On Thu, 30 May 2019 05:59:08 -0700 (PDT), Tim R
wrote:

It's probably worse than with CFLs because of the concentrated area
where the heat is produced, right?

Maybe. LED's are certainly more efficient.
Lumens/watt
LED 100
CFL 60
Incandescent 16
For equal amounts of light output (lumens), a CFL lamp takes 1.7 times
as much power to produce that light as does an LED. That's a big
difference, but not as spectacular as the 6.3 times jump from
incandescent to LED.

You're concerned about "heat" with measured in Joules where
1 joule = 1 watt/second
What I think you want is the final operating temperature of the device
to make sure that it doesn't melt plastic, degrade the LED's, destroy
electronic components, or set fire to the vicinity. The final
temperature has many parameters, most of which are NOT the same for
LED and CFL. For radiation loss, the surface area of a CFL lamp is
larger than the equivalent LED, and is therefore a more efficient heat
radiator. While the LED might waste fewer watts than the CFL light
heating up the room, the CFL will remove the heat from the lamp more
efficiently because it has a larger surface area. The LED compensates
for its smaller size by using aluminum heat sinks, while the CFL has a
larger thermal mass by using ceramics. Lots of other differences
making a general conclusion rather difficult.

What isn't obvious to me, having looked at the internals, is why the
orientation would have more than a trivial effect. That heat looks
pretty trapped no matter where the base faces.

Also maybe. The various lamps will move heat using conduction,
radiation, and convection. All three mechanisms are involved in
determining the final temperature of a lamp. In a light fixture, the
ability of move air through the fixture to remove the heat via
convective air currents is restricted. Without air flow the
temperature of the lamp will rise. If the air flow is uneven, there
will be hot spots on the lamp surface. Some lamps are more tolerant
to heating than others. My plastic case MR16 LED lamp was probably
the least tolerant. High temperature halogen incandescent lamps are
quite happy at much higher temperatures. LEDs lose half their light
output going from room temp (25C) to operating temperature (100C)
https://www.lrc.rpi.edu/programs/nlpip/lightinganswers/led/heat.asp
which is why LED heat sinks are much better and larger than CFL which
can tolerate higher temperatures.


--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Thursday, May 30, 2019 at 1:36:48 PM UTC-4, Jeff Liebermann wrote:

Maybe. LED's are certainly more efficient.
Lumens/watt
LED 100
CFL 60
Incandescent 16
For equal amounts of light output (lumens), a CFL lamp takes 1.7 times
as much power to produce that light as does an LED. That's a big

I doubt that's true. Flash power, the first 3 seconds of operation for an LED, might be 100 lumens per watt, but I think 60 is more realistic.


LED and CFL. For radiation loss, the surface area of a CFL lamp is
larger than the equivalent LED, and is therefore a more efficient heat
radiator. While the LED might waste fewer watts than the CFL light
heating up the room, the CFL will remove the heat from the lamp more
efficiently because it has a larger surface area. The LED compensates

I'm not sure that's the case. Surface area makes a difference for both radiation and convection, but the temperature difference is what really drives the heat transfer. The plastic globe area of an LED equivalent lamp runs much hotter than the curlicues of a CFL. (I haven't measured, but that's what my fingers tell me when changing both while hot.) (and that's probably because an LED doesn't have vacuum inside the globe, it has air that's in contact with the emitters)

I don't think the majority of the heat dissipation from a CFL is from the curlicues. But if even a portion is, then the orientation will make no difference. Air will flow through them base up or down about the same.

On Thu, 30 May 2019 12:51:20 -0700 (PDT), Tim R
wrote:

On Thursday, May 30, 2019 at 1:36:48 PM UTC-4, Jeff Liebermann wrote:

Maybe. LED's are certainly more efficient.
Lumens/watt
LED 100
CFL 60
Incandescent 16
For equal amounts of light output (lumens), a CFL lamp takes 1.7 times
as much power to produce that light as does an LED. That's a big

I doubt that's true. Flash power, the first 3 seconds of operation
for an LED, might be 100 lumens per watt, but I think 60 is more realistic.

That's correct, if you include:
1. Losses through any lenses or reflectors.
2. Losses in the switching power supply.
3. Losses in any power factor correction circuitry.
4. AC power is power factor corrected.

Easy enough to measure. I have enough junk from my flashlight
tinkering to measure lumens, but not with any great accuracy. Here's
how I do it with flashlights, bicycle headlights, spot lights, and any
light that generates a round spot on a wall.
https://groups.google.com/forum/#!original/rec.bicycles.tech/UJdJQFTDgl8/NgOZUloVCwAJ

I found new LED "flood" light in my collection.
https://www.feit.com/products/bulbs/flood-and-spot/led_lampsreflectorr_ledsrr_and_brbr30-927-led-can/
750 lumens 12.5w
Notice that there's plastic diffuser in front of the LEDs.

