WO2002103412A2

WO2002103412A2 - Continually adjustable birefringence optic

Info

Publication number: WO2002103412A2
Application number: PCT/US2002/021644
Authority: WO
Inventors: Jason P. Sokoloff
Original assignee: Stratus Photonics Inc.
Priority date: 2001-01-29
Filing date: 2002-01-29
Publication date: 2002-12-27
Also published as: AU2002332408A1; WO2002103412A3

Abstract

A new optical unit, called a continually adjustable birefringence optic (Fig. 1, element 1) has an effective birefringence that can be continually adjusted from zero to a maximum value. It can be used to build devices that require adjustable time delays between pulses, such as an optical autocorrelator for short pulse lasers. It also has applications in devices that require a time delay to be generated between pulses of orthogonal polarization, such as polarization mode dispersion (PMD) generation or relief devices.

Description

CONTINUALLY ADJUSTABLE BIREFRINGENCE OPTIC

FIELD OF THE INVENTION The present invention relates in general to optical communications and more particularly to an optical apparatus having an effective birefringence that can be continually adjusted from zero to a maximum value.

CROSS REFERENCE TO RELATED APPLICATION This application claims priority to pending provisional application Serial No. 60/264299 filed January 29, 2001, herein incorporated by reference in its entirety.

BACKGROUND OF THE INVENTION Optical communications systems have an increasing need for greater bandwidth, speed and flexibility. Components such as tunable optical delays, autocorrelators, wave samplers, and Polarization Mode Dispersion (PMD) compensation and measurement devices play an important role in the implementation of optical communications systems. Limitation of such components, in terms of capacity, cost and efficiency often define the limitations of the overall system. The autocorrelators known in the prior art, for example, are difficult to align and operate. Such systems typically include a beam-splitter to create two resulting beams, or "arms," one of which must be adjusted or calibrated within a few tens of microns to cause the pulses from the two beams to overlap. The present invention overcomes significant limitations of prior art optics using a heretofore unknown method and apparatus.

SUMMARY OF THE INVENTION In one aspect the present invention is a continuously adjustible birefringent optical element that rotates about a surface normal such that light impinging on the surface normal is oriented such that for only one rotational orientation of optical element, the incident light beam propagates through the transmissive element in a direction parallel to the optic axis regardless of the polarization state of the device.

According to another aspect of the present invention, an optical autocorrelator includes a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom. The incident surface is oriented such that a linearly polarized incident optical beam impinges on the incident surface at an angle of 45 degrees to the surface normal, with the optically transmissive element having an optic axis oriented at a predetermined angle from the surface normal. The optically transmissive element is rotatable about the surface normal, and for one rotational orientation of said transmissive element, the incident optical beam which impinges on the optically transmissive element, propagates through the optically transmissive element in a direction parallel to the optic axis, regardless of the polarization state of the incident optical beam. There is a polarization rotator interposed between the incident optical beam and the optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator wherein the initial rotational orientation of the polarization rotator is offset by π/8 relative to the optically transmissive element such that two orthogonally polarized equal intensity beams exit from the birefringent optically transmissive element.

According to another aspect of the present invention, an optical waveform sampler includes a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom. The incident surface is oriented such that an incident optical beam impinges on the incident surface at an angle of 45 degrees to the surface normal. The optically transmissive element has an optic axis oriented at a predetermined angle from the surface normal. The optically transmissive element is rotatable about the surface normal, and for one rotational orientation of the optically transmissive element, incident light impinging on the optically transmissive element propagates through the optically transmissive element in a direction parallel to said optic axis, regardless of the polarization state of the incident optical beam. There is a polarization rotator interposed between and aligned with the incident light beam and the optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator whereby the incident light beam propagates through the optically transmissive element as a pure extraordinary beam. There is also an apparatus for generating a sampling waveform optical beam having the same frequency and orthogonal polarization relative to the incident optical beam. The sampling waveform optical beam impinges on the incident surface and is oriented so as to propagate through the optically transmissive element as an ordinary beam, whereby the incident and sampling beams temporally overlap to provide cross-correlation of their respective waveforms.

