shirriff@sprite.berkeley.edu (Ken Shirriff) (03/17/91)
Today's "Dear Abby" states that it only saves electricity to turn off a light bulb if it's going to be off more than 10 minutes. This is because of the power surge when you turn it on. Is this as bogus as I think it is? 10 minutes' power in a brief surge is serious current. Ken Shirriff shirriff@sprite.Berkeley.EDU
valley@uchicago (Doug Dougherty) (03/17/91)
shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >Today's "Dear Abby" states that it only saves electricity to turn off a >light bulb if it's going to be off more than 10 minutes. This is because >of the power surge when you turn it on. >Is this as bogus as I think it is? 10 minutes' power in a brief surge >is serious current. >Ken Shirriff shirriff@sprite.Berkeley.EDU I bet it has more to do with the wear & tear on the bulb turning it off & on than the cost of the electricity. The cost of the bulb thus has to be figured into the equation.
russ@bbx.basis.com (Russ Kepler) (03/17/91)
In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >Today's "Dear Abby" states that it only saves electricity to turn off a >light bulb if it's going to be off more than 10 minutes. This is because >of the power surge when you turn it on. > >Is this as bogus as I think it is? 10 minutes' power in a brief surge >is serious current. I read this one as well and my wife thought I was going to die from laughter. As to the truth of the statement: Bogus as all hell. But interesting nonetheless. It looks to me a cross of the "lights burn out faster if turned off and on repeatedly" and "cars use 10 minutes idling gas when started" lines. I believe the first (should I?) and partially believe the second (for older non-injected cars and not 10 minutes in any case). -- Russ Kepler - Basis Int'l SNAIL: 5901 Jefferson NE, Albuquerque, NM 87109 UUCP: bbx.basis.com!russ PHONE: 505-345-5232
pete@ohm.york.ac.uk (-Pete French.) (03/18/91)
in article <1268@bbx.basis.com>, russ@bbx.basis.com (Russ Kepler) says: > > In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >>Today's "Dear Abby" states that it only saves electricity to turn off a >>light bulb if it's going to be off more than 10 minutes. This is because >>of the power surge when you turn it on. > > I read this one as well and my wife thought I was going to die from > laughter. As to the truth of the statement: Bogus as all hell. But > interesting nonetheless. I think you'll find that it is true for fluorescent light fittings - thesew do take a considerable amount of starting current (hence the need for those nice little starter thingys...). Utter crap for a normal bulb of course. -bat. -- -Pete French. / "Two wrongs don't make a right, Adaptive Systems Engineering / - but three lefts do !"
dahls@elab-runit.sintef.no (Joern Yngve Dahl-Stamnes) (03/18/91)
In article <1268@bbx.basis.com>, russ@bbx.basis.com (Russ Kepler) writes: >In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >>Today's "Dear Abby" states that it only saves electricity to turn off a >>light bulb if it's going to be off more than 10 minutes. This is because >>of the power surge when you turn it on. >> >>Is this as bogus as I think it is? 10 minutes' power in a brief surge >>is serious current. > >I read this one as well and my wife thought I was going to die from >laughter. As to the truth of the statement: Bogus as all hell. But >interesting nonetheless. > >It looks to me a cross of the "lights burn out faster if turned off >and on repeatedly" and "cars use 10 minutes idling gas when started" >lines. I believe the first (should I?) and partially believe the >second (for older non-injected cars and not 10 minutes in any case). When you turn on a neon light it use more energy the first 1 or 2 minutes until it became warm. The light is also less before it became warm. So you can save energy only if you turn the neon light off and keep it off for a longer time. Turning it on again (after it has cooled down) cause it to use more energy. So it is possible to save energy by keeping the light on in some cases (read: don't turn it off if you are turning it on in 5 or 10 minutes). For normal bulbs - turn it off even if you are to turn it on again 1 minute later. BTW, Norway is the land where everybody turn on the light and nobody turn it off! Joern Yngve Dahl-Stamnes, The University of Trondheim, The Norwegian Institute of Technology, Division of Physical Electronics Please note: The address in the header is incorrect. My correct e-mail address is "dahl-stamnes@delab.sintef.no" * "Let chaos storm! Let cloud shapes swarm! I wait for form." * - - - - o o o o - - - -
mcovingt@athena.cs.uga.edu (Michael A. Covington) (03/18/91)
I never heard of anybody using NEON light for illumination. Do you mean FLUORESCENT (mercury vapor in a tube with white phosphor)? -- ------------------------------------------------------- Michael A. Covington | Artificial Intelligence Programs The University of Georgia | Athens, GA 30602 U.S.A. -------------------------------------------------------
leland@cbnewse.att.com (leland.m.kornhaus) (03/19/91)
For a 60 watt bulb to use 10 minutes worth of energy when first turned on it would have to draw 300 Amperes for a full second. I doubt many homes have much more capacity than this.
