End of life conditions of incandescent light bulbs
Greg
I am interested in some of the conditions that occur within an
incandescent light bulb when it fails when running at its rated
voltage. More specifically, I am looking for the characteristics of
the arc that forms between the ends of the broken filament. My main
interest lies in the peak current of the arc, the peak impedance of
the arc, the voltage across the bulb, and the length of time before
the filament ends melt to the point where the arc cannot be sustained.
I am looking for information on specific light bulbs (mainly 100w GE
and Philips products, all 120v), but I am assuming that the similar
light bulbs I am testing will have similar arc characteristics, so any
general information that you could provide would be great. Thanks for
your time and any help you can provide.
Regards,
Greg Morton
Wallbox Engineer
Lutron Electronics, Inc.
gmo...@lutron.com
Victor Roberts
I don't have your answer and suspect that the people who may have known
at one time are all retired or perhaps worse <g>. In any case, do you
know that one of the leads in the stem of an incandescent lamp is a fuse
wire. It is there to limit the current drawn by the arc, which can have
a very low impedance under the proper conditions. So --- you may be able
to get the information you need by measuring the characteristics of a
bunch of those fuse wires. Of course, an easier way would be to contact
GE and/or Philips Engineering and see if they can and will tell you the
peak current that the lamps are allowed to draw during a filament arc or
perhaps even give you design parameters of the fuse.
If you can't get through to the right person, perhaps Terry McGowan, who
frequents this group, can point you to the right person at Nela Park. he
has much better contacts there than I do.
Vic Roberts
Andrew Gabriel
gmo...@lutron.com (Greg) writes:
> Hi all.
>
> I am interested in some of the conditions that occur within an
> incandescent light bulb when it fails when running at its rated
> voltage. More specifically, I am looking for the characteristics of
> the arc that forms between the ends of the broken filament. My main
> interest lies in the peak current of the arc, the peak impedance of
> the arc, the voltage across the bulb, and the length of time before
> the filament ends melt to the point where the arc cannot be sustained.
The arc has an impedance which drops dramatically as the current
increases - I suspect you can regard it as a short circuit for
the purposes of working out peak current flow (which is then
pretty much just limited by the supply impedance).
The arc only initially forms between the filament ends. The arc
ends very rapidly move back to the lead-in wires as the arc is
much lower impedance than the filament. The filament is not involved
in the high arc current - the arc is directly between the lead-in
wires. The bulb blows up long before the lead-in wires give up,
but normally some overcurrent protective device operates before
that happens, such as the fuse in the lead-in wire in the lamp
base blows (that's why it's there). It is almost unheard of for
GLS (what I believe you call A-line) lamps to blow up in the UK.
I only ever experienced one instance.
> I am looking for information on specific light bulbs (mainly 100w GE
> and Philips products, all 120v), but I am assuming that the similar
> light bulbs I am testing will have similar arc characteristics, so any
> general information that you could provide would be great. Thanks for
> your time and any help you can provide.
I can give you some figures from UK (240V) mains supply lamps.
A 40W 240V lamp which arcs on failure has no difficulty tripping
a 60A magnetic circuit breaker. Often, the breaker trips before
or as well as the fuse in the lamp base, and the breaker's tripping
time is half a mains cycle. That doesn't tell you what the peak
current or duration would be in the absence of the breaker, but it
gives you a minimum data point of at least 60A for 1/100 second.
Peak current will depend largely on supply impedance, which could
easily be under 1 ohm on the UK, so you might well be looking at
half cycle peak currents into the 100's of amps, but I have no
measurements of this.
--
Andrew Gabriel
Consultant Software Engineer
Andrew Gabriel
AC/DCdude17 <Je...@prontoREMOVETHISmail.com> writes:
> X-No-Archive: Yes
>
> Andrew Gabriel wrote:
>
>> I can give you some figures from UK (240V) mains supply lamps.
>> A 40W 240V lamp which arcs on failure has no difficulty tripping
>> a 60A magnetic circuit breaker. Often, the breaker trips before
>> or as well as the fuse in the lamp base, and the breaker's tripping
>> time is half a mains cycle.
>
> breaker pop on me when a lightbulbs failed.
Since two-component breakers became common, yes. Lighting circuits
in the UK are protected with an MCB (minature circuit breaker) rated
at 6A (Type C in my case). This contains two tripping components;
a thermal breaker designed to pass 6A and trip within an hour at
around 8-9A (and faster with higher overloads obviously), and a
magnetic component which trips at about 10 times the MCB rating
within half a mains cycle. Light bulbs blowing quite often trip
the magnetic component of these breakers.
I believe the minature circuit breakers in US panels only contain
the thermal tripping component. That was the case here in the UK
too some ~30 years ago. Introduction of the magnetic tripping
component allowed the current rating of our standard cable sizes
to be uprated (although in the case of standard lighting circuits,
the cable isn't the limiting factor - it's the lampholders which
are).
>> That doesn't tell you what the peak
>> current or duration would be in the absence of the breaker, but it
>> gives you a minimum data point of at least 60A for 1/100 second.
