LED Apparent Brightness

[RST Engineering (jw)]

I started a thread similar to this one over in SEC and didn't really get
anywhere. Let's try it here and see how it goes.


We have a 5 volt source from which we want to drive an LED. For the sake of
discussion, let's presume that with 20 mA into the LED that it drops 2.0
volts. Calculating the resistor isn't difficult -- 150 ohms ought to work.
Calculating the power consumed by the diode isn't any more difficult -- 40
mW.

Now let's put a one-shot in between the source and the LED so that the duty
cycle is 25% at some reasonable frequency above the eye's flicker rate.
Again for discussion, we'll presume 100 Hz.. Assuming that the LED can take
the current, we can pump 80 mA into the LED with this duty cycle, at which
point the LED's voltage rises to 2.5 volts (actual data from the HP data
book). Resistor? 33 ohms ought to work. Power? 200 mW peak, 50 mW
average.

With my limited biotechnical knowledge of the human eye, it would seem as
though the second LED should be brighter. Yet in (admittedly few) bench
observations the second LED was a bit dimmer, if anything. The LED wasn't
getting overly warm, nor did the voltage and current change as time went on,
indicating some moderate thermal stability.

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

I'm not understanding something.

Jim

 

[John Larkin]

On Sat, 7 May 2005 07:48:41 -0700, "RST Engineering \(jw\)"
<jim@rstengineering.com> wrote:

I started a thread similar to this one over in SEC and didn't really get
anywhere. Let's try it here and see how it goes.


We have a 5 volt source from which we want to drive an LED. For the sake of
discussion, let's presume that with 20 mA into the LED that it drops 2.0
volts. Calculating the resistor isn't difficult -- 150 ohms ought to work.
Calculating the power consumed by the diode isn't any more difficult -- 40
mW.

Now let's put a one-shot in between the source and the LED so that the duty
cycle is 25% at some reasonable frequency above the eye's flicker rate.
Again for discussion, we'll presume 100 Hz.. Assuming that the LED can take
the current, we can pump 80 mA into the LED with this duty cycle, at which
point the LED's voltage rises to 2.5 volts (actual data from the HP data
book). Resistor? 33 ohms ought to work. Power? 200 mW peak, 50 mW
average.

With my limited biotechnical knowledge of the human eye, it would seem as
though the second LED should be brighter. Yet in (admittedly few) bench
observations the second LED was a bit dimmer, if anything. The LED wasn't
getting overly warm, nor did the voltage and current change as time went on,
indicating some moderate thermal stability.

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

I'm not understanding something.

1. Human eyes respond to average illumination above the flicker rate.

2. LEDs have a most efficient current; past that peak, they're less
efficient.

John

 

[Genome]

"RST Engineering (jw)" <jim@rstengineering.com> wrote in message
news:117pla67qq9hs0a@corp.supernews.com...

I started a thread similar to this one over in SEC and didn't really get
anywhere. Let's try it here and see how it goes.

Oooooh, let's not.

DNA

 

[John Popelish]

RST Engineering (jw) wrote:
(snip)

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

I'm not understanding something.

Yes. They tell you how to drive the LEDs with pulses, not to increase
the perceptible brightness, but because that technique is necessary to
multiplex LEDs to save driver hardware when lots of LEDs must be
driven. Think large screen TV and message boards. These days, the
best you will see about brightness is that pulse driving up to some
multiplexing factor does not degrade the brightness by more than X%.

 

[Rich Grise]

On Sat, 07 May 2005 08:25:17 -0700, John Larkin wrote:

On Sat, 7 May 2005 07:48:41 -0700, "RST Engineering \(jw\)"

We have a 5 volt source from which we want to drive an LED. For the sake of
discussion, let's presume that with 20 mA into the LED that it drops 2.0
volts. Calculating the resistor isn't difficult -- 150 ohms ought to work.
Calculating the power consumed by the diode isn't any more difficult -- 40
mW.

