Strange CFL Failure Mode

[Samuel M. Goldwasser]

See: http://repairfaq.cis.upenn.edu/Misc/cflhole1.jpg

This is a ~1 mm hole in the glass near one of the filaments.
Something got hot enough for the glass to melt, and after
that, as they say, the rest was history.

I've seen this on 3 CFLs in 3 different lamps/fixtures. There are no known
problems that could account for such nasty behavior. They were all high
mileage, so perhaps the filament at that end of the lamp opened resulting
in the discharge going to one post, near the glass, or something.

The CFLs were all from GE but I don't know if they are of the same
ballast/lamp design.

Comments welcome.

--
sam | Sci.Electronics.Repair FAQ: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Sites: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: Anything sent to the email address in the message header above is
ignored unless my full name AND either lasers or electronics is included in the
subject line. Or, you can contact me via the Feedback Form in the FAQs.

 

[TKM]

See: http://repairfaq.cis.upenn.edu/Misc/cflhole1.jpg

This is a ~1 mm hole in the glass near one of the filaments.
Something got hot enough for the glass to melt, and after
that, as they say, the rest was history.

I've seen this on 3 CFLs in 3 different lamps/fixtures. There are no
known
problems that could account for such nasty behavior. They were all high
mileage, so perhaps the filament at that end of the lamp opened resulting
in the discharge going to one post, near the glass, or something.

The CFLs were all from GE but I don't know if they are of the same
ballast/lamp design.

Comments welcome.

--
sam | Sci.Electronics.Repair FAQ: http://www.repairfaq.org/
Repair | Main Table of Contents: http://www.repairfaq.org/REPAIR/
+Lasers | Sam's Laser FAQ: http://www.repairfaq.org/sam/lasersam.htm
| Mirror Sites: http://www.repairfaq.org/REPAIR/F_mirror.html

Important: Anything sent to the email address in the message header above
is
ignored unless my full name AND either lasers or electronics is included
in the
subject line. Or, you can contact me via the Feedback Form in the FAQs.

Sounds like a hot blob of tungsten or emission material was ejected from the
lamp's cathode. An arc could certainly do that. It's common in
incandescent filament lamps that arc upon failure; but I wouldn't have
thought that there's enough energy available in a low-wattage CFL to do it.
What were the wattages of the lamps?

But, there were some arcing problems with T5 and some T8 lamps a few years
ago when operated on instant-start ballasts. At end-of-life, the cathodes
failed and an arc started chewing away at the cathode mount wires. Things
got hot, lamps cracked and lamp holders were charred. This all resulted in
the development of ballasts that shut the lamp off if an arc was sensed so
as to avoid the mess of cleaning up broken lamps and the concern that
sputtering lamps, smoke and the smell of overheated plastic caused.

Terry McGowan

 

[Andrew Gabriel]

See: http://repairfaq.cis.upenn.edu/Misc/cflhole1.jpg

This is a ~1 mm hole in the glass near one of the filaments.
Something got hot enough for the glass to melt, and after
that, as they say, the rest was history.

I've seen this on 3 CFLs in 3 different lamps/fixtures. There are no known
problems that could account for such nasty behavior. They were all high
mileage, so perhaps the filament at that end of the lamp opened resulting
in the discharge going to one post, near the glass, or something.

The CFLs were all from GE but I don't know if they are of the same
ballast/lamp design.

Comments welcome.

Caused by control gear which fails to detect when the tube has reached
end of life (emission material all sputtered off), and and provides enough
voltage headroom to continue driving the tube as a cold cathode tube,
which generates too much heat at the tube electrodes. Dead thermionic
tubes don't last long when driven as cold cathode tubes, because the
filaments and support wires are quickly burned away, until the tube cracks
and vents to the atmosphere. If the support wires are very close to the
glass tube wall, as is likely with thin tubes, it can melt the glass.
With linear tubes, it can cause the tube to break such that it drops out
of the fitting - I've had one case of this. This was in a set of T4 tubes
which come in various lengths from 6W to 24W, and all use the same control
gear. I had mostly 16W ones, and when the tube reaches end of life, the
control gear simply up's the tube voltage to maintain the current, and the
excess power is dumped into the (now) cold cathode electrodes, which start
glowing as bright red hot dots, clearly visible through the dark sputtered
coating which is now round the tube ends. The heat also did enough damage
to the plastic lampholder that it was no longer usable. I did think this
could be a fire risk too, particularly if there had been any flammable
material nearby. (I've since been phasing these lamps out of use, which
is now forced as the manufacturer has gone and no spare tubes available.)

