Free-Space Electro-Optic Modulators

Lithium Niobate Electro-Optic Modulators

Thorlabs free-space electro-optic (EO) amplitude and phase lithium niobate modulators combine our experience with crystal growth and electro-optic materials. Our EO modulators use MgO-doped lithium niobate for high power operation. The modulators have an SMA RF input, which is directly compatible with our HVA200 High Voltage Amplifier (shown below). We offer both broadband DC-coupled and high Q resonant models.

Click to Enlarge
The required half-wave voltage needed to drive the EO modulators
is a function of the wavelength of the optical signal. The curves labeled
Amplitude Modulators and Phase Modulators describe the broadband
(non-resonant) EO Modulators.

Features

• High Performance in a Compact Package
• Broadband DC Coupled or High-Q Resonant
• 2 mm Diameter Clear Aperture
• MgO-Doped Lithium Niobate Crystal
• SMA RF Modulation Input Connector
• DC to 100 MHz
• Custom OEM Versions Available

DC-Coupled Broadband Modulators

Our broadband EO modulators consist of an EO crystal packaged in a housing optimized for maximum RF performance. The RF drive signal is connected directly to the EO crystal via the SMA RF input. An external RF driver supplies the drive voltage for the desired modulation. The crystal may be modulated from DC up to the frequency limits of the external RF driver. Please see the Specs tab for amplifier requirements. These broadband amplitude and phase modulators are offered with a -C4* (400 - 600 nm), -C1 (600 - 900 nm), -C2 (900 - 1250 nm), or -C3 (1250 - 1650 nm) AR coating. See the Graphs tab for more information on the reflectance of these coatings.

High Q Resonant Modulators

Resonant EO Modulators are operated at a fixed frequency, and Thorlabs' design includes a high-Q resonant circuit. When compared with the voltage required to drive broadband EO modulators, the operating voltage of resonant EO modulators is dramatically reduced. Our resonant amplitude and phase modulators operating at 20 MHz are offered with a -C1 AR coating (600 - 900 nm). See the Graphs tab for more information on the reflectance of this coating. Standard and custom AR coating options are available, as well as versions with user-specified resonance frequencies from 0.1 to 100 MHz. Please contact Tech Support to discuss your application needs. Please see the Specs tab for more information on these modulators.

Driver and Accessories

Our broadband EO phase modulators produce phase shifts from -180° to +180° when the drive voltage varies from -V_π to +V_π. While the EO phase modulators require the entire ±V_π range to fully modulate an optical signal, broadband EO amplitude modulators are able to fully modulate an amplitude signal when the drive voltage varies from 0 to either +V_π or -V_π. Please refer to the EO Modulator Intro tab for an illustration.

A standard laboratory function generator is needed in conjunction with a high-voltage amplifier to drive our modulators. Our HVA200 high-voltage amplifier, which features a ±200 V output at 100 mA of continuous current, a 1 MHz bandwidth, and low noise, is ideal for driving our broadband EO modulators when certain operating conditions are met. Generally speaking, EO modulators driven by the HVA200 can fully modulate signals when the required half-wave voltage is ≤200 V. As illustrated in the plot above, this 200 V maximum allows the HVA200 to modulate our EO amplitude modulators and our non-resonant EO phase modulators, at full depth, for wavelengths up to ~620 nm and ~900 nm, respectively. However, the effective voltage range of the amplifier can be extended using the technique described in the Lab Facts tab, which enables our EO amplitude modulators to be fully modulated at wavelengths up to 1000 nm.

Please contact Tech Support if you require more information about driving the resonant or non-resonant broadband EO modulators. 

Thorlabs also offers a Glan-Thompson Polarizer mounting adapter, as well as an EO modulator mount for integration into our fiberbench collimation hardware (see below).

Modulators

These plots show the reflectance of each of our four dielectric coatings for a typical coating run. The shaded region in each graph denotes the spectral range over which the coating is highly transmissive.

