LED lighting info
John R.
I am currently conducting some tests on these lights. I have posted some
information at this website: http://www.mp3brick.com/marine/
Just follow the "Lighting Project" link.
John R.
nosmatt
Acroreef
>I am currently conducting some tests on these lights.
BRAVO !
Excellent Job!!
Keep up the good work, this looks very promising.
Regards,
Robert
BioStress
>
> A month ago I posted some questions reguarding LED lighting. As a followup,
> information at this website: http://www.mp3brick.com/marine/
> Just follow the "Lighting Project" link.
>
> John R.
John,
I applaud your efforts in "light experimentation." Keep us posted.
Bio
--
-------
Biostress Marine Aquarium
75 Gallon Reef
http://www.geocities.com/biostress/reef.htm
Nano Reef Project
http://www.geocities.com/biostress/nano_reef_project.htm
Leftovers
in use and cost per LEd etc.... inquiring minds wanna no...
"John R." wrote:
>
> A month ago I posted some questions reguarding LED lighting. As a followup,
> I am currently conducting some tests on these lights. I have posted some
> information at this website: http://www.mp3brick.com/marine/
> Just follow the "Lighting Project" link.
>
> John R.
--
+++++++++++++++++++++++++++++++++++++++++++++++
remove spamfree to reply -yes the address works
so flame away
If fish were like dogs I wonder if we would
kill so many of them...
+++++++++++++++++++++++++++++++++++++++++++++++
John R.
soon.
"Leftovers" <leftovers...@home.com> wrote in message
news:3955E7F6...@home.com...
Wayne Sallee
does it feal?
Wayne Sallee
http://members.aol.com/waynesallee/weblink.htm
Andreas
mounted?
JBortz6469
directly from the source to be effective or for example could you create a
mirror system to help spread the light emmited form the source whatever that
might be say even a Mh source. I am not sure why I am asking just curious I
guess.
Jamin
Tim Houghton
Keep us posted! :)
Tim
John R.
heat.
John R.
Michael H.
<wsf55.10256$aB6....@typhoon.hawaii.rr.com>...
> A month ago I posted some questions reguarding LED lighting. As a
followup,
> I am currently conducting some tests on these lights. I have posted some
> information at this website: http://www.mp3brick.com/marine/
> Just follow the "Lighting Project" link.
>
> John R.
>
18" (70-90 gallon) tank with a similar setup to what you currently have, it
would consume only about 250 watts. If the LEDS were, say, 4 times as
efficient as a metal halide, then this would be equivalent to 1000 watts of
MH lighting.
Cost wise, the LED's only would be approx: 576 x $5 = $2880 for blue
and 1024 x $2.80 = $2867 for white (also, more blue, less white may be
required to obtain the correct colour)
So, for $5747 + reflectors + 12 DC power supply (s) etc. you would be set.
Still a bit on the expensive side, it seems.
Yrrep34
commented on this in your original post and if so, I apologize, but I wondered
how many LED's would you need to achieve 3-5 watt per gallon (and I know that
is just a rough guideline, but I am wondering about the cost to use these
lights in a large tank I am planning.) Secondly, do you get the wonderful
shimmer that MH lighting gives you? I would think LED could be a terrific
option, even with the higher set up costs, because of their efficiency and low
heat output.
Thanks,
Perry R Branson
Phil Tomson
John R. <mp3b...@haNOSPAMwaii.rr.com> wrote:
>A month ago I posted some questions reguarding LED lighting. As a followup,
>I am currently conducting some tests on these lights. I have posted some
>information at this website: http://www.mp3brick.com/marine/
>Just follow the "Lighting Project" link.
>
>John R.
>
>
Looks promising!
Maybe I'll have to fire up the old soldering iron and give it a go!
Phil
nosmatt
what would you need over say a 15 g nano????
or better yet, what did u use /sq " for white & blue ???
Acroreef
>>Just follow the "Lighting Project" link.
I see there's a new update which includes schematic. After reviewing the
schematic, I can easily see a solid state dimming device that could provide
"ramp-up" and "ramp-down" abilities.
I also wonder how this type of lighting would stand up to some scientific
testing of the Dana Riddle or Sanjay Joshi type? You know, just to see how the
intensity and spectrum compare to metal halide.
This looks very encouraging. Keep up the good work.
Regards,
Robert
Chris Fanning
> John R. <mp3b...@haNOSPAMwaii.rr.com> wrote in article
> <wsf55.10256$aB6....@typhoon.hawaii.rr.com>...
> followup,
>> I am currently conducting some tests on these lights. I have posted some
>> Just follow the "Lighting Project" link.
>>
>>
> If my calculations are correct, if you filled the entire hood of a 48" x
> 18" (70-90 gallon) tank with a similar setup to what you currently have, it
> would consume only about 250 watts. If the LEDS were, say, 4 times as
> efficient as a metal halide, then this would be equivalent to 1000 watts of
> MH lighting.
> Cost wise, the LED's only would be approx: 576 x $5 = $2880 for blue
> and 1024 x $2.80 = $2867 for white (also, more blue, less white may be
> required to obtain the correct colour)
> So, for $5747 + reflectors + 12 DC power supply (s) etc. you would be set.
> Still a bit on the expensive side, it seems.
