Mach-Zehnder Interferometer (MZI)

We use SiEPIC EBeam library in this tutorial.

This notebook walks through the process of setting up and simulating a mach-zehnder interferometer device using the OPICS package.

A mach-zehnder interferometer is a basic waveguide interference device. It consists of two couplers (or Y branches) connected by two waveguides of different length (see below). The difference between the two waveguide lengths causes differential delay, which contributes to the frequency dependent interference pattern.

import time
import warnings
import numpy as np
from opics.globals import c
from opics import Network
import opics

freq = np.linspace(c * 1e6 / 1.5, c * 1e6 / 1.6, 2000)

OPICS version 0.1.6

Import component library

Import ebeam library from libs module.

ebeam = opics.libraries.ebeam

Define network

Create an instance of Network class, which is used to add, connect, and simulate circuit components.

circuit_name = "mzi"
circuit = opics.Network()

Add circuit components

Add grating couplers, 3dB power splitters (e.g. Y-splitter or Y-branch), and waveguides to circuit. You can define custom frequency data points for a component as well (see the example for output_GC).

#define component instances
input_gc  = circuit.add_component(ebeam.GC(freq))
y1 =   circuit.add_component(ebeam.Y(freq))
wg1 =  circuit.add_component(ebeam.Waveguide(freq,50e-6))
wg2 =  circuit.add_component(ebeam.Waveguide(freq,150e-6))
y2 =  circuit.add_component(ebeam.Y(freq))
output_gc = circuit.add_component(ebeam.GC(freq))

Define circuit connectivity

In this section, we define the component connections. The connections are defined using Network.connect, e.g.

Network.connect(component1, component1_port, component2, component2_port)

#define circuit connectivity
circuit.connect(input_gc, 1, y1, 0)
circuit.connect(y1, 1, wg1, 0)
circuit.connect(y1, 2, wg2, 0)
circuit.connect(y2, 0, output_gc, 1)
circuit.connect(wg1, 1, y2, 1)
circuit.connect(wg2, 1, y2, 2)

Simuate the circuit

warnings.filterwarnings('ignore') #ignore all/complex number warnings from numpy or scipy
sim_start = time.time()

#simulate network
circuit.simulate_network()

print("simulation finished in %ss"%(str(round(time.time()-sim_start,2))))

simulation finished in 0.06s

Visualize the simulation result

circuit.sim_result.plot_sparameters(show_freq = False, scale="log")