LED Specification Question

[Remove SPAM From Address to Reply]

I have noticed that, on LED apecification sheets, the optical output
power is sometimes (but not always) given directly in milliwatts.
When not given as absolute power, the favored unit seems to be
"mw/sr".

I assume mw is still milliwatts, but what is "sr"? I suspect the r
may be radians, as the "output power" (specified using these units)
for similar devices seems inversely proportional to beamwidth. Help?

tbanks,

Martin

 

[Boris Mohar]

I have noticed that, on LED apecification sheets, the optical output
power is sometimes (but not always) given directly in milliwatts.
When not given as absolute power, the favored unit seems to be
"mw/sr".

I assume mw is still milliwatts, but what is "sr"? I suspect the r
may be radians, as the "output power" (specified using these units)
for similar devices seems inversely proportional to beamwidth. Help?

tbanks,

Martin

Steradian

http://whatis.techtarget.com/definition/0,,sid9_gci528813,00.html



Regards,

Boris Mohar

Got Knock? - see:
Viatrack Printed Circuit Designs http://www3.sympatico.ca/borism/

 

[R.Lewis]

I have noticed that, on LED apecification sheets, the optical output
power is sometimes (but not always) given directly in milliwatts.
When not given as absolute power, the favored unit seems to be
"mw/sr".

I assume mw is still milliwatts, but what is "sr"? I suspect the r
may be radians, as the "output power" (specified using these units)
for similar devices seems inversely proportional to beamwidth. Help?

tbanks,

Martin

mW/sr tend to be given for visible leds toward the ends of the visible
spectrum (reds and blues) because the figures look so bad in candela (due to
the relative inefficiency of the eye).

 

[Don Klipstein]

I have noticed that, on LED apecification sheets, the optical output
power is sometimes (but not always) given directly in milliwatts.
When not given as absolute power, the favored unit seems to be
"mw/sr".

I assume mw is still milliwatts, but what is "sr"? I suspect the r
may be radians, as the "output power" (specified using these units)
for similar devices seems inversely proportional to beamwidth. Help?

Sr is a steradian. A one steradian portion of the surface of a sphere
has an area equal to the square of the radius. The surface of a sphere is
4-pi steradians. A cone 65.54 degrees wide has a solid angle of 1
steradian.

- Don Klipstein ([email protected])

 

[Remove SPAM From Address to Reply]

Sr is a steradian. A one steradian portion of the surface of a sphere
has an area equal to the square of the radius. The surface of a sphere is
4-pi steradians. A cone 65.54 degrees wide has a solid angle of 1
steradian.

- Don Klipstein ([email protected])

Thanks, I suspected that it would be some sort of area/angle related
specification. So, for an LED with a 50-degree viewing angle, the
output power in mW/sr is also the "approximate" absolute output power.

regards,

Martin

 

[R.Lewis]

[email protected] (Don Klipstein) wrote in message

Thanks, I suspected that it would be some sort of area/angle related
specification. So, for an LED with a 50-degree viewing angle, the
output power in mW/sr is also the "approximate" absolute output power.

No.
The total output power ( your absolute power?) may be determined from the
polar diagram and the known output in a given direction.
Frequently the output, candela or mW/sr, is quoted on axis, or wherever the
figure is the largest and most impressive, and the angle quoted is the angle
from the axis where the output has fallen to 1/2 of the impressive (on axis)
value.

Note that, in this context, one mW per steradian does not mean that over one
steradian one milliwatt is radiated.
It could be, for example, that on axis the radiated power is 1mw/sr but at 1
degree off axis, the power radiated( in that direction) is 0.5mW/sr, at 2
degrees it could be 0.25mW/sr and so on.
This is what the polar diagram shows.

Hope this helps.

regards

 

[Remove SPAM From Address to Reply]

So you're saying that each manufacturer who quotes a power density in
mw/sr, possibly (or probably) collects that data over a small
incremental area where the peak output is concentrated.

If so, then in the absense of a polar plot (which no one seems to
supply), it's really impossible to approximate total IR output for
purposes of other calculations.

Martin

 

[R.Lewis]

So you're saying that each manufacturer who quotes a power density in
mw/sr, possibly (or probably) collects that data over a small
incremental area where the peak output is concentrated.

Normally yes (and normally this is the on axis radiation).
Most also quote the total radiated power (in mW) for a given input current

If so, then in the absense of a polar plot (which no one seems to
supply), it's really impossible to approximate total IR output for
purposes of other calculations.