Click to Enlarge
Raw Data

Click to Enlarge
Raw Data

Click to Enlarge
Raw Data

Click to Enlarge
Raw Data

Click to Enlarge
Figure 2: Electro-Optic Amplitude Modulator

Click to Enlarge
Figure 1: Electro-Optic Phase Modulator

Click to Enlarge
Figure 3: EO Phase Modulator Output as a Function of Applied Voltage (Listed in Terms of V_π)
The X marks on the curve correspond to the phase shifts listed in the table above the graph.

Click to Enlarge
Figure 4: EO Amplitude Modulator Output as a Function of Applied Voltage (Listed in Terms of V_π)
The X marks on the curve correspond to conditions in the table columns.

Introduction to EO Modulators

Electro-optic (EO) modulators are designed to modulate the phase or intensity (amplitude) of electromagnetic radiation (light) propagating through them. They are used in the Q-switching and mode locking of lasers, the generation of optical pulses, side-band generation, and other applications. EO modulators employ the Pockels effect, which is a linear (to first order) electro-optic phenomenon in which the induced birefringence of certain crystals is proportional to the amplitude of the electric field (voltage) applied across the crystal. Birefringent crystals exhibit one refractive index (n_||) for light polarized parallel to the optic axis of the crystal and a different refractive index (n_⊥) for light polarized perpendicular to the optic axis. For the crystals used in EO modulators, the optic axis is parallel to the direction of the applied electric field.

Generally speaking, the optical path length (OPL) through a medium is the product of the geometrical length of the path through a medium and the refractive index (n_r) of the medium. The refractive index of, and therefore the OPL through, a birefringent crystal is determined by the polarization of the light referenced to the optic axis of the crystal. The OPL through a medium is directly related to the phase accumulated by light travelling through it. When the Pockels effect applies, it is possible to modulate the accumulated phase (phase shift) of the light that propagates through the crystal by modulating the applied voltage. Longer crystals produce larger phase shifts, and the OPL and accumulated phase shift are both functions of the wavelength of the light. EO modulators exploit the changes in the crystal's refractive indices that occur as a response to an applied voltage signal.

EO Phase Modulator

An EO phase modulator applies a voltage-controlled phase shift to the light without affecting its polarization. Figure 1 shows one configuration used to accomplish this: the input light is linearly polarized and aligned with the optic axis of the crystal. As is stated above, the optic axis (parallel to the z-axis in this diagram) is determined by the orientation of the applied electric field. The phase accumulated by the light is a function of the wavelength of the light, the geometric width of the crystal (measured along the y-axis, which is parallel to the direction of propagation), and the refractive index of the optic axis, n_||. The modulation of n_|| controls the phase modulation of the light; the value of n_⊥ is not a factor. Figure 3 lists and plots the phase shift as a function of the applied voltage, which is expressed in terms of V_π. In the plot, the X marks correspond to the values listed in the table above the graph. V_π is called the half-wave voltage and is defined as the voltage required to shift the output phase by pi radians.

EO Amplitude Modulator

An example of an EO amplitude modulator is shown in Figure 2. In this configuration, the input light is linearly polarized and oriented at a 45° angle to the crystal's optic axis, which is again determined by the direction of the applied electric field. The input light in this case can be equivalently described as the sum of two equal-amplitude and perpendicularly polarized components: one polarized parallel to the optic axis (along z-axis) and one polarized perpendicular to the optic axis (along the x-axis). These two components accumulate phase differently as they propagate through the crystal, which is a consequence of the birefringence of the crystal; the OPL of the component of light polarized parallel to the optic axis is determined by n_||, and the OPL of the perpendicularly polarized component is determined by n_⊥. As the applied voltage is varied, the difference between the values of the crystal's two refractive indices varies; therefore, the relative phase shift between the two components also varies.

The polarization of the total electromagnetic field (the sum of the components) at the output of the crystal changes as the two components move in and out of phase with one another. The polarization of the light at the output of the crystal is linear when the two components are in or 180° out of phase, circular when they are π/2 radians out of phase, and elliptical otherwise. The amplitude of the light transmitted by the 45° polarizer is a function of the polarization of the light incident on it; when the applied voltage is modulated, the intensity of the output beam is modulated. The light transmitted by the polarizer, and therefore the EO amplitude modulator, is linearly polarized and oriented parallel to the axis of the polarizer.