Now let's put a one-shot in between the source and the LED so that the duty
cycle is 25% at some reasonable frequency above the eye's flicker rate.
Again for discussion, we'll presume 100 Hz.. Assuming that the LED can take
the current, we can pump 80 mA into the LED with this duty cycle, at which
point the LED's voltage rises to 2.5 volts (actual data from the HP data
book). Resistor? 33 ohms ought to work. Power? 200 mW peak, 50 mW
average.

1. Human eyes respond to average illumination above the flicker rate.

2. LEDs have a most efficient current; past that peak, they're less
efficient.

John's right. At 80 mA, you get way less than 4X the instantaneous
luminosity. There's no mystery about it.

Now, if you kept the current at, say, 20 mA, and pulsed it, then the
luminosity should decrease proportionally, but our eyes have a logarithmic
response (pretty much the same as all of our other senses), so "half as
bright" could be very much different from "half the power".

Hope This Helps!
Rich

 

[R.Lewis]

"John Popelish" <jpopelish@rica.net> wrote in message
news:bYydneepguG6deHfRVn-3Q@adelphia.com...

RST Engineering (jw) wrote:
(snip)
The HP apps note on the subject (AN-1005) goes into a long song and
dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and
say
that it will be brighter if you do this. But why would they take the
time
to show how to do it if it didn't give any benefit?

I'm not understanding something.

Yes. They tell you how to drive the LEDs with pulses, not to increase
the perceptible brightness, but because that technique is necessary to
multiplex LEDs to save driver hardware when lots of LEDs must be
driven. Think large screen TV and message boards. These days, the
best you will see about brightness is that pulse driving up to some
multiplexing factor does not degrade the brightness by more than X%.

In one of the agilent app notes they say it is always more efficacious to
drive their leds with a puredc current rather than modulate it.

 

[Spehro Pefhany]

On Sat, 7 May 2005 07:48:41 -0700, the renowned "RST Engineering
\(jw\)" <jim@rstengineering.com> wrote:

I started a thread similar to this one over in SEC and didn't really get
anywhere. Let's try it here and see how it goes.


We have a 5 volt source from which we want to drive an LED. For the sake of
discussion, let's presume that with 20 mA into the LED that it drops 2.0
volts. Calculating the resistor isn't difficult -- 150 ohms ought to work.
Calculating the power consumed by the diode isn't any more difficult -- 40
mW.

Now let's put a one-shot in between the source and the LED so that the duty
cycle is 25% at some reasonable frequency above the eye's flicker rate.
Again for discussion, we'll presume 100 Hz.. Assuming that the LED can take
the current, we can pump 80 mA into the LED with this duty cycle, at which
point the LED's voltage rises to 2.5 volts (actual data from the HP data
book). Resistor? 33 ohms ought to work. Power? 200 mW peak, 50 mW
average.

With my limited biotechnical knowledge of the human eye, it would seem as
though the second LED should be brighter. Yet in (admittedly few) bench
observations the second LED was a bit dimmer, if anything. The LED wasn't
getting overly warm, nor did the voltage and current change as time went on,
indicating some moderate thermal stability.

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

I'm not understanding something.

Jim

At well above the frequency at which flicker is perceptible, the eye
averages the brightness. So, if the LED is linear and has no internal
resistance (ie. to a first approximation), it's a wash.

 

[Don Klipstein]

In article <117pla67qq9hs0a@corp.supernews.com>, RST Engineering \(jw\) wrote:

I started a thread similar to this one over in SEC and didn't really get
anywhere. Let's try it here and see how it goes.


We have a 5 volt source from which we want to drive an LED. For the sake of
discussion, let's presume that with 20 mA into the LED that it drops 2.0
volts. Calculating the resistor isn't difficult -- 150 ohms ought to work.
Calculating the power consumed by the diode isn't any more difficult -- 40
mW.

Now let's put a one-shot in between the source and the LED so that the duty
cycle is 25% at some reasonable frequency above the eye's flicker rate.
Again for discussion, we'll presume 100 Hz.. Assuming that the LED can take
the current, we can pump 80 mA into the LED with this duty cycle, at which
point the LED's voltage rises to 2.5 volts (actual data from the HP data
book). Resistor? 33 ohms ought to work. Power? 200 mW peak, 50 mW
average.

