CFL in ceiling fixture

[Norris Smith]

Am I likely to have heat problems in using 20W or 26W fluorescent bulbs
in a ceiling fixture? The fixture is surface mounted 12" dia and 5"
deep. The fixture uses two 75W incandescent bulbs and is made of metal
and glass.

Thanks, Norris

 

[Andrew Gabriel]

Am I likely to have heat problems in using 20W or 26W fluorescent bulbs
in a ceiling fixture? The fixture is surface mounted 12" dia and 5"
deep. The fixture uses two 75W incandescent bulbs and is made of metal
and glass.

I've put a 21W 2D fluorescent with separate electronic control
gear inside a couple of fixtures that size. They are running on
the max temperature for the control gear, so that would seem
to be the limit. My fixtures, although designed for ceiling
use, are actually on a wall though, which probably improves
cooling. So I guess you might get away with one 26W CFL,
but not two, and don't expect full quoted life at such elevated
temperatures -- you may well get an early control gear failure.

 

[Victor Roberts]

Am I likely to have heat problems in using 20W or 26W fluorescent bulbs
in a ceiling fixture? The fixture is surface mounted 12" dia and 5"
deep. The fixture uses two 75W incandescent bulbs and is made of metal
and glass.

Probably, especially if you use two lamps and if the fixture is
closed. But even with an open recessed fixture you would probably have
problems.

CFLs are life tested in free air at 25C. The general rule of thumb for
electronics is that life is cut in half for each 10C rise in
temperature. So, if you measure the ambient air temperature in your
fixture with the lamps installed you can get an idea of how long they
might last.

On the other hand, I use a 15-watt Philips SLS lamp a desk lamp where
the shade prevents air flow over the ballast compartment and I'm sure
the ambient is over 25C and the SLS has lasted for at least two years
of heavy use.

--
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.

 

[Jeff Waymouth]

Just a quick clarification (not a disagreement!)

life is cut in half for every 10C rise OVER NOMINAL in temperature.

and that life is actualy ballast life

Jeff Waymouth


{unless, of course, you want to start predicting life extensions over
rating as the temperature runs cooler than nominal, too!)

 

[Don Klipstein]

Probably, especially if you use two lamps and if the fixture is
closed. But even with an open recessed fixture you would probably have
problems.

CFLs are life tested in free air at 25C. The general rule of thumb for
electronics is that life is cut in half for each 10C rise in
temperature. So, if you measure the ambient air temperature in your
fixture with the lamps installed you can get an idea of how long they
might last.

True, but I would not go to the extent of expecting only half the rated
life at 35 C, 1/4 the rated life at 45 C, etc. I am under the impression
that most failures at 25 C should be from the lamp wearing out rather than
electronics. I expect life to decrease with increasing temperature to an
extent less than halving per 10 degrees C until the temperature is high
enough for failures to be mostly from the electronics.
But once the temperature gets high enough for most failures to be from
the electronics, I consider it reasonable for the life expectancy to half
per 10 deggres C beyond that point.

Meanwhile, don't underestimate how hot things get in a ceiling fixture,
or even in any fixture enclosed at the top and sides, even if open at the
bottom.

On the other hand, I use a 15-watt Philips SLS lamp a desk lamp where
the shade prevents air flow over the ballast compartment and I'm sure
the ambient is over 25C and the SLS has lasted for at least two years
of heavy use.

I remember looking at the package of one of these, and it said that it
was rated for use in recessed ceiling fixtures. Most CFLs don't say this.
Last time I checked, Philips SLS being rated for use in recessed ceiling
fixtures was only for models up to 20 watts.

ALSO - CFLs produce more non-radiant heat than incandescents of the same
wattage, although generally less than incandescents of the same light
output. This is because incandescents produce plenty of IR - which
escapes the fixture, although mostly becoming heat somewhere in the room.
I recently got a "Raytek" remote thermometer that works by sensing
thermal infrared, and found an 8 inch globe in mid-air to reach at its top
(in ambient varying from 21 to 23 C):

Incandescents:
41 C with a 40 watt T10 (vacuum)
59 C with a 40 watt A19 (gas filled)
69 C with a 60 watt A19 (gas filled)
82 C with a 100 watt A19 (gas filled)

CFLs:
50 C with a 20 watt spiral
57 C with a 25 watt Philips SLS
70 C with a 23 watt Sylvania Dulux EL (probably from having its top closer
to the top of the globe)
70 C with a 42 watt spiral, (presumably with the sides of the globe hotter
than with the 23 watt Sylvania Dulux EL)

- Don Klipstein ([email protected])

 

[Victor Roberts]

Just a quick clarification (not a disagreement!)

life is cut in half for every 10C rise OVER NOMINAL in temperature.

and that life is actualy ballast life

Jeff Waymouth


{unless, of course, you want to start predicting life extensions over
rating as the temperature runs cooler than nominal, too!)

