An artist friend of mine asked me to help her with a project she is
designing.

The project is an atrium wall (ten foot above the floor) with an array
of blown glass elements(each about ten inches across). The elements
are to be lit from behind and the lights are to dim and brighten in a
random pattern.

She asked me about a way to hinge the glass elements so people could
replace the light bulbs, but she wanted the mounting to be hidden so
that the elements would appear to be floating.

I thought about this and came to the conclusion that it would be
easier, to never have to replace the bulbs.

I am wondering if a cluster of LEDs could be used in place of the
incandescent bulbs. It would cost more, but that might be offset by
much lower wiring and maintainance costs.

Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

Are they dimable?

How much extra would be involved in the control circuitry to drive the
LEDs?

Any help would be appreciated.

Paul K. Dickman

wrote:
An artist friend of mine asked me to help her with a project she is
designing.
snip
Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

Yes.


Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

The control circuitry is practically free.

Of course, the LEDs will come in at around $100/$160 per bulb equivalent.
And will probably need a fan, or heatsink per.

Any help would be appreciated.

Paul K. Dickman

On Sun, 30 Jan 2005 14:20:22 -0600, the renowned
wrote:

An artist friend of mine asked me to help her with a project she is
designing.

The project is an atrium wall (ten foot above the floor) with an array
of blown glass elements(each about ten inches across). The elements
are to be lit from behind and the lights are to dim and brighten in a
random pattern.

She asked me about a way to hinge the glass elements so people could
replace the light bulbs, but she wanted the mounting to be hidden so
that the elements would appear to be floating.

I thought about this and came to the conclusion that it would be
easier, to never have to replace the bulbs.

I am wondering if a cluster of LEDs could be used in place of the
incandescent bulbs. It would cost more, but that might be offset by
much lower wiring and maintainance costs.

Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

You may need a bunch of them. But they could change colors as well as
just brightness. Using RGB LEDs, you could get a large number of
effects. Or just white. The color temperature of white LEDs tends to
be higher than that of even halogen incandescents.

Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

Maybe less, they'd need a low voltage supply (like a LV Halogen) and
some control circuitry. Depends on what this "random" thing is all
about, but it's not difficult to get that kind of effect with LEDs.

Just for fun, I whipped up a green and white LED "spooky light" for
inside a ghostly-white pumpkin just before Halloween. It would be
similar, but on a larger scale. I just used a microprocessor (running
a little program I wrote) and some transistors to drive the (about)
half a dozen super-bright 5mm LEDs I used.

Any help would be appreciated.

Paul K. Dickman


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

On Sun, 30 Jan 2005 14:20:22 -0600, wrote:

An artist friend of mine asked me to help her with a project she is
designing.

The project is an atrium wall (ten foot above the floor) with an array
of blown glass elements(each about ten inches across). The elements
are to be lit from behind and the lights are to dim and brighten in a
random pattern.

She asked me about a way to hinge the glass elements so people could
replace the light bulbs, but she wanted the mounting to be hidden so
that the elements would appear to be floating.

I thought about this and came to the conclusion that it would be
easier, to never have to replace the bulbs.

I am wondering if a cluster of LEDs could be used in place of the
incandescent bulbs. It would cost more, but that might be offset by
much lower wiring and maintainance costs.

Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

Are they dimable?

How much extra would be involved in the control circuitry to drive the
LEDs?

Luxeons are the brightest LED's by far. They're about twice as
efficient as indandescants, so a 3-watt Luxeon would be equivalent to
about 5 or 6 watts of incandescant light -- if it's white light. If
it's colored light (red, green, blue, etc) then the Luxeons are more
effective: a 1-watt red lux is about equivalent to the 12-watt 1157
brakelight on a Chevy truck. A 3-watt red Luxeon probably would be
comparable to a color-filtered 25-watt incandescant. Color filters
significantly reduce the brightness of incandescants, but colored
Luxeons emit all of their lumens as light of the particular color at
hand -- no filter.