According to still another aspect of the present invention, an optical apparatus for determining Pulse Mode Dispersion includes a birefringent optically transmissive element with predetermined thickness and predetermined birefringence and has an incident surface with a surface normal extending therefrom. The incident surface is oriented such that an incident optical beam impinges on the incident surface at an angle of 45 degrees to the surface normal. The optically transmissive element also has an optic axis oriented at a predetermined angle from the surface normal wherein the optically transmissive element is rotatable about the surface normal, and for one rotational orientation of the transmissive element, incident optical beam impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to the optic axis, regardless of the polarization state of the incident optical beam such that the incident optical beam propagates as an extraordinary beam and and ordinary beam temporally separated, dependent upon the predetermined thickness and the predetermined birefringence of the optically transmissive element. There is also a polarization rotator interposed between the incident light beam and the optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator, whereby the polarization planes of the propagated extraordinary and ordinary beams remain stationary.

BRIEF DESCRIPTION OF THE DRAWINGS

Figs, la is a schematic illustration of the Continually Adjustable Birefringence Optic (CABO) of the present invention.

Fig. lb is a schematic illustration of the CABO shown in Fig. la, further depicting the ordinary and extraordinary beams for arbitrary c -axis.

Fig. 2a is a schematic illustration of the index ellipsoid of the index function n_e( ).

Fig. 2b is a schematic illustration of the extraordinary and ordinary beams at the surface of the CABO of the present invention.

Fig. 3 is a schematic illustration of the CABO of the present invention showing dimensions and angles used to calculate propagation delay.

Fig. 4a is a schematic illustration of a single pass tunable optical delay line using the CABO of the present invention.

Fig. 4b is a schematic illustration of a tunable optical delay line with a double pass configuration.

Fig. 5a is a schematic illustration depicting input and output beam relative positions for the double pass configuration of Fig. 4b. Fig. 5b is a schematic illustration depicting input and output polarization states for the double pass configuration of Fig. 4b.

Fig. 5c is a schematic illustration of the induced time delay for single beam inputs and outputs for the double pass configuration of Fig. 4b.

Fig. 6 is a schematic illustration of an autocorrelator using the CABO of the present invention.

Fig. 7 is a graphical representation of an oscilloscope sweep depicting two autocorrelation traces 30ms apart.

Fig. 8 is a schematic illustration of the temporal separation of orthogonally polarized pulses as a function of rotation angle.

Fig. 9 is a graphical illustration of the dependence of induced Polarization Mode Disperson on rotation angle.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT Referring to Fig. la, the continually adjustable birefringent optic (CABO) 1, of the present invention includes a birefringent crystal, preferably a flat calcite disk 3 having c-axis 5, which disk 3 is cut, oriented and mounted to rotate about its surface normal axis 7 such that for only one rotational orientation of the disk 3 incident light 8 travelling through the crystal has a propagation direction parallel to the c-axis 5 regardless of the polarization state of the incident light 8. The birefringent crystal may alternatively be rutile (titanium dioxide) or any doubly refracting crystal that separates a light beam into two diverging beams, commonly known to those skilled in the art, as ordinary beams (o-beams) 13 and extraordinary beams (e-beams) 15. (See Fig. lb). The disk 3 is oriented so that an incoming light beam 8 will have an angle of incidence, 9, to the surface normal 7 of 45 degrees. The orientation of the CABO internal c-axis 5 is determined by the extraordinary and ordinary indices of the crystal, n_eo and n₀ respectively, and is neither parallel nor perpendicular to the disk's surface 11. As indicated in Figs, la and lb, the angle 17 that the incident beam 8 makes with the c-axis 5 varies as the disk 3 is rotated about the surface normal 7.

Calculation of the angle 19 that the c-axis should be cut relative to the surface normal 7 is determined from the extraordinary and ordinary indices of the crystal. For calcite, and for light traveling along a principle axis (or optical axis) at 1550 nm the extraordinary and ordinary indices are respectively n_eo=1.4822 and n₀=1.649. However for a light beam travelling at an angle α relative to the c-axis 5 the extraordinary index is given by the index function n_e(α).

n_e(α) = { cos²(α)/n₀ ² + sin²(α)/n, 2 1-1/2

The index ellipsoid 2 of this function is shown in figure 2a. Referring to Fig. 2a, the c-axis 5 is depicted as the 0-180 axis and α, 17, is the angle that the extraordinary or ordinary beam makes relative to the c-axis 5. For a birefringent crystal there are two distinct speeds with which light can propagate, depending on the direction of propagation. When propagating along the c-axis, however, light has only one index of refraction (or speed).