I imagine you'd get a significant price break on 1000s.
For a 1980 typ mcd blue led, in singles it goes for 2.95 and 1000 for 1.95
from Digikey. For blue that's 1000 x 1980 mcd = 1980 cd @ 20A @ 5.0V
or 100W. Take white, for one 3.95, for 1000 2.75. It's another 3700 cd
at 72W. This is from Jameco and there are better blue LEDs.
Anyone convert cd into lumens? LEDs can be horribly efficient. Aside from
the cost I wonder how the brightness compares?
72W of cool running lighting to replace 500W of hot stuff would sure
sound good.
Chris
John R.
re-itterate that I'm not an expert in lighting.
> How many LEDs to achieve 3-5 watt/gal?
Not sure. What I understand you cannot rate LED in the same watt catagory as
other light sources. White LEDs produce much more light-per-watt rather than
light/heat-per-watt. Yes to the shimmering effect question.
> How many LEDs for a 55gal tank?
Micheal H,
I love your quick calculation, but I don't think one would need that many
LEDs. You must consider lumens and candelas and not the watts. I'm not sure
how many times more efficient these LEDs are but Chris Fanning has the right
idea.
>"I can easily see a solid state dimming device that could provide "ramp-up"
and "ramp-down" abilities", Robert.
Right-on Robert! I was thinking along the same lines. The nice thing about
these LEDs is that they do not change color when dimmed.
>what did u use /sq " for white & blue ???
Well I managed about 32 whites and 16 blues with reflectors in a 3" X 6"
circiut board.
On a side note, I need someone who is willing to do some real spectrum
analysis on these LEDs. I'm not trying to give up anyone's hopes here but
I'm worried that the whites are giving off too much red and yellow light
(not good). Perhaps higher ratio of blues to whites would be needed. Can
anyone verify this?
Regards,
John R.
fan_de_ranger
have a pretty low dispersion.
Rob
"John R." <mp3b...@haNOSPAMwaii.rr.com> wrote in message
news:wsf55.10256$aB6....@typhoon.hawaii.rr.com...
John R.
by 50%. Of course I thought the same as you did. But after playing around
with those little reflectors, I did see a noticable difference. Seems they
redirect some of the stray light that escape out the sides (not going
through the LED lens) giving a more uniform glow. LEDs are not perfect and
have various "hot spots" even color temps. These variations seem less
evident as you increase distance between the object and the LEDs.
"fan_de_ranger" <nos...@twcny.rr.com> wrote in message
news:hCV55.11408$Bc.3...@typhoon.nyroc.rr.com...
Chris Fanning
> I'll do my best to address some of the very good questions here. I have to
> re-itterate that I'm not an expert in lighting.
>> How many LEDs to achieve 3-5 watt/gal?
> Not sure. What I understand you cannot rate LED in the same watt catagory as
> other light sources. White LEDs produce much more light-per-watt rather than
> light/heat-per-watt. Yes to the shimmering effect question.
>> How many LEDs for a 55gal tank?
> Micheal H,
> I love your quick calculation, but I don't think one would need that many
> LEDs. You must consider lumens and candelas and not the watts. I'm not sure
> how many times more efficient these LEDs are but Chris Fanning has the right
> idea.
>>"I can easily see a solid state dimming device that could provide "ramp-up"
> and "ramp-down" abilities", Robert.
> Right-on Robert! I was thinking along the same lines. The nice thing about
> these LEDs is that they do not change color when dimmed.
Why dim them? If you've got 1000 LEDs just shut a proportion of them off
entirely.
>>what did u use /sq " for white & blue ???
> Well I managed about 32 whites and 16 blues with reflectors in a 3" X 6"
> circiut board.
> On a side note, I need someone who is willing to do some real spectrum
> analysis on these LEDs. I'm not trying to give up anyone's hopes here but
> I'm worried that the whites are giving off too much red and yellow light
> (not good). Perhaps higher ratio of blues to whites would be needed. Can
> anyone verify this?
Yes, this is going to be the interesting bit. Don't manufacturers have
these specs published? I'm going to have to check.
Chris
John
LED's are made by actually putting three different LED's on one tiny chip,
one blue, one red, and one green[see www.nichia.com]. Single LED's emit
monochromatic light - diode lasers are made using the same semiconductor
crystals that are used in LED's. Therefore the spectrum of a white LED has
three sharply defined peaks at the wavelengths corresponding to
blue/green/red. So while the spectrum of your lighting system may "look
good," I would bet that it it scores low in PAR and will not be as "healthy"
as lighting with a more diverse spectrum. But I could be wrong because if
those peaks fall in exactly the right places they may do quite well. I tend
to think the more you simulate sunlight (a broad well covered spectrum) the
better results you'll have. Adding a little blue to a "faux-sunlight-lit"
tank to make it look better certainly has no ill effects as long as you have
all the other wavelengths present. Another concern is the outrageous cost
of constructing an LED system, if it indeed works well with animals, for a
much larger tank. Just my $.02 - john.
> From: acro...@aol.com (Acroreef)
> Organization: AOL http://www.aol.com
> Newsgroups: rec.aquaria.marine.reefs
> Date: 27 Jun 2000 00:59:34 GMT
> Subject: Re: LED lighting info
>
>> http://www.mp3brick.com/marine/
>>> Just follow the "Lighting Project" link.