jcav@quads.uchicago.edu (john cavallino) (03/19/91)
In article <1268@bbx.basis.com> russ@bbx.basis.com (Russ Kepler) writes: >In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >>Today's "Dear Abby" states that it only saves electricity to turn off a >>light bulb if it's going to be off more than 10 minutes. This is because >>of the power surge when you turn it on. >> >>Is this as bogus as I think it is? 10 minutes' power in a brief surge >>is serious current. > >I read this one as well and my wife thought I was going to die from >laughter. As to the truth of the statement: Bogus as all hell. But >interesting nonetheless. > >It looks to me a cross of the "lights burn out faster if turned off >and on repeatedly" and "cars use 10 minutes idling gas when started" >lines. I believe the first (should I?) and partially believe the >second (for older non-injected cars and not 10 minutes in any case). Another one I heard about cars is that if you never leave the engine running for more than xx minutes at a time you will eventually discharge the battery, because the alternator needs at least xx minutes to recharge the battery from the drain of starting the engine. (The value I heard for xx was 15). Does this make sense? -- John Cavallino | EMail: jcav@midway.uchicago.edu University of Chicago Hospitals | USMail: 5841 S. Maryland Ave, Box 145 Office of Facilities Management | Chicago, IL 60637 "Opinions, my boy. Just opinions" | Telephone: 312-702-6900
jack@cs.glasgow.ac.uk (Jack Campin) (03/19/91)
shirriff@sprite.berkeley.edu (Ken Shirriff) writes: > Today's "Dear Abby" states that it only saves electricity to turn off a > light bulb if it's going to be off more than 10 minutes. This is because > of the power surge when you turn it on. Perhaps she said "energy" rather than "electricity", in which case it would have been perfectly true. Tungsten-filament light bulbs carry a much heavier current when switched on than when running, since their conductivity is highest when cold. The result is a heat pulse that damages the filament more than steady use. I have often seen a figure in the 10 to 20 minute range for the cost of this, either in cash to replace the bulb from it burning out faster or in energy used to make the replacement. On the other hand, the surge can't eat up more than a few milliseconds' worth of power or your fuses would blow. -- -- Jack Campin Computing Science Department, Glasgow University, 17 Lilybank Gardens, Glasgow G12 8QQ, Scotland 041 339 8855 x6854 work 041 556 1878 home JANET: jack@dcs.glasgow.ac.uk BANG!net: via mcsun and ukc FAX: 041 330 4913 INTERNET: via nsfnet-relay.ac.uk BITNET: via UKACRL UUCP: jack@glasgow.uucp
carter@mcs.anl.gov (Richard Carter) (03/19/91)
In article <1268@bbx.basis.com>, russ@bbx.basis.com (Russ Kepler) writes: > >It looks [like] the "lights burn out faster if turned off >and on repeatedly" [legend] > This is an interesting side point. When I was being examined for my pilot's license last year, I was grilled for a while about how fast various systems wear out and what you should do to extend their life. The examiner came up with this one (in reference to lifespan of Halogen landing lights on a plane), which I had never heard of before. I believed it at the time, since flight instructors have a religion about knowing the characteristics and limitations of equipment. Now that I think about it, though, it seems less reasonable. Anyone have the real scoop? If this is a legend, it may have arisen from the "most engine wear occurs during first 15 seconds after ignition, before the oil heats up and circulates" story. Now, this one I personally believe, having seen it stated so often, but I don't recall seeing any references to an actual study. -- Richard Carter | In dentibus anticus frustrum MCS Division | magnum spinaciae habes. Id tibi Argonne National Lab | praebet speciem lepidissimam! Argonne IL 60439 |
iho@cac.washington.edu (Il Oh) (03/19/91)
carter@mcs.anl.gov (Richard Carter) writes: >It looks [like] the "lights burn out faster if turned off >and on repeatedly" [legend] I can't really comment on the power consumption of the spike when you turn it on (and off, too), but I can tell you this. Light bulbs burn out faster when used with an A/C power source than when used with a D/C power source. This is because A/C is actually like turning the power on and off really fast (120 times each second -- half of the time, reversing the flow of current). One of the easiest ways of extending the lifetime of a light bulb is to simply rectify the power before it gets to the bulb. In fact, I've seen these "mystery devices" advertised to extend the life of your light bulb, which I'm sure are very simple rectifying circuits. Given this information, I don't think turning it off and on just one more time will make that big of a difference. Afterall, the power company is doing it 120 times every second. -- "I'm sorry. | Il Hwan Oh If you were right, | University of Washington, Tacoma I would agree with you." | iho@cac.washington.edu -- Dr. Malcolm Sayer, _Awakenings_ |
richg@locus.com (Rich Greenberg) (03/19/91)
In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes: >Today's "Dear Abby" states that it only saves electricity to turn off a >light bulb if it's going to be off more than 10 minutes. This is because >of the power surge when you turn it on. >Is this as bogus as I think it is? 10 minutes' power in a brief surge >is serious current. Sorry to inject some truth into this newsgroup, but. . . The bulb doesn't use 10 mins of power in a moment. But Abby is right. The reason is that a cold bulb has aproximately 1/10 of the resistance that it has at normal temprature/brightness. When you turn it on, for a few milliseconds, until it gets hot, it draws about 10 times normal current, which puts a strain on the filament. (Both from thermal expansion and magnetic force). This is why bulbs usually burn out when first turned on. BTW, if you have doubts on the magnetic part, hold a magnet near an unfrosted bulb, and watch the filament. If the magnet is strong enough, it will destroy the filament. Ten minutes of use costs about the same as the cost of the replacement bulb prorated over the many times it is turned on & off in its normal lifetime. The above is for incandescent bulbs ONLY. I don't know enough about flouresents to comment on them. -- ------------------------------------------------------------------------- Disclaimer: The above writings are the ramblings of one human being and have nothing what-so-ever to do with Locus Computing Corp. ---> Rich Greenberg, richg@locus.com TinsleTown, USA 213-337-5904