3-watt Luxeons are about $8.50 each. Lifetime to 50% brightness is
20,000 hours if run at 1 amp, 50,000 hours to 70% brightness if run
at 0.7 amps. Yes, with electronics they are dimmable -- and they
don't change color when dimmed, liked incandescants do. With
multiple colors, both brightness and color could be varied.
Electronics for use on mains could be very inexpensive if optimal
efficiency isn't a big deal.

Depending on the situation, you may find that the light can be
directed well enough with available inexpensive ($3.50) molded
collimators that you don't need as many watts (lumens) as you think
-- particularly if you use colors.

There is an upscale apartment in London (Chelsea) that did some
interesting illumination with Luxeons. Check out
http://www.thevospad.com/gallery.htm

Email me if you'd like to discuss Luxeons and/or elex to drive them
further. I'm fascinated with them!

The easiest way to dim a LED in your case would be to use a light dimmer
module in the wall and use that to drive a low voltage transformer that will
put out only enough voltage to light the LED to full brightness at line
voltage input. You need to turn the output of the transformer into DC
voltage but that is easily done with a bridge rectifier and a large cap. if
you desire. If you do this, lower voltages will end up making the LED
dimmer.

--
Why isn't there an Ozone Hole at the NORTH Pole?

"Bob May" wrote: (clip) You need to turn the output of the transformer into
DC (clip)
^^^^^^^^^^^^^
Just out of curiosity, what would happen if you applied AC? Would the light
go dark on the "wrong" half cycle, or act like a short and burn out?

On 30 Jan 2005 20:40:31 GMT, Ian Stirling
wrote:

wrote:
An artist friend of mine asked me to help her with a project she is
designing.
snip
Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

Yes.


Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

The control circuitry is practically free.

Of course, the LEDs will come in at around $100/$160 per bulb equivalent.
And will probably need a fan, or heatsink per.

Thanks;
Googling around the web I was coming up with similar numbers.

But, this is for a very public space here in Chicago and the array
would contain 60 or 70 elements. In incandescent that would probably
get listed as 60 of 70 separate electrical fixtures. Half that cost
could be saved in permit costs alone.

Plus, it is gonna cost them at least a hundred bucks every time they
have to get somebody on a lift to change the bulbs.

Paul K. Dickman


Any help would be appreciated.

Paul K. Dickman

On Sun, 30 Jan 2005 15:56:55 -0500, Spehro Pefhany
wrote:

On Sun, 30 Jan 2005 14:20:22 -0600, the renowned
wrote:

replace a 25 0r 40 watt incandescent bulb.?

You may need a bunch of them. But they could change colors as well as
just brightness. Using RGB LEDs, you could get a large number of
effects. Or just white. The color temperature of white LEDs tends to
be higher than that of even halogen incandescents.


An interesting idea. She did not even know what an LED was when I
suggested them. I think she'll stay with white at this point, just to
keep her confusion down, but it would increase the possibilities for
future projects.

Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

Maybe less, they'd need a low voltage supply (like a LV Halogen) and
some control circuitry. Depends on what this "random" thing is all
about, but it's not difficult to get that kind of effect with LEDs.

I think she just wants to have about 50% lit at any one time, each
randomly selected piece would go from off to fully lit to off over a
period of 15 to 20 seconds.


Best regards,
Spehro Pefhany

Thanks
Paul K. Dickman

On Sun, 30 Jan 2005 15:15:01 -0600, Don Foreman
wrote:

On Sun, 30 Jan 2005 14:20:22 -0600, wrote:



Luxeons are the brightest LED's by far. They're about twice as
efficient as indandescants, so a 3-watt Luxeon would be equivalent to
about 5 or 6 watts of incandescant light -- if it's white light. If
it's colored light (red, green, blue, etc) then the Luxeons are more
effective: a 1-watt red lux is about equivalent to the 12-watt 1157
brakelight on a Chevy truck. A 3-watt red Luxeon probably would be
comparable to a color-filtered 25-watt incandescant. Color filters
significantly reduce the brightness of incandescants, but colored
Luxeons emit all of their lumens as light of the particular color at
hand -- no filter.


