B Optical interference

1. Apr 28, 2017

Zachariah

Hi, i have a question about the optical interference, supposing we have two LEDs which spectra are partially overlapped, can we have interferences?? is there any chance to have a partial coherence between the two lights??

2. Apr 28, 2017

Drakkith

Staff Emeritus
Hmmm. I believe LED's emit non-coherent light, so I don't think you will see interference with light from another LED no matter what the spectrum of each LED is. To see an interference pattern, I think you'd first need to pass the light from each one through something like a very small slit.

3. Apr 28, 2017

Zachariah

Thanks for your fast replay, actually i am not intending to visualize interference pattern as with Young slits, but just looking for an evidence for the existence of a constructive or destructive superposition of waves.

4. Apr 28, 2017

Staff: Mentor

The waves from two sources always superpose, but you don't get a visible interference pattern unless the sources are coherent.

If you could take an instantaneous "snapshot" of the light arriving at a screen from two incoherent sources, you would see an interference pattern. However, it changes very rapidly and randomly, on a time scale on the order of the period of oscillation of the waves, i.e. about 10-15 second. So it simply "washes out" as far as the eye or any photographic device (that I know of, anyway) are concerned.

5. Apr 28, 2017

Zachariah

I agree, in this case we cannot visualize the pattern, and even if we take an image of it with the camera, we will get the mean intensity which is the sum of the individual intensities, but in my case, the result intensity is always below that sum, that pushed me to suppose the existence of destructive interference, even if i cannot visualize it. would that be true?

6. Apr 28, 2017

Drakkith

Staff Emeritus
It's true that the light interferes both constructively and destructively, but you would not be able to see evidence of either of these just by looking at the incoherent light. How far off is the total measured intensity vs the sum of the individual intensities?

7. Apr 28, 2017

Zachariah

To be clear, i tried two LEDs with spectra totally separated, i took the image of the incoming light via an optical fiber, the mean intensity of a ROI is equal to the sum of individual intensities, then i used the two LEDs mentioned above, and the resulted intensity is less than the sum by more than 10% depending on their power, the more i increase their power the more the gap go larger.

8. Apr 29, 2017

Tom.G

Is you detector linear with respect to the incident power?
If you are measuring the sum of the LEDs with an optical fiber are their angles equally off-axis from the fiber entrance angle?

9. Apr 29, 2017

Zachariah

Yes, the detector is linear with respect to the incident power.
I actually mix the two lights inside an integrant sphere coupled with the optical fiber, all measures are taken without changing any parameter, i only turn on and off LEDs.

10. Apr 29, 2017

SlowThinker

Still must be a measurement error. Either the power source is shared for both LEDs and cannot supply both at high power, or there's a problem measuring the light intensity.
You can't have interference from two independent sources of light.

11. Apr 29, 2017

Zachariah

To be sure i dont have a measurement error i tried several configurations:
- Two LEDs with separated spectra --> good result
- Two LEDs with overlapped spectra, LEDs are similar (fabricated from same semiconductor materiel) --> intensity decreases
- Two LEDs with overlapped spectra, LEDs are not similar (fabricated from different semiconductor materiel) --> good result
I even made the output pass through a diffraction grid and plotted the substraction of the sum of the individual spectra and the mixture spectrum, i found that the region responsible for this phenomena is located within the overlapping region.

12. Apr 29, 2017

SlowThinker

Still it has to be an error somewhere. It definitely is not interference. All optical phenomena are linear, at least at the intensities a mere mortal can access (I know that very strong lasers use nonlinear mirrors).

This sounds like the detector can't take light that is too intense.
What happens if you switch one light source on and off at say 1 Watt (or whatever typical value it has) and increase the other from 0 to 10 Watts? Is the step the same at first then flattens out? Or is it 10% less throughout the range? Also what is the accuracy of the detector, can you actually perform this experiment?

If you have enough equipment around, you can try to split the light in two using a piece of glass at around 45 degrees, measure the two images (straight and reflected) and add the intensities.

13. Apr 29, 2017

Zachariah

The imager was not saturated in any of those measurements, i worked with moderated powers.
Using low powers leads to good result, and as long as i increase the power, this phenomena becomes critical.
The detector accuracy is taken into consideration.

I think the experiment u suggested is similar to using 2 identical LEDs having same spectrum, i already tried it and i still have a slight lose in power.

14. Apr 30, 2017

Tom.G

Since the usual problems have been largely addressed, here is another track to follow.

The LEDs could be pumping each other when their spectrums overlap. There is a known similiar problem with LASER diodes. A possible way to test this would be:

• Place an 90%/10% beamsplitter in front of each LED. 90% beam going to your present integrating sphere & detector
• Monitor each of the 10% beams with additional detectors and readouts
• If the individual, 10%, detectors track with the power anomaly in your main setup, then the LEDs are talkin' to each other... either optically, thru a common power supply, or thermally.

15. Apr 30, 2017

Zachariah

Ok i will try this, i keep u tuned.
Is this protocol has an appellation so i can read more about it? Link?

16. Apr 30, 2017

sophiecentaur

IS it possible that the two LEDs are sharing a power supply and its output power is limited?
Edit - sorry, that's been thought of already.

Last edited: May 1, 2017
17. Apr 30, 2017

Zachariah

LEDs need 3V and consume some milliamperes, the power supply output is 3V/1A

18. Apr 30, 2017

sophiecentaur

OK but is there any common series resistance in the circuit? Power supplies and lack of decoupling are very common faults. Sometimes they are very subtle.
I don't see the cause as being due to interference because you would have to expect fringe patterns that would move about as the equipment is physically distorted. You would need a fantastic level of alignment to see partial cancellation all across the field of view. This is in addition to the other reasons from other people, above.
A diagram (complete) would be an advantage in chasing something as odd as this.

19. Apr 30, 2017

Tom.G

Here is one that's along the same lines as, but not exactly identical to, the optical crosstalk conjecture. When the spectrums of the two LEDs overlap, that could be similiar to a 3-mirror optical cavity when used with a LASER source.

Quote from: http://opticalengineering.spiedigitallibrary.org/article.aspx?articleid=2195791

"The power spectrum of the laser changes due to interaction between the lasing field and the small backscattered field, which re-enters the laser cavity with results described by Eq. (1)."

20. Apr 30, 2017

Zachariah

Well actually i dont say this is definitely an interference, i encountered this phenomena and i tried to know why it occurs and i did not find any other explanation but the possibility of having an optical interaction. Same electrical circuit is used in all other combination i experimented and gave good results. The existence of the interference does not imply that we must have a pattern of it, in case of partial coherence we will have a blurred screen instead of fringes.

Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted

Similar Threads - Optical interference Date
I Using multiple mirrors in Michelson Interferometer Jul 17, 2017
I Sweat on Screen, Interference or Refraction? Apr 26, 2017
B Thin film interference - why thin, exactly? Mar 10, 2017