LED aging...

[Don Y]

<https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing>

While the lamps are spec\'d for ~14 years of continuous use (at 25C... \"yeah,
right!\"), there\'s nothing that indicates how the output changes, over time
(i.e., use).

The color of the individual emitters shouldn\'t change -- just their optical
power output, right?

Can one assume that equal use will degrade that output equally? I.e., the
color produced by *combinations* of emitters won\'t shift, over time (?)

If it does, is there any practical way to compensate for this (e.g., by
tracking the amount of \"use\" each emitter has seen)? Or, is it too
component specific?

 

[John Larkin]

On Tue, 18 Aug 2020 15:00:15 -0700, Don Y
<blockedofcourse@foo.invalid> wrote:

https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C... \"yeah,
right!\"), there\'s nothing that indicates how the output changes, over time
(i.e., use).

The color of the individual emitters shouldn\'t change -- just their optical
power output, right?

Different color LEDs, with different chemistry, will probably degrade
at different rates.

Higher currents and higher temperatures make LEDs degrade faster.

 

[Don Y]

On 8/18/2020 3:10 PM, John Larkin wrote:

On Tue, 18 Aug 2020 15:00:15 -0700, Don Y
blockedofcourse@foo.invalid> wrote:

https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C... \"yeah,
right!\"), there\'s nothing that indicates how the output changes, over time
(i.e., use).

The color of the individual emitters shouldn\'t change -- just their optical
power output, right?

Different color LEDs, with different chemistry, will probably degrade
at different rates.

Yes, but will that be related to operating current, duration, duration x
current, etc. I.e., if you can watch (control) how each is used, can
you make reasonable predictions as to what to expect regarding their
output (without having to directly *sense* that)?

Higher currents and higher temperatures make LEDs degrade faster.

 

[Jon Elson]

Don Y wrote:

https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C...
\"yeah, right!\"), there\'s nothing that indicates how the output changes,
over time (i.e., use).

I did my first retrofit in our laundry/pantry in 2013. I then did all the
dual fluorescent 48\" fixtures in the kitchen from 2014 - 2019. I made up my
own system, using long strips of copper-clad PC board material to mount SMT
LEDs and act as the interconnect as well as the heat sink. I made my own
constant current supply for the first one, and commercial LED lighting
supplies for the others. They are all working quite nicely, and I don\'t
think they have dimmed at all. We use those kitchen lights a LOT,
especially in the winter, too.

Jon

 

[Don Y]

On 8/18/2020 4:16 PM, Jon Elson wrote:

Don Y wrote:

https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C...
\"yeah, right!\"), there\'s nothing that indicates how the output changes,
over time (i.e., use).

I did my first retrofit in our laundry/pantry in 2013. I then did all the
dual fluorescent 48\" fixtures in the kitchen from 2014 - 2019. I made up my
own system, using long strips of copper-clad PC board material to mount SMT
LEDs and act as the interconnect as well as the heat sink. I made my own
constant current supply for the first one, and commercial LED lighting
supplies for the others. They are all working quite nicely, and I don\'t
think they have dimmed at all. We use those kitchen lights a LOT,
especially in the winter, too.

But you probably are only relying on your day-to-day recollection of
their brightness -- do you have a light that you have NEVER/rarely powered
to compare against, as a reference?

A neighbor installed some 100W LED fixtures in his front yard. (note that this
is roughly comparable to the fixture I cited -- though his are \"monochrome\")
As I\'ve been wanting to replace the 1500W of Halogens that we have in our BACK
yard (which see a lot of use and, thus, \"operating cost\"), I asked him where
he\'d purchased them, cost, etc. (cheap chinese products from eBay).

As I\'d not been able to find a package that I like (no exposed cable to
deteriorate from prolonged exposure to heat and UV), I\'ve put off purchasing
and installing any. Easier to just pay the electric bill than to spend
hours chasing down suitable fixtures!

Last week, I was out, at night, and noticed his lights seemed considerably
dimmer. And, I\'m 103% sure he doesn\'t have intensity control on those! :>

So, I asked him about it, the other day. He frowned and nodded that they
are, in fact, noticeably dimmer. He said, when he suspected it, he pulled
the \"spare\" (never rely on chinese replacement products to be available
\"long term\") that he had purchased and \"plugged it in\" (fixture has a long
pigtail). Knowing that they\'ve dimmed, he really has little recourse -- buy
yet another inexpensive chinese product and hope??

 

[Jon Elson]

Don Y wrote:

But you probably are only relying on your day-to-day recollection of
their brightness -- do you have a light that you have NEVER/rarely powered
to compare against, as a reference?

No, I did the first one in early 2013, and did the last one in Dec. 2019,
using the same LED part #. If they had seriously dimmed, I would have
noticed the difference. I can\'t really tell any difference between the
first and last unit built. Now, because these LED strips are completely
exposed above the diffuser, they run pretty cool -- about 45 C at the
copper-clad boards. Many packaged lamps have to be seriously sealed because
they are not isolated from mains voltage. That causes a huge temperature
rise. My retrofits are about 44 inches long by 2\" wide, that spreads out
the heat.