The data tabulated in Figure 4 shows the effect of the applied voltage on the phase shift and polarization of the total light field at the output of the crystal. The plot shows the amplitude of the light transmitted by the polarizer as a function of the applied voltage over a modulation range of 2V_π. (The X marks on the curve correspond to the values listed in the above table.) For the configuration shown in Figure 2, the applied voltage signal must vary from 0 V to V_π (or from -V_π to 0 V) to fully modulate the amplitude of input light; in order to capture both a maxima and minima, the modulator must pass through 0 V and either +V_π or -V_π. This restriction can be modified when a quarter-wave plate is added to the setup (please see the Lab Facts tab).

Driving EO Modulators

EO modulators may be configured so that the voltage is applied in a direction perpendicular to the propagation of light, as shown in Figures 1 and 2, or in a direction parallel to the propagation of light. An advantage of applying the voltage in the transverse orientation, as shown in these figures, is that it is not necessary for the light to travel through the electrodes. This generally results in better light transmission through the modulator and allows electrode materials that are opaque at the operating wavelength to be used. Additionally, in this transverse configuration, the OPL is proportional to the product of the applied field and crystal length.

Variable high voltage supplies, such as a function generator paired with a high voltage amplifier and delivering hundreds of volts, are typically used to drive free-space EO modulators. As shown by the graph included in the Overview tab, the voltage required to drive the EO modulator increases with the wavelength of the light being modulated. Because of this, at longer wavelengths the voltage needed to achieve full modulation of the optical signal can exceed the voltage output range of the high voltage supply. As an example, the HVA200 high voltage amplifier has a range of ±200 V, which is ideal for driving the broadband EO modulators EO-AM-NR and EO-PM-NR when the wavelength of the light is below approximately 610 and 900 nm, respectively. At higher wavelengths, this amplifier does not supply the voltage needed to achieve full-scale modulation of the optical signals. However, when pairing the HVA200 with EO amplitude modulators such as the EO-AM-NR products, it is possible to extend the available voltage range of the HVA200 by placing a quarter-wave plate before the input of the modulator. This procedure is detailed in the Lab Facts tab.

The driving voltage requirements of fiber-coupled EO modulators are often significantly less than those of free-space EO modulators. In many cases, these modulators have half-wave voltages on the order of 4 V, and one challenge of driving them is locating voltage drivers capable of producing a modulating frequency high enough to fulfill the needs of the application. A basic setup built around a function generator and an amplifier that was successfully used to drive a fiber-coupled EO phase modulator to 1 GHz is described here. 

Click to Enlarge
Figure 1: Electro-Optic Amplitude Modulator Without the Quarter-Wave Plate

Click to Enlarge
Figure 2: Electro-Optic Amplitude Modulator with a Quarter-Wave Plate at the Input

Pairing an EO Amplitude Modulator with a Quarter-Wave Plate to Reduce Drive Voltage Requirements

Variable high voltage supplies, such as a function generator paired with a high voltage amplifier delivering hundreds of volts, are typically used to drive free-space EO modulators. As shown by the graph included in the Overview tab, the voltage required to drive the EO modulator increases with the wavelength of the light being modulated. Because of this, at longer wavelengths the voltage needed to achieve modulation over the full amplitude or phase range can exceed the output voltage range of the high voltage supply. Please see the EO Modulators tab for more information about EO modulators.

In the case of EO amplitude modulators such as the EO-AM-NR products, placing a quarter-wave plate at the input of the EO amplitude modulator can reduce the maximum voltage necessary to drive the modulator over its full range; when the high voltage supply is the limiting factor, the use of a quarter-wave plate makes available more, or all, of the operational wavelength range of the driven EO amplitude modulator. Figures 1 and 2 show an EO amplitude modulator with and without, respectively, the quarter-wave plate in place. Experiments conducted in our laboratory using the HVA200 high voltage amplifier, which has an output voltage range of ±200 V, and the EO-AM-NR-C1 electro-optic amplitude modulator demonstrate and document this approach. A summary of that work is presented in this tab. Please click on the Lab Facts icon for details.