>
>
Jimmy Chen
> how many LED's would you need to achieve 3-5 watt per gallon (and I know
that
> is just a rough guideline, but I am wondering about the cost to use these
> lights in a large tank I am planning.)
Judging from the LED web site, these lights seems like the common "desk" lamp
now that is sold just about everywhere nowadays. One of those little sucker is
almost a replacement of a common 60W incandescent. If you happen to live near
an Ikea, you surely can go to their lighting dept and see plenty of LED lights
fixtures there.
Secondly, do you get the wonderful
> shimmer that MH lighting gives you?
It seems like it wont be able to produce that kind of effect as now you will
have many source points, more like a VHO, rather then one source point like a
MH. Of course, this MAY change if there are larger LEDs abailable.
--
jc
Reefkeepers ML (http://www.reefkeepers.org)
Archived @ http://www.escribe.com/pets/reefkeepers/
Jimmy Chen
White LED Specifications: Type T-1 3/4, 5mm diameter; DC forward voltage of
3.5 to 4.0V; 5600mcd (millicandela) @ 20mA; 120mW power dissipation; Approx.
15 to 16 Lumens/watt; Estimate of 6000 to 6500 Kelvin; 85 CRI.
> If my calculations are correct, if you filled the entire hood of a 48" x
> 18" (70-90 gallon) tank with a similar setup to what you currently have, it
> would consume only about 250 watts. If the LEDS were, say, 4 times as
> efficient as a metal halide, then this would be equivalent to 1000 watts of
> MH lighting.
The math: ;-)
So, 36 White LEDs approx. 360 milliamps (4.32 watts) and outputs 540W. Based
on your numbers, 576 will consume ~70W, but produces ~8640W. A little too
sunny, I'll say. So we will only need around ~67 LEDs to produce ~1000Ws.
> Cost wise, the LED's only would be approx: 576 x $5 = $2880 for blue
> and 1024 x $2.80 = $2867 for white (also, more blue, less white may be
> required to obtain the correct colour)
>
> So, for $5747 + reflectors + 12 DC power supply (s) etc. you would be set.
> Still a bit on the expensive side, it seems.
The 36 LED (~500Ws) setup is quoted at $280, so it is still cheaper then a
400W MH fixture + bulb.
Jimmy Chen
> have various "hot spots" even color temps
Here is the graph on the color temps ...
http://www.theledlight.com/images-tech/spectrum-2.jpg
Jimmy Chen
> 3.5 to 4.0V; 5600mcd (millicandela) @ 20mA; 120mW power dissipation;
Approx.
> 15 to 16 Lumens/watt; Estimate of 6000 to 6500 Kelvin; 85 CRI.
Judging from a little more research on my part, it doesnt look like those LEDs
will be able to produce 15w, or at least not the same W we use for MH lights.
So, my "math" was totally off as I need to find the true "watt" comparison
with this light vs MH.
Okay, I will shut up now ... ;-)
John R.
Here's a quote from a reputable LED dealer......
>quote<
White Light
When light from all parts of the visible spectrum overlap one another, the
additive mixture of colors appears white. However, the eye does not require
a mixture of all the colors of the spectrum to perceive white light. Primary
colors from the upper, middle, and lower parts of the spectrum (red, green,
and blue), when combined, appear white. To achieve this combination with
LEDs requires a sophisticated electro-optical design to control the blend
and diffusion of colors. Variations in LED color and intensity further
complicate this process.
Presently it is possible to produce white light with a single LED using a
phosphor layer (Yttrium Aluminum Garnet) on the surface of a blue (Gallium
Nitride) chip. Although this technology produces various hues, white LEDs
may be appropriate to illuminate opaque lenses or backlight legends.
However, using colored LEDs to illuminate similarly colored lenses produces
better visibility and overall appearance.
http://www.theledlight.com/technical.html
>end quote<
The newer white LEDs do not use the "tri-chip" design as you mention.
Instead, they use a typical blue LED with a special phosphor layer on a
single chip. One problem is that depending on how thick and pure the
phosphor layer will dictate a slight color shift in the white LED. Now, the
question is weather or not this technology reduces any spectrum peaks. I
wish someone had a spectrometer to conduct some real testing..
Thanks,
John R.
"John" <john.mu...@home.com> wrote in message
news:B57EB86F.2308%john.mu...@home.com...
> There is a definite problem with using LED's to simulate sunlight. White
> LED's are made by actually putting three different LED's on one tiny chip,
> one blue, one red, and one green[see www.nichia.com]. Single LED's emit
> monochromatic light - diode lasers are made using the same semiconductor
> crystals that are used in LED's. Therefore the spectrum of a white LED has
> three sharply defined peaks at the wavelengths corresponding to
> blue/green/red. So while the spectrum of your lighting system may "look
> good," I would bet that it it scores low in PAR and will not be as
"healthy"
> as lighting with a more diverse spectrum. But I could be wrong because if
> those peaks fall in exactly the right places they may do quite well. I
tend
> to think the more you simulate sunlight (a broad well covered spectrum)
the
> better results you'll have. Adding a little blue to a "faux-sunlight-lit"
> tank to make it look better certainly has no ill effects as long as you
have
> all the other wavelengths present. Another concern is the outrageous cost
> of constructing an LED system, if it indeed works well with animals, for a
> much larger tank. Just my $.02 - john.