Fig. 2b, depicts refraction of the extraordinary and ordinary beams at the surface of the CABO of the present invention where φ₀ and φ_e (20 and 22 respectively) are the angles of refraction of the ordinary and extraordinary beams.

Using Snell's law φ₀ = arcsin(sin(45)/n₀)) = 25.392 deg. It can be seen that light propagates only as ordinary beams when α = 0 because that is when n_e(α) = n₀.

When n_e(α) = n₀ Snell's law for the extraordinary beam is φ_e = arcsin(sin(45)/ n_e(α))) = arcsin(sin(45) /n₀)) = 25.392 deg. For α=0 the extraordinary beam is parallel to the c-axis. In other words, when the c-axis is cut at 25.392 and the disk is rotationally aligned so that the c-axis is parallel to the ordinary beam, then it is also parallel to the extraordinary beam. In this case all light will propagate through the crystal as an "ordinary" beam with polarization perpendicular to the c-axis. For all other rotational orientations of the disk, it is possible for light to break up into an ordinary and extraordinary beam which two components will take different paths through the disk and travel at different speeds. Rotating the crystal will change the orientation of the c-axis relative to the incoming beam, consequently the optical refractive index experienced by the ordinary beam won't change but the refractive index (and therefore the speed, optical path and optical path length) experienced by the extraordinary beam will change. Thus the effective birefringence (EB) is adjusted as the CABO of the present invention is rotated about the normal 7.

When the CABO is made from calcite the angle that the c-axis makes with the surface normal is about 25.4 degrees and it is slightly wavelength dependent.

As described below, for a one cm thick disk the induced propagation delay possible for a pulse at wavelength 1550 nm is about 2.4ps. Referring to Fig. 3, the propagation delay is determined by the optical path length difference between extraordinary 30 and ordinary 32 beams. Planes A (34) and B (36) are arbitrary starting and finishing planes for the measured rays. For a CABO of thickness L, 39, the optical path length of the refracted beam is determined analytically by the function

where θ is the angle of incidence 24, φ_t is the refracted angle φ₀ or φ_e, (20 or 22 respectively) and n_ray is the refractive index of the beam. For the ordinary beam 32, φ and n are calculated directly from Snell's law. For the extraordinary beam however, φ and n depend on the orientation of the c-axis 5 and, for a given orientation, may be calculated from n_e , φ_e , and φ_ray, where φ_e is the propagation direction of the extraordinary beam electric field; φ is the propagation direction of the extraordinary beam given by the Poynting vector; and n_e is the index of refraction of the extraordinary beam 30. As noted above, n_e is dependent upon α, the angle, 17, between the incident light and the c-axis. By way of example:

let θ = 45 deg, L = 1 cm, n₀ = 1.649, n_eo = 1.482. then φ₀ = 25.4 deg and the largest φ_ray is found to be 31.8 deg with a n_ray of 1.52. This gives a single pass time delay of F(n₀, φ₀) - F(n_ray, φ_ray) = 2.5 ps.

Thus, by adjusting the CABO's birefringence the transit time of an extraordinary beam through the CABO disk is adjusted. The time delay properties of the CABO have several useful applications including in tunable optical delays, autocorrelators and PMD compensation as will be more fully described below.