3-watt Luxeons are about $8.50 each. Lifetime to 50% brightness is
20,000 hours if run at 1 amp, 50,000 hours to 70% brightness if run
at 0.7 amps. Yes, with electronics they are dimmable -- and they
don't change color when dimmed, liked incandescants do. With
multiple colors, both brightness and color could be varied.
Electronics for use on mains could be very inexpensive if optimal
efficiency isn't a big deal.

Depending on the situation, you may find that the light can be
directed well enough with available inexpensive ($3.50) molded
collimators that you don't need as many watts (lumens) as you think
-- particularly if you use colors.

I looked at the luxeon info. It was pretty impressive. Their flood
arrays looked like they had the right jazz and their dispersion
pattern looked like a good match for the shape of her objects.

I think the translucency of the glass used for the objects and the
fact that the sculpture is behind a south facing window will require
as many lumens as we can muster.



There is an upscale apartment in London (Chelsea) that did some
interesting illumination with Luxeons. Check out
http://www.thevospad.com/gallery.htm

Very interesting stuff

Email me if you'd like to discuss Luxeons and/or elex to drive them
further. I'm fascinated with them!

I probably will.

Thanks
Paul K. Dickman

On Sun, 30 Jan 2005 14:17:08 -0800, "Bob May"
wrote:

The easiest way to dim a LED in your case would be to use a light dimmer
module in the wall and use that to drive a low voltage transformer that will
put out only enough voltage to light the LED to full brightness at line
voltage input. You need to turn the output of the transformer into DC
voltage but that is easily done with a bridge rectifier and a large cap. if
you desire. If you do this, lower voltages will end up making the LED
dimmer.

An interesting concept, and one I'll keep around for my own use (maybe
I'll come out of this with a pocketful of LEDs) but, as I mentioned
in another post this is for a very public building in a big city, I
imagine we'll have EE's from the dept of construction and permits,
crawling all over this.

Thanks
Paul K. DIckman

On Sun, 30 Jan 2005 17:26:09 -0600, wrote:

On 30 Jan 2005 20:40:31 GMT, Ian Stirling
wrote:

wrote:
An artist friend of mine asked me to help her with a project she is
designing.
snip
Is this possible, do these new high output LED have enough jazz to
replace a 25 0r 40 watt incandescent bulb.?

Yes.


Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

The control circuitry is practically free.

Of course, the LEDs will come in at around $100/$160 per bulb equivalent.
And will probably need a fan, or heatsink per.

Thanks;
Googling around the web I was coming up with similar numbers.

But, this is for a very public space here in Chicago and the array
would contain 60 or 70 elements. In incandescent that would probably
get listed as 60 of 70 separate electrical fixtures. Half that cost
could be saved in permit costs alone.

Plus, it is gonna cost them at least a hundred bucks every time they
have to get somebody on a lift to change the bulbs.

Paul K. Dickman


Any help would be appreciated.

Paul K. Dickman
Greetings Paul,
After reading the above post about permit costs it made me wonder if
the permitting would be easier and/or cheaper because the LEDs use low
voltage. It sure makes a difference where I live when it comes to
putting lights along the driveway and similar situations. With the low
voltage I don't think it even needs UL approval. The power supply,
which has a high voltage input needs UL approval, but that is store
bought. Everything that's connected to the output is not considered a
shock hazard. Like those halogen track lights that are suspended from
uninsulated wires.
ERS

On Sun, 30 Jan 2005 16:14:04 -0800, Eric R Snow
wrote:

On Sun, 30 Jan 2005 17:26:09 -0600, wrote:

On 30 Jan 2005 20:40:31 GMT, Ian Stirling
wrote:

wrote:
An artist friend of mine asked me to help her with a project she is
designing.
snip

But, this is for a very public space here in Chicago and the array
would contain 60 or 70 elements. In incandescent that would probably
get listed as 60 of 70 separate electrical fixtures. Half that cost
could be saved in permit costs alone.

