Infrared LED Lighting: How IR LEDs Work and How to Specify Them
An infrared LED is a semiconductor emitter whose output sits just beyond the red end of the visible spectrum, in the band where image intensifier tubes and many camera sensors still respond but the unaided eye does not. That is what makes it useful on a military vehicle: the driver on night vision goggles sees a lit scene, and an observer without them sees darkness.
This article explains how IR LEDs produce light, how the peak wavelength decides both what the night vision device sees and what an observer may see, and which figures to ask for when specifying an emitter for a vehicle or platform: radiant intensity, beam angle, thermal behaviour and the IEC 62471 risk group.
How an IR LED Produces Near-Infrared Light
An LED is a semiconductor junction. When current flows through it, electrons recombine with holes and release the energy as photons, and the band gap of the semiconductor sets the photon energy and therefore the wavelength. Infrared emitters use gallium arsenide (GaAs) and aluminium gallium arsenide (AlGaAs, also written GaAlAs), which emit above about 760 nm. Vishay’s 850 nm VSMY12850, for example, is built on a GaAlAs surface emitter chip, and its 940 nm TSAL6200 on GaAlAs multi quantum well technology.
The output is a band, not a single line. A typical emitter has a spectral bandwidth of 30 to 35 nm around its peak: 35 nm for the VSMY12850 and 30 nm for the TSAL6200. The ams OSRAM IR LED range, for example, is concentrated at two peaks, 850 nm and 940 nm, with smaller families at other wavelengths.
ICNIRP places the visible band at 380 to 780 nm and infrared from 780 nm to 1 mm, divided into IR-A (780 nm to 1.4 µm), IR-B (1.4 to 3 µm) and IR-C (3 µm to 1 mm). Emitters for night vision work sit in IR-A, close to the visible edge. Thermal imagers work far from there: they detect heat radiated by the scene in the long-wave band, roughly 8 to 14 µm, and need no illumination at all. The practical differences are covered in infrared vs thermal.
Because an LED is an incoherent, broadband source, its eye safety is assessed under the lamp standard IEC 62471 rather than under the laser rules. That is covered below.
Common peak, well inside Gen II and Gen III tube response
Common peak with reduced red glow
Typical spectral bandwidth of one emitter
End of the visible band, start of IR-A
What the Unaided Eye Sees
IR LEDs are often described as invisible. For a 940 nm emitter that is close to true in practice. For an 850 nm emitter it is not quite: the visible band ends at 780 nm, and an emitter with a 35 nm bandwidth puts part of its output not far beyond that edge. An observer looking straight at an 850 nm emitter at close range may see a dull red point. ams OSRAM recommends its 940 nm parts where an application needs reduced red glow compared to 850 nm.
On a vehicle the glow is a property of the emitter, not of the beam. The illuminated ground stays dark to the eye; what can give the vehicle away is a red point on the lamp face, seen by someone close to it and in line with it. Mounting position, the direction the lamp faces and the choice of wavelength for each function decide how much of that an observer gets.
Matching the Emitter to the Night Vision Device
An emitter is only useful if the device in service responds at its wavelength. A US Air Force Armstrong Laboratory paper on night vision devices (Task, 1992, published in NATO AGARD-LS-187) gives second generation tubes a sensitivity from about 400 nm to about 900 nm and third generation tubes from about 600 nm to a little over 900 nm, against about 400 to 700 nm for the human eye.
850 nm sits well inside both tube ranges. 940 nm sits at or beyond the upper edge of a third generation tube’s response, so the same radiant power produces a dimmer image, and more output is needed for the same viewing distance. Digital night vision and cameras vary with the sensor and with any infrared cut filter fitted, so compare the published spectral response curve with the emitter’s peak and bandwidth before buying. Wavelengths beyond the reach of image intensifiers, such as 1,064 nm and SWIR, are covered in IR wavelengths.
the driver or observer needs the brightest image through Gen II or Gen III goggles, the vehicle is moving and needs distance, and the lamp is unlikely to be seen head-on at short range by observers without night vision.
the emitter may be seen directly at close range, the task is short range, and a dimmer image or more output power is an acceptable price for reduced red glow.
Radiant Intensity, Beam Angle and Range
Two figures describe an emitter’s output. Radiant flux, in milliwatts, is the total optical power. Radiant intensity, in milliwatts per steradian (mW/sr), is the power sent in a given direction, so it shows how concentrated the beam is. The same chip behind narrower optics gives a higher intensity on the axis and a narrower lit area.
| Emitter (datasheet) | Peak | Beam | Radiant intensity (typical) |
|---|---|---|---|
| Vishay VSMY12850 | 850 nm | ±40° half intensity | 16 mW/sr at 70 mA |
| Vishay TSAL6200 | 940 nm | ±17° half intensity | 72 mW/sr at 100 mA |
| ams OSRAM SFH 4715AS | 860 nm | 80° beam angle | 900 mW/sr at 1,000 mA (1,530 mW flux) |
Range is where the physics bites. Irradiance on a target falls with the square of the distance, so lighting a scene to the same level at twice the distance takes four times the radiant intensity, and only part of that light is reflected back to the goggles. A wide flood suits near-field driving and manoeuvring; a narrower beam suits distance. Our covert illuminator, for example, is offered with 15, 25 and 45 degree view angles at 850 nm and 180 mW radiant power, for a stated visibility of 50 to 100 m.