>> Peak current will depend largely on supply impedance, which could
>> easily be under 1 ohm on the UK, so you might well be looking at
>> half cycle peak currents into the 100's of amps, but I have no
>> measurements of this.
Actually, I thought of another data point I have. The 32A MCB's used
on final ring circuits (for socket outlets) never trip when light
bulbs blow in my experience. The magnetic part of a 32A Type B MCB
trips at around 100-150A, so this would tend to point to the arc
current being bounded by the region 60-150A for a 240V GLS lamp
for half a mains cycle.
Greg
Thank you Vic and Andrew for your input. I have already contacted the
US technical lines of both GE and Philips, and neither could provide
the information I was looking for. I have gotten several leads to
more information on this subject (mainly papers that were published in
the 1970’s), but most have turned up only dead ends.
I now feel that it is unlikely I will find exactly what I’m
looking for. So, in the meantime while searching for this
information, I have been designing a test setup to help gather this
information. I am going for a low impedance setup to simulate worse
case scenarios for this data. If I were to use a scope to trigger
recording during a failure, I should be able to get some of the
information I need.
Is this feasible? Any input/advice would be greatly appreciated.
Thanks again.
Greg
Victor Roberts
Yes, it is feasible, but I have two suggestions. To make sure that you
capture the arc event, I would use a digital scope that gathers data
continuously and then stops after the arc even has taken place. That was
the arc event will be stored in memory. This is even easier to set up if
you use a so-called "deep memory" scope.
Your other problem is getting enough arc events to analyze in a
reasonable amount of time. I suggest that you vibrate the lamps when
they are on. Hot tungsten is rather brittle, and as has been mentioned
in another thread in this newsgroup, vibrating incandescent lamps when
they are hot will greatly shorten their life <g>
Vic Roberts
Andrew Gabriel
Victor Roberts <V...@RobertsResearchInc.com> writes:
> Greg wrote:
>>
>> I now feel that it is unlikely I will find exactly what I’m
>> looking for. So, in the meantime while searching for this
>> information, I have been designing a test setup to help gather this
>> information. I am going for a low impedance setup to simulate worse
>> case scenarios for this data. If I were to use a scope to trigger
>> recording during a failure, I should be able to get some of the
>> information I need.
>>
>> Is this feasible? Any input/advice would be greatly appreciated.
>> Thanks again.
>
> Yes, it is feasible, but I have two suggestions. To make sure that you
> capture the arc event, I would use a digital scope that gathers data
> continuously and then stops after the arc even has taken place. That was
> the arc event will be stored in memory. This is even easier to set up if
> you use a so-called "deep memory" scope.
>
> Your other problem is getting enough arc events to analyze in a
> reasonable amount of time. I suggest that you vibrate the lamps when
> they are on. Hot tungsten is rather brittle, and as has been mentioned
> in another thread in this newsgroup, vibrating incandescent lamps when
> they are hot will greatly shorten their life <g>
If you are going to the effort of setting this up, it would also
be really interesting (at least to me;-) to get some high-speed
film of the event, if possible. I searched around some time back,
but never found any such images.
Victor Roberts
> If you are going to the effort of setting this up, it would also
> be really interesting (at least to me;-) to get some high-speed
> film of the event, if possible. I searched around some time back,
> but never found any such images.
>
Before the advent of digital video it would have been almost impossible
to "film" this event unless you could initiate the filament failure, and
then you would not be sure that the filament failure was the same as a
naturally occurring one. It should be possible now with a video camera
using digital memory that could run continuously, reusing the same video
memory, and then stop acquiring images a short time after the filament
has failed.
Back in the late 1960's or early 1970's a scientist at the GE Research
and Development Center took high speed movies of the starting process in
HID lamps. This took a special camera and a LOT of film, but at least he
could synchronize the camera to the event he was looking for - more or
less <g> They were really interesting films.
Vic Roberts
William H. Hathaway
I may have mentioned this before, but my partner and I have been wanting
to get high speed movies of electrostatic discharge formation (caused
by dielectric breakdown) in clear PMMA. The problem of synchronization
is only one difficulty. The possible difference between a spontaneous
discharge and an induced one is another. The effect of the high radiation
environment (several kilo-Gray of beta) on the (expensive) recording equipment
is another. Coordinating the various equipment is another (harder to get
access to an e-beam post-anthrax fer example). And the time resolution
(few nano-seconds per frame) is one more. I have one analog video of
one spontaneous discharge. The secondary sparks lasted several minutes,
but our time resolution is no-no-nano. It does make nice light emission
though. A spectral resolution would be the next challenge. Paying
for it all is the kicker. No one else on earth cares.
Any references to the GE movies and/or researcher? Be nice to see some.
Wm. Hathaway
Victor Roberts
> Any references to the GE movies and/or researcher? Be nice to see some.
I do not believe that the movies were ever published. If
they were, the author would have been Harald Witting and
they probably would have been part of a talk at one of the
annual Gaseous Electronic Conferences in the 1960's or
1970's.
