5
$\begingroup$

In the Taylor series expansion of the mode-propagation constant $𝛽$ in optical fibers, we have multiple terms known as $𝛽_0$, $𝛽_1$, $𝛽_2$, etc. The terms $𝛽_1$, $𝛽_2$, and higher can be measured by different techniques to find the fiber dispersion and dispersion slope. My question is whether there is a technique to accurately measure $𝛽_0$ which is a common phase shift. I have searched literature extensively and have not come across any such technique.

$𝛽(𝜔)=𝑛(𝜔)𝜔/𝑐=𝛽_0+𝛽_1(𝜔−𝜔_0)+1/2 𝛽_2(𝜔−𝜔_0)^2+...$

where

$𝛽_𝑚=(𝑑^𝑚𝛽/𝑑𝜔^𝑚)𝜔=𝜔_0(𝑚=0,1,2,...)$

$\endgroup$
5
  • 1
    $\begingroup$ measuring $\beta_0=\beta(\omega_0)$ at $\omega=\omega_0$ is no different from measuring $\beta(\omega)$ at any other frequency, you need what is called an IQ detector which a beam splitter and a $\pi/2$ phase shifter at whatever frequency you measure; create $\sin(\omega t+\phi)$ and $\cos(\omega t+\phi )$ and then correlate each with $\cos(\omega t)$ to get $\cos \phi$ and $\sin \phi $, $\endgroup$
    – hyportnex
    Commented Jul 29 at 14:56
  • 1
    $\begingroup$ $\phi(\omega)=\beta(\omega) \ell$ where $\ell$ is the length of the waveguide (fiber). $\endgroup$
    – hyportnex
    Commented Jul 29 at 15:11
  • $\begingroup$ @hyportnex that feels to me like it should be an answer. $\endgroup$
    – A Nejati
    Commented Jul 29 at 15:19
  • $\begingroup$ @hyportnex Thank you for your comment. Can you please direct me to a paper or a resource to fully understand the details of the measurement? In the interferometric techniques I use, I usually get a phase difference which only gives the dispersion which is the $beta_1$ value. $\endgroup$ Commented Jul 29 at 16:31
  • 1
    $\begingroup$ this is called "IQ" detection/measurement in RF: if you have $\cos\phi$ and $\sin\phi$ then you get $\phi$ unambiguously $\mod 2\pi$ and knowing $\ell$ then get $\beta$. $\cos\phi$ and $\sin\phi$ are the result of a multiplication. $\endgroup$
    – hyportnex
    Commented Jul 29 at 19:13

1 Answer 1

1
$\begingroup$

An "I/Q" detector, also called a coherent detector, is a four-quadrant detector. Its purpose is unambiguously to measure the phase of a sinusoid relative to another of the same frequency. It is to measure the unknown phase $\phi$ of a sinusoid $x(t)=A\cos(\omega t+\phi)$ relative to a known reference, say $r(t)=R\cos\omega t$.

The measurement involves splitting $x(t)$ into two equal amplitude signals, multiplication and delaying one of it by exactly $\pi/2$ resulting in an "I/Q" pair, called in-phase and quadrature, say $x_I(t)=B\cos(\omega t+\phi)$ and $x_Q(t)=B\sin(\omega t +\phi)$ for some $B=kA$. Here $k=k(A,B,\omega)$ is a known calibrated function of $A,B,\omega$, characteristic of the splitter, but is independent of $\phi$.

Next each "arm" is multiplied by the reference $r(t)$ followed by a low-pass filter to remove the double frequency term $2\omega$. For example, when the product $$x_I r(t)=B\cos(\omega t +\phi)R\cos(\omega t)= \frac{1}{2}B \cos\phi + \frac{1}{2}B \cos(2\omega t+\phi)$$ is low-pass filtered you get $$y_I=[x_Ir(t)]_{LP}=\left[\frac{B}{2}\cos(\omega t +\phi)R\cos(\omega t)\right]_{LP}= mB \cos\phi,$$ and similarly for the other arm $$y_Q=[x_Qr(t)]_{LP}=\left[\frac{B}{2}\sin(\omega t +\phi)R\cos(\omega t)\right]_{LP}= mB \sin\phi,$$ where $m=m(R,B,\omega)$ is a factor characteristic of the multiplier and filter, depending on $R,B,\omega$, but is independent of $\phi.$

From $y_I$ and $y_Q$ one gets unambiguously $\phi =\beta \ell=\arctan (y_I,y_Q).$

The actual optical or RF "multiplication" can be mechanized several ways but one possible method is based on the identity $(a+b)^2+(a-b)^2=4ab$ and consists of a pair of "square-law" or intensity detectors.

Of course, instead of $x(t)$ one could just as well split and phase shift the "reference" signal $r(t).$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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