4

I've just got access to spectrometer and measuring spectrum of everything. But when I took a look at 365nm UV led from chinese UV flashlight I got something really strange - it had some IR at exactly half the energy. Any suggestions on what it might be?

I double checked that it is actually IR, and not a grating glitch - IR-cut filter does remove it.

UV peak: 370.5nm

IR peak: 741.4nm

enter image description here

dmckee --- ex-moderator kitten
  • 86,062
BarsMonster
  • 2,371
  • 1
    Most green laser pointers are actually IR with a frequency double. Or so I've read. Might be connected. – dmckee --- ex-moderator kitten Jul 28 '15 at 19:24
  • As @dmckee pointed out, green lasers are not as simple as red. Check out this question: http://physics.stackexchange.com/questions/173677/does-my-green-laser-pointer-emit-three-distinct-frequencies-or-is-my-measuremen It is likely something similar here. – tpg2114 Jul 28 '15 at 19:30
  • Physics of 1064nm and 808nm leakage in 532nm lasers is completely clear to me (808nm diode pump, 1064nm from Nd:YVO4 not converted by non-linear crystal). In this case there is an LED, not a laser, so I am in doubt there might be something similar. – BarsMonster Jul 28 '15 at 20:19
  • 2
    How did you remove the possibility of grating harmonics? What does your spectrometer setup look like? – boyfarrell Jul 29 '15 at 13:53
  • @boyfarrell You was right, apparently strong signals make sensor non-linear, so software processing was removing harmonics not perfectly. – BarsMonster Jul 30 '15 at 04:26

1 Answers1

1

As boyfarrell pointed out, the IR peak that you observe is due to second order reflection from the diffraction grating of the spectrometer. Strong spectral lines are often seen in the second order (with the advantage of better resolution). You can easily verify it by inserting a long-pass filter that blocks 370 nm (some orange glass or even sunglasses) between the LED and the spectrometer. You will see that both peaks disappear simultaneously.

gigacyan
  • 4,690
  • That was quite unusual... I was able to replicate the situation where IR peak was removed by IR-cut filter, but IR-pass (680nm+) filter does not show anything behind it... – BarsMonster Jul 30 '15 at 04:11
  • It appears there is indeed significant non-linearities involved when suppressing strong harmonics. Weaker signal somehow does not produce this "ghost" IR... – BarsMonster Jul 30 '15 at 04:12
  • It seems harmonic suppression is in software? Due to different sensitivity between IR and UV strong signals might saturate the sensor and make correct software processing impossible... – BarsMonster Jul 30 '15 at 04:17