>
>snip<
Acroreef
>question is weather or not this technology reduces any spectrum peaks. I
>wish someone had a spectrometer to conduct some real testing..
I sent an inquisitive Email to Sanjay Joshi wondering if he might be interested
in doing this testing, but he seems to be away from his computer until July
6th.
Regards,
Robert
John
change with age just as in fluorescent lamps - I think it's a great idea
worth investigating but I'm skeptical because of cost and the longevity of
the lamps spectrum. The lamps may burn for for 40,000 hours but are they the
same at the end of their life as in the beginning? I would expect that the
LED's would in fact become more blue as they age because the phosphor will
not be making as much light allowing the blue LED to shine through. Who
knows? I have a Trek 7 LED flashlight that runs an array of 7 white Nichia
LED's (said to be the brightest on the market) and it is quite bright and
runs for 40 hours on three C-cells. This same light is water proof to 1000ft
and is used as standard issue to U.S. navy divers. I have not had it long
enough to tell if there has been any color change with age. The manufacturer
of the light claims that this will not happen. - john.
> From: "John R." <mp3b...@haNOSPAMwaii.rr.com>
> Newsgroups: rec.aquaria.marine.reefs
> Date: Wed, 28 Jun 2000 02:05:09 GMT
> Subject: Re: LED lighting info
>
> John, you are part right here. That's the way they did it in the old days.
> Here's a quote from a reputable LED dealer......
>
>> quote<
> White Light
> When light from all parts of the visible spectrum overlap one another, the
> additive mixture of colors appears white. However, the eye does not require
> a mixture of all the colors of the spectrum to perceive white light. Primary
> colors from the upper, middle, and lower parts of the spectrum (red, green,
> and blue), when combined, appear white. To achieve this combination with
> LEDs requires a sophisticated electro-optical design to control the blend
> and diffusion of colors. Variations in LED color and intensity further
> complicate this process.
>
> Presently it is possible to produce white light with a single LED using a
> phosphor layer (Yttrium Aluminum Garnet) on the surface of a blue (Gallium
> Nitride) chip. Although this technology produces various hues, white LEDs
> may be appropriate to illuminate opaque lenses or backlight legends.
> However, using colored LEDs to illuminate similarly colored lenses produces
> better visibility and overall appearance.
> http://www.theledlight.com/technical.html
>> end quote<
>
> The newer white LEDs do not use the "tri-chip" design as you mention.
> Instead, they use a typical blue LED with a special phosphor layer on a
> single chip. One problem is that depending on how thick and pure the
> phosphor layer will dictate a slight color shift in the white LED. Now, the
> question is weather or not this technology reduces any spectrum peaks. I
> wish someone had a spectrometer to conduct some real testing..
>
Tim Houghton
watts of MH. Isn't that like comparing 55 watts of PC with 55 watts of
MH - surely a meaningless comparison, unless you're interested in how
much your electricity bill will be. Why do we care about watts when we
have lumens/candela/colour temp etc to compare?
Tim
James Ronald
light bulb.
http://www.theledlight.com/led-assemblies.html
Cool, a $210.00 40 watt light bulb that will last for 100,000 hours.
So I guess I would need like 30 of the 36 White LED assemblies w/reflectors.
That's $6300.00!!!
James Ronald
Jimmy Chen <jimmy...@mindspring.com> wrote in message
news:8jbq0b$tf2$1...@slb7.atl.mindspring.net...
> The Specs:
>
> White LED Specifications: Type T-1 3/4, 5mm diameter; DC forward voltage
of
> 3.5 to 4.0V; 5600mcd (millicandela) @ 20mA; 120mW power dissipation;
Approx.
> 15 to 16 Lumens/watt; Estimate of 6000 to 6500 Kelvin; 85 CRI.
>
> > If my calculations are correct, if you filled the entire hood of a 48" x
> > 18" (70-90 gallon) tank with a similar setup to what you currently have,
it
> > would consume only about 250 watts. If the LEDS were, say, 4 times as
> > efficient as a metal halide, then this would be equivalent to 1000 watts
of
> > MH lighting.
>
> The math: ;-)
>
> So, 36 White LEDs approx. 360 milliamps (4.32 watts) and outputs 540W.
Based
> on your numbers, 576 will consume ~70W, but produces ~8640W. A little too
> sunny, I'll say. So we will only need around ~67 LEDs to produce ~1000Ws.
>
> > Cost wise, the LED's only would be approx: 576 x $5 = $2880 for blue
> > and 1024 x $2.80 = $2867 for white (also, more blue, less white may be
> > required to obtain the correct colour)
> >
> > So, for $5747 + reflectors + 12 DC power supply (s) etc. you would be
set.
> > Still a bit on the expensive side, it seems.
>
> The 36 LED (~500Ws) setup is quoted at $280, so it is still cheaper then a
> 400W MH fixture + bulb.
>
Jimmy Chen
would
> still be shining as intensly as the day they were born I bet.
