Exam2 April2023
Exam2 April2023
Exam2 April2023
Instructions
1. Submit your answers as a soft copy by 5:00PM, 20th April 2023. Even high resolution Photos
are fine.
2. The examination is open book
3. Provide sufficient details to understand the thought process. Part marks will be given for the
ideas
4. Make any approximations as appropriate but list it.
5. No discussions of any type is allowed between people (all people, not just your course
mates). That’s the only constraint.
Q1. (9 marks)
Pulses which are unaffected by nonlinearity and dispersion
In class, we discussed Solitons, which propagated unaffected in the presence of nonlinearities.
Here, let us look at a different form of a pulse which does not change shape in the presence of
nonlinearities (but can potentially change pulse width)
a. What do you think would the functional form of that pulse be? Derive (Figure it out,
guess etc) (3 points) (Hint: assume a general functional form, look at the effects of SPM
and dispersion independently and guess)
b. Suggest an optical scheme (architecture with components) to generate such a pulse. (3
points)
c. Read up on the internet and suggest some uses for such pulse shapes in optical systems.
For each application, tell me why it is useful (2 points)
Q2. (7 marks)
Z-scan technique to measure intensity dependent refractive index
This is an easy question, but its an important diagnostic tool. Z-scan is a technique used to
measure the value of the intensity dependent refractive index of a sample. The common
experimental setup to implement z scan is shown below.
The sample is translated along the z-axis around the focus and the output power after the
aperture is measured. If the sample were purely linear, then no change in output power is
anticipated. Nonlinear behavior causes this change. From the plot between z and measured
power, the value of nonlinear refractive index can be inferred.
a. How does the technique work? Explain qualitatively.
b. For a fixed position of the sample after the focus (closer to the detector), the laser power
is increased. Plot qualitatively the dependence of measured power vs laser intensity for
i. n2 > 0
ii. n2 < 0
In each case, indicate the line for the measured power in case of no nonlinearity.
c. Plot the anticipated dependence of measured power vs z for both positive and negative
values of z (starting near the lens to all the way till the detector). Indicate the line for
measured power vs z in case of no nonlinearity. The plot should be for both positive n2
and negative n2 on the same plot.
(Hint: First investigate the focusing distance (effective focal length) of self-focusing as a
function of beam-size and power. Use an equivalent focal length lens component instead
of the lens to see the behavior. )
Q4. (9 marks)
Four-wave mixing and self-phase modulation
Two signals, one a higher power perfect CW (carrier) (exp(𝑗𝜔0 𝑡)) and another modulated lower
power signal at a slightly offset frequency 𝑎(𝑡)exp(𝑗(𝜔0 + ∆𝜔)𝑡)) where a(t) is a complex
baseband envelope (and the bandwidth of a(t) being much lower than ∆𝜔) is combined in the
same polarization and sent through optical fiber at sufficiently higher power to trigger the
intensity dependent refractive index. For this problem, ignore dispersion effects.
(a) If output power is not a limitation, what are the series of new frequencies generated at
the output (on either side of the carrier).
(b) What are the envelope functions of these new frequencies? Tabulation until the second
order sidebands is sufficient.
(Hint: In band four wave mixing as in this case can be easily captured as self-phase
modulation of the net signal (sum of all components). To get upto 2 nd order terms, you
can make the assumption that exp(jx) ~ 1+jx-x^2/2 )
(c) At the output of the fiber, we now implement a tunable filter which can select individual
frequency bands. If originally, a(t) was a bandwidth limited Gaussian pulse with a pulse
width of 10 picoseconds, what is the envelope (time domain picture, pulse width etc) of
the envelope of
i. The signal at (𝜔0 − ∆𝜔)
ii. The signal at (𝜔0 + 2∆𝜔)
(d) For a(t) being a bandwidth limited Gaussian pulse (as above), plot in frequency domain,
the output spectrum from the fiber. The figure should capture the spectral changes due
to the different time domain envelopes of the different sidebands.