Plus, it is gonna cost them at least a hundred bucks every time they
have to get somebody on a lift to change the bulbs.

Paul K. Dickman


Any help would be appreciated.

Paul K. Dickman
Greetings Paul,
After reading the above post about permit costs it made me wonder if
the permitting would be easier and/or cheaper because the LEDs use low
voltage. It sure makes a difference where I live when it comes to
putting lights along the driveway and similar situations. With the low
voltage I don't think it even needs UL approval. The power supply,
which has a high voltage input needs UL approval, but that is store
bought. Everything that's connected to the output is not considered a
shock hazard. Like those halogen track lights that are suspended from
uninsulated wires.
ERS

That's my thought. Plus wiring would be a lot simpler. No running
conduit or bx, just running low voltage wires through the walls.

And the things have a low enough profile, that the fixtures themselves
could be surface mounted.

As an added bonus, this is for a corporate "Green building" The
energy efficiency of the LEDs could help sell it.


Paul K. Dickman

On Sun, 30 Jan 2005 16:14:04 -0800, the renowned Eric R Snow
wrote:

Greetings Paul,
After reading the above post about permit costs it made me wonder if
the permitting would be easier and/or cheaper because the LEDs use low
voltage. It sure makes a difference where I live when it comes to
putting lights along the driveway and similar situations. With the low
voltage I don't think it even needs UL approval. The power supply,
which has a high voltage input needs UL approval, but that is store
bought. Everything that's connected to the output is not considered a
shock hazard. Like those halogen track lights that are suspended from
uninsulated wires.
ERS

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

wrote in message
...
On Sun, 30 Jan 2005 15:15:01 -0600, Don Foreman
wrote:

On Sun, 30 Jan 2005 14:20:22 -0600, wrote:

....


I think the translucency of the glass used for the objects and the
fact that the sculpture is behind a south facing window will require
as many lumens as we can muster.


Which reminds me of a friend of mine who got a contract to build
a bunch of those signs with the flashing lights that the road works
people use to warn drivers to merge lanes and such.

You know the ones. Big yellow arrows flashing in sequence...

He did them using leds and everything worked great until the
sun came out.

They ended up in the dumpster, and I don't recall what
my buddy said about who paid for what!

Now this is some time ago and there were no super brights at
the time but...

DOC

wrote:

An interesting concept, and one I'll keep around for my own use (maybe
I'll come out of this with a pocketful of LEDs) but, as I mentioned
in another post this is for a very public building in a big city, I
imagine we'll have EE's from the dept of construction and permits,
crawling all over this.
Paul, I've been doing some research on low voltage lighting for one of
my clients and "IF" I remember correctly, low voltage circuits (50v or
less) have much less stringent requirements as far as conduit and
fixtures than "mains" voltage. My sources are ICC and NEC. In the next
couple of days I will get chapter and verse for you if you like. I am
currently working on several low voltage lighting projects and the
company I am working with are quoting a light output ratio of
incandescent to LED of about 8 or 9 to 1 (their claim not mine). If you
like, I would be happy to discuss some options with you off group. They
might be interested in your project, they are currently going a huge
light sculpture in a shopping center in Asia.

In article , Spehro Pefhany says...

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.

Mooo.

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

there is a device you can buy that is called "sauce" - you can buy it at
some retail stores - a bunch of LEDs and a microcontroller that put out
varying color light.

see here for a bunch of similar ideas http://colorkinetics.com/products/pro/
wrote in message
...
On Sun, 30 Jan 2005 15:56:55 -0500, Spehro Pefhany
wrote:

On Sun, 30 Jan 2005 14:20:22 -0600, the renowned
wrote:

replace a 25 0r 40 watt incandescent bulb.?

You may need a bunch of them. But they could change colors as well as
just brightness. Using RGB LEDs, you could get a large number of
effects. Or just white. The color temperature of white LEDs tends to
be higher than that of even halogen incandescents.


An interesting idea. She did not even know what an LED was when I
suggested them. I think she'll stay with white at this point, just to
keep her confusion down, but it would increase the possibilities for
future projects.