1. C ABO-Tunable Optical Delay Line (C ABO-TDL)

Tunable optical delay lines are used, for example, in systems which require precise timing or synchronization adjustments such as femtosecond and picosecond time resolution experiments, and in measurement devices which contain an optical delay line such as an optical autocorrelator. Fig. 4a shows a single pass tunable optical delay line 40 using the CABO of the present invention. Referring to Fig. 4a, if the input beam 8 to the CABO disk 3 is linearly polarized (depicted by arrows 18) and propagates as a pure extraordinary beam, then by rotating the CABO disk 3 about the surface normal 7 it can be seen that the CABO of the present invention functions as a tunable optical delay line. As will be understood by those skilled in the art, as the CABO rotates, the input beam polarization plane must also rotate in order for the light within the disk 3 to propagate as an e-beam. This can be done by using a polarization rotator 10 (such as a half -wave plate) placed before the CABO disk 3 to set the input beam to the proper polarization. If desired, a second half-wave plate 12 can be used at the output to restore the input polarization state 18.

Figure 4b shows a double pass tunable optical delay line using the CABO of the present invention. The double pass configuration uses a retroreflector 14 to give the CABO-TDL a fixed output beam position. Referring to Figs. 5a and 5b the retroflector 14 is made from a penta prism 60 and a right angle prism 62 which arrangement laterally displaces a pair of orthogonal beams (See Fig. 5a) while maintaining individual beam polarizations 55 (Figl. 5b) and preserving relative beam positions 63 and 64.. (See Fig. 5a.) As depicted schematically in Fig. 5c, when the inputs and outputs are single beams, the induced time delay 55 is double that of a single pass delay line using the CABO of the present invention.

In both the single pass configuration of Fig. 4a and the double pass configuration of Fig. 4b, the initial orientation of the CABO disk 3 and the halfwave- plate(s) 10 (and 12) are such that they all have their c-axes in a vertical plane and the input polarization state 18 is either horizontal or vertical. For both configurations the rotation angle 4 of the CABO disk 3 is twice that of the halfwave plate(s) 10. .

CABO-Optical Autocorrelator An optical autocorrelator measures the temporal width of a short optical pulse. In an optical autocorrelator a pulse stream is split into two equal intensity beams, and one is temporally delayed relative to the other. The time delay varies continuously from zero to some maximum. As will be more fully described below, both the process of splitting a pulse into two equal intensity components, and of generating a continuous time delay between components can be done easily by rotating the CABO of the present invention at some angular velocity.

In one embodiment, splitting a pulse into two equal intensity components with continuous time delay between components may be accomplished using the configuration shown in Fig. 4a (the second halfwave plate 12 is not needed) where the initial orientation of the halfwave plate(s) 10 is π/8 rather than zero. This will result in equal intensity o-and e-beams propagating through the CABO disk as long as the rotation rate of the CABO disk 3 is twice that of the halfwave plates 10 and 12. As the CABO disk 3 and halfwave plates 10 and 12 rotate, the propagation time of only the e-beam changes, and a time delay between e- and o-beams is generated which varies from zero to some maximum value in a continuous way. The output of this arrangement is two orthogonally polarized beams, (not shown) whose orientations rotate at the same rate as the CABO disk.

A second and simpler embodiment requires just a rotating CABO disk. In this second embodiment, shown in Fig. 6, the input 8 to the disk 3 must be a pulse train with circular polarization 61. This insures that there will always be equal intensity o- and e-beams within the disk 3. To prepare a circularly polarize'd input, a fixed non-rotating quarter wave plate 62 (or properly oriented Fresnel Rhomb) can be placed just before the CABO disk 3. The two output beams 65 and 66 of the CABO autocorrelator 60 will then always be orthogonal, and their polarization planes will rotate at the same rate as the CABO disk 3 spins. As understood by those skilled in the art, a polarization insensitive detector 67 based on two-photon absorption, such as an LED or semiconductor detector is preferable to a polarization sensitive SHG crystal.

The CABO autocorrelator embodiment shown in Fig. 6, may be demonstrated using a Ti:S laser operating at about 790nm, that outputs a polarized 82 MHz pulse train of several hundred fs pulses as a light source. For simplicity and convenience a Fresnel rhomb, rather than a quarter wave plate, can be used to circularly polarize the input beam. A CABO crystal 3mm thick, AR coated for 800nm and cut with the c-axis at 25.4 degrees to the surface normal, can be rotated around its surface normal 7 at about 33 Hz at 45 degrees to the input beam 8. A

Hamamatsu GaAsP diffusion detector 67 with a band edge at 680 nm³ that is largely transparent to the 790nm pulse train can be used to generate a larger signal due to two-photon absorption when the two orthogonal pulses (65 and 66) exiting the CABO crystal 3 temporally and spatially overlap within the detector 67 A n oscilloscope trace of the autocorrelation signal is shown in Fig. 7.