The experimental setup used the previous-generation HRS015 stabilized HeNe laser (633 nm) and LP980-SF15 fiber-pigtailed laser (980 nm) as light sources. A linear polarizer (LPVISB100 or LPNIR100) placed after the source ensured the polarization of the input light was at a 45° angle to the optic axis of the EO-AM-NR-C1 electro-optic amplitude modulator. A second linear polarizer placed at the output of the modulator was aligned parallel to the first. The HVA200, which was used to amplify the 100 kHz triangle wave output by a function generator, produced the variable voltage signal (±200 V maximum) driving the EO amplitude modulator. A triangle waveform was chosen to drive the EO amplitude modulator so that the light transmitted by the modulator would be sinusoidally modulated. Data were taken with and without a quarter-wave plate (WPMQ05M-633 or WPMQ05M-980) placed in front of the EO amplitude modulator. When used, the quarter-wave plate was oriented with its axis parallel to the optic axis of the crystal. (Figure 2 shows the placement and orientation of the quarter-wave plate with respect to the modulator).

Click to Enlarge
Figure 3: The output of an EO amplitude modulator as a function of applied voltage when the setup does not include a quarter wave plate, diagrammed in Figure 1. The X marks on the curve correspond to conditions in the table columns.

As discussed in the EO Modulators tab and shown in Figure 3, when a quarter-wave plate is not placed at the input of the EO amplitude modulator (Figure 1), the light input to the modulator is linearly polarized and the applied voltage signal must vary from 0 V to V_π (or -V_π to 0 V) to fully modulate the amplitude of input light. The value of V_π was measured for the EO-AM-NR-C1 and a wavelength of 633 nm using the common method of overmodulating the modulator (apply a peak-to-peak voltage greater than V_π). V_π was measured to be 220 V at 633 nm, and V_π increases linearly with wavelength. As the voltage output of the HVA200 is capped at ±200 V, it cannot supply the voltage range necessary to fully modulate the output signal at this wavelength. If output voltage of the amplifier is saturated in an attempt to drive the modulator, the amplitude of the modulated optical signal will suffer distortion, as shown in Figure 4.

When a quarter-wave plate is placed in front of the modulator, the light entering the EO amplitude modulator is circularly polarized (as shown in Figure 2). The quarter-wave plate essentially adds an optical bias to the light input to the modulator. With this bias, it is not necessary for the voltage signal to vary between 0 and +V_π or -V_π in order fully modulate the amplitude of the input light: a drive voltage range of only -1/2 V_π to 1/2 V_π is needed to achieve an amplitude maximum and minimum of the optical signal. The table in Figure 5 illustrates the polarization at the output of EO crystal for a range of drive voltages, and the plot shows the amplitude of the light after the polarizer as a function of drive voltage. When a quarter-wave plate is used and the wavelength of the light source is 633 nm, a drive voltage range of ±110 V is required to fully modulate the EO amplitude modulator. This is well within the range of the HVA200, and the fully modulated optical signal produced using this technique is shown in Figure 6.

By including a quarter-wave plate in the optical setup, the available operating range of the amplifier is effectively doubled. With the bias supplied by the quarter-wave plate, the HVA200 is able to drive the EO-AM-NR-C1 across its entire wavelength operating range. This technique also enables the HVA200 to fully modulate the EO-AM-NR-C2 for wavelengths up to 1000 nm. As the price of amplifiers rises dramatically with the maximum voltage output, using this technique can result in significant cost savings. For details on the experimental setup employed and the results obtained, please click here.

Click to Enlarge
Figure 4: This plot shows the distorted modulation response that results when the amplifier is saturated in an attempt to fully modulate a signal whose V_π is greater than the maximum driving voltage that can be supplied by the amplifier. The wavelength of this optical signal is 633 nm.

Click to Enlarge
Figure 5: Inserting a quarter-wave plate (QWP) before the EO amplitude modulator, illustrated in Figure 2, shifts the relationship between the output amplitude and the applied voltage. For a bias voltage of 0 V, the output amplitude is halfway between min and max. The X marks on the curve correspond to conditions in the table columns.

Click to Enlarge
Figure 6: This plot shows the undistorted and fully-modulated 633 nm optical signal achieved by inserting a quarter-wave plate into the optical setup as shown in Figure 2, which makes available the full driving voltage range of the amplifier.