Are they dimable?

Sure.

How much extra would be involved in the control circuitry to drive the
LEDs?

Maybe less, they'd need a low voltage supply (like a LV Halogen) and
some control circuitry. Depends on what this "random" thing is all
about, but it's not difficult to get that kind of effect with LEDs.

I think she just wants to have about 50% lit at any one time, each
randomly selected piece would go from off to fully lit to off over a
period of 15 to 20 seconds.


Best regards,
Spehro Pefhany

Thanks
Paul K. Dickman

jim rozen wrote in
:

In article , Spehro
Pefhany says...

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.

Mooo.

Nah, it (McCormick Place) burned to the ground while a housewares show was
being held there.
http://www.chipublib.org/004chicago/...mick_fire.html
One of the worst fires in the cities history.

Dan

On 31 Jan 2005 04:35:49 GMT, the renowned Dan Murphy
wrote:

jim rozen wrote in
:

In article , Spehro
Pefhany says...

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.

Mooo.

Nah, it (McCormick Place) burned to the ground while a housewares show was
being held there.
http://www.chipublib.org/004chicago/...mick_fire.html
One of the worst fires in the cities history.

Dan

Thanks for that link.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

On Sun, 30 Jan 2005 19:49:41 -0600, Greg Postma
wrote:

(snip)

company I am working with are quoting a light output ratio of
incandescent to LED of about 8 or 9 to 1 (their claim not mine). If you
like, I would be happy to discuss some options with you off group. They
might be interested in your project, they are currently going a huge
light sculpture in a shopping center in Asia.

That claim must be based on LEDs of various colors compared to
incandescants working thru color filters. It's not true with white.
2.5 : 1 would be more in the ballpark.

A typical 25-watt halogen lamp produces about 200 lumens. Ref:
http://www.con-techlighting.com/prod...oductTypeID=32
That's about 8 lumens/watt. A white Luxeon III nominally produces
80 lumens at about 3.9 w (1 amp at 3.9 volts). That's about 20 lumens
per watt, giving a ratio of about 2.5 : 1.

Bottom line: three white Luxeon III's would be brighter than an
unfiltered 25-watt halogen incandescant light, while using less
than half as much energy.

On Sun, 30 Jan 2005 20:16:33 -0500, Spehro Pefhany
wrote:

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.

Invite the inspector to touch his nose to a operating 3W LED, and
then to an operating 12-volt halogen bulb. Better yet, in the
interest of safety of course, invite him to use a more southern and
more sensitive temp-test probe to protect eyes from bright light....

Don Foreman wrote:
On Sun, 30 Jan 2005 20:16:33 -0500, Spehro Pefhany
wrote:

I remember from doing many trade shows in Chicago that they are really
fussy about anything that could cause a fire. They've had problems in
the past, as I understand it.

Invite the inspector to touch his nose to a operating 3W LED, and
then to an operating 12-volt halogen bulb. Better yet, in the
interest of safety of course, invite him to use a more southern and
more sensitive temp-test probe to protect eyes from bright light....

They actually produce about the same amount of heat.
It's just that the LED has to be heatsinked, or it'll explode.
If LEDs worked at 500C, then they wouldn't be heatsinked so well, and
would get just as hot.

In article , Spehro Pefhany says...

Mooo.

http://www.chipublib.org/004chicago/disasters/mccormick_fire.html

Thanks for that link.

Don't forget the *other* big conflagration:

http://www.chipublib.org/004chicago/disasters/great_fire.html

It may be that the first fire also tends to polarize the
chicago fire inspectors!

As I said before, Mooooooooo. translation: "oops."

Jim


--
==================================================
please reply to:
JRR(zero) at pkmfgvm4 (dot) vnet (dot) ibm (dot) com
==================================================

If you want to get really fancy you could use clusters of 3 Luxeon IIIs with
PWM drivers. #18 wire would carry 1 amp @ 12V (to drive 3 3.9V LEDs) up to
40' so the drivers could all be mounted on one centrally located board with
the microcontroller. The drivers could have digital pots so the lighting
sequences can be controlled by a microcontroller. The microcontroller could
set some random pattern and speeed based on something like the room
temperature or noise level. Maybe even a capacitance sensor pad so the
viewer could participate.