When comparing offers, ask for radiant intensity and beam angle at the stated drive current, not electrical watts. Electrical power describes the input: the same covert illuminator consumes 2.4 W to deliver 180 mW of radiant power.
Heat, Drive Current and Wavelength Drift
Most of the electrical input ends up as heat at the junction, and IR LED output depends on temperature. Vishay gives a radiant power temperature coefficient of -0.19 %/K for the VSMY12850 and -0.6 %/K for the TSAL6200. The peak also moves: by 0.25 nm/K and 0.2 nm/K respectively. A 50 K rise in junction temperature therefore shifts an 850 nm peak about 12 nm towards longer wavelengths, and the output drops at the same time.
Forward voltage falls as the junction warms (-1.7 mV/K for the VSMY12850). Under a fixed voltage the current would rise and heat the junction further, which is why IR LEDs are driven at constant current.
On a vehicle that turns into three procurement checks. The housing has to carry heat away, so look at the rated operating temperature: our infrared driving light is rated -40 °C to 105 °C and the infrared headlight -40 °C to +65 °C. The driver has to accept the vehicle’s supply range, 7 to 28 V DC on the driving light. And the switching driver must not disturb radios and vehicle electronics: our infrared driving light and covert illuminator are specified to MIL-STD 461G, and the infrared headlight to ECE R10 and CISPR 25 Level 5.
Eye Safety Under IEC 62471
IEC 62471:2006 (edition 1.0, published 26 July 2006, prepared as CIE S 009:2002), Photobiological safety of lamps and lamp systems, sets exposure limits, the measurement technique and a classification scheme for incoherent broadband sources from 200 nm to 3,000 nm. It covers LEDs and excludes lasers.
Two of its hazards apply to near-infrared emitters: the infrared radiation hazard to the cornea and lens (780 to 3,000 nm) and the retinal thermal hazard for a weak visual stimulus (780 to 1,400 nm). Products are classified as Exempt, Risk Group 1, 2 or 3. The reason infrared gets attention is spelled out in ams OSRAM’s application note on IRED eye safety: IR-A gives the eye very little visual stimulus, so the aversion response that normally protects it does not work, the pupil does not contract, and the assessment assumes a fully open 7 mm pupil.
For a buyer, the practical steps are to ask the supplier for the IEC 62471 risk group and the distance at which it was assessed (the ams OSRAM note reports hazard values at 200 mm), carry it into the labelling and maintenance instructions, and make sure crews know that a high-output array can be switched on while looking unlit.
Covert Does Not Mean Undetectable
An IR emitter is visible to every device that responds at its wavelength, friendly or not. The US Army’s NVG driving training circular, TC 21-305-2 (September 1998), makes the point about the goggle’s own illuminator: using it makes the NVG an active system that can be detected by the threat’s night vision devices. The same applies to every IR lamp on a vehicle. Covert means hidden from the unaided eye; it does not mean undetectable.
Three consequences follow. Switch IR functions on only when the task needs them. Point and shield them so the light goes where the crew needs it and not towards likely observers. And treat 940 nm as a way to reduce the visible glow, not as protection from night vision: it remains within reach of image intensifiers and cameras that respond to it.
Lighting designed for night vision imaging systems (NVIS) works the other way round. MIL-STD-3009 (2 February 2001) sets the emission characteristics of aircraft lighting intended for use with NVIS, so that crew lighting does not degrade the goggle image. An IR emitter is designed to be seen through the goggles. Where a platform carries both, specify them as separate functions with separate requirements.
Where Betalight Uses IR LEDs
Our infrared lights apply these principles around the vehicle: the infrared driving light (a wide flood at 770 to 840 nm), the infrared headlight with its 850 nm high beam and 940 nm low beam, the covert illuminator (850 nm, selectable view angle), the convoy brake light and tail light at the rear, and the IR beacon for identification. How these functions work together on a vehicle is covered in infrared vehicle lighting.
Outside defence, the same 850 nm and 940 nm parts serve security cameras, machine vision and driver monitoring. Where a platform needs a wavelength, beam angle or connector the catalogue does not carry, we build to specification.
Frequently Asked Questions
Look at it through a night vision device or a camera that responds at its wavelength: the emitter should show as a bright point and the area in front of it as lit. An 850 nm emitter may also show a dull red point to the naked eye at close range, but a 940 nm emitter usually shows nothing, so the absence of a glow proves nothing. For acceptance testing, measure radiant output with a calibrated detector at the stated drive current, and do not look into a high-output array at close range while it is energised.
No. An LED is an incoherent source with a spectral bandwidth of tens of nanometres and a beam shaped by its optics. A laser emits a very narrow line in a tightly collimated beam, which is why it reaches further and why it is regulated differently: IEC 62471 covers LEDs and excludes lasers, which are classified under the laser safety standard IEC 60825-1:2014. LED emitters suit flood illumination, marking and signalling; lasers suit pointing and long-range spot illumination.
Because the peak varies between parts as well as with temperature. The Vishay VSMY12850 is sold as an 850 nm emitter, but its datasheet allows a peak anywhere from 830 nm to 870 nm, on top of a 35 nm bandwidth and a drift of 0.25 nm/K. That is why lamp specifications often state a band rather than a single value: our infrared driving light, for example, is specified at 770 to 840 nm. When comparing products, ask for the peak, its tolerance and the bandwidth at operating temperature, and check them against the response of the goggles or sensors in service.