage lamp when working in series for some reason such as in a strip or
cyc light for some reason does not have problems with under 10% dimming having an
effect on the lamp life - no matter what brand of lamp is used. It no doubt is much in an electrical way due to how the
system works, but it’s still not well defined as to why you can dim while low
voltage, a lamp in series wiring - say 12 of them at 12v in a 120v
system without the
halogen effect problems on life. Or in the Euro
system, two 120v
PAR 64 lamps without problems especially with the now in theory 10% warming
current to the lamp.
Consider this abstraction however more exception as to it’s function of a bunch of small filaments acting as one 120v
filament however as opposed to any difference of the rule in general. Don’t worry much about your 300 hr/115v lamp lasting only now 50 hours. Be concerned however about the 12v lamp at 10% on the
dimmer most of it’s life.
Finally in response to Steve B’s very good post, lamp life is dependant upon the brand and as shown in IES #LM46-98, “IESNA Approved Method for Photometric Testing of Indoor Luminaries Using High
Intensity Discharge or
Incandescent Filament Lamps,” the in general lattitude and longitudnal and atmospheric conditions of the testing faculty. What’s tested for life in Southern China won’t in a absolute sense be what the same lamp in Northern Germany will find in output and life of the lamp. Atmosphere and elevation
play a factor even one country to the next in Europe as this boring document studied.
In addition to this, there is no even if using the same test equipment - noted above to have different data dependant upon location, there is certain testing data that is company specified. Sometimes 10% of the lamps in sample are thrown out before that 50% in lamp life is stated. They don’t mention the 10% in saying how well their lamps test nor in many ways do they need to. So you might have 110 lamps in the test, or might have 100 but only after 60% have died do you state a lamp life. Other testing companies have other data on “average” lamp life - or what they consider average. Even Philips today has two figures for their moving light lamps. Both an expected lamp life, and replace before lamp life data published. There is also initial lumens and sustained lumens and
color temperature as different and normal.
In testing lamps, it’s also a question of how average these lamps are that are tested. Is it just lamps pulled from one lot of them - initial specification grade of them or to what extent of mixed and un-known in lot the lamps are. Is of 100 lamps, it all of the same lot in manufacturing or some and one of each in being tested? Is on the other
hand, it a question of an average of average of testing of these lamps continuously or the average of one test? Lamp specs constantly change in published tested specification. How new and up to date is your source on what to expect with a new lamp verses that of an old lamp? Specifications are important but very much subject to question in changing or accuracy. Within the same lot number of lamps, dependant upon the time of day (how early in the morning, near vacation, near lunch or quitting time) you will get some that will last longer, and others that will if rated for say 750 hours die after 360 and others that will last over 1,580 hours without problems. About 80% of listed life - given you can accurately
track this is warrantied. After and including this, it’s subject to manufacturer inspection of the lamp for warranty and approval of it. You pay shipping at times both ways. Buyer beware in getting what you pay for in a quality control type of way.
Finally,
“Even if using DC would allow lamps to last longer it is much more dangerous. Your muscles work on very low levels of DC
current (along with lots of chemicals and the such). Defibrilaters use DC
current and we all know how a human body reacts to that. If you were to touch a DC source at 120v 20A your body would freeze into place and you would not be able to let go, it can happen with AC as well but generally it doesn't as the changing
current will at some
point send your muscles the other direction causing you to
release the source of
voltage (hopefully before 0.6A passes through your heart).” - M Swan
Yep I touched stupidly the 20 amp
circuit.
Dimmer never tripped,
conduit melted into the palm of my
hand, show lighting browned out and I could not let go up until a continuos effort to do so. The actual
effect is as follows:
Effects of Electrical
Current on the Human Body
0.005-2mA = Just noticeable
2-10mA = Slight to strong muscular reaction
5-25mA = Strong
shock, inability to let go
25-50mA = Violent muscular contractions
50-100mA = Irregular twitching of the hart muscles no pumping action (ventricular fibrillation)
100- >100mA = Paralysis of breathing
(PGS
Power Guard Systems, LLC.)
Note it does not specify
voltage type nor actual
voltage.
This is my field of study in the industry, but I’m not an absolute expert - more home study into it. Questions or something not covered so well?
Sorry, but in being a good question, the explained why part of the answer is very long.