Cost-benefit of LEDs ?

[gentlegreen]

Am I right in supposing that, with LEDs now competing with small CFLs in
terms of lumens per watt, the key disadvantage is the high initial cost /
inconvenient form factor / spectrum / light distribution and the key
advantage their long life ?

I thought I might start playing with them for domestic lighting until I
calculated that I would need 150 LEDs at 20p = £30. to replace a 20 watt
CFL.

So the LEDs would need to last 30 times as long as a CFL from Aldi .....

(White ultrabright LEDs 3.3 volts, 20mA, 8 lumens 0.066 watts = 121 lumens
per watt.
CFL 20 watts 1200 lumens = 60 lumens per watt.)

 

[Victor Roberts]

Am I right in supposing that, with LEDs now competing with small CFLs in
terms of lumens per watt, the key disadvantage is the high initial cost /
inconvenient form factor / spectrum / light distribution and the key
advantage their long life ?

I thought I might start playing with them for domestic lighting until I
calculated that I would need 150 LEDs at 20p = £30. to replace a 20 watt
CFL.

So the LEDs would need to last 30 times as long as a CFL from Aldi .....

(White ultrabright LEDs 3.3 volts, 20mA, 8 lumens 0.066 watts = 121 lumens
per watt.
CFL 20 watts 1200 lumens = 60 lumens per watt.)

Is your LED data typical, or minimum guaranteed? Is it at a
junction temperature of 25C or the junction temperature you
expect to have in an application? What is the lumen output
after 10,000 hours of operation?

--
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]

WOW!!

Please please please -where can you get 8 lumen 20mA white leds - even
at 0.35 US !!

Ebay apparently. Primarily because Chinese sellers feel the need to
beat their competitors intensity figures, so use a little artistic
licence. I've got some 50 billion candela* 5mm whites if anyone's
interested. They run at 20mA and are brighter than the sun itself.

I'm selling them in quantities of 10 at just 1 penny each (plus 50 quid
shipping).

(*When struck directly by lightning.)

 

[Don Klipstein]

You might want to check your lumen figure there. For the brightest 5
mm packages I'm aware of, they might have 0.8 lumens at 20 ma not 8
lumens. Which is only 12 lumens per watt, using your calculation,
which would be more typical.

The high power LED packages do better, but even the best ones of those
don't achieve 60 lumens per watt under normal operation unless you have
a really really good heat sink and reasonable ambient temperature.

I am under the impression that Nichia's NSPW-xxxCS ones produce around
2-3 lumens apiece at 20 mA.

- Don Klipstein ([email protected])

 

[Victor Roberts]

You might want to check your lumen figure there. For the brightest 5
mm packages I'm aware of, they might have 0.8 lumens at 20 ma not 8
lumens. Which is only 12 lumens per watt, using your calculation,
which would be more typical.

The high power LED packages do better, but even the best ones of those
don't achieve 60 lumens per watt under normal operation unless you have
a really really good heat sink and reasonable ambient temperature.

Small LED chips are actually more efficient than "power"
LEDs. Back in June Cree issued a press release announcing a
131 lm/W LED, with the performance validated by NIST. I
think we discussed that device in this group. When I
contacted them they said it was a small device intended for
use in cell phones and such. They said that their higher
power LEDs were not yet that efficient.


--
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.

 

[gentlegreen]

Is your LED data typical, or minimum guaranteed? Is it at a
junction temperature of 25C or the junction temperature you
expect to have in an application? What is the lumen output
after 10,000 hours of operation?

Sorry I didn't get back sooner.

It seems my gut reaction was right about LED manufacturers' claims then ;-)
You would think that claims relating to something as vital as energy
efficiency would be subject to legal penalties ....

------------------------------

I've been buying LEDs from this supplier :-

http://cgi.ebay.co.uk/WHITE-Ultra-B...5915932QQihZ020QQcategoryZ66954QQcmdZViewItem

I'm very pleased with the results when used as bicycle lights :-

I'm using a 7x7 grid of red LEDs on the back and 24 white ones on the
front.
(I used to have 12 volt 21 watt incandescents front and back and a huge
battery.)

I also have a 9 watt minispiral fluorescent mounted with a minimal amount of
focusing which I reckon gives a better bang for the buck at that power
level, and I'm considering switching from a 20 watt MR16 to a 2 watt LED
version or two for my "dog poo / broken glass detector".

 

[Victor Roberts]

Sorry I didn't get back sooner.

It seems my gut reaction was right about LED manufacturers' claims then ;-)
You would think that claims relating to something as vital as energy
efficiency would be subject to legal penalties ....

Light output and power input claims of lamps are subject to
Federal Trade Commission in the US, but this has not
migrated won to LEDs, which are still considered electronic
components.

However, without knowing the details of your case, most
manufacturers don't lie, they just fail to give all the
information. If they give an accurate value for "typical"
lumen output at a junction temperature of 25C and never show
minimum lumen output or the lumen output vs. junction
temperature curve and never give lumen maintenance data,
have they lied to the customer?

I've been buying LEDs from this supplier :-

http://cgi.ebay.co.uk/WHITE-Ultra-B...5915932QQihZ020QQcategoryZ66954QQcmdZViewItem

I looked at this page and can't find any lumen output claim.
If you converted their brightness data (mcd) to flux
(lumens) please describe your calculations, since I don't
see enough data there to perform that calculations.

[snip]

--
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.

 

[gentlegreen]

Light output and power input claims of lamps are subject to
Federal Trade Commission in the US, but this has not
migrated won to LEDs, which are still considered electronic
components.

However, without knowing the details of your case, most
manufacturers don't lie, they just fail to give all the
information. If they give an accurate value for "typical"
lumen output at a junction temperature of 25C and never show
minimum lumen output or the lumen output vs. junction
temperature curve and never give lumen maintenance data,
have they lied to the customer?


5915932QQihZ020QQcategoryZ66954QQcmdZViewItem

I looked at this page and can't find any lumen output claim.
If you converted their brightness data (mcd) to flux
(lumens) please describe your calculations, since I don't
see enough data there to perform that calculations.

embarassingly you have exposed my ignorance :-(

I just googled for a convertor and it turns out they're only 1.191 lumens
for 0.0666 watts - so 28.3 lumens per watt ...

So I would need a vast number (soldered into a sphere perhaps ?) to replace
a CFL !

http://led.linear1.org/lumen.wiz

 

[Victor Roberts]

embarassingly you have exposed my ignorance :-(

I just googled for a convertor and it turns out they're only 1.191 lumens
for 0.0666 watts - so 28.3 lumens per watt ...

So I would need a vast number (soldered into a sphere perhaps ?) to replace
a CFL !

http://led.linear1.org/lumen.wiz

Probably even lower, since the converter you used assumes it
is being fed the average intensity while your supplier seems
to list the peak.

--
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.