bender@oobleck.Eng.Sun.COM (I want to be eating rich soup in another town) (03/19/91)
In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: ->carter@mcs.anl.gov (Richard Carter) writes: -> ->>It looks [like] the "lights burn out faster if turned off ->>and on repeatedly" [legend] -> ->I can't really comment on the power consumption of the spike when you ->turn it on (and off, too), but I can tell you this. Light bulbs burn ->out faster when used with an A/C power source than when used with a D/C ->power source. This is because A/C is actually like turning the power ->on and off really fast (120 times each second -- half of the time, reversing ->the flow of current). One of the easiest ways of extending the lifetime ->of a light bulb is to simply rectify the power before it gets to the bulb. ->In fact, I've seen these "mystery devices" advertised to extend the life ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ->of your light bulb, which I'm sure are very simple rectifying circuits. ^^^^^^^^^^^^^^^^^^^ Heh, heh, heh...! It's time for the "mystery-device light-bulb-life-extenders-burn-down-your-house" flame war again! Larry Lippman, to your corner (after you get back from the DMV and state beurau of records :-), and John D'Armond to your corner... DING! THE BELL HAS RUNG! COME OUT WITH YOUR THEORIES, MUD-SLINGING AND SCIENTIFIC EVIDENCE FIGHTING, AND MAY THE BEST POSTER WIN!! -- Won't look like rain, Won't look like snow, | DOD #000007 Won't look like fog, That's all we know! | AMA #511250 We just can't tell you anymore, We've never made oobleck before! | MSC #298726 "NeXT is like Sun was 4 years ago & Sun is becoming like DEC" -misc
tneff@bfmny0.BFM.COM (Tom Neff) (03/19/91)
In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: >carter@mcs.anl.gov (Richard Carter) writes: >Given this information, I don't think turning it off and on just one more >time will make that big of a difference. Afterall, the power company is >doing it 120 times every second. Repetitive heating and cooling weakens the filament. This is not an issue with AC oscillation.
sbeagle@kennels.actrix.gen.nz (Sleeping Beagle) (03/19/91)
iho@cac.washington.edu (Il Oh) writes: > carter@mcs.anl.gov (Richard Carter) writes: > > >It looks [like] the "lights burn out faster if turned off > >and on repeatedly" [legend] > > Given this information, I don't think turning it off and on just one more > time will make that big of a difference. Afterall, the power company is > doing it 120 times every second. While I'm possibly talking through a hole in my head (just wait for the bright idea lightbulb to pop out and turn on!) I had assumed that the damage done to a lightbulb turning on and off was from the resulting heating and colling of the filament and surrounding bits. Therefore, AC current could be cycling it (is it?) but it's so fast that there wouldn't be time for cooling/heating damage to be done. "Just another attempt to spread fear and ignorance ma'am." -- Sleeping Beagle (aka Thomas Farmer) sbeagle@kennels.actrix.gen.nz The Kennels Ph. +64-4-796306 (voice) 25 Awarua St, Ngaio, Wellington, New Zealand. "You ain't nothin' but a Hound Dog."
rwb@vi.ri.cmu.edu (Bob Berger) (03/19/91)
In article <18627@milton.u.washington.edu>, iho@cac.washington.edu (Il Oh) writes: > > Light bulbs burn > out faster when used with an A/C power source than when used with a D/C > power source. This is because A/C is actually like turning the power > on and off really fast (120 times each second -- half of the time, reversing > the flow of current). I wonder if the author of this was serious. The turn on surge occurrs because the filament is still cold, and therefore has a low resistance. The resistance of the tungsten increases as it heats up. At 120 times/second, the filament does not have time to cool down, so the alternating current does not reduce the life.
anachem@silver.ucs.indiana.edu (|mehcana| (undersampled)) (03/19/91)
In article <12405@pt.cs.cmu.edu> rwb@vi.ri.cmu.edu (Bob Berger) writes: >In article <18627@milton.u.washington.edu>, iho@cac.washington.edu (Il Oh) writes: > >> >> Light bulbs burn >> out faster when used with an A/C power source than when used with a D/C >> power source. This is because A/C is actually like turning the power >> on and off really fast (120 times each second -- half of the time, reversing >> the flow of current). > >I wonder if the author of this was serious. The turn on surge occurrs because >the filament is still cold, and therefore has a low resistance. The resistance >of the tungsten increases as it heats up. > >At 120 times/second, the filament does not have time to cool down, so >the alternating current does not reduce the life. It was my understanding that it is the 120Hz hum that physically fatigues the filament leading to a breakage sooner on ac than on dc supplies. Listen closely to your light bulbs new and old. Hear it? If as the first reply to this thread said - it is the cost of the bulb that makes it cheaper to not be turning the bulb off and on (not the cost of electricity) then the physical stress on the filament at turn on/off = heat up/down should be the major fatigue mechanism to consider.
lamb@brahms.udel.edu (Richard E Lamb) (03/20/91)
> It was my understanding that it is the 120Hz hum that physically
????