With my limited biotechnical knowledge of the human eye, it would seem as
though the second LED should be brighter. Yet in (admittedly few) bench
observations the second LED was a bit dimmer, if anything. The LED wasn't
getting overly warm, nor did the voltage and current change as time went on,
indicating some moderate thermal stability.

There is a bit of a myth that the human eye to some extent is a peak
detector even if the flicker rate is high enough for the light to appear
continuous. This is not true. If it was true, then there would have been
plenty of effort put into pulsing fluorescent lamps, mercury lamps, metal
halide lamps, etc.

What is true is that many LEDs are nonlinear, and back in the 1980's it
was noticed that the type of LED chip mostly used in digital displays then
were much brighter with high instantaneous current than with low
instantaneous current with the average current being the same in both
cases.
A display with a few digits and 7 segments per digit and two or a few
LED chips per segment eats a lot of current, even with only a few mA per
chip. It was noticed that with only a couple mA average current per chip,
they were brighter when pulsed with higher instantaneous current than when
fed steady DC at the same low average current. Many people believed that
a quirk in human vision was responsible, but actually the reason was that
these LEDs were less efficient at low instantaneous current.

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

They actually recommend pulsing some of their LEDs to achieve
instantaneous current of at least 10 mA if average current is less than 10
mA.

Some of their LEDs with a particular chemistry have degradation rate
not only a function of junction temperature, but also a function of duty
cycle, proceeding only while current is flowing (or maybe if voltage is
across the junction). I suspect some sort of electrochemicaly motivated
diffusion.
I also suspect that some of their LEDs may have minor flaws, probably
mainly as a result of aging, causing excessively low efficiency (or maybe
no output at all) when instantaneous current is low.

Another thing: If you have LEDs with adjustable brightness, especially
in an application involving color mixing, it is often recommended to vary
duty cycle rather than current. This is because LEDs have performance
specified at a particular instantaneous current, and variations in color
and in efficiency from one lot to another or from one piece to another can
be greater when current is greatly different from that which performance
is specified at. Some LEDs have color varying with instantaneous
current. In addition, different chemistries have different nonlinearities
- most InGaN blue LEDs have efficiency maximized when current is somewhere
around 10-20% of maximum rated, while most red LEDs have efficiency
maximized when current is at least 30% sometimes over 50% of maximum
rated.

- Don Klipstein (don@misty.com, http://www.misty.com/~don/ledp.html)

 

[John Larkin]

On Sun, 8 May 2005 03:31:52 +0000 (UTC), don@manx.misty.com (Don
Klipstein) wrote:

Some of their LEDs with a particular chemistry have degradation rate
not only a function of junction temperature, but also a function of duty
cycle, proceeding only while current is flowing (or maybe if voltage is
across the junction). I suspect some sort of electrochemicaly motivated
diffusion.
I also suspect that some of their LEDs may have minor flaws, probably
mainly as a result of aging, causing excessively low efficiency (or maybe
no output at all) when instantaneous current is low.

Not to change the subject (never!) but here in San Francisco, at least
half of the LED traffic lights seem to be failing. It's almost always
the green, but I did just see a bad red. Strangly-shaped
blobs/rows/polygons within the array go black, often accompanied with
other regions that flicker weakly. On Van Ness Avenue, maybe 70% of
the green lights are defective.

As a cost saving measure maybe they should use incandescents.

John

 

[Don Klipstein]

In article <1d3r71lqb3nur7nmne645n24jsrjup3ikv@4ax.com>, John Larkin wrote:

On Sun, 8 May 2005 03:31:52 +0000 (UTC), don@manx.misty.com (Don
Klipstein) wrote:

Some of their LEDs with a particular chemistry have degradation rate
not only a function of junction temperature, but also a function of duty
cycle, proceeding only while current is flowing (or maybe if voltage is
across the junction). I suspect some sort of electrochemicaly motivated
diffusion.
I also suspect that some of their LEDs may have minor flaws, probably
mainly as a result of aging, causing excessively low efficiency (or maybe
no output at all) when instantaneous current is low.

