Re: Two questions - LED Lighting and archive for this group.

On Fri, 23 Sep 2005 17:34:31 GMT, Clive Mitchell
<bigclive1@xxxxxxxxxxxx> wrote:

>Remember and allow for the poor power factor and rapid lumen fall-off
>from CFL's. I bet the intensity of white LEDs doesn't drop at the same
>rate as a CFL.

OK - some real data. First, I can find only a very limited
amount of LED lumen maintenance data. Actually only one
published experiment, and one forward-looking statement from
Lumileds.

A few years ago Narandren and Deng of the LRC ran a lumen
maintenance experiment on 5mm white LEDs and unspecified
white "high power" LEDs. Their data was presented at an IES
conference and is available at www.lrc.rpi.edu, but the same
data is also published in a Lumileds Application Brief AB07:
http://www.lumileds.com/pdfs/AB07.PDF

Figure 3 of AB07 shows the LRC data. The 5mm devices dropped
to 40% of initial output after 9000 hours while the "high
power" LEDs had dropped to only 90% of initial output.
Figures 4 and 5 of this same Application Brief show data
obtained by Lumileds at case temperatures of 25C and 85C.
The 85 C data shows light output decreasing to approximately
93% after 2000 hours of operation. This is all rather old
data, but the 5mm data is probably typical of many of the
inexpensive LEDs being sold today.

I cannot find any additional LED lumen maintenance data,
though I must admit I have not done anything other than a
short search. Note that it would take almost 6 years to
accumulate 50,000 hours of life test data and none of the
best LEDs have been around that long. However, there is an
interesting forward-looking statement on page 18 of the
Lumileds DS23 data *** for the Luxeon Star devices.

http://www.lumileds.com/pdfs/DS23.pdf

"Lumileds projects that Luxeon products will deliver on
average 70% lumen maintenance at 50,000 hours of operation.
This performance is based on independent test data, Lumileds
historical data from tests run on similar material systems,
and internal Luxeon reliability testing. This projection is
based on constant current 350 mA operation with junction
temperature maintained at or below 90°C. Observation of
design limits included in this data *** is required in
order to achieve this projected lumen maintenance."

Note that Lumileds does not make this same forward-looking
statement for their Luxeon V or SuperFlux LEDs.

Now, lets take a look at some CFL data. Mainline CFL
manufacturers give both the Initial Lumen value, taken at
100 hours, and a Mean Lumen value, taken at 40% of rated
life. However, lumen maintenance curves for CFLs are a bit
harder to find - but there are a few. GE published the lumen
maintenance curve for their 42-watt spiral CFL. See:

http://www.gelighting.com/na/business_lighting/education_resources/literature_library/sell_sheets/downloads/cfl/28306_42w_spiral.pdf

or http://tinyurl.com/a57fh if the long link is broken.

This data shows that the output of this lamp has dropped to
70% of its initial value after 10,000 hours, which gives it
distinctly worse performance than Lumileds projects for
their white LEDs. But this is a highly loaded CFL (high
power per unit area) which degrades the phosphor faster than
a lamp with lower loading and the spiral shape degrades
lumen maintenance.

Another CFL that has a published maintenance curve is the GE
32-watt triple BIAX. See:

http://www.gelighting.com/na/business_lighting/education_resources/literature_library/sell_sheets/downloads/cfl/biax_32.pdf

or http://tinyurl.com/8g9z3 if the long link is broken.

This is a 4-pin CFL, but the same wire lamp would be used to
make a screw-base lamp, and the same technology is used to
make lower power screw base GE BIAX lamps. The curve on page
5 of this document shows a lumen maintenance of about 83% at
12,000 hours. which is quite good. GE also provides the
following data:

Hours Maintenance
500 0.96
1000 0.95
2000 0.90
4000 0.85

If this data is plotted in Excel (along with the 0.83 point
at 12,000 hours) and then a exponential trendline is added
in Excel I get the following equation for the lumen
maintenance:

M = 1.2365 * T^-0.0423

This gives 78% at 50,000 hours, but the extrapolation to
50,000 is obviously suspect and the results are highly
sensitive to the data point at 12,000 hours, which looks a
bit too good based on my own experience.

The CFL obviously does not last for 50,000 hours (because
its electrodes will die before than) and I'm not trying to
imply it will. However, since I only have the 50,000 hour
projection from Lumileds and not the whole curve, this seems
to be the best way to compare the efficacy of this CFL to
the Luxeon Star LED.

The best I can say now is that good CFLs seem to have lumen
maintenance that is better than any published LED data I
have seen, and probably as good as Lumileds projections for
their Luxeon Star devices.

In response to those who will post far worse lumen
maintenance data on cheap CFLs, I agree that there is no
limit to how poor a device can be made, and that includes
CFL and LEDs. That is why I am trying to compare only high
quality products.

Also, the LED data above applies only to white LEDs using
phosphor. LEDs made with three or more "monochromatic" LEDs
without the use of any phosphor should have much better long
term lumen maintenance - but the long term color balance
problem must be solved.

Finally, when comparing light source efficacy, we should be
discussing Mean Efficacy, not Initial Efficacy. Based on the
lumen maintenance curve for older fluorescent lamps, Mean
Lumens has been defined as the output at 40% of rated life.
The 40% point no longer represents actual Mean Lumens for
some fluorescent lamps, but continues to be the standard. As
stated above, we don't have lumen maintenance curves or Mean
Lumen data for LEDs, but if LED life is going to be defined
as the point where their light output has dropped to 70% of
the initial value, then even if LEDs have the same Initial
Efficacy as fluorescent lamps, (not likely any time soon in
my opinion) they will have lower Mean Efficacy since good
fluorescent lamps have end-of-life maintenance of over 80%
at the present time, and some are over 85% at end of life.

--
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.

.

Loading