Somehow, I always thought this was 60 Hz?slootman@dri.nl (Paul Slootman) (03/20/91)
In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: >carter@mcs.anl.gov (Richard Carter) writes: > >>It looks [like] the "lights burn out faster if turned off >>and on repeatedly" [legend] > >Light bulbs burn >out faster when used with an A/C power source than when used with a D/C >power source. This is because A/C is actually like turning the power >on and off really fast (120 times each second -- half of the time, reversing >the flow of current). One of the easiest ways of extending the lifetime >of a light bulb is to simply rectify the power before it gets to the bulb. >In fact, I've seen these "mystery devices" advertised to extend the life >of your light bulb, which I'm sure are very simple rectifying circuits. > >Given this information, I don't think turning it off and on just one more >time will make that big of a difference. Afterall, the power company is >doing it 120 times every second. Somehow, I think I must disagree. The reason that lightbulbs (the filament type) usually blow when they're turned on, is because at that point, the resistance of the filament is extremely low (go and measure the resistance of a 25 watt bulb...). If you're unlucky enough to switch on the power at exactly the peak of the sine wave, the bulb gets hit by 1.4142 * volts RMS, i.e. 1.4 times the rated voltage. If the bulb is aged, chances are THAT'S the time it blows. The gadgets sold to extend the lifetime of a bulb probably consist of a zero-crossing detection, such that a bulb only gets switched on at the point that the voltage is zero. The filament thus gets a (short!) time to heat up, thus building up resistance. Rectifying the voltage (if you add a capacitor to make a nice, steady voltage!) may kill your bulb sooner; the effective voltage is higher... (Main story is about right, I may have the details wrong. I'm not an electrical engineer...) Paul. -- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= : slootman@dri.nl : Don't hit the keys so hard, : : ...!hp4nl!dri500!slootman : it hurts : =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
jeh@dcs.simpact.com (03/20/91)
In article <10026@exodus.Eng.Sun.COM>, bender@oobleck.Eng.Sun.COM (I want to be eating rich soup in another town) writes: > In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: > ->carter@mcs.anl.gov (Richard Carter) writes: > ->>It looks [like] the "lights burn out faster if turned off > ->>and on repeatedly" [legend] > ->I can't really comment on the power consumption of the spike when you > ->turn it on (and off, too), but I can tell you this. Light bulbs burn > ->out faster when used with an A/C power source than when used with a D/C > ->power source. This is because A/C is actually like turning the power > ->on and off really fast (120 times each second -- half of the time, reversing > ->the flow of current). One of the easiest ways of extending the lifetime > ->of a light bulb is to simply rectify the power before it gets to the bulb. > ->In fact, I've seen these "mystery devices" advertised to extend the life > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > ->of your light bulb, which I'm sure are very simple rectifying circuits. > ^^^^^^^^^^^^^^^^^^^ > > Heh, heh, heh...! It's time for the "mystery-device > light-bulb-life-extenders-burn-down-your-house" flame war again! Larry > Lippman, to your corner (after you get back from the DMV and state beurau of > records :-), and John D'Armond to your corner... I don't know about "burning down your house", but a "very simple rectifying circuit" would *not* extend the life of the bulb via the mechanism described. These "mystery devices" are placed in just one leg of the circuit, can therefore only be a single diode, and therfore provide the bulb with half- wave-rectified DC. The bulb is now being hit with a half-sine-wave lasting 1/120 second, followed by 1/120 second of nothing, instead of a half of a sine wave every 1/120 second. The bulb therefore has MORE of a chance to cool down between "lightings" when run on the "mystery device". If anything, we would expect this to reduce the life of the bulb, not extend it! If indeed DC is better than AC, you need *filtered* DC, and you are not gonna get 100 watts of filtered DC out of a device about the size of two quarters. Rather, by providing the bulb with a half-wave rectified power source, the bulb is simply provided with a lot less RMS voltage. So it lasts a lot longer. (It also gets a lot dimmer. TANSTAAFL.) The mechanism that causes incandescant lamps to wear out faster if turned off and on a lot is twofold: First, there is a thermal shock each time the bulb is powered on, which causes mechanical stress to the filament. Second, the cold filament has a lower resistance and there is therefore an inrush of high current for a short time (milliseconds) before the filament warms up. Both of these mechanisms depend on the filament cooling down to room temperature when it's powered off. An ordinary light bulb filament run on 60 Hz AC does not have enough time for the filament to cool down "between cycles" for AC vs. DC to be a factor in bulb life by either of these mechanisms. (Again, if it did, the "mystery devices" would hurt bulb life A LOT by providing a whole half-cycle for the filament to cool down during each cycle.) --- Jamie Hanrahan, Simpact Associates, San Diego CA Internet: jeh@dcs.simpact.com, or if that fails, jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!jeh
dahls@elab-runit.sintef.no (Joern Yngve Dahl-Stamnes) (03/20/91)
In article <1991Mar18.153504.15818@athena.cs.uga.edu>, mcovingt@athena.cs.uga.edu (Michael A. Covington) writes: >I never heard of anybody using NEON light for illumination. >Do you mean FLUORESCENT (mercury vapor in a tube with white phosphor)? Yes, sorry for using the wrong name, but in Norway we call them neon lights. Don't know why, but that's the way it is. Joern Yngve Dahl-Stamnes, The University of Trondheim, The Norwegian Institute of Technology, Division of Physical Electronics Please note: The address in the header is incorrect. My correct e-mail address is "dahl-stamnes@delab.sintef.no" * "Let chaos storm! Let cloud shapes swarm! I wait for form." * - - - - o o o o - - - -
whit@milton.u.washington.edu (John Whitmore) (03/20/91)