 

[GregS]

Yep, soon they'll be considered as dangerous as incandescent lamps
that burn fingers, start fires, explode, send molten glass and
tungsten shooting about and electrocute those who try to change the
bulb. ;-)

Terry

My my, Terry. You do live an interesting and dangerous life ... During my
many years of shuffling around on this little orb of ours, I don't think I
have ever had a domestic incandescent light bulb start a fire, explode,
reach a temperature where the glss could become molten, shoot molten
tungsten at me, or try to electrocute me (how does that one work then ??). I
would have to admit to having scorched (rather than burnt) my fingers a time
or two, but that was down to me employing stupid tactics for handling them
.... d

Arfa

No fires here yet.

I just had one infantile failure out of a batch of 6 I bought.
9 watts at 40 watts effective halogen brightness. I don't think being outside
was a problem even thought its not for outside. These are flood
or spots. What I have found, I can get moe light out of a 1.5
watt LED spot vs a 9-12 watt CFL. The reason being, CFL's are
poor choice for spots due to the poor beam width.

A lot of CFL's say, no dimmers or relays. The same for LED's most
of the time. I can't really figure the relay bit.

greg

 

[William R. Walsh]

Hi!

This is a ~1 mm hole in the glass near one of the filaments.
Something got hot enough for the glass to melt, and after
that, as they say, the rest was history.

Comments welcome.

I've seen the effect on some very narrow GE tubes in use with conventional
overhead fixtures. When the tubes die, they oftentimes crack a ring in the
glass or blow a slightly larger hole in it at either edge. This makes
changing them "fun".

I'm not sure if it is specific to the tubes or not. The GE tubes have not
proven long lived. I started dating them after some tubes from a different
maker with the wrong color temperature were purchased and installed during
late 2005. All of the misordered tubes are still going strong, nine hours a
day in most locations. The oldest GE tubes date from 2007, and there are
very few left. All fixtures are identical.

William

 

[bud--]

No fires here yet.

I just had one infantile failure out of a batch of 6 I bought.
9 watts at 40 watts effective halogen brightness. I don't think being outside
was a problem even thought its not for outside. These are flood
or spots. What I have found, I can get moe light out of a 1.5
watt LED spot vs a 9-12 watt CFL. The reason being, CFL's are
poor choice for spots due to the poor beam width.

A lot of CFL's say, no dimmers or relays. The same for LED's most
of the time. I can't really figure the relay bit.

Likely "solid state" relays. Some X-10 devices, for example, are solid
state and may not work. A mechanical relay would work fine.

 

[TKM]

Yep, soon they'll be considered as dangerous as incandescent lamps
that burn fingers, start fires, explode, send molten glass and
tungsten shooting about and electrocute those who try to change the
bulb. ;-)

Terry

My my, Terry. You do live an interesting and dangerous life ... During my
many years of shuffling around on this little orb of ours, I don't think I
have ever had a domestic incandescent light bulb start a fire, explode,
reach a temperature where the glss could become molten, shoot molten
tungsten at me, or try to electrocute me (how does that one work then ??).
I would have to admit to having scorched (rather than burnt) my fingers a
time or two, but that was down to me employing stupid tactics for handling
them ... d

Arfa

All that stuff has happened usually because (full disclosure) I treated the
lamp rather badly. A drop or two of water on a hot 100 watt GLS lamp, for
example, is sure to cause damage and, probably, fireworks.

I was electrically shocked when I tried to put a 60-watt lamp into a
portable lamp socket with the socket turned on and didn't realize it was an
old portable which didn't have a polarized plug. It's easy to touch the
threaded part of the base when you reach under the shade with the lamp and
try to position it in the socket.

Terry McGowan

 

[Don Klipstein]

The (foot high internally assembled with nuts and bolts) tubes on these
7 foot tall amplifier cabinets, often used to run so hot that the glass was
actually glowing dull red. I have seen dents in the glass where the vacuum
has started sucking on areas that have got *so* hot that they have softened.
The HT supply on these beasts was about 2.5Kv as I recall, derived from
mercury vapour rectifier tubes.

Given these facts, if the glass had much conductivity when hot, you would
pretty soon be into having leakage paths between the pins, and to the
grounded retainer metalwork, that holds the devices into their sockets,
wouldn't you ?

The glass around the pins is not as hot as the glass on the sides and
the top.

Meanwhile, even soda lime glass is not that much of a conductor at
200-300 degrees C.

I have heard of HID lamps requiring any metal supports to be isolated,
to avoid problems with electrolysis of the glass bulb. I imagine this
problem involves amounts of current low enough to not be a shock hazard.

On the other hand, I have seen glass heated to the point of being soft
conduct impressive amounts of current (several milliamps, possibly more)
at a few kilovolts. I do not know what kind of glass was involved -
either soda lime or a cheaper flint glass, probably soda lime.