Luxeon IIIs are not cheap but I would lay odds that you could get a deal on
150 to 200 of them. Any electronics geek with a little patience could build
an addressable 50 circuit controller board for under $500.

Another alternative would be to use regular 12cd hi-bright LEDS mounted
directly to the glass. The light would enter the glass and come out at the
edges and imperfections with no light wasted. Also by combining red green
and blue LEDs you could change the colors as well as the brightness. At
$.50 each retail those would be a lot cheaper and the total amout of light
required would be less. I have used that method for some etched Lucite
restroom signs and it works really well.

--
Glenn Ashmore

I'm building a 45' cutter in strip/composite. Watch my progress (or lack
there of) at: http://www.rutuonline.com
Shameless Commercial Division: http://www.spade-anchor-us.com

On 31 Jan 2005 05:02:38 -0800, the renowned jim rozen
wrote:

In article , Spehro Pefhany says...

Mooo.

http://www.chipublib.org/004chicago/disasters/mccormick_fire.html

Thanks for that link.

Don't forget the *other* big conflagration:

http://www.chipublib.org/004chicago/disasters/great_fire.html

It may be that the first fire also tends to polarize the
chicago fire inspectors!

As I said before, Mooooooooo. translation: "oops."

Jim

They wanted to see UL stickers on everything that plugged in. And they
have a lot of authority. We certainly felt cowed.


Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

On 31 Jan 2005 11:21:56 GMT, Ian Stirling
wrote:

They actually produce about the same amount of heat.
It's just that the LED has to be heatsinked, or it'll explode.
If LEDs worked at 500C, then they wouldn't be heatsinked so well, and
would get just as hot.

Yes, but surface temps are markedly different. Few materials ignite
at 75C but ya don't want a halogen bulb in the powder magazine!

On Mon, 31 Jan 2005 08:45:40 -0500, "Glenn Ashmore"
wrote:

Luxeon IIIs are not cheap but I would lay odds that you could get a deal on
150 to 200 of them. Any electronics geek with a little patience could build
an addressable 50 circuit controller board for under $500.

Another alternative would be to use regular 12cd hi-bright LEDS mounted
directly to the glass. The light would enter the glass and come out at the
edges and imperfections with no light wasted. Also by combining red green
and blue LEDs you could change the colors as well as the brightness. At
$.50 each retail those would be a lot cheaper and the total amout of light
required would be less. I have used that method for some etched Lucite
restroom signs and it works really well.

Cost per lumen is about the same. White Luxeon III's were $8.50 ea
last time I bought. I think the price break on Luxeons is at about
2500 pcs.

Yep. LEDs respond quite quickly to the voltage level, usually in the
megahertz region and 60 hertz is a very low frequency.
You can backbias a LED to several times the forward voltage (if you're
worried about this, put another LED in there pointing the other way and it
will be on when the other is off) without problems. There is a spec on how
much back voltage that any particular LED will take without problems.
If you use a regular light dimmer control for brightness, the brightness
level will be rather nonlinear as the dimmers do their work by being off
until a certain time is attained in the sine wave period of the incoming
power. This will mean that the low levels will be very nonlinear due to the
low voltage that comes out. The short time period plus the low voltage is
where the problem (if you want to call it that) happens. Using DC plus the
cap will average out the voltage and that will tend to linearize the
brightness vs. position of the dimmer knob.

--
Why isn't there an Ozone Hole at the NORTH Pole?

One thing to remember is that with LEDs, the light is directional to a fair
degree so they don't need reflectors to reflect the light to the area of
interest. This will magnify the lumens that are viewed by some ratio.
In the relativey early days of using LEDs for indicators, the problem with
the LEDs as that they didn't have the wide angle light production so there
had to be some piece of material to widen the light so that it could be seen
brightly way off-axis from the direct on viewing position. The LED
indicator was brighter on-axis than the incadescent bulbs were but the
off-axis light output caused the LEDs to not be used so much as they were
dimmer in those directions. Eventually light output got high enough in LEDs
that they could use frosted plastic in the construction and thus be used for
that application.