In article <72R3y3w164w@kennels.actrix.gen.nz> sbeagle@kennels.actrix.gen.nz (Sleeping Beagle) writes: >While I'm possibly talking through a hole in my head (just wait for the >bright idea lightbulb to pop out and turn on!) I had assumed that the >damage done to a lightbulb turning on and off was from the resulting >heating and colling of the filament and surrounding bits. >Therefore, AC current could be cycling it (is it?) but it's so fast >that there wouldn't be time for cooling/heating damage to be done. I always thought it was another effect; the filament is coiled in most bulbs, and the magnetic force attracting adjacent coils when current flows is cycling at twice the AC line frequency (i.e. 120 Hz). I know for certain that triac light dimmers cause some bulbs to 'sing' at higher frequencies, and a little acoustic energy can definitely break a brittle metal filament. I also assume this force (which is maximum when the turnon surge occurs) is the reason for lamps breaking at power-up time. John Whitmore
forbes@aries.scs.uiuc.edu (Jeff Forbes) (03/20/91)
In article <22743@oolong.la.locus.com> richg@locus.com (Rich Greenberg) writes: > >BTW, if you have doubts on the magnetic part, hold a magnet near an > unfrosted bulb, and watch the filament. If the magnet is strong > enough, it will destroy the filament. > I presume that you mean that the bulb should be powered, since tungsten is not magnetic. Jeff Forbes "....I have not failed. I've just found 10,000 ways that won't work." Thomas Edison
farren@sat.com (Michael J. Farren) (03/21/91)
jeh@dcs.simpact.com writes: >In article <10026@exodus.Eng.Sun.COM>, bender@oobleck.Eng.Sun.COM >> ->In fact, I've seen these "mystery devices" advertised to extend the life >> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> ->of your light bulb, which I'm sure are very simple rectifying circuits. >> ^^^^^^^^^^^^^^^^^^^ >> >> Heh, heh, heh...! It's time for the "mystery-device >> light-bulb-life-extenders-burn-down-your-house" flame war again! > >These "mystery devices" are placed in just one leg of the circuit, can >therefore only be a single diode, and therfore provide the bulb with half- >wave-rectified DC. Precisely so. >The bulb is now being hit with a half-sine-wave lasting >1/120 second, followed by 1/120 second of nothing, instead of a half of a sine >wave every 1/120 second. The bulb therefore has MORE of a chance to cool down >between "lightings" when run on the "mystery device". If anything, we would >expect this to reduce the life of the bulb, not extend it! The thermal inertia of the filament effectively eliminates this as one of the possible failure modes. The filament isn't going to change temperature to any appreciable extent in 1/60 second - it takes several full seconds for the temperature of a filament to drop below red heat when you turn it off. >Rather, by providing the bulb with a half-wave rectified power source, the >bulb is simply provided with a lot less RMS voltage. So it lasts a lot >longer. (It also gets a lot dimmer. TANSTAAFL.) Not a LOT dimmer. Some dimmer, to be sure. The way I believe these work is by reducing the actual wattage used by the bulb without reducing the voltage. It would appear that the lifetime curve of a lightbulb is not linear with respect to watts, but closer to exponential - and the lifetime increase is significantly greater than the light decrease, at least subjectively. My personal experience with these devices has been very good. One fixture at the foot of the stairs, which required a new bulb every few months before using the device, was still using the same bulb three years after installation. The same thing was true of the light on the porch. You can postulate some other factor coming into play, but to me it didn't matter - all I observed was that the average lifespan of the bulbs used in fixtures with the device installed far exceeded the lifespan of bulbs in the same fixtures before installation. The cost of the diodes wasn't very high, and I think it's worth the cost to do a little experimentation on your own behalf, especially if you've got one of those fixtures that seems to eat lightbulbs as if they were salted peanuts. -- +-----------------------------------------------------------------------+ | Michael J. Farren farren@sat.com | | He's moody, but he's cute. | +-----------------------------------------------------------------------+
mercer@npdiss1.StPaul.NCR.COM (Dan Mercer) (03/21/91)
In article <72R3y3w164w@kennels.actrix.gen.nz> sbeagle@kennels.actrix.gen.nz (Sleeping Beagle) writes: :iho@cac.washington.edu (Il Oh) writes: : :> carter@mcs.anl.gov (Richard Carter) writes: :> :> >It looks [like] the "lights burn out faster if turned off :> >and on repeatedly" [legend] :> :> Given this information, I don't think turning it off and on just one more :> time will make that big of a difference. Afterall, the power company is :> doing it 120 times every second. : :While I'm possibly talking through a hole in my head (just wait for the :bright idea lightbulb to pop out and turn on!) I had assumed that the :damage done to a lightbulb turning on and off was from the resulting :heating and colling of the filament and surrounding bits. No. Rapid change of direction in curernt causes fatigue from a variety of causes, heating and cooling being only one. Oh, yes, and it does heat and cool, which can be seen by high speed photography. If you listen closely, you can hear the coil hum. Under direct current, bulb life is much longer. One of Edison's original bulbs has been burning steadily for decades using DC. BTW, electric current actually carries away some of the energy that might ordinarily be translated into heat. One the bulb is turned off, there is an additional heat spike, explaining why bulbs burn out when you turn them off. When I was in computer school, we were told of a SAC tube based computer behind three foot thick concrete walls where staff were warned that if the current ever failed, the resulting heat spike would not only melt hte computer but cause the surrounding concrete walls to catch fire. Needless to say, their backup power supplies (multiple redundancy) got everyone's close attention. : :Therefore, AC current could be cycling it (is it?) but it's so fast :that there wouldn't be time for cooling/heating damage to be done. : :"Just another attempt to spread fear and ignorance ma'am." : : :-- : Sleeping Beagle (aka Thomas Farmer) sbeagle@kennels.actrix.gen.nz : The Kennels Ph. +64-4-796306 (voice) : 25 Awarua St, Ngaio, Wellington, New Zealand. : "You ain't nothin' but a Hound Dog." -- Dan Mercer NCR Network Products Division - Network Integration Services Reply-To: mercer@npdiss1.StPaul.NCR.COM (Dan Mercer) "MAN - the ultimate one word oxymoron in the English Language"
brendan@otc.otca.oz (Brendan Jones) (03/21/91)
Along the lines of the current discussion re incandescents, the old story
goes that you shouldn't turn off a fluoro if you are going to turn it on again
in x minutes (usually 5 < x < 30 in folklore).