- Don Klipstein ([email protected])

 

[Andrew Gabriel]

Ordinary incandescent lamps are made from soda lime glass,
which has enough conductivity at 100C to maintain a low
frequency (DC or 60Hz) discharge in T12 fluorescent lamps
using electrodes on the outside of the bulb. The place
where the wires are sealed in an incandescent lamp (the
pinch seal) is made from lead glass, that has much lower
conductivity than soda lime glass. Vacuum tubes are usually
made from borosilicate glass (Pyrex(R)) that has even lower
conductivity than lead glass, and much lower than soda lime
glass.

Thanks, interesting.

However note that Pyrex(R) no longer implies any particular
glass type, and in particular, it's no longer borosilicate
glass in the US.

 

[TKM]

There's an interesting demonstration that can be done with a
rod of soda lime glass. If it is connected to a source with
the appropriate voltage output and current capability, and
then heated, it will conduct enough current to sustain the
temperature required for conduction when the external
heating supply is removed. I'm sorry to say I don't know
how much voltage is required to do this on any particular
size rod of soda lime glass.


--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

This information is provided for educational purposes only.
It may not be used in any publication or posted on any Web
site without written permission.

Fascinating. I wonder if there are glass furnaces that switch from gas to
electrical glass-conductive heating once the glass is molten.

Terry McGowan

Terry McGowan

 

[Don Klipstein]

There's an interesting demonstration that can be done with a
rod of soda lime glass. If it is connected to a source with
the appropriate voltage output and current capability, and
then heated, it will conduct enough current to sustain the
temperature required for conduction when the external
heating supply is removed. I'm sorry to say I don't know
how much voltage is required to do this on any particular
size rod of soda lime glass.

I do remember seeing somewhere on the Web how to melt a soda bottle (or
was it a beer bottle?) in a microwave oven. The glass bottle to be melted
required a hot spot pre-heated with a "blowtorch" / "propane torch" or the
like, to temperature that I visually estimate to be in the 800's degrees C.
It appears to me that soda lime glass that hot achieves conductivity and
resistivity suitable for being sufficiently receptive for heating by a
microwave oven.

=========================

Disclaimer: I sense that this is adventurous usage of a microwave oven,
attempt to do such only at your own risk of damaging your microwave oven or
burning down the real estate that such microwave oven is in/on and
injuring/killing life forms in/on such real estate including loved-ones,
friends and pets by starting a fire.

- Don Klipstein ([email protected])

 

[Don Klipstein]

Pyrex seems to still be a registered trademark of Corning,
at least in the US, though I do see that the use of the
Pyrex name has expanded well beyond borosilicate glass, for
example, to metal cookware.

This makes me wonder if the name "Pyrex" "expanded" to glass types other
than borosilicate, as opposed to merely being "expanded" to non-glass
items.

- Don Klipstein ([email protected])

 

[Andrew Gabriel]

This makes me wonder if the name "Pyrex" "expanded" to glass types other
than borosilicate, as opposed to merely being "expanded" to non-glass
items.

In the US, Pyrex is currently tempered sola lime glass:
http://www.pyrexware.com/thetruthaboutpyrex/manu.htm
A factory in Europe also licenses the name, and apparently
produces borosilicate glass.

Pyrex is now simply used as a respected brand name, still
owned by Corning, but licensed out to whoever wants to use the
name. It doesn't imply glass, or any particular type of glass.
Corning no longer make anything under the Pyrex name themselves.

 

[Franc Zabkar]

During my
many years of shuffling around on this little orb of ours, I don't think I
have ever had a domestic incandescent light bulb start a fire, explode,

I had a bad batch of incandescent light bulbs where the glass
separated from the base and fell, or was ejected, onto the floor.

- Franc Zabkar

 

[Jeff Jonas]

Yep, soon they'll be considered as dangerous as incandescent lamps

that burn fingers, start fires, explode, send molten glass and
tungsten shooting about and electrocute those who try to change the bulb.

Please don't say that too loud, lest the electrician's union
place a seal and/or lock on every fixture
to assure only a licensed union electrician ever changes a lamp.
Don't laugh. With all the microcontrollers embedded into everything,
it's possible to add lockout codes so only authorized people may restart it,
just like you car's "service required" light and reset code.

It's bad enough my TV, monitor, radio and such all say
"No user servicable parts inside".
Pray, don't let that happen to my home too.

My my, Terry. You do live an interesting and dangerous life ... During my
many years of shuffling around on this little orb of ours, I don't think I
have ever had a domestic incandescent light bulb start a fire, explode ...

I had a night light bulb shatter, fresh from the package.
I'd say that counts as an explosion.
I swore off dollar store bulbs ever since.

As to fires, I refuse to use lamps over 100 watts at home,
thus avoiding the 300-500 watt halogen torchiere fires
that were terribly common for a while.
The closest I came to a lamp induced fire was a plastic figurine
melting and smoking from being too close to a reflector halogen lamp.