Plugging it into my Kill-A-Watt watt-guesser, I measu
12watts 15VA 0.78PF (power factor)
which seems about right.
Luminous Efficacy = 750lumens / 12watts = 63 lumens/watt
So, you're correct if I include all the losses.

However, the numbers I provided were for the raw LED at room
temperature and does NOT include all the losses.

Might as well grind the numbers for CFL.
I found one of these in my collection:
http://www.greenliteusa.com/en/lights/1684-26w-t2-ultra-mini-spiral.html
1700 lumens 26 watts
Measuring with the Kill-a-Watt meter:
26watts 37VA 0.70PF
Luminous Efficacy = 1700lumens / 26watts = 65 lum/w

Hmmm... looks like the LED (with the diffuser) is almost as bad as
CFL. I suspect the diffuser loss is what's causing the lower efficacy
for the LED, but I don't have LED bulb handy to prove it. I'll see if
I can find or borrow one, at least with a glass lens, not plastic.
According to this chart:
https://en.wikipedia.org/wiki/Luminous_efficacy#Lighting_efficiency
the LED floodlight should have been about 100 lum/W.

Surface area makes a difference for both radiation and convection, but
the temperature difference is what really drives the heat transfer.
The plastic globe area of an LED equivalent lamp runs much hotter than
the curlicues of a CFL. (I haven't measured, but that's what my
fingers tell me when changing both while hot.) (and that's probably
because an LED doesn't have vacuum inside the globe, it has air
that's in contact with the emitters)

Temperature (C or F) is not the same as heat (calories) which is not
the same power (watts) which is not the same as energy (joules or
watts/second). Also, there are several different types of
efficiencies and efficacy:
"Energy Efficiency of White LEDs" (2009)
https://www.fcgov.com/utilities/img/site_specific/uploads/led-efficiency.pdf

https://en.wikipedia.org/wiki/Luminous_efficacy#Lighting_efficiency
Overall luminous Overall luminous
efficacy lum/W efficiency
LED screw base lamp (120 V) Up to 102 Up to 14.9%
9- 32 W compact fluorescent 46 - 75 8 - 11.45%
(with ballast)

Kinda looks like the LED produces about twice the light output of the
CFL, for the same amount of input power. However, my measurements say
they're about the same.

I don't think the majority of the heat dissipation from a CFL is from the curlicues. But if even a portion is, then the orientation will make no difference. Air will flow through them base up or down about the same.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

On Thursday, 30 May 2019 18:36:48 UTC+1, Jeff Liebermann wrote:

You're concerned about "heat" with measured in Joules where
1 joule = 1 watt/second
What I think you want is the final operating temperature of the device
to make sure that it doesn't melt plastic, degrade the LED's, destroy
electronic components, or set fire to the vicinity. The final
temperature has many parameters, most of which are NOT the same for
LED and CFL. For radiation loss, the surface area of a CFL lamp is
larger than the equivalent LED, and is therefore a more efficient heat
radiator. While the LED might waste fewer watts than the CFL light
heating up the room, the CFL will remove the heat from the lamp more
efficiently because it has a larger surface area. The LED compensates
for its smaller size by using aluminum heat sinks, while the CFL has a
larger thermal mass by using ceramics. Lots of other differences
making a general conclusion rather difficult.

What isn't obvious to me, having looked at the internals, is why the
orientation would have more than a trivial effect. That heat looks
pretty trapped no matter where the base faces.

Also maybe. The various lamps will move heat using conduction,
radiation, and convection. All three mechanisms are involved in
determining the final temperature of a lamp. In a light fixture, the
ability of move air through the fixture to remove the heat via
convective air currents is restricted. Without air flow the
temperature of the lamp will rise. If the air flow is uneven, there
will be hot spots on the lamp surface. Some lamps are more tolerant
to heating than others. My plastic case MR16 LED lamp was probably
the least tolerant. High temperature halogen incandescent lamps are
quite happy at much higher temperatures. LEDs lose half their light
output going from room temp (25C) to operating temperature (100C)
https://www.lrc.rpi.edu/programs/nlpip/lightinganswers/led/heat.asp
which is why LED heat sinks are much better and larger than CFL which
can tolerate higher temperatures.


Heat radiation is not significant at LED operating temp. Ditto for most of a CFL, but the end filaments do run hot.

All heat produced by both does get dissipated, it can't be trapped else the thing would get endlessly hotter & self destruct.

Why does orientation matter? With a 20w CFL, the ballast silicon is vulnerable to high temps. Ballast down it gets cool air flowing slowly upward past it. Ballast up it gets hot air past it, reducing life expectancy. The tube OTOH doesn't care either way.