Not to change the subject (never!) but here in San Francisco, at least
half of the LED traffic lights seem to be failing. It's almost always
the green, but I did just see a bad red. Strangly-shaped
blobs/rows/polygons within the array go black, often accompanied with
other regions that flicker weakly. On Van Ness Avenue, maybe 70% of
the green lights are defective.

As a cost saving measure maybe they should use incandescents.

Or find out what green units are used in Upper Darby, PA. They have
gone a few years already quite well. I have not noticed any failures.

I have seen a few reds in Philadelphia with partial failures, but they
could be less-good models (Philadelphia tried several different models, at
least in pilot programs on a few streets) and also most of these could be
about 10 years old already.

A modern red or green LED traffic signal consumes typically about 12
watts maybe a bit less nowadays, and the incandescent it replaces is
probably a 92 watt or 116 watt one.

- Don Klipstein (don@misty.com)

 

[Ban]

RST Engineering (jw) wrote:

I started a thread similar to this one over in SEC and didn't really
get anywhere. Let's try it here and see how it goes.

Instead of trying again, when your preconcepts are not met, wouldn't it be
better to drop or modify your expectations?

We have a 5 volt source from which we want to drive an LED. For the
sake of discussion, let's presume that with 20 mA into the LED that
it drops 2.0 volts. Calculating the resistor isn't difficult -- 150
ohms ought to work. Calculating the power consumed by the diode isn't
any more difficult -- 40 mW.

Now let's put a one-shot in between the source and the LED so that
the duty cycle is 25% at some reasonable frequency above the eye's
flicker rate. Again for discussion, we'll presume 100 Hz.. Assuming
that the LED can take the current, we can pump 80 mA into the LED
with this duty cycle, at which point the LED's voltage rises to 2.5
volts (actual data from the HP data book). Resistor? 33 ohms ought
to work. Power? 200 mW peak, 50 mW average.

With my limited biotechnical knowledge of the human eye, it would
seem as though the second LED should be brighter. Yet in (admittedly
few) bench observations the second LED was a bit dimmer, if anything.
The LED wasn't getting overly warm, nor did the voltage and current
change as time went on, indicating some moderate thermal stability.

The HP apps note on the subject (AN-1005) goes into a long song and
dance as to HOW to achieve pulse driving an LED, but doesn't come
right out and say that it will be brighter if you do this. But why
would they take the time to show how to do it if it didn't give any
benefit?

If you would have listened to the advice all this unnecessary experiments
would have been saved. Can you see how your belief (in an urban myth)
completly overrides even the outcome of your own experiments? You seem to be
unable to work scientifically. An intelligent person continuously questions
his beliefs, in fact he is clear about where his opinions come from and
doesn't identify with either side. Truth is so powerful, that it will win
ultimately, so why extend this period unnecessarily?

I'm not understanding something.

Even if you meant that just as an empty phrase, it hits right on the spot.

--
ciao Ban
Bordighera, Italy

 

[John Popelish]

Don Klipstein wrote:
(snip)

I have seen a few reds in Philadelphia with partial failures, but they
could be less-good models (Philadelphia tried several different models, at
least in pilot programs on a few streets) and also most of these could be
about 10 years old already.

A modern red or green LED traffic signal consumes typically about 12
watts maybe a bit less nowadays, and the incandescent it replaces is
probably a 92 watt or 116 watt one.

I wish they would lengthen their life by dimming them in proportion to
the ambient light level. The green ones around here (central
Virginia) are very annoyingly bright at night. If it is foggy or
drizzling rain, they light the mist on the windshield so bright you
can't see past it.

 

[RST Engineering (jw)]

"Ban" <bansuri@web.de> wrote in message
news:1fhfe.870128$b5.38522689@news3.tin.it...

RST Engineering (jw) wrote:
I started a thread similar to this one over in SEC and didn't really
get anywhere. Let's try it here and see how it goes.


Instead of trying again, when your preconcepts are not met, wouldn't it be
better to drop or modify your expectations?

Excuse me? I got three or four answers in s.e.c., none of which spoke
directly to the question. This morning, a week and a half after I asked the
question, Don Klipstein (thank you, sir) answered the question with some
intelligent comments, which I take to heart.