--
Why isn't there an Ozone Hole at the NORTH Pole?

On Mon, 31 Jan 2005 11:53:00 -0800, "Bob May"
wrote:

One thing to remember is that with LEDs, the light is directional to a fair
degree so they don't need reflectors to reflect the light to the area of
interest. This will magnify the lumens that are viewed by some ratio.
In the relativey early days of using LEDs for indicators, the problem with
the LEDs as that they didn't have the wide angle light production so there
had to be some piece of material to widen the light so that it could be seen
brightly way off-axis from the direct on viewing position. The LED
indicator was brighter on-axis than the incadescent bulbs were but the
off-axis light output caused the LEDs to not be used so much as they were
dimmer in those directions. Eventually light output got high enough in LEDs
that they could use frosted plastic in the construction and thus be used for
that application.

Luxeons are not very directional. The radiation pattern of the
Lambertian model approximates a cosine curve, with luminosity down to
50% at about +/- 50 degrees off axis. The batwing and sideshooter
models emit more of their light sideways, less on-axis. The
sideshooter is visible over 360 degrees, is designed for use in a
parabolic reflector.

A bare Luxeon acts more like a "bare bulb" in front of a flat
reflector. A reflector or collimator is necessary if directionality
is desired. Good molded acrylic collimators are available for
about $3.50 in unit qty. Beam divergence angles are available from
"flashlight" (10 degrees) up to about 45 degrees.

The lumen is a measure of total light flux; lenses or reflectors
don't magnifiy lumens, but just govern where the lumens go. You did
say "lumens that are viewed", so I mention this merely for clarity.

They do magnify luminous intensity, measured in candela (lumens per
steradian) or lux (lumens per square meter). The luminous
intensity of a 1-watt Luxeon collimated to a beam with 10 degree
divergence is 342 cd, as compared to an "ultrabright" LED of perhaps
12 cd. A Luxeon III running at 1 amp in the same situation would be
1368 cd. Do not view directly with remaining good eye....

On Mon, 31 Jan 2005 14:54:53 -0600, the renowned Don Foreman
wrote:

They do magnify luminous intensity, measured in candela (lumens per
steradian) or lux (lumens per square meter). The luminous
intensity of a 1-watt Luxeon collimated to a beam with 10 degree
divergence is 342 cd, as compared to an "ultrabright" LED of perhaps
12 cd. A Luxeon III running at 1 amp in the same situation would be
1368 cd. Do not view directly with remaining good eye....

At 1A, you've got around 3.5W that has to be dissipated. Have you had
any problems keeping the die temperature down on the Luxeons at full
current?



Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

On Mon, 31 Jan 2005 16:15:03 -0500, Spehro Pefhany
wrote:


At 1A, you've got around 3.5W that has to be dissipated. Have you had
any problems keeping the die temperature down on the Luxeons at full
current?

No, but they do need a heatsink. Theta j-c is 13C/W for the
emitter, max rated die temp is 135C, so theoretically if the case is
kept below 70C you're OK. I don't like to run them that hot, though.
The star is 17C/W.

I have a 3W emitter (not a star) mounted on a 3/8" dia aluminum post
about 1.1" long with a 2" square piece of 0.100" thick aluminum on the
other end of the post. With the thermal epoxy, I probably have about
17C/W, junction-to-post. (The post is so I can poke it into a
reflector)

At 1.000 amps the voltage on this particular part is running 3.517
volts at temperature. The post settled out at 55.2C right near the
LED. The plate is at 52.3C. Ambient in the lab is 19.5C

Aavid Thermalloy has a selection of heatsinks specifically recommended
for Luxeons.
http://tinyurl.com/6k6f6

One model 1.81" square x 1.25" high does about 5C/W, so 20C rise at 4
watts.