Maker Pro
Maker Pro

Why to multiple bulbs in series ALL die?

L

lens

Jan 1, 1970
0
I've seen this phenomenon 1st hand, so this is not an urban legend. In
one case I have a clock with a face that is lit from behind by several
series-connected bulbs that are powered from 110VAC. Most times one
bulb dies and replacing that bulb is all I need to do. But maybe twice
in 8 years or so, the light goes out and I find ALL the bulbs blown!
Another similar case is a dimmer for a floor lamp that has a fuse.
Just yesterday I find the bulb AND the fuse blown. What is going on
here? You would think that when one part of the series circuit dies,
the rest would be fine, but something else if happening.
 
E

Eeyore

Jan 1, 1970
0
lens said:
I've seen this phenomenon 1st hand, so this is not an urban legend. In
one case I have a clock with a face that is lit from behind by several
series-connected bulbs that are powered from 110VAC. Most times one
bulb dies and replacing that bulb is all I need to do. But maybe twice
in 8 years or so, the light goes out and I find ALL the bulbs blown!
Another similar case is a dimmer for a floor lamp that has a fuse.
Just yesterday I find the bulb AND the fuse blown. What is going on
here? You would think that when one part of the series circuit dies,
the rest would be fine, but something else if happening.

You've got poltergeists. A Priest may help.

Graham
 
L

legg

Jan 1, 1970
0
I've seen this phenomenon 1st hand, so this is not an urban legend. In
one case I have a clock with a face that is lit from behind by several
series-connected bulbs that are powered from 110VAC. Most times one
bulb dies and replacing that bulb is all I need to do. But maybe twice
in 8 years or so, the light goes out and I find ALL the bulbs blown!
Another similar case is a dimmer for a floor lamp that has a fuse.
Just yesterday I find the bulb AND the fuse blown. What is going on
here? You would think that when one part of the series circuit dies,
the rest would be fine, but something else if happening.

It's called a cascading failure mode, when the failure of one section
inreases the stress on on others, whether they are shorts occuring in
voltage-source-fed series loads or open circuits appearing in
current-fed parallel loads. The same failure mode can occur in series
and parallel-connected sources, under the influence of constant-power
loading.

If the single lamp failure produced an open circuit in the voltage-fed
series string, others would normally be protected and there would be
no cascading effect, but the single point of failure would be
difficult to locate immediately.

There is a mechanism involving low-breakdown-voltage oxide contacts
which, when located in parallel with low voltage lamps, will tend to
'fuse' as a ~short, intentionally, if a large voltage occurs across a
failed-open lamp filament. In a series string, with the short in
effect, only the failed lamp would fail to illuminate, but the
remaining lamp stresses increase proportionally while the condition
persists. If the increased stress causes other failures, a cascade
occurs until one mechanism fails to short circuit as intended or more
points finally fuse open circuit due to abnormal current levels.

The resulting dark string is not necessarily composed entirely of dead
lamps, due to the imprecise and untested function and limitations of
the bypass mechanism, though a high percentage of failures (including
100%) is possible.

When the mechanism fails to break down as intended, one fast method of
restoring continuity for trouble-shooting is to stress the string with
a higher voltage source; one that is internally limited to low power
or current (something like a hipot tester, if you've got one lying
around). This may produce the intended conduction in the open circuit
location, allowing it's identification when diagnostic (low) voltages
are slowly re-applied to the string, to illuminate surviving lamps.

A common repair procedure, dictated by economic reality, presently
involves the scrapping and replacement of the entire product. The
replacement may use entirely different materials and methods for
illuminaton.

RL
 
Another similar case is a dimmer for a floor lamp that has a fuse.

If feasible, try the more expensive dimmable fluorescent variant if
you were using an incandescent bulb. Also, avoid dimmers that allow on/
off at any brightness level.
 
L

lens

Jan 1, 1970
0
I'm not asking how to fix it. I've already done that in the clock by
rewiring all bulbs in parallel to a 12V xformer. I only made the post
to try and understand what causes multiple series-connected bulbs to
fail at once. These were normal low watt (maybe 2W) 12V tungsten (not
halogen) bulbs wired in series and connected to the AC line. Thanks
for the replies.
 
Yes. I haven't messed with that. Though so far the fuse has blown
twice when the bulb blew....or the bulb died when the fuse blew.
 
A

Adrian Jansen

Jan 1, 1970
0
lens said:
I've seen this phenomenon 1st hand, so this is not an urban legend. In
one case I have a clock with a face that is lit from behind by several
series-connected bulbs that are powered from 110VAC. Most times one
bulb dies and replacing that bulb is all I need to do. But maybe twice
in 8 years or so, the light goes out and I find ALL the bulbs blown!
Another similar case is a dimmer for a floor lamp that has a fuse.
Just yesterday I find the bulb AND the fuse blown. What is going on
here? You would think that when one part of the series circuit dies,
the rest would be fine, but something else if happening.

I have seen failure modes on incandescent lamps where the filament
opens, and an arc forms, the arc then migrates back down to the support
leads, and sustains there until the support wire vaporises. If that
happens in a series string, I could guess that the increased voltage
across the rest of the string causes a similar failure on each globe in
succession. 12 volt globes certainly have short spacings across the
support leads. They were not designed to withstand mains voltage.

--
Regards,

Adrian Jansen adrianjansen at internode dot on dot net
Design Engineer J & K Micro Systems
Microcomputer solutions for industrial control
Note reply address is invalid, convert address above to machine form.
 
Top