If you would have listened to the advice all this unnecessary experiments
would have been saved. Can you see how your belief (in an urban myth)
completly overrides even the outcome of your own experiments? You seem to
be unable to work scientifically. An intelligent person continuously
questions his beliefs, in fact he is clear about where his opinions come
from and doesn't identify with either side. Truth is so powerful, that it
will win ultimately, so why extend this period unnecessarily?

Perhaps your newsreader got some posts mine didn't. Or perhaps you had best
lay off the schnapps before posting. I'm not sure which. I got nothing of
benefit from s.e.c..

I'm not understanding something.

Even if you meant that just as an empty phrase, it hits right on the spot.

No, it meant exactly what I said. I did an experiment. It didn't turn out
like what I was led (you should pardon the pun) to believe would happen.
Klipstein explained it elegantly. Case closed.

Jim

 

[Ban]

RST Engineering (jw) wrote:

"Ban" <bansuri@web.de> wrote in message
news:1fhfe.870128$b5.38522689@news3.tin.it...
RST Engineering (jw) wrote:
I started a thread similar to this one over in SEC and didn't really
get anywhere. Let's try it here and see how it goes.


Instead of trying again, when your preconcepts are not met, wouldn't
it be better to drop or modify your expectations?

Excuse me? I got three or four answers in s.e.c., none of which spoke
directly to the question. This morning, a week and a half after I
asked the question, Don Klipstein (thank you, sir) answered the
question with some intelligent comments, which I take to heart.



If you would have listened to the advice all this unnecessary
experiments would have been saved. Can you see how your belief (in
an urban myth) completly overrides even the outcome of your own
experiments? You seem to be unable to work scientifically. An
intelligent person continuously questions his beliefs, in fact he is
clear about where his opinions come from and doesn't identify with
either side. Truth is so powerful, that it will win ultimately, so
why extend this period unnecessarily?

Perhaps your newsreader got some posts mine didn't. Or perhaps you
had best lay off the schnapps before posting. I'm not sure which. I
got nothing of benefit from s.e.c..



I'm not understanding something.

Even if you meant that just as an empty phrase, it hits right on the
spot.

No, it meant exactly what I said. I did an experiment. It didn't
turn out like what I was led (you should pardon the pun) to believe
would happen. Klipstein explained it elegantly. Case closed.

Jim

Jim, this theme has been discussed to death in this NG, always the same
urban myth, maybe it was not you, I also didn't check your posts in other
NG. Don maintains his patience and he also maintains this nice site where he
explaines the second myth about higher efficiency
http://members.misty.com/don/lede.html Good source of information from an
unbiased technical person.
--
ciao Ban
Bordighera, Italy

 

[RST Engineering (jw)]

I've been a regular on this ng for about 8 years and don't remember this
subject being discussed, at least not in detail.

I perused Don's site for about two hours before I asked these questions. I
just went back and reread every word on the page you referenced below.
Perhaps you would be so kind as to point out the error in my reading and
give me a couple of words from the paragraph that answers my question
directly. I'll go back and see how I could have missed the subject twice.

It is only an urban myth if you don't know any better. If you are designing
something and it doesn't come out the way you expected it, it is a mystery.
I don't mind solving my own mysteries, but I also choose not to reinvent the
light bulb.

Jim

Jim

Jim, this theme has been discussed to death in this NG, always the same
urban myth, maybe it was not you, I also didn't check your posts in other
NG. Don maintains his patience and he also maintains this nice site where
he explaines the second myth about higher efficiency
http://members.misty.com/don/lede.html Good source of information from an
unbiased technical person.
--
ciao Ban
Bordighera, Italy

 

[Ban]

RST Engineering (jw) wrote:
corrected top posting:

Jim, this theme has been discussed to death in this NG, always the
same urban myth, maybe it was not you, I also didn't check your
posts in other NG. Don maintains his patience and he also maintains
this nice site where he explaines the second myth about higher
efficiency http://members.misty.com/don/lede.html Good source of
information from an unbiased technical person.
--
ciao Ban
Bordighera, Italy

I've been a regular on this ng for about 8 years and don't remember
this subject being discussed, at least not in detail.