First, a sanity check should tell you that if it takes 3 seconds for a fluro
to come on, and it uses 5 minutes worth of power, then the fluoro is consuming
power at 100 times the steady state rate at turn on, or 600 times the steady
state rate if x=30. You'd blow a fuse for sure with the currents required!!
Secondly, when I was at Uni, I measured the turn-on energy required for a
40W fluoro and compared it to its steady state energy use using a Watt-Hour
meter.
The result? In the 3 seconds it took for the fluoro to turn on, it used 3
seconds worth of steady state power (40W), ie 33 mW-h.
Another myth debunked (sorry to introduce facts to a.f.u :-)
--
Brendan Jones | ACSnet: brendan@otc.otca.oz.au | What does
R&D Contractor | UUCP: {uunet,mcvax}!otc.otca.oz.au!brendan | your
Services R&D | Phone: (02)2873128 Fax: (02)2873299 | company
|||| OTC || | Snail: GPO Box 7000 Sydney 2001, AUSTRALIA | export?rando@skipper.dfrf.nasa.gov (Randy Brumbaugh) (03/21/91)
In article <2530@otc.otca.oz>, brendan@otc.otca.oz (Brendan Jones) writes: > Along the lines of the current discussion re incandescents, the old story > goes that you shouldn't turn off a fluoro if you are going to turn it on again > in x minutes (usually 5 < x < 30 in folklore). > > [. . . Calculations Deleted . . .] > > Another myth debunked (sorry to introduce facts to a.f.u :-) The way I've always understood this is that the lifetime of a fluoroscent lamp is determined almost entirely by the number of times it is turned on. The lamps are not cheap, so the lowest cost of operation is a trade-off between the cost of electricity and the cost of the lamp. (And sometimes more efficient lamps are more expensive!). Most calculations I've seen of this type determine that a flourescent lamp is generally best used in areas where it can be left on for more than 3 hours at a time, otherwise cost / hr of operation is higher than incandescents. Randy Brumbaugh rando@skipper.dfrf.nasa.gov
richg@locus.com (Rich Greenberg) (03/21/91)
In article <944@dri500.dri.nl> slootman@dri.nl (Paul Slootman) writes: >Somehow, I think I must disagree. The reason that lightbulbs (the filament >type) usually blow when they're turned on, is because at that point, the >resistance of the filament is extremely low (go and measure the resistance >of a 25 watt bulb...). If you're unlucky enough to switch on the power at >exactly the peak of the sine wave, the bulb gets hit by 1.4142 * volts RMS, >i.e. 1.4 times the rated voltage. If the bulb is aged, chances are THAT'S >the time it blows. This is partly correct, the thermal inertia is several cycles of 60hz, but the point in the cycle that contact is made is a minor (if any) consideration. >The gadgets sold to extend the lifetime of a bulb probably consist of a >zero-crossing detection, such that a bulb only gets switched on at the >point that the voltage is zero. The filament thus gets a (short!) time to >heat up, thus building up resistance. Not quite. These devices are called varistors or thermistors or some such name. They have thermal vs resistance characteristics just the opposite of a light bulb, i.e. high resistance when cold, low when hot. This limits the initial surge of current to a much lower value, and then warm up so they have a negligible voltage drop. NOT zero though, and this also adds to the bulb life by reducing the voltage (and light output) by a small amount. >(Main story is about right, I may have the details wrong. I'm not an >electrical engineer...) I am, although I haven't practiced it since I moved into computers. -- ------------------------------------------------------------------------- Disclaimer: The above writings are the ramblings of one human being and have nothing what-so-ever to do with Locus Computing Corp. ---> Rich Greenberg, richg@locus.com TinsleTown, USA 213-337-5904
jgk@osc.COM (Joe Keane) (03/21/91)
In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: >I can't really comment on the power consumption of the spike when you >turn it on (and off, too), but I can tell you this. Light bulbs burn >out faster when used with an A/C power source than when used with a D/C >power source. This isn't true. Light bulbs run on AC actually last longer than on DC. This assumes you're using the same amount of power in each case. >This is because A/C is actually like turning the power >on and off really fast (120 times each second -- half of the time, reversing >the flow of current). Yes, but the relevant parameter is the temperature of the filament. For practical purposes, this is constant even with AC power. A filament can't go from hot to cold in 1/240 second. >One of the easiest ways of extending the lifetime >of a light bulb is to simply rectify the power before it gets to the bulb. This has been discussed before on `sci.electronics'. It turns out that the main effect of the rectifier is simply to reduce the average amount of power delivered to the light bulb. So the light bulb runs cooler and lasts longer, but it's also less efficient. -- Joe Keane, amateur mathematician
lc@goanna.cs.rmit.oz.au (leon) (03/21/91)