Thanks Jeff, that is correct. And yes, I do expect that the life of
the ballast would increase as the temperature is dropped. However, the
life of the "wire lamp" portion of the system would stay constant.

Vic

 

[Victor Roberts]

True, but I would not go to the extent of expecting only half the rated
life at 35 C, 1/4 the rated life at 45 C, etc. I am under the impression
that most failures at 25 C should be from the lamp wearing out rather than
electronics. I expect life to decrease with increasing temperature to an
extent less than halving per 10 degrees C until the temperature is high
enough for failures to be mostly from the electronics.
But once the temperature gets high enough for most failures to be from
the electronics, I consider it reasonable for the life expectancy to half
per 10 deggres C beyond that point.

A good point - and probably explains my experience with the SLS.

Meanwhile, don't underestimate how hot things get in a ceiling fixture,
or even in any fixture enclosed at the top and sides, even if open at the
bottom.


I remember looking at the package of one of these, and it said that it
was rated for use in recessed ceiling fixtures. Most CFLs don't say this.
Last time I checked, Philips SLS being rated for use in recessed ceiling
fixtures was only for models up to 20 watts.

I'm never sure what "rated for use in recessed ceiling fixtures"
means. If the life is measured at 25C in free air, then it WILL be
lower when installed in a recessed ceiling fixture. Since the life of
many CFLs is 10,000 hours, some of these "rated for use in recessed
fixture" CFLs have a life of 15,000 hours under standard test
conditions and will be expected to have a life of at least 10,000
hours in recessed fixtures.

--
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.

 

[Clive Mitchell]

I note that the nearest 3 dissasembled ballasts (2 philips, 1 unknown
cheap make) all use 105C rated capacitors. I believe these ones are
rated at 1000 hours at 105C, so that'd be about 16000 hours at
105-40=65C. Fingering my nearest horizontally mounted 20W CFL, I get a
skin temp of the ballast at maybe 45C, so it's reasonable that the
electronics, are at about 65C.

Oh dear, so I'm not the only one with a box of dismembered CFL bases.


It's surprising that many of the CFL bases don't have even a basic form
of ventilation. The spiral lamps do, but many don't. This may be to
contain mini electro-explosions when things go wrong. I'm guessing it
can't be to reduce the risk of stuff getting poked through the
ventilation slots, since removing the lamp from it's socket reveals a
much larger area of exposed live metalwork.

I've got a spherical light fixture in the close where I live and drilled
some holes in both the base of the plastic globe and the top of the
fixture to allow air flow through the fixture. I reckon it will extend
the life of the CFL installed in it.

Talking about exposed live metalwork in lamp holders.... It's 2005.
Isn't it time we started switching all lamp holders over to something
shrouded like the GU10 base? The traditional lamp holders are very
Victorian. There's not much else on the market that has huge chunks of
live metal exposed for easy finger contact.

 

[Andrew Gabriel]

Talking about exposed live metalwork in lamp holders.... It's 2005.
Isn't it time we started switching all lamp holders over to something
shrouded like the GU10 base? The traditional lamp holders are very
Victorian. There's not much else on the market that has huge chunks of
live metal exposed for easy finger contact.

Lampholders are the one exception allowed to violate IP2X when doing
a PAT test of an electrical appliance.

Whilst you can say "gosh, this must be dangerous", there simply aren't
any significant incidents involving lampholders. If you try to put
together a risk/cost assessment for changing to some other style, then
you will just find it's all cost and no benefits. (Yea, I know that
hasn't stopped our current government with the likes of Part P, etc.)