In Fig. 7 a time sweep that includes two autocorrelation traces 30 ms apart, (71 and 72) - one centered at about .006 s and one at .036 s - is shown. The prominent features of the traces are that each trace has a small background and large autocorrelation signal. The fringes at the top of each peak are due to the interferometric nature of the device. These fringes can either be filtered out by an appropriate RC circuit at the detector, or enhanced by rotating the CABO crystal slower.

Although the time scale in Fig. 7 is not scaled to "femtoseconds of time delay", the trace can still be used to crudely estimate the optical pulse width. As indicated above, for a 180-degree rotation of the crystal, a 1 cm CABO crystal generates a 2.4 ps delay between the two orthogonally polarized components. Therefore the single pass autocorrelator shown in Fig. 6, with its 3 mm crystal displaces one component relative to the other by 2.4ps/cm x 0.3cm =0.72 ps for a half rotation. The second half of the rotation moves the pulses 0.72 ps in the other direction, to bring them back together again. Since an autocorrelation trace is symmetric, a full rotation is equivalent to a 1.44 ps scan. The autocorrelation peak in Fig. 7 is about 9ms wide (FWHM) so an estimate of the pulse width is 9/30 x 1.44 x 2/3 = 288 fs. (The factor of 2/3 is due to the sinh² shape of the optical pulse generated by the laser.) This experimental value of 288 fs is slightly more than the value of 200 fs estimated based on the spectral width of the optical pulse and may be due to broadening of the pulse by chromatic dispersion that takes place as the pulse travels through the rather lengthy Fresnel rhomb.

It will be appreciated by those skilled in the art that the CABO autocorrelator provides significant advantages over autocorrelators which have two separate optical paths and a scanned linear delay line. In these prior art types of autocorrelators the two "arms" of the autocorrelator must first be matched to within a few tens of microns. Then one arm is repetitively scanned over some small distance by varying that arm's optical path length. Since the scanned length is not known, a calibration procedure must be performed. In the CABO autocorrelator of the present invention perfect temporal overlap occurs for a particular rotational orientation of the CABO crystal which can be known in advance, or just found by spinning the CABO. Furthermore, no calibration is required since the temporal scan of the CABO autocorrelator depends on the thickness of the CABO unit, and can be measured to a very high degree of accuracy.

3. CABO-Optical Waveform Sampler

Another application of the CABO of the present invention is as an optical waveform sampler. It is sometime necessary to know the shape of a pulse that has been distorted by propagation through some material. Providing the CABO with two optical pulse trains of equal frequency and orthogonal polarization can do this. One input is made of pulses of the waveform which is to be investigated. The second input includes narrow pulses that will "sample" the waveform of the first input. These two inputs to the CABO disk are oriented so that one propagates as the ordinary beam and the other as the extraordinary beam. As the CABO disk spins, one beam is delayed and the pulse in this beam will overlap temporally with a different part of the measured waveform. The intensity at the two-photon- absorption-sensitive detector will be a product of the overlap of these two pulses, and the photocurrent produced by the detector will be proportional to this intensity. If this electrical signal is sent to an oscilloscope, then the scope screen will trace out the cross-correlation of the waveform under investigation.

Except that the waveform sampler has two optical inputs, the optical configuration of the CABO-Waveform Sampler will be the same as the CABO-TDL. (See Fig. 4a) There will be an initial halfwave plate that spins to keep the beams properly oriented within the CABO disk.

4. CABO Polarization Mode Dispersion (PMD)

The CABO disk can be used to delay a pulse of one polarization with respect to an orthogonally polarized pulse. It can also be used to generate a relative delay between the orthogonal polarization components of a single pulse. These features make the CABO of the present invention advantageous for compensation of Polarization Mode Dispersion (PMD). The most widespread application of the CABO may well be for PMD mitigation for high speed (40 Gbit/s ) optical communication systems.