In article <1991Mar20.185201.10187@sat.com> farren@sat.com (Michael J. Farren) writes: >jeh@dcs.simpact.com writes: >>In article <10026@exodus.Eng.Sun.COM>, bender@oobleck.Eng.Sun.COM >>> ->In fact, I've seen these "mystery devices" advertised to extend the life >>> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >>> ->of your light bulb, which I'm sure are very simple rectifying circuits. >>> ^^^^^^^^^^^^^^^^^^^ >>> >>> Heh, heh, heh...! It's time for the "mystery-device >>> light-bulb-life-extenders-burn-down-your-house" flame war again! >> >>These "mystery devices" are placed in just one leg of the circuit, can >>therefore only be a single diode, and therfore provide the bulb with half- >>wave-rectified DC. > >Precisely so. > This is some blurb on filament lamps from the RadioSpares (English) catalogue Filament lamps will perform to specification and give optimum life when : a) they are powered at the rated voltage from a stabilised 50 Hz ac supply b) ambient temperature is in the range 20 C to 25 C c) mechanical vibration is present Deviations from these criteria can have a significant effect on lamp life, particularly with low current filaments (<80 mA) Operation from a dc supply at rated voltage can reduce specified life by up to 50% The effect of variation from the rated operating voltage should be noted: 5% above will reduce lamp life by up to 50% 5% below will significantly increase lamp life Current consumption can vary by up to +/- 10% on stated values at rated supply voltage. Care therefore should be exercised when running a number of lamps in series. In such applications it is recommended that the rated voltage be reduced by 10% to prevent failure of individual lamps.
jeh@dcs.simpact.com (03/21/91)
In article <937@npdiss1.StPaul.NCR.COM>, mercer@npdiss1.StPaul.NCR.COM (Dan Mercer) writes: > If you listen closely, you can hear the coil hum. certainly. But this does not establish that the filament is undergoing significant mechanical shock due to heating and cooling. It just means that the filament is humming, no doubt due to its own electromagnetic field. This vibration will hurt filament life (and is the real reason that bulbs last longer on pure DC). > Under direct current, bulb life is much longer. One of Edison's > original bulbs has been burning steadily for decades using DC. this is a great UL! But it needs to be dressed up with folks who have been paid off by GE and Westinghouse not to reveal the secrets of the long-lived bulbs... Sorry, but Edison's original bulbs were lucky to last a hundred hours, given enough voltage to produce a reasonable amount of light. By reducing the voltage, ANY incandescant lamp can be made to last for decades, even on AC. > BTW, electric current actually carries away some of the energy that > might ordinarily be translated into heat. care to explain the mechanism whereby this occurs??? Don't tell me that the electrons "vibrate" the way molecules do and carry some of the heat away; for one thing you're at the wrong scale of matter. > One the bulb is turned off, > there is an additional heat spike, explaining why bulbs burn out when > you turn them off. no. Once the bulb is turned off, there is rapid cooling and contraction of the filament. This is as bad for the filament as the turn-on inrush but it is in no way a "heat spike". however, if you put a thermometer inside much computer and other electronic equipment you will be able to measure a "heat spike" shortly after power-off... due to the shutdown of the fans. > When I was in computer school, we were told of a SAC tube based > computer behind three foot thick concrete walls where staff were > warned that if the current ever failed, the resulting heat spike > would not only melt hte computer but cause the surrounding concrete > walls to catch fire. Needless to say, their backup power supplies > (multiple redundancy) got everyone's close attention. I don't doubt (much) that you were told this, but the veracity of the story is another matter. Of course, it is not beyond the realm of possibility that this story was in fact told to the computer's maintenance techs to get them to pay close attention to the backup supplies. Doesn't mean that it's true, though. --- Jamie Hanrahan, Simpact Associates, San Diego CA Internet: jeh@dcs.simpact.com, or if that fails, jeh@crash.cts.com Uucp: ...{crash,scubed,decwrl}!simpact!jeh
rando@skipper.dfrf.nasa.gov (Randy Brumbaugh) (03/22/91)
I just wanted to point out that lamps which are operated on a dimmer typically last longer than those operated on full line voltage. And most dimmers operate by "chopping" the AC waveform, which creates all sorts of nasty frequency components. (as anyone who has worked with audio gear near light dimmers knows). The point is that the rapid turning on and off doesn't matter in the life expectancy of the lamp, and can actually extend it. Dimmers rely on this rapid turning on/off and thermal inertia, and can extend bulb life. Another reason lamps on dimmers last longer is the reduction of inrush current when the lamp is faded-on instead of turned-on. It is generally a bad idea to "hot patch" - plug a cold lamp into a dimmer which is set to 100%. The lamp is much more likely to fail, and the dimmer may be damaged by the large current spike. Randy
farren@sat.com (Michael J. Farren) (03/22/91)