I perused Don's site for about two hours before I asked these
questions. I just went back and reread every word on the page you
referenced below. Perhaps you would be so kind as to point out the
error in my reading and give me a couple of words from the paragraph
that answers my question directly. I'll go back and see how I could
have missed the subject twice.

"Most such white LEDs will be slightly more efficient when moderately
underpowered and will be less efficient when overpowered."

"Another significant difference is that incandescent lamps operate much less
efficiently when the filament is not as hot, while most LEDs used in
flashlights ususally operate slightly more efficiently than normal at
reduced power."

These two sentences say the highest efficiency is measured with a little
less than the nominal current, so when you try to increase the current the
efficiency goes down. Maybe you didn't understand it this way? There is a
lot of useful information on these pages, but not written in spectacular but
rather technical manner.

--
ciao Ban
Bordighera, Italy

 

[RST Engineering (jw)]

"Most such white LEDs will be slightly more efficient when moderately
underpowered and will be less efficient when overpowered."

Which speaks directly to white LEDs, but not to LEDs in general. And what
is "slightly" more efficient? If I can put in an average 25% more power by
pulsing and the LED goes 1% less efficient, I've achieved a 24% gain with no
more battery power consumed.

"Another significant difference is that incandescent lamps operate much
less efficiently when the filament is not as hot, while most LEDs used in
flashlights ususally operate slightly more efficiently than normal at
reduced power."

Same argument. Back to the question of what "slightly" means.

These two sentences say the highest efficiency is measured with a little
less than the nominal current, so when you try to increase the current the
efficiency goes down. Maybe you didn't understand it this way? There is a
lot of useful information on these pages, but not written in spectacular
but rather technical manner.

"efficiency" can be measured in a whole bunch of ways. If lamp output goes
down a couple of percent in efficiency with more current, but the current
can be delivered more efficiently, then total efficiency can be discussed as
lamp output as a function of total power taken from the driving source.
That was my original question, but Klipstein elaborated to the point that I
think I understand that lamp efficiency falls faster than source efficiency
when you climb above nominal current. At least that's what I THINK he said.

Then we get into the question of what "nominal" current means and how it is
specified. I don't think I've seen that number on any of the data sheets
I've perused, be it HP, Kingbright, Toshiba ... or Mouser. THere are a lot
of "recommended" currents, but what is the basis for recommendation?

Too many questions, not enough answers.

Jim

 

[Don Klipstein]

In article <117sh7od541seb@corp.supernews.com>, RST Engineering \(jw\) wrote:

I've been a regular on this ng for about 8 years and don't remember this
subject being discussed, at least not in detail.

I perused Don's site for about two hours before I asked these questions. I
just went back and reread every word on the page you referenced below.

and the page referenced below was:

http://members.misty.com/don/lede.html

That one has no direct link to my page that mentions pulsing and this
business about it supposedly making LEDs appear brighter. My page that
discusses this is:

http://members.misty.com/don/ledp.html

My lede.html page has a link to my LED top page (ledx.html) which has a
link to ledp.html, but this is roundabout and I do consider it possible to
spend 2 hours reading the other stuff (including linked pages) before
getting that far unless you read all of ledx.html before following any
links.

- Don Klipstein (don@misty.com, http://www.misty.com/~don/ledx.html,
AKA http://members.misty.com/don/ledx.html)

 

[RST Engineering (jw)]

Perfect. Thanks.

Jim

That one has no direct link to my page that mentions pulsing and this
business about it supposedly making LEDs appear brighter. My page that
discusses this is:

http://members.misty.com/don/ledp.html

 

[Ol' Duffer]

In article <117pla67qq9hs0a@corp.supernews.com>, "RST Engineering \(jw
\)" <jim@rstengineering.com> says...

The HP apps note on the subject (AN-1005) goes into a long song and dance as
to HOW to achieve pulse driving an LED, but doesn't come right out and say
that it will be brighter if you do this. But why would they take the time
to show how to do it if it didn't give any benefit?

Two come to mind without much thought:

1) A means of varying apparent brightness without changing
supply voltage or limiting resistor.

2) A means of achieving similar apparent brightness with a
multiplexed drive scheme as compared to steady state.