The web site mentioned something about if it was under heat conditions, the
intensity will decrease to 1/2 in the same time frame of the LED's life.
jc
Boomer
Started to read this post....Nice job. You said somewhere LED's have come
along way. I believe a lot of the new stoplights on streets are also LED's
--
Boomer
WCW...@BresnanLink.Net
\ \ \ | / / /
-(@ @)-
------------oOO--(_)--OOo--------------------------
o
If you See Me Running You Better Catch-Up
"John R." <mp3b...@haNOSPAMwaii.rr.com> wrote in message
news:wsf55.10256$aB6....@typhoon.hawaii.rr.com...
: A month ago I posted some questions reguarding LED lighting. As a
followup,
: I am currently conducting some tests on these lights. I have posted some
: information at this website: http://www.mp3brick.com/marine/
: Just follow the "Lighting Project" link.
:
: John R.
:
:
Alan Funk
covered by reflector assemblies, maybe even the back of the tank, with solar
collectors (which usually deliver 12VDC) you could use the stray/reflected
light to generate voltage to run the lights/equipment. There really wouldn't
be a great deal of power generated but if the cost of the panels are cheap
enough then they power savings could add up. Even something as low as 5%
less power per year would add up nicely. It has been a long time since I
have read any articles on solar collectors and at that time they were
ridiculously expensive and not very effecient. Anyone know what improvements
have been made in the technology ?
Alan
"Better to keep your mouth shut and appear a fool, then open it and remove
all doubt !" Samuel Clemens
"James Ronald" <jjro...@mail.flashcom.net> wrote in message
news:slllkj...@corp.supernews.com...
> Chris, All,
>
> I did some reading and calculations and came up with the following...
>
> 5,600 millicandelas = 5.6 candelas
> 1 Candela = 12.56 Lumens
> 5.6 * 12.56 = 70.336 Lumens/LED
> 36 LEDs * 70.336 Lumens/LED = 2532 lumens / 36 LED Panel
>
> 4 Volts * 0.020Amps = .08 watts
> 70.336 Lumens / .08 Watts is 879.2 LPW (Lumens per Watt)
> An standard 100 watt Incandescent light is about 17.5 LPW
> A 250/DX MH Bulb is 90 LPW
> Compact FL in the 80 range LPW
> STD 40 FL in the 50 - 55 range LPW
>
> The LED's are about 10 times more efficient then MH assuming a ballast
> efficiency of 100%.
> I would guess the actual core type ballast efficiency to be in the 80%
> range.
>
> BTW, someone check my math it's getting kind of late.
>
> Later,
> James Ronald
>
>
>
> Chris Fanning <cfan...@jingoro.prevmed.sunysb.edu> wrote in message
> news:sll9c1...@news.supernews.com...
> > James Ronald <jjro...@mail.flashcom.net> wrote:
> > > The online catalog says that 36 White LEDs are about as bright a
40watt
> > > light bulb.
> > > http://www.theledlight.com/led-assemblies.html
> >
> > > Cool, a $210.00 40 watt light bulb that will last for 100,000 hours.
> > > So I guess I would need like 30 of the 36 White LED assemblies
> w/reflectors.
> > > That's $6300.00!!!
> >
> > Well, first of all, you wouldn't buy their outrageously priced
assemblies.
> > They're charging you $5.80 for each LED when you buy the assembly.
Plus,
> > you really don't want that assembly anyway.
> >
> > Complications, that's 30W of incandescent light. Flourescent is much
more
> > efficient, and MH is more efficient than that. I wish I knew the value,
> > but lets say MH is 4x more efficient than incandescent.
> >
> > To get my 500W of MH+VHO I would need 2000W of incandescent, or 1800
LEDs.
> > Going for likely less than $2.50 in that type of quantity we're talking
> > $4500. There's blue green ones that put out 4x the light of these that
> > we might like to throw in there. They could bring down the cost a bit.
> > Not sure about the green tint though.
> >
> > The nice thing is that it's all for 130W ... Wow.
> >
> > At my electric rates it would take only... 20 years to pay for itself.
:)
> > Of course, this does not include electric for running my chiller. Add
in
> > the cost of my chiller and the electric to run it and it would probably
> > only take 2-3 years to pay for itself. Scary huh?
> >
> > Chris
>
>
Chris Fanning
Acroreef
Not to mention the MH bulbs you'd go through in those 20 years. The LED's would
still be shining as intensly as the day they were born I bet.
Regards,
Robert
Jimmy Chen
> 3.5 to 4.0V; 5600mcd (millicandela) @ 20mA; 120mW power dissipation;
Approx.
> 15 to 16 Lumens/watt; Estimate of 6000 to 6500 Kelvin; 85 CRI.
>
> > If my calculations are correct, if you filled the entire hood of a 48" x
> > 18" (70-90 gallon) tank with a similar setup to what you currently have,
it
> > would consume only about 250 watts. If the LEDS were, say, 4 times as
> > efficient as a metal halide, then this would be equivalent to 1000 watts
of
> > MH lighting.
>
> The math: ;-)
Okay, NEW math ... Hope I am at least a little more on the bullseye this time.
