LEDS seem to have been a constant source of questions and confusion ‘round here. Most of us don’t have ample IR LEDS laying around and must buy them, which raises questions of which are best suited, and where are they cheapest. There is already one excellent post that covers the basic electronics quite nicely, I’m explaining the OTHER considerations...mostly here are a few definitions and concepts which may help you to choose and implement your LEDs without my pretending to know the best for your use or where to get them.
Now to roll out the long winded explanations:
OUTPUT: This goes hand in hand with beam angle in measurement. The real consideration is how much light your
LED die produces at it’s rated voltage and current and where it gets pointed by the lens. In visible light this
is expressed as lumens or footcandles but with IR it is usually called the ‘radiant energy’ and more IS better!
But realistically just like Watts in a stereo system, different manufacturers use varying methods to measure it,
and may use different terms to express it. So comparing apples to oranges is an art. There are some proven LEDs
posted by a variety of people, and there are lots of options that will work as yet unexplored.
As a rule of thumb, if it is dirt cheap (ebay anyone?) and has no specs listed, it is probably not an
overachiever...it still may work for you, but you won’t know unless you try it. So if you don’t want to gamble
on your results, stick with a purchase from a source that can tell you the specs of your device. If you are
building a large display, then you had best use something that has a known vicious output specification.
BEAM ANGLE: The definition of where the majority of the LEDs output is focused. If you draw a centerline from
a 20 degree LED, it will (theoretically) put it’s rated radiant intensity of IR in a cone that is 10 degrees
from the centerline all the way around (conveniently called the half-beam angle). So a wider angle LED that has
the same radiant intensity actually puts out more IR...assuming the specs you are looking at are defined that
way. Some manufacturers shortcut by measuring the intensity on their standard lens/die combination and copying
that info to the other models that share that die. You can see this by looking at the specs on different beam
patterns of the same LED. Usually manufacturer’s part numbers will append the beam angle of that devices’ lens
...If the radiant intensity for the same LED doesn’t vary with beam angle, then it is the measurement for the
die, not that particular LED package and your focused output will instead be higher on a narrower device.
For DI anything that allows you to flood your entire surface with IR while not creating a direct reflection
pointing BACK into your camera, works.
With FTIR the objective is to get as much IR striking your faces on your acrylic at a 40 degree angle as you
can, anything steeper will escape the surface increasing your need to mask the edge of the display, anything
shallower will basically bounce around lengthwise until it finds an escape path or is attenuated by the acrylic.
So with FTIR the two schools of thought are:
1)plan your entry angle with a narrow beam LED to focus as great of a bounce as you can, the short
method is to aim the centerline at the flat edge of your acrylic so that it hits the display surface
at 40 degrees, the long method is to do the trigonometry to calculate the half beam angle’s maximum
spread that will hit at or near 40 degrees...they really don’t vary much on results. The other thing
to consider with narrow beam angle LEDs for FTIR is, if you don’t pay attention to the how the beam
angles interact between LEDS on their various bounces around the screen you can end up with areas of
the screen that simply aren’t served by any LEDs and will be dim, unresponsive, etc.
2)throw a buttload of wide angle IR directly at the edge and mask off the part that glows at your
camera. (I opted for the latter)
WAVELENGTH: The primary wavelength of light output by an LED is expressed as a distance measurement. 880nm,
920nm, 940nm, etc...all are readily available, each has it’s own considerations. Which is best for you is not
a factor of the chosen display technology (DI, FTIR, Other) but rather a matter of the other materials you
are using i.e. your acrylic, camera sensor, and visible light filters available. Try to find out the
manufacturers transmissive specifications for your acrylic if you can before choosing LEDs, this is frequently
overlooked but most acrylic is manufactured for use as a window material. Often acrylic glazing will be
intentionally designed to NOT pass far infrared well (bigger nm numbers) to increase it’s insulation value,
mine begins attenuating transmission at 900mn so it would have obviously impacted the effectiveness of a 940mn
LED array in my FTIR. On the other hand, if you are using exposed film as a visible light filter, then it
will by default be more transparent to far IR than the near which is closer to visible light. If you can find
one within budget, a ‘real’ IR passing, visible light blocking filter of some sort really will improve your
results (ideally a 20nm bandpass centered on the wavelength of your LEDs, but a cold mirror is also an option)
but many projects do just fine with exposed film or the like. There are also differences in sensitivity to IR
between different camera sensor types, but I’m not well enough versed in those to quote them accurately...the
sensor (CMOS, etc), not the connection type (USB/etc)
Most IR illuminators used in the security camera industry use 920-940mn LEDs.