mercer@npdiss1.StPaul.NCR.COM writes: >it does heat and cool, which can be seen by high speed >photography. If you listen closely, you can hear the coil hum. This is a magnetic effect, not thermal. Quite similar to the things we saw in very high-power transformers for XRay equipment (1 amp at 100KV = a hell of a lot of watts, albeit not for very long) - in those units, all internal wiring was taut point-to-point. If there was any slack at all, the magnetic fields created by the current would rip the wiring right off of its mooring as it "snapped" into the new position. The same thing happens, but on a much smaller scale, in a light bulb. -- +-----------------------------------------------------------------------+ | Michael J. Farren farren@sat.com | | He's moody, but he's cute. | +-----------------------------------------------------------------------+
johne@hp-vcd.HP.COM (John Eaton) (03/22/91)
<<< < BTW, electric current actually carries away some of the energy that < might ordinarily be translated into heat. One the bulb is turned off, < there is an additional heat spike, explaining why bulbs burn out when < you turn them off. < < When I was in computer school, we were told of a SAC tube based < computer behind three foot thick concrete walls where staff were < warned that if the current ever failed, the resulting heat spike < would not only melt hte computer but cause the surrounding concrete < walls to catch fire. Needless to say, their backup power supplies < (multiple redundancy) got everyone's close attention. ---------- Are you sure they didn't mean if the power and COOLING failed that the residual heat would cause permanant damage? John Eaton !hp-vcd!johne
good@pixar.com (Craig Good: One brick shy of a full load.) (03/22/91)
In article <1991Mar16.185608.11969@agate.berkeley.edu> shirriff@sprite.berkeley.edu (Ken Shirriff) writes:
:Today's "Dear Abby" states that it only saves electricity to turn off a
:light bulb if it's going to be off more than 10 minutes. This is because
:of the power surge when you turn it on.
:
:Is this as bogus as I think it is?
If Dear Abby said it, it's probably wrong. But in this case possibly
not as wrong as you think. In general, starting things up uses a lot
of energy and wears them out faster. I believe that the 10-minute
figure is probably more accurate for flourescent bulbs which have to
run a small heater as part of the turn-on cycle. Incandescent bulbs
will certainly wear out faster if you flick them on and off a lot
during their life.
BTW, I believe that starting a *cold* engine could use enough gas
to idle an engine for 10 minutes, but a warm engine can probably
be restarted with on the vapor that's still floating around in the
manifold. The main cost should be wear and tear on the starter.
--
...{ucbvax,sun}!pixar!good
No legitimate government fears an armed citizenry.good@pixar.com (Marsey doats and doesey doats but little lambsie divey) (03/24/91)
In article <18627@milton.u.washington.edu> iho@akbar.UUCP (Il Oh) writes: :carter@mcs.anl.gov (Richard Carter) writes: : :Given this information, I don't think turning it off and on just one more :time will make that big of a difference. Afterall, the power company is :doing it 120 times every second. The exfoliation and mechanical stress are related to heating up a cold filament. In AC operation I rather suspect the temperature is quite stable. If you let the bulb cool off (ie: stop producing light) and switch it back on, it's more likely to break down the filament. -- ...{ucbvax,sun}!pixar!good No legitimate government fears an armed citizenry.
shirriff@sprite.berkeley.edu (Ken Shirriff) (03/24/91)
I did some more research on turning on lights. From the IES Lighting Handbook: A 100 watt general service incandescent light at 120V has normal current of .835A. It has theoretical inrush current of 17.9A. Current reaches the normal value in 0.10 seconds. (They compute the theoretical inrush current from the hot-to-cold resistance ratio. They say the real inrush current is less due to circuit reactance.) So as an absolute worst case, the inrush uses 17.9Ax.1s/.835A = 2.1 seconds worth of electricity. (Assuming peak inrush current for the entire time until current falls to normal level. This isn't realistic, but I'm looking for an upper bound.) Therefore, the power surge when you turn on an incandescent light bulb uses less than 2 seconds electricity. (Probably a lot less.) So "Dear Abby" is only off by a factor of 300. Ken Shirriff shirriff@sprite.Berkeley.EDU
urbanf@tuura.UUCP (Urban Fredriksson) (03/26/91)
brendan@otc.otca.oz (Brendan Jones) writes: >Along the lines of the current discussion re incandescents, the old story >goes that you shouldn't turn off a fluoro if you are going to turn it on again >in x minutes (usually 5 < x < 30 in folklore). The value of x depends on how much you pay for electricity and how much you pay for flourocent light. Every time you turn it on, it loses some hours of life. If you pay nothing for the flouro, save electricity and turn it off whenever you can. If you pay nothing for the electricity, don't turn it off if you expect to turn it on again in less than 3 hours. (Which is the life you lose when you turn it on, according to the makers.) Here in Stockholm, x = 30 min. | Urban Fredriksson | I speak ONLY |"The best way to get rid of an | | Stockholm, Sweden | for myself! | enemy is to make him a friend."| | I can't read mail 6 Apr - 31 May |