I was thinking of building a DIY IR Spectroscope but for this, I need to find an IR diode that can do variable wavelengths. So far I have looked at Farnell and Mouser but all the diodes I have looked at seem to be fixed on a wavelength.

Is this possible? or would I need some sort of filters and optical equipment, if so which parts would I need?


  • 2
    \$\begingroup\$ The wavelength that LEDs emit is determined by the materials used. As far as I know, that wavelength only varies a tiny bit depending on (maybe) temperature and current. I think physics is against you on this. \$\endgroup\$ – JRE Jan 22 at 19:31

There may be an near-infrared narrow band wavelength optical converter that would do the job. But not likely an affordable solution.

The more common NIR light source would be to use a broadband (200-2000 nm) Quartz Tungsten-Halogen (QTH) light source.

StellarNet has some inexpensive calibrated light sources in the $400 to $2,000 range.

You may want to use a QTH light lamp for under $200. Or if a very tight budget buy a QTH light bulb for about $10. ThorLabs QTH Lamps


There are broadband IR LED's available that go from 650 to 1050nm

If I wanted two different IR wavelengths in one product, I would use two different SMT diodes with a light guide (make sure you check the tranmisitivity of the material of the light guide in the IR)

Some of these light guides could probably fit two 0603's SMT LED's as they are enter image description here

  • \$\begingroup\$ I see you inherently kind of answered that, but: OP's looking for variable wavelength LEDs; broadband isn't variable. \$\endgroup\$ – Marcus Müller Jan 22 at 21:49
  • \$\begingroup\$ A broadband source can be desirable in a spectrometer application \$\endgroup\$ – Voltage Spike Jan 22 at 22:43
  • 1
    \$\begingroup\$ @MarcusMüller except the OP is looking for what does not exist. Spectroscopy light sources generally use an incandescent broadband light source. While this LED has very little radiant intensity, it is still a very good answer. \$\endgroup\$ – Misunderstood Jan 22 at 23:26

The wavelength of the photons emitted by an LED is defined by the bandgap of the semiconductor junction; that's a material composition constant. It can't be adjusted, far as I can see, by means of anything but a temperature change which probably isn't feasible for an LED potted in plastics connected with solder.

So, nope, you won't get a variable-wavelength LED: LEDs are, by principle, fixed wavelength.

If you need an adjustable range of wavelengths: Get a source of broadband radiation (e.g. a broadband IR LED, which is essentially a pretty impure LED, or, even simpler, an incandescent bulb), get a prism adequate to diffract a beam of that light (think making an infrared rainbow) and a mechanism that rotates that prism so that you can select the wavelength you desire with e.g. a slot (followed, typically, by some focussing optics).

There's tunable lasers but I'm more or less certain these are out of your price range of interest.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.