Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!usc!apple!agate!pasteur!libra.Berkeley.EDU!e221a-bx
From: e221a-bx@libra.Berkeley.EDU (Charlie Sullivan)
Newsgroups: sci.electronics
Subject: Re: incandescent light bulb life extender
Summary: diodes and PTCs are used in different versions.  Both are stupid.
Message-ID: <9709@pasteur.Berkeley.EDU>
Date: 13 Dec 90 02:48:24 GMT
References: <5232@rsiatl.Dixie.Com> <4225@kitty.UUCP> <1990Dec7.142843.1870@rodan.acs.syr.edu> <1990Dec11.163848.21840@isis.cs.du.edu>
Sender: news@pasteur.Berkeley.EDU
Reply-To: charless@cory.Berkeley.EDU
Lines: 37

In article <1990Dec11.163848.21840@isis.cs.du.edu> whester@isis.UUCP (William R. Hester) writes:

>I'm not sure, however, how reduced voltage will affect the operation of
>halogen cycle types...I guess it stops the tungsten replacement cycle and
>actually reduces bulb life...???

At significantly reduced voltage, the halogen cycle stops working, and the
bulb wall starts to blacken, due to evaporated tungsten.  This could lead to
shortened life than at nominal voltage.  But it would take a very low voltage,
like from having the lamp on a dimmer.  If it starts to happen, you can
crank the voltage back up and fix the bulb.

About bulb life extenders and the argument about whether they are PTCs or
diodes:  Surprise! You're both right.  Both types exist.

In addition to the fire hazards that have been mentioned, I have heard
from a researcher at one of the major lamp companies that stress at
turn-on is not significant in decreasing lamp life.  True, lamps usually
die at turn-on, but only when they are within eight hours of failure
anyway.  So a perfect soft start would only gain you about 4hrs. average
life.  The way these things actually increase lamp life is by dropping
some voltage in normal operation.  As others have pointed out, life is
very sensitive to operating voltage.

A diode is a more efficient (and safer) way to drop the voltage than a resistor.
Even so, the efficiency of the lamp goes down at reduced voltage, and
a diode cuts the voltage quite a bit.  

A dimmer is similar to a diode in that (unlike a resistor) it doesn't dissipate
much power.  It has the advantage that the voltage reduction is continuously
adjustable.

Instead of fussing over the optimum in the life/efficiency tradeoff in 
incandescants, why not have have it both ways?  Buy compact fluorescents.

Charlie Sullivan                                   charless@cory.Berkeley.EDU
EE grad student--U.C. Berkeley                     usual disclaimers