--
Vic Roberts
Terry McGowan
with a laser. I don't know the tye of laser or how it was done just that a
lamp engineer told me once that the technique was used to study failures and
the resulting arcs.
Terry McGowan
Victor Roberts <V...@RobertsResearchInc.com> wrote in message
news:3D2B6A8A...@RobertsResearchInc.com...
Victor Roberts
>
> Another technique used to force the failure of the filament was to zap it
> with a laser. I don't know the tye of laser or how it was done just that a
> lamp engineer told me once that the technique was used to study failures and
> the resulting arcs.
It seems this would work by over heating a small section of
the filament to cause rapid evaporation at that point.
--
Vic Roberts
Greg
I've been "aging" some bulbs now and I think I am at the point where I
should start getting some failures. I just need to figure out how
exactly to gather the data. I have several LeCroy scopes that have
been used for other projects, but I've been told that they may not be
fast enough or able to record enough of the data to be helpful to me.
I was looking at some very inexpensive options, but they are not
nearly fast enough to catch the failures (1 sample/sec...). The
Yokogawa PZ4000 has been recommended to me, and it looks as if it will
work perfectly, but it has a very hefty pricetag. Can anyone
recommend a power analyzer or other data acquisition system that I
could use to gather and record information that isn't too expensive?
Thanks!
Regards,
Greg
Victor Roberts
> I've been "aging" some bulbs now and I think I am at the point where I
> should start getting some failures. I just need to figure out how
> exactly to gather the data. I have several LeCroy scopes that have
> been used for other projects, but I've been told that they may not be
> fast enough or able to record enough of the data to be helpful to me.
> I was looking at some very inexpensive options, but they are not
> nearly fast enough to catch the failures (1 sample/sec...). The
> Yokogawa PZ4000 has been recommended to me, and it looks as if it will
> work perfectly, but it has a very hefty pricetag. Can anyone
> recommend a power analyzer or other data acquisition system that I
> could use to gather and record information that isn't too expensive?
> Thanks!
>
> Regards,
> Greg
What s your price range? "Too expensive" for a company is
far different than "too expensive" for an individual. There
may be some good digital scopes in the $3K to $7K range. Is
this too expensive?
--
Vic Roberts
Greg
> far different than "too expensive" for an individual. There
> may be some good digital scopes in the $3K to $7K range. Is
> this too expensive?
Maybe I shouldn't have thrown in that "too expensive" part. Right
now, I am mainly looking for options (companies, models, anything). I
don't want to go overboard and spend a large sum of money on something
that isn't completely necessary. If $10k+ (such as the Yokogawa
PZ4000) needs to be spent, then so be it. However, if there is
something out that that will do what I need it to do for less, I would
like to know about it.
Thanks for all of your help!
Greg
Terry McGowan
watt A-line as the lamp of interest.
That's certainly a popular lamp; but it may not be the type that is the most
likely to arc when it fails. Also, if it does arc, it may not generate the
most damaging arc from the dimmer standpoint.
Back when electronic dimmers were first used, for example, we found that
30R20 lamps, probably because of their filament support construction, caused
dimmer failures a high percentage of time when the lamp failed while a-line
lamps did not. Perhaps that's no longer true; but it's likely that lamps
vary in their ability to take out a dimmer when they fail and the
probability doesn't necessarily correlate with lamp wattage -- if that's of
importance to your study.
Terry McGowan
Greg <gmo...@lutron.com> wrote in message
news:c22a900f.02071...@posting.google.com...
Victor Roberts
Sorry - I don't have time to check the Yokogawa site this
morning, but I assume you are referring to a power meter. I
was suggesting a digital oscilloscope with a large amount of
memory. Those are made by Agilent, Tektronix and a few
others.
--
Vic Roberts
Greg
I have already done some worst case testing (Philips Tubular bulbs
turned out to be the worst bulb that I tested if anyone is
interested...) and now I am focusing on more popular bulbs. Also, the
testing will involve other lamps as well (recessed floods, vanity,
etc), but the majority of them are 100 watt A lamps.
I am actually at the point where the bulbs that I've been aging have
begun to fail. I don't want to interrupt the testing, but I may have
to until I get the measuring equipment in place.
I am looking into the Agilent and Tektronix scopes recommended below,
but I haven't decided on anything yet.
Any input would be greatly appreciated!
Regards,
Greg
"Terry McGowan" <ligh...@ieee.org> wrote in message news:<_FMZ8.110294$UT.70...@bgtnsc05-news.ops.worldnet.att.net>...
Terry McGowan
solution to the problem (complaints from lamp-buying customers that dimmers
burned out when the lamp burned out). We found that a 1-ohm impedance - or
even a bit less - held the current in check long enough for the arc to burn
itself out before the dimmer electronics did.
The high-volume standard household lamps 40, 60 and 100 watts generated the
fewest complaints. Reflector lamps the most.
Terry McGowan
Greg <gmo...@lutron.com> wrote in message
news:c22a900f.02072...@posting.google.com...