Can I use my modulator in both directions ?
Genarally no. Most of LiNbO3 modulators have “travelling wave” electrodes that are designed so as the RF signal travels at the same speed along them than the light in the optical waveguide. At high data rate or high frequency, this velocity matching is a necessary condition for effective modulation.
Only low frequency modulators (LN-0.1 series with 100 MHz bandwidth) can operate in both directions.
Should I order my modulator with SMF or PMF ?
You have to know that LiNbO3 modulators are polarization sensitive, because the LiNb03 material is bi-refringent and because the electric field is applied along one direction only by the electrodes.
So the light has to be launched in the waveguide along a specific direction, as accurately as possible. Thus a PM input fiber is often more convenient, especially if the light source comes with a PMF output fiber. If a SM fiber is selected, one must use a polarization controller to obtain the desired light orientation at the input of the modulator chip.
As far the output fiber is concerned, it depends on the set up and and the other components along the optical path. There is generally no drawback in using a PM output fiber, even if connected to a SM fiber.
iXblue Photonics modulators come standard with PM fibers in and out.
What happen to the portion of the light that is not aligned in the correct orientation ?
One must distinguish to cases there :
- with modulators processed with APE (Annealed Proton Exchange) and with modulators featuring a polarizer : this portion of the light is not transmitted.
- with modulators processed with Ti-diffusion and not featuring a polarizer : this portion of the light is weakly modulated. It is not modulated at the same level than the main polarization, and it is not in phase because of the LiNb03 bi-refringence. One must be careful with this small portion of light because it can affect the quality of the modulated signal.
Will my modulator operate at lower bandwidth than the nominal bandwith ?
All iXblue Photonics modulators are wideband modulators. That means they will work between DC and at least the frequency given as the “-3 dB bandwidth”. Actually they will also work at higher frequencies because their response curve does not drop suddenly after the -3 dB bandwidth.
How should I select the RF driver for my modulator ?
The RF driver is often more application specific than the modulator. iXblue Photonics offer RF drivers for digital modulation, analog modulation and pulse modulation. So, you have to know what kind of modulation will be applied, and then determine the required electrical bandwidth, or rise time for a pulse application.
What are the DC electrodes for ?
One can find DC electrodes with intensity modulators. Those modulators are also called MZ (“Mach-Zehnder”) modulators, because of their waveguide design : they are basically a Mach-Zehnder interferometer with two arms that recombine. Consequently their transfer function is :
Io/2 × (1 + cos[∏(V + Vbias) /Vπ + Φo],
Where Φo is a phase term caused by the intrinsic imbalance of the interferometer and Vbias is the DC bias voltage.
The DC electrodes allow to apply the Vbias voltage, that means to adjust the phase term in the transfer function so as to select the desired operating point and compensate for any drift of the modulator.
What is the internal photodiode for ?
iXblue Photonics Mach-Zehnder intensity modulators are offered with an optional internal monitoring photodiode. This photodiode receives a part of the light diffracted at the output waveguide level and the generated photocurrent can be used in a bias control circuit intended to lock the operating point of the modulator and compensate for any modulator drift.
Why do intensity modulators drift ?
The answer is yes, and here is the explanation. Intensity modulators are waveguide Mach-Zehnder interferometers and feature two arms. Any non-symetric variation of the optical path within the two arms translates into a variation of the output of the interferometer. If this variation is not desired and not resulting from an applied modulation signal, it is considered as drift.
In spite of the high purity and homogeneity of the LiNb03 wafers used to produce the modulators and in spite of the highly controlled waveguide process, aging and temperature variations, among other reasons, will cause slight non-symetric optical path differences between the two arms of the modulator, and consequently a drift of the modulator. That’s why, when used in systems, all intensity modulators must be equipped with a bias control circuit.