5. CABO-PMD Generator/ Emulator/ Measurement Device

Inputting a pulse to the CABO, which always has a mixed polarization state within the CABO disk, insures that the o-beam and e-beam will exit the crystal temporally separated. As noted above, this temporal separation can be continually adjusted from zero to a maximum dependent on the thickness and birefringence of the material used in the CABO. This effectively introduces a PMD into an optical pulse or continuous wave beam of a controllable amount. An adjustable PMD can be used as a reference to characterize and measure the PMD of an unknown device. It can also be used to generate a PMD which is in turn used to test another system's tolerance or reaction to PMD.

The optical setup for this CABO device is similar to that shown in Fig. 4b. A retroreflecting scheme is best used because the double pass doubles induced time delays and also produces an optical output which has both exiting o- and e-beams on the same path. This is necessary if the output is to be fiber coupled. If the polarization planes at the device output must remain stationary, then a half wave plate which rotates through half the angle of the CABO disk, must be used as a retarder. Otherwise, a stationary quarter wave plate can be used.

The oscilloscope trace shown in Fig. 7 can also be interpreted as the result of generating a PMD that varies with rotation angle. This is shown schematically in Fig. 8, which illustrates the temporal separation, τ, 85, of orthogonally polarized pulses 81 and 82 as a function of rotation angle. When the rotation angle is small and the components partially overlap, a two-photon signal is generated at the detector 67. Increasing the rotation angle increases τ, and decreases the overlap signal.

Fig. 9 shows the dependence of induced PMD on the rotational angle. It can be seen that a maximum PMD of amount τ_max can be emulated by a π rotation of the CABO crystal. Other amounts of PMD can be continuously introduced and, as shown in Fig. 9, there is a direct analytical dependence of PMD on the rotational angle.

6. CABO-PMD Mitigation Device /O-beam Pulse and E-beam Pulse Combiner

As temporally short optical pulses travel through optical fiber (or other optical media), optical birefringence causes a temporal walk-off between orthogonally polarized components of the pulses. This in turn leads to errors in the optically encoded data transfer. Orthogonally polarized, temporally separated pulses can have their temporal separation reduced or eliminated by a CABO PMD device made up of a halfwave plate, CABO disk, and retroref lector (see Fig.4b). The CABO disk 3 is rotated to a position so that its effective birefringence compensates for the temporal displacement between the two orthogonal components. The halfwave plate 10 is then oriented so that the leading polarization component is aligned on the slow axis of the CABO disk. The retroreflector 14 doubles the available PMD relief possible, and permits the output to be fiber coupled.

The foregoing applications based on a new device called a CABO, - which is an optic with an adjustable birefringence wherein the birefringence is simply adjusted by rotation around the surface normal such that only the extraordinary index changes due to this rotation - . are meant to be illustrative, not exhaustive. Those skilled in the art will appreciate the advantages of the present invention in allowing the CABO to either delay an extraordinary polarized beam, or a light wave's extraordinary component relative to its ordinary component. This delay is not necessarily a linear function of rotation angle, but this can be compensated for in any of the above devices since the delay as a function of rotation angle can be either directly calculated or measured. While the present invention has been described with reference to the preferred embodiment and applications thereof, it will be understood by those skilled in the art that various obvious changes may be made, and equivalents may be substituted for elements thereof, without departing from the essential scope of the present invention. Therefore it is intended that the invention not be limited to the particular embodiments disclosed, but that the invention includes all embodiments falling within the scope of the appended claims.

Claims

WHAT IS CLAIMED:

1. An optical apparatus comprising

a birefringent optically transmissive element having first and second opposed surfaces each of said surfaces with a surface normal extending therefrom, the optically transmissive element having a optic axis oriented at a select angle from said surface normal such that an optical beam received at said first opposed surface will be split into extraordinary and ordinary optical beams transiting through the optically transmissive element at different speeds; an apparatus for positioning said optically transmissive element to receive said optical beam at a predetermined angle relative to said first opposed parallel surface; and an apparatus for moving said optical beam relative to said incident optical surface to present said optical beam to said optically transmissive element at a plurality of presentation angles relative to said optic axis, thereby creating an effective birefringence whose value varies with said presentation angle from zero to a maximum value.