NT

The Dubai Lamp range is comprised of four LED bulbs, each of which is available in €œcool daylight€� and €œwarm white€� colors. Theres a 1-W, 200-lm E14 candle bulb, 400-lm E27 classic bulb and a 600-lm E27 classic bulb. Philips says it designed the filament LED bulbs to replace 25-W, 40-W, and 60-W incandescent bulbs, respectively. The bulbs run off Dubais 220-240-V mains voltage.

The above article is from very nearly two years ago. 150 - 175 l/W lamps are common in the commercial market, and readily available in to consumers with a just a bit of effort. A bit more costly, perhaps - but if one is in a region where the Utility is subsidizing prices, you may not notice.

That Home Depot or whatever passes for a Big-Box store at whatever location will not be selling either the latest, nor certainly not cutting-edge technology. They WILL be selling whatever may be mass-produced at the lowest cost with the highest margins.

Most of the discussions here are based on assumptions that are - at least - three years out of date.

Suffice it to understand:

LED Lamp drivers get HOT. These are drivers, not ballasts.
The amount of heat generated is in direct proportion to the amount of light generated as function of emitting surface. Linear emitters

https://images-na.ssl-images-amazon....1yQl0egsTL.jpg

spread the emitter heat out, and allow (relatively) tiny drivers making not-much heat. Most of these lamps are also "universal" inasmuch as they may be run in any position. And, their lumens-per-watt is not massive, either.

Point-source emitters such as the CREE XHP35 will make up to 706 lumens at 350 ma. - which translates to 183 lumens per watt. Really. And that was introduced in 2018, and is commonly found in your MagLite, if you want a "real world" application. It runs at 150C at the junction - which is tiny, so that heat is easily managed. And as it is a direct DC device, there is no separate driver in a flashlight application. In a lamp application, that driver gets quite complex as that same emitter may run anywhere from 5500K with a CRI of 50 to 2700K with a CRI of 90.

The technology is still evolving. And it is NOT where it was even a year ago.

Tabby, for the record, you give us all a deeper understanding of the term "invincible ignorance". Thank you for that!

Peter Wieck
Melrose Park, PA

On Friday, 31 May 2019 13:12:03 UTC+1, wrote:

The Dubai Lamp range is comprised of four LED bulbs, each of which is available in €œcool daylight€� and €œwarm white€� colors. Theres a 1-W, 200-lm E14 candle bulb, 400-lm E27 classic bulb and a 600-lm E27 classic bulb. Philips says it designed the filament LED bulbs to replace 25-W, 40-W, and 60-W incandescent bulbs, respectively. The bulbs run off Dubais 220-240-V mains voltage.

The above article is from very nearly two years ago. 150 - 175 l/W lamps are common in the commercial market, and readily available in to consumers with a just a bit of effort. A bit more costly, perhaps - but if one is in a region where the Utility is subsidizing prices, you may not notice.

That Home Depot or whatever passes for a Big-Box store at whatever location will not be selling either the latest, nor certainly not cutting-edge technology. They WILL be selling whatever may be mass-produced at the lowest cost with the highest margins.

Most of the discussions here are based on assumptions that are - at least - three years out of date.

Suffice it to understand:

LED Lamp drivers get HOT. These are drivers, not ballasts.
The amount of heat generated is in direct proportion to the amount of light generated as function of emitting surface. Linear emitters

https://images-na.ssl-images-amazon....1yQl0egsTL.jpg

spread the emitter heat out, and allow (relatively) tiny drivers making not-much heat. Most of these lamps are also "universal" inasmuch as they may be run in any position. And, their lumens-per-watt is not massive, either.

Point-source emitters such as the CREE XHP35 will make up to 706 lumens at 350 ma. - which translates to 183 lumens per watt. Really. And that was introduced in 2018, and is commonly found in your MagLite, if you want a "real world" application. It runs at 150C at the junction - which is tiny, so that heat is easily managed. And as it is a direct DC device, there is no separate driver in a flashlight application. In a lamp application, that driver gets quite complex as that same emitter may run anywhere from 5500K with a CRI of 50 to 2700K with a CRI of 90.

The technology is still evolving. And it is NOT where it was even a year ago.

Tabby, for the record, you give us all a deeper understanding of the term "invincible ignorance". Thank you for that!

Peter Wieck
Melrose Park, PA

They may be available, but they're not what's in the shops, so not what you find in people's homes. As I said. That you think a driver/ballast produces more heat than the LEDs borders on funny.


NT

They are found in both our houses. That you wish to make a virtue of ignorance is what is funny in a sad sort of way.

Peter Wieck
Melrose Park, PA

On Sun, 2 Jun 2019 14:26:16 -0700 (PDT), "
wrote:

They are found in both our houses. That you wish to make a virtue of
ignorance is what is funny in a sad sort of way.

To whom are you replying and about what?

Peter Wieck
Melrose Park, PA