 

[Victor Roberts]

I note that the nearest 3 dissasembled ballasts (2 philips, 1 unknown cheap
make) all use 105C rated capacitors.
I believe these ones are rated at 1000 hours at 105C, so that'd be about
16000 hours at 105-40=65C.
Fingering my nearest horizontally mounted 20W CFL, I get a skin temp of
the ballast at maybe 45C, so it's reasonable that the electronics,
are at about 65C.

There are higher rated caps - my nearest electronics catalog says up to
5000 hour at 105C are available - for a small premium.
That'd live either 5 times longer, or allow temperatures of 25C or so hotter.

I can tell you that when I was at GE the electrolytic caps were the
most carefully chosen and carefully tested component in their CFLs. I
can't describe the testing, but can say that the ratings provided by
the manufacturers are practically worthless when you are designing a
device to last 10,000 hours or more. All that matters are tests that
are run to _your_ specifications.

--
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.

 

[Victor Roberts]

Oh dear, so I'm not the only one with a box of dismembered CFL bases.


It's surprising that many of the CFL bases don't have even a basic form
of ventilation. The spiral lamps do, but many don't. This may be to
contain mini electro-explosions when things go wrong. I'm guessing it
can't be to reduce the risk of stuff getting poked through the
ventilation slots, since removing the lamp from it's socket reveals a
much larger area of exposed live metalwork.

No, it is to eliminate the risk of some one poking a paper clip or
similar device into a ventilation hole. The socket and base are
"grandfathered". UL would not allow them if they did not already
exist, but you cannot create any NEW hazards.

I've got a spherical light fixture in the close where I live and drilled
some holes in both the base of the plastic globe and the top of the
fixture to allow air flow through the fixture. I reckon it will extend
the life of the CFL installed in it.

Talking about exposed live metalwork in lamp holders.... It's 2005.
Isn't it time we started switching all lamp holders over to something
shrouded like the GU10 base? The traditional lamp holders are very
Victorian. There's not much else on the market that has huge chunks of
live metal exposed for easy finger contact.

--
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.

 

[Clive Mitchell]

One idiot sticks his fingers in an E27 base while standing in a bath
full with water and bath salts, and people start screaming for
standards.

Ooh! That sounds fun. Tell me more.

 

[David Lee]

Clive Mitchell wrote...

Ooh! That sounds fun. Tell me more.

I have this bizarre mental picture of a European standard specifying the
correct concentration of salts required for water intended for intentional
self-electrocution.

Actually - as EU directives go - that probably isn't so bizarre!

David

 

[Victor Roberts]

Too true! And the big name electrolytic manufacturers could be amongst
the most disappointing.

I didn't say that, and don't agree.

A wise old engineer once pointed out to me that
data sheets were marketing documents, created to define a marketing
intention not to inform engineers of anything useful.

I don't agree with this either. The data sheets from reputable
manufacturers are usually correct, but obviously for the conditions
under which the devices were tested. However, if your operating
conditions are different then you will need to run your own tests. My
specific comment was about the temperature rating of electrolytic
capacitors, since there is often no definition of what that rating
means - it is often not presented in the form of a specification. Data
shown in the data sheets along with test conditions is usually
reliable.

A decade or so ago Motorola (I think it was) acknowledged the general
uselessness of conventional data sheets and produced "engineering" data
sheets with real information in.

Data sheets fill the full range from superficial to rather complete.
If the data is not there it is obvious and I do applaud those
companies that provide a more compete set of data. However, I read
your earlier statement to mean that data that was supplied was not to
be trusted, and I disagree with that, at least from any of the
reputable manufacturers.

I think the attack of honesty must
have been seriously damaging to their sales because the idea appeared
to die shortly afterwards with no other vendors following suit.

Do you have any examples of these so-called "engineering" data sheets
vs. the "normal" ones. And I again disagree with your assertion that
the majority of data provided is dishonest. It is often incomplete,
but that data that is provided is usually correct, as long as you
understand the conditions under which the data was obtained.

Back on Caps. Having found your ideal electrolytic, how did you monitor
and maintain their quality from your supplier?

Ran tests on each production run.

Did you insist on
appropriate tests by your supplier?

I believe that was included.

Or perhaps some guarantees from
them on their suppliers and processes??

Don't know. And I also do not know how the process works today. There
may be a greater reliance on tests by the capacitor manufacturer
rather then by GE.

--
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.