Which IR led manufacturers do not offer polar diagramms of their devices?

 

[Remove SPAM From Address to Reply]

"Which no one seems to supply" was a reference to resellers, not
manufacturers. Most of the surplus catalogs I shop list rudimentary
specs such as angle, case size, and mw/sr, but no manufacturer ID.

Martin

 

[R.Lewis]

"Which no one seems to supply" was a reference to resellers, not
manufacturers. Most of the surplus catalogs I shop list rudimentary
specs such as angle, case size, and mw/sr, but no manufacturer ID.

Martin

Unless you are looking for UV or IR leds the mw/sr data is a strange thing
to quote.
Since the majority led manufacturers quote outputs in 'light' terms (lumen,
candela) for their visible leds, with only a few also adding the radiometric
(wattage) data, your surplus store must be transposing the data.

Doesn't seem too likely on the surface.

 

[Remove SPAM From Address to Reply]

Unless you are looking for UV or IR leds the mw/sr data is a strange thing
to quote.
Since the majority led manufacturers quote outputs in 'light' terms (lumen,
candela) for their visible leds, with only a few also adding the radiometric
(wattage) data, your surplus store must be transposing the data.

Doesn't seem too likely on the surface.

Looking at near-IR LEDs in the 850 to 880 range. Don't know what you
mean by transposing data, they just don't publish much data at all
when you buy surplus.

 

[Don Klipstein]

Thanks, I suspected that it would be some sort of area/angle related
specification. So, for an LED with a 50-degree viewing angle, the
output power in mW/sr is also the "approximate" absolute output power.

Usually, very roughly. Not always close - the amount of radiation
outside the viewing angle and how much the radiation intensity varies
throughout the viewing angle can mess that up.

I once purchased some red LEDs by Chicago Miniature that produced only
about half as much light as expected from the mcd rating (1000) and
viewing angle (60 degrees). These figures would indicate .84 lumen
assuming 1000 mcd throughout the viewing angle and no light outside it.
Actual output was close to .4 lumen. It appeared to me that the light
intensity 30 degrees off axis was a lot less than half what it was
on-axis. Since "viewing angle" is usually defined as the angle including
the range at which light intensity is at least half the on-axis (or
brightest-part-of-the-beam) value, I think they either used a different
definition or the figure was optimistic.

I have seen a lot of optimistic figures, especally for some Radio Shack
LEDs.

- Don Klipstein ([email protected])

 

[Steve J. Noll]

Usually, very roughly. Not always close - the amount of radiation
outside the viewing angle and how much the radiation intensity varies
throughout the viewing angle can mess that up.

I once purchased some red LEDs by Chicago Miniature that produced only
about half as much light as expected from the mcd rating (1000) and
viewing angle (60 degrees). These figures would indicate .84 lumen
assuming 1000 mcd throughout the viewing angle and no light outside it.
Actual output was close to .4 lumen. It appeared to me that the light
intensity 30 degrees off axis was a lot less than half what it was
on-axis. Since "viewing angle" is usually defined as the angle including
the range at which light intensity is at least half the on-axis (or
brightest-part-of-the-beam) value, I think they either used a different
definition or the figure was optimistic.

I have seen a lot of optimistic figures, especally for some Radio Shack
LEDs.

- Don Klipstein ([email protected])

I'm not surprised at the optimistic figures.

My day job is at an optoelectronics manufacturer. I'm shocked at
the state of the poor measurement techniques used by some LED
fabricators. One major LED house uses a 15-year-old obsolete
EG&G radiometer for their measurements. A test lab tested
visible LEDs for me for in a sphere, applied a "correction" to get
a candela reading resulting in a 2:1 error. Another LED manufacturer
measured some LEDs for me in their best sphere measurement system -
when the readings appeared way off I contacted the measurement
system manufacturer only to find out that their instrument was never
calibrated for that kind of measurement.

These measurements are much more difficult than measuring a
voltage or current where we're all used to better that 1% accuracy
from any common handheld DMM. I'm thrilled with 10% to 20%
for opto measurements. But it's still no excuse for the sloppiness
I see in the industry.

I'm adopting the CIE 127 Averaged LED Intensity standard for visible
LED candela measurements. I only wish I could get some others to do
the same so we wouldn't see so many "optimistic" figures!

Steve J. Noll | Ventura California |
| The Used High-Tech Equipment Dealer Directory
| http://www.big-list.com
| The Peltier Device Information Site:
| http://www.peltier-info.com