Re: Does an automatic shutoff energy saver light switch exist?

[don@xxxxxxxxxxxxxx (Don Klipstein)]

In article <RRtth.14592$yx6.3244@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx>,
VWWall wrote:
>Steve Lockridge wrote:
>>
>> I have a question about turning lights on and off.  I have been told
>> that leaving a light on maximizes its life and turning it on and off
>> frequently will shorten its life.  Is this true?  My background is in
>> transformers and it is better to leave them running than to shut them
>> off.  Of course, we're talking about several thousand volts versus a
>> 120V light bulb.  Thanks.
>
>A tungsten filament lamp almost always expires the last time you turn it
>on. (Sort of like why you always find something in the last place you
>look.)  The current surge in a cold filament and rapid expansion causes
>the failure.  The total time before this last gasp is probably about the
>same even if the lamp is cycled on and off.
>
>So if it were left burning, it's total hours would be longer than if you
>had not turned it off, and then on, the last time.

  Strange thing few people seem to realize about most incandescents - a
cold start does not cause much actual wear, but kills a filament that
would have survived a little longer if it was not turned off.

  The usual cold start burnout is from a temperature overshoot in a thin
spot in the filament.  The thin spot is a hot spot, and on a cold start
the thin spot can overshoot in temperature before the remainder of the
filament has warmed up.  Resistance of the filament material varies
directly with temperature, so a thin spot overshoots in temperature
from the current surge before the remainder of the filament has full
resistance and the current drops to normal.

  However, a thin spot is is hotter than the remainder of the filament
even during steady operation (but to a lesser extent than during a cold
start).  The hotspot will evaporate faster and get thinner at a faster
rate, making it hotter and thinner.  This downward spiral accelerates in a
manner worse than exponentially.  By the time the thin spot cannot survive
a cold start, the filament's hours are numbered!

>Fluorescent lamps rely on a starter mechanism which degrades with use.
>They may benefit from fewer on-off cycles.

  The bulbs also suffer actual wear from starting.  How much life is lost
varies with the type of ballast and starting means and this has been
improved over the decades.  I am under the impression that a typical
figure is 10 minutes of life lost per start.
  As for how much off time is needed before the electricity saved from
turning off a fluorescent exceeds the fraction of the bulb's cost
associated with life reduction of a start - this varies with electricity
cost, bulb (lamp, tube, whatever) cost, and ballast and starting method,
but is typically a couple minutes for 4-footers, ballpark of 10-30 minutes
for lower wattages 9-22 watts, and ballpark of 30-60 minutes for wattages
8 watts or less.

>As an aside, in the UK and other 240V lighting countries, an expiring
>lamp will often cause a large enough "plasma arc" to take out the
>circuit protection.  Hence the lamps commonly include an "internal
>fuse".  This not needed at 120V.

  My experience in the USA is that 120V gas-filled incandescent lamps often
have an internal fuse.  In the USA that bright blue flash from a cold
start burnout is this "plasma arc".  At 120V that arc tends to die at the
next end of a half cycle and some incandescents lacked this fuse and
usually got away with it, but sometimes things got nastier and one thing
that sometimes happened (I would guess mainly when this fuse was lacking)
was that wires between the glass bulb and the base would either fuse
explosively or vaporize adjacent glue - and the glass bulb pops from the
base!
  These "burnout arcs" sometimes also blow dimmers with the large amount
of current that they sometimes draw even at 120V.

 - Don Klipstein (don@xxxxxxxxx)