Jon

 

[Don Y]

On 8/18/2020 8:01 PM, Jon Elson wrote:

Don Y wrote:

But you probably are only relying on your day-to-day recollection of
their brightness -- do you have a light that you have NEVER/rarely powered
to compare against, as a reference?
No, I did the first one in early 2013, and did the last one in Dec. 2019,
using the same LED part #. If they had seriously dimmed, I would have
noticed the difference. I can\'t really tell any difference between the
first and last unit built.

OK. I\'ll attribute the neighbor\'s \"problem\" to \"Made In Chine\".

Now, because these LED strips are completely
exposed above the diffuser, they run pretty cool -- about 45 C at the
copper-clad boards. Many packaged lamps have to be seriously sealed because
they are not isolated from mains voltage. That causes a huge temperature
rise. My retrofits are about 44 inches long by 2\" wide, that spreads out
the heat.

These are 100W units in sealed (waterproof) enclosures. They are intended
for outdoor use (see slideshows):

<https://www.lumenpulse.com/projects/182/sasktel-center>

<https://www.lumenpulse.com/projects/174/grand-river-penmans-dam-portage>

My concern is that if one \"element color\" sees more use than another -- and,
thus has more opportunity to degrade -- that the color(s) formed in concert
with other elements will significantly change.

In particular, if mainly showing \"green\" (and, thus, aging the green emitters)
then the light output in contrast with an eventual *red* (signalling an
anomaly) could be degraded in such a way that the red and green states are
less distinguishable (by an individual with red-green colorblindness).

Repeat for other -- less common -- types of colorblindness.

[If you can\'t separate indicators so that the position of the indication
conveys meaning (regardless of color) -- much like a traffic light -- then you
need to rely on other \"encodings\" to convey the color information.

\"Blinking\" is a poor way of differentiating state in a color-independent
manner so you want to rely on other aspects of the presentation -- imagine
if every (US) traffic light was represented by a single indicator that used
color and blink rates to convey just those three states (red/yellow/green)
instead of standardized \"positions\"!]

 

[Martin Brown]

On 18/08/2020 23:00, Don Y wrote:

https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C...
\"yeah,
right!\"), there\'s nothing that indicates how the output changes, over time
(i.e., use).

The color of the individual emitters shouldn\'t change -- just their optical
power output, right?

The lattice might degrade a bit with time so the FWHM of the wavelengths
output might slowly broaden as well. Not noticeable to the human eye...

Can one assume that equal use will degrade that output equally?  I.e., the
color produced by *combinations* of emitters won\'t shift, over time (?)

The yellow phosphors on the White emitters will be the thing that
degrades first. Though some systems now use phosphors for each of RGBW.

If it does, is there any practical way to compensate for this (e.g., by
tracking the amount of \"use\" each emitter has seen)?  Or, is it too
component specific?

Not worth the effort in such lamps since they are generally being used
for putting swathes of slowly changing colours onto buildings. And if
they were in a theatre they would be on a control console that could if
necessary be recalibrated as they faded to correct systematic errors.

I\'d expect the blue emitter to be the most fragile but not by much.

--
Regards,
Martin Brown

 

[Don Y]

On 8/19/2020 2:42 AM, Martin Brown wrote:

On 18/08/2020 23:00, Don Y wrote:
https://www.lumenpulse.com/products/1385/lumenbeam-grande-color-changing

While the lamps are spec\'d for ~14 years of continuous use (at 25C... \"yeah,
right!\"), there\'s nothing that indicates how the output changes, over time
(i.e., use).

The color of the individual emitters shouldn\'t change -- just their optical
power output, right?

The lattice might degrade a bit with time so the FWHM of the wavelengths output
might slowly broaden as well. Not noticeable to the human eye...

Can one assume that equal use will degrade that output equally? I.e., the
color produced by *combinations* of emitters won\'t shift, over time (?)

The yellow phosphors on the White emitters will be the thing that degrades
first. Though some systems now use phosphors for each of RGBW.

In my case, I only use the white emitters for general/supplemental lighting.
I *could* use other fixtures for that but it just complicates the control
and installation.

If it does, is there any practical way to compensate for this (e.g., by
tracking the amount of \"use\" each emitter has seen)? Or, is it too
component specific?

Not worth the effort in such lamps since they are generally being used for
putting swathes of slowly changing colours onto buildings.

Yes. Unfortunately, there are no *other* products that are designed to throw
a LOT of \"colorable\" light (to illuminate large *spaces*)

And if they were in
a theatre they would be on a control console that could if necessary be
recalibrated as they faded to correct systematic errors.

In a theater, you really want precise control over light color (witness
the variety of \"gels\" that have traditionally been used). In my case,
I only want to make sure that the \"color choices\" (i.e., operating
configurations) retain their distinguishability over the life of the lamp.
Imagine if your \"emergency exit\" signs had to be recolored as the RED
faded over the years... :>

> I\'d expect the blue emitter to be the most fragile but not by much.