2. The apparatus of claim 1 wherein said optically transmissive element further comprises calcite crystal.

3. The apparatus of claim 1 wherein said optically transmissive element further comprises rutile (titanium dioxide).

4. The apparatus of claim 1 further comprising: an apparatus for linearly polarizing said input optical beam and wherein said optically transmissive element is in a substantially disk form; and a means for rotating said optically transmissive element and said polarization apparatus such that said optical beam propagates within said disk as an extraordinary beam.

5. The apparatus of claim 4 wherein said polarization apparatus comprises a half wave plate.

6. The apparatus of claim 5 further comprising a second half wave plate at an output of said disk positioned to receive an extraordinary beam exiting said disk and restore an input beam polarization state.

7. The apparatus of claim 1 wherein said optically transmissive element is further characterized in that there is one rotational orientation of said optically transmissive element relative to said surface normal such that said optical beam propagates through said optically transmissive element in a direction parallel to said optic axis regardless of the polarization state of the incident light.

8. An optical apparatus comprising: a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom, said incident surface oriented such that an optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal, wherein said optically transmissive element is rotatable about said surface normal, and for one rotational orientation of said transmissive element, said incident optical beam propagates through said optically transmissive element in a direction parallel to said optic axis regardless of the polarization state of said incident optical beam.

9. The optical apparatus of claiml further comprising: a means for presenting said optical beam to said incident surface with linear polarization; and a polarization rotator interposed between said incident optical beam and said optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator whereby said incident optical beam propagates through the optically transmissive element as a pure extraordinary beam.

10. The optical apparatus of claim 9 further comprising a second polarization rotator aligned with said propagated extraordinary beam to restore the polarization state of the incident optical beam.

11. The optical apparatus of claim 9 further comprising a retroflector aligned with said propagated extraordinary beam to maintain said propagated extraordinary beam in a fixed position and polarization state relative to said incident light beam.

12. An optical autocorrelator comprising: a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom, said incident surface oriented such that a linearly polarized incident optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal; wherein said optically transmissive element is rotatable about said surface normal, and for one rotational orientation of said transmissive element, incident optical beam impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to said optic axis, regardless of the polarization state of the incident optical beam; and a polarization rotator interposed between said incident opticla beam and said optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator; and wherein the initial rotational orientation of said polarization rotator is offset by π/8 relative to said optically transmissive element , whereby two orthogonally polarized equal intensity beams exit from the birefringent optically transmissive element.

13. An optical autocorrelator comprising a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom, said incident surface oriented such that a circularly polarized incident optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal, wherein said optically transmissive element is rotatable about said surface normal, and for one rotational orientation of said transmissive element, incident light impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to said optic axis, regardless of the polarization state of the incident optical beam; and a non-rotating quarter wave plate interposed between said circularly polarized incident optical beam and said optically transmissive element whereby two orthogonally polarized equal intensity optical beams exit from the birefringent optically transmissive element.

14. An optical waveform sampler comprising: a birefringent optically transmissive element having an incident surface with a surface normal extending therefrom, said incident surface oriented such that an incident optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal; wherein said optically transmissive element is rotatable about said surface normal, and for one rotational orientation of said transmissive element, incident light impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to said optic axis, regardless of the polarization state of the incident optical beam; a polarization rotator interposed between and aligned with said incident light beam and said optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator whereby said incident light beam propagates through the optically transmissive element as a pure extraordinary beam; and a means for generating a sampling waveform optical beam having the same frequency and orthogaonal polarization relative to said incident optical beam, said sampling waveform optical beam impinging on said incident surface and oriented so as to propagate through said optically transmissive element as an ordinary beam, whereby said incident and sampling beams temporally overlap to provide cross-correlation of their respective waveforms.

15. An optical apparatus for determining Pulse Mode Dispersion comprising a birefringent optically transmissive element with predetermined thickness and predetermined birefringence and having an incident surface with a surface normal extending therefrom, said incident surface oriented such that an incident optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal; wherein said optically transmissive element is rotatable^' about said surface normal, and for one rotational