;-)
Assumption: MH 250 ~ 20,000-30,000 lumens
So, a 36 LED will produce 540 lumens and consumes ~5W (for easy calc), so in
order to produce 30k lumens, we will need ~55 of these 36 LEDs for a total of
275W used, or ~2000 LEDs. In respect to the current MH's cost, which is ~500
for a nice 250W setup with all the goodies, these LEDs may not see to be such
a good deal if we factor in the initial cost of the equal LED setup at ~$4000
per 30,000 lumens of output.
jc
James Ronald
I did some reading and calculations and came up with the following...
5,600 millicandelas = 5.6 candelas
1 Candela = 12.56 Lumens
5.6 * 12.56 = 70.336 Lumens/LED
36 LEDs * 70.336 Lumens/LED = 2532 lumens / 36 LED Panel
4 Volts * 0.020Amps = .08 watts
70.336 Lumens / .08 Watts is 879.2 LPW (Lumens per Watt)
An standard 100 watt Incandescent light is about 17.5 LPW
A 250/DX MH Bulb is 90 LPW
Compact FL in the 80 range LPW
STD 40 FL in the 50 - 55 range LPW
The LED's are about 10 times more efficient then MH assuming a ballast
efficiency of 100%.
I would guess the actual core type ballast efficiency to be in the 80%
range.
BTW, someone check my math it's getting kind of late.
Later,
James Ronald
Chris Fanning <cfan...@jingoro.prevmed.sunysb.edu> wrote in message
news:sll9c1...@news.supernews.com...
> James Ronald <jjro...@mail.flashcom.net> wrote:
> > The online catalog says that 36 White LEDs are about as bright a 40watt
> > light bulb.
> > http://www.theledlight.com/led-assemblies.html
>
> > Cool, a $210.00 40 watt light bulb that will last for 100,000 hours.
> > So I guess I would need like 30 of the 36 White LED assemblies
w/reflectors.
> > That's $6300.00!!!
>
> Well, first of all, you wouldn't buy their outrageously priced assemblies.
> They're charging you $5.80 for each LED when you buy the assembly. Plus,
> you really don't want that assembly anyway.
>
> Complications, that's 30W of incandescent light. Flourescent is much more
> efficient, and MH is more efficient than that. I wish I knew the value,
> but lets say MH is 4x more efficient than incandescent.
>
> To get my 500W of MH+VHO I would need 2000W of incandescent, or 1800 LEDs.
> Going for likely less than $2.50 in that type of quantity we're talking
> $4500. There's blue green ones that put out 4x the light of these that
> we might like to throw in there. They could bring down the cost a bit.
> Not sure about the green tint though.
>
> The nice thing is that it's all for 130W ... Wow.
>
> At my electric rates it would take only... 20 years to pay for itself. :)
> Of course, this does not include electric for running my chiller. Add in
> the cost of my chiller and the electric to run it and it would probably
> only take 2-3 years to pay for itself. Scary huh?
>
> Chris
John R.
posting a lot of my research notes on my website. I just put up some new
stuff regarding this "# of LED required" question. Take another
look.....(original post)
John R.
Mike Noren
: The Specs:
:
: White LED Specifications: Type T-1 3/4, 5mm diameter; DC forward voltage of
: 3.5 to 4.0V; 5600mcd (millicandela) @ 20mA; 120mW power dissipation; Approx.
: 15 to 16 Lumens/watt; Estimate of 6000 to 6500 Kelvin; 85 CRI.
Hm... They're efficient, but the total output is important here.
A typical 250W MH gives out about 19000 lumen; according to the above
the leds give out 16 lumen/watt, so we can calculate that one'd need
19000/16 = 1187 W of LED lighting to equal the light output of a 250W
metal halide....
Each LED apparently uses 4 volts at 0.020 amps = 0.08 W, meaning that
one'd need 1187 / 0.08 = 14800 LED's to equal the light put out by a
single 250W MH.
Serious rain on the parade. Or have I got the numbers wrong here?
: > If my calculations are correct, if you filled the entire hood of a 48" x
: > 18" (70-90 gallon) tank with a similar setup to what you currently have, it
: > would consume only about 250 watts. If the LEDS were, say, 4 times as
: > efficient as a metal halide, then this would be equivalent to 1000 watts of
: > MH lighting.
:
: The math: ;-)
:
: So, 36 White LEDs approx. 360 milliamps (4.32 watts) and outputs 540W.
According to the above specs 4.32 watts of LED light has an output of
69 lumen, equalling the output of ONE WATT of metal halide lighting...
(Besides, if each LED draws 360mA at 4V, that's 1.44W, not 4.32W..)
: The 36 LED (~500Ws) setup is quoted at $280, so it is still cheaper then a
: 400W MH fixture + bulb.
Surely this can't be right?
I don't think there exists LEDs that bright - each LED would have to
equal a 60W incandescent bulb in brightness...
When I read electronics it was said that LED's were useless for
general purpose lighting despite their extreme efficiency due to the
fact that their total brightness was so low.. Things may have changed
since then, but...
Michael Norén, Doctoral student,
Stockholm University and Tel: Int +46 (0)8 5195 4236
Swedish Museum of Natural History, Fax: Int +46 (0)8 5195 4125
"Nihil umquam facile"
Kevin M
news:sll9c1...@news.supernews.com...
>
You can't compare LED's to other types of lighting, by wattage, because of
their MASSIVE efficiency. They are orders of magnitude more efficient than
even the BEST electronic MH ballast.
--
Kev
http://www.flash.net/~kevmit/main.htm
reply address spoofed to dot com, make it net
nosmatt
technology as it is far more efficient than the standard
halogen/incandecent bulb arrangement they currently have. in fact, every
major automaker has concept cars that use LEDs for their lighting
systems. if this trend continues, i think well see the price on them
drop substantially.
matt
Chris Fanning
> "Chris Fanning" <cfan...@jingoro.prevmed.sunysb.edu> wrote in message
> news:sll9c1...@news.supernews.com...
>>
>> The nice thing is that it's all for 130W ... Wow.
> You can't compare LED's to other types of lighting, by wattage, because of
> their MASSIVE efficiency. They are orders of magnitude more efficient than
> even the BEST electronic MH ballast.
Maybe an order of magnitude. I was going by their approximation
(theledlight) of 40W incandescent being equivalent to 36 of their white
LEDs. I extrapolated from there ... If they are an order of magnitude
more efficient than what I calculated then we'd only need 180 of them.
I *seriously* doubt that. No way we're getting by with 13W of light.
Have to do the viewing angle vs. cd vs. lumens math to give more than an
approximation.
Chris
Acroreef
>major automaker has concept cars that use LEDs for their lighting
>systems. if this trend continues, i think well see the price on them
>drop substantially.
And perhaps the technology will get better. ie; brighter, larger,
whatever......
Regards,
Robert
Mike Noren
: > You can't compare LED's to other types of lighting, by wattage, because of
: > their MASSIVE efficiency. They are orders of magnitude more efficient than
: > even the BEST electronic MH ballast.
:
: Maybe an order of magnitude. I was going by their approximation
: (theledlight) of 40W incandescent being equivalent to 36 of their white
: LEDs. I extrapolated from there ... If they are an order of magnitude
: more efficient than what I calculated then we'd only need 180 of them.
: I *seriously* doubt that. No way we're getting by with 13W of light.
:
: Have to do the viewing angle vs. cd vs. lumens math to give more than an
: approximation.
But... A 40W incandescent gives out just 300 to 450 lumens. A 250W
metal halide gives out _19000_ lumen.
From that it follows one'd need (19000/450)*36 = 1520 LED's of this
brightness to equal one 250W MH with regards to brightness.
: Chris
Boomer
--
Boomer
WCW...@BresnanLink.Net
\ \ \ | / / /
-(@ @)-
------------oOO--(_)---------------------------
o
If you See Me Running You Better Catch-Up
"nosmatt" <nos...@socal.rr.com> wrote in message
news:395BB76C...@socal.rr.com...
: and FORD motor company is designing tailamps and headlamps that use LED
: technology as it is far more efficient than the standard
: halogen/incandecent bulb arrangement they currently have. in fact, every
: major automaker has concept cars that use LEDs for their lighting
: systems. if this trend continues, i think well see the price on them
: drop substantially.
: matt
:
Alan Funk
Old Reconditioned Dodge
Alan
Escort driver (head hung in shame)
"Boomer" <wcw...@bresnanlink.net> wrote in message
news:slp85nd...@corp.supernews.com...
Acroreef
>Old Reconditioned Dodge
F*&%ed Over, Ready to Die
Ching Kong
Well, I don't know where did JC found the lm/W figure, but my calculation
based on www.theledlight.com/catalog.html agreed with J. Ronald's 875_lm/W
(8 cd is around 100 lm, as I remembered)
I think the wrong lm/W ratio lead to calculation that "proofs" that LED is
less efficient than MH. Please check the number before using them.
One complication is that LEDs have a narrow view angle. This may decrease
the lm/W ration... we need a lighting engineering here. Can anybody tell
me how much is one einstein? TIA.
bye -ck-
nosmatt
everybody knows it First On Race Day!
DUH!
Acroreef
Robert
Jimmy Chen
> based on www.theledlight.com/catalog.html agreed with J. Ronald's 875_lm/W
Hehee, which figures?
The spec for the LED is from http://www.theledlight.com/led-assemblies.html
The first set of math is based on my screwed up calculation comparing lm and
W, the 2nd set of math is based on a 250 MH light, using the max lm I found,
which was ~30K lm.
jc
Ching Kong
70_lm, which I immediately realize that I will need more than 400 LEDs to
get 30_klm. Now that's expensive! However, the confusing part is the
wattage. It says 120_mW power dissipation per lamp, but I always think
that to get watt, you multiply current with voltage. So 20_mA at 4_V
should be 80_mW, which is not exact match with previous number... but the
specs at http://www.theledlight.com/led-specs.html say that the "DC
Forward current is 30_mA" which matches the 120_mW.
At any rate, LED is way too expensive for hi-intensity application for me.
Maybe I just wait until the efficiency goes up, or the price goes down.
bye -ck-
Mike Noren
: hi.
: Well, I don't know where did JC found the lm/W figure, but my calculation
: based on www.theledlight.com/catalog.html agreed with J. Ronald's 875_lm/W
Just one more thing - Osram has a concept/experiment/whatever where
they used leds for lighting a room.
Check the article "The first LED room lighting" at:
http://www.osram.com/frameset.htm?hauptnavi1.htm-sekund1.htm%3F5._http%3A//www.osram.com/chap_1/1_5/index.htm
The text says that the luminosity of each individual LED is 1-2 lumen
with a power consumption of 0.1W.
That equates to 10-20 lumen/watt.
I looked at the site you mention, and they say that light output is 15
lumen/watt.
: bye -ck-
:
Chris Fanning
> Replying to Chris Fanning <cfan...@jingoro.prevmed.sunysb.edu> :
> : > You can't compare LED's to other types of lighting, by wattage, because of
> : > their MASSIVE efficiency. They are orders of magnitude more efficient than
> : > even the BEST electronic MH ballast.
> :
> : Maybe an order of magnitude. I was going by their approximation
> : (theledlight) of 40W incandescent being equivalent to 36 of their white
> : LEDs. I extrapolated from there ... If they are an order of magnitude
> : more efficient than what I calculated then we'd only need 180 of them.
> : I *seriously* doubt that. No way we're getting by with 13W of light.
> :
> : Have to do the viewing angle vs. cd vs. lumens math to give more than an
> : approximation.
> But... A 40W incandescent gives out just 300 to 450 lumens. A 250W
> metal halide gives out _19000_ lumen.
> From that it follows one'd need (19000/450)*36 = 1520 LED's of this
> brightness to equal one 250W MH with regards to brightness.
Yeah, I fudged this number based on PC lighting. The bulbs they sell
in the stores as 40, 60W equivalent with 17W power consumption, etc.
From that I estimated flourescent to be about 4X more efficient than
incandescent and gave MH slightly more of an efficiency boost.
I didn't use indcandescent lighting efficiencies to calculate my estimate.
I don't mean to sound nasty, but I wish people would have read that
post more carefully before pointing out what looked like errors.
Chris
Jim Ivey
Just one quick point. The conversion from candela to lumens assumes
that the light coming from the LED is perfectly distributed about a
sphere. A candela is the light at a given 3D angle from a point on the
surface of the LED. Since the light distribution from the LED is not
spherical, then the correct amount of lumens would be less. If the
light emitted is in an hemispherical pattern the the lumens would equal
to 2Pi*candelas. The figure would be half the value using the 4Pi
converson factor.
The light measurement hand book at http://www.intl-light.com/handbook/>
can be a great resource for these calculations.
BTW, if anyone would like to test these lamps we have a number of optics
sensors (radiance sensors, Licor 1800's, ect) and I could measure the
light output. I don't have the time or funds to purchase LEDS and test
them, but I could run some tests on any sent to me and then return
them. Drop me a line and don't forget to delete "eatspam" from my
e-mail address.
Jim
--
Jim Ivey
University of South Florida
Department of Marine Science
Ocean Optics Group
E-mail address needs correction before replying.
Jim Ivey
Just another point. I remembered some discussion on LED efficiency in
the sci.engr.lighting news group. Go to
http://www.deja.com/home_ps.shtml and search for LED efficiency. Don
Klipstein <http://www.misty.com/~don/light.html> who is kind of a
lighting guru posted that the highest efficiency he had ever seen from
an LED was 20 lumens/watt. Still fantastic since lumens are weighted to
the spectrum visible by the human eye and LEDs have a very narrow color
spectrum of output. That means that a bright blue LED probably would
appear brighter to the corals at wavelengths that we don't see as well.
If you find light output measured in watts or Quanta it would probably
be a better comparison than Lumens for an LED.
Jim
Mike Noren
: Yeah, I fudged this number based on PC lighting. The bulbs they sell
: in the stores as 40, 60W equivalent with 17W power consumption, etc.
: From that I estimated flourescent to be about 4X more efficient than
: incandescent and gave MH slightly more of an efficiency boost.
:
: I didn't use indcandescent lighting efficiencies to calculate my estimate.
:
: I don't mean to sound nasty, but I wish people would have read that
: post more carefully before pointing out what looked like errors.
No need to estimate anything.
According to Osram and that LED place cited in this thread a LED has
an output of 1-2 lumen, depending on colour, with a power consumption
of 0.1W. That equates to 10 - 20 lumen/watt.
A 250W metal halide has an output of approximately 20000 lumen, or 80
lumen/watt.
A 55W Power Compact has an output of about 3000-3800 lumen, or 55 to
70 lumen/watt. New PC's with 3 tubes released this autumn will improve
this further, allowing PC's to approach MH efficiency.
A 40W incandescent has an output of about 450 lumen, or 11.5
lumen/watt.
Meaning that LEDs are more efficient, watt for watt, than
incandescents, but *far* less efficient than metal halides or
VHO/Power Compacts.
And at the extreme top-of-the-line we have the ridiculously efficient
Solar1000 sulphur bulb, which outputs 140000 lumen @ 1450W, or a
whopping 96.5 lumen/watt.
Although the details change with colour temperature, the math is quite
straight-forward, and the conclusion is that LED lighting isn't
presently a viable alternative for reef aquaria due to price and
efficiency.
It seems my electronics teachers were right: LEDs remain useless for
general purpose lighting - at least until prices drop drastically.
They are simply not competitive with more conventional forms of
lighting.
It also seems like the chief advantage of LED's is their small
footprint and extreme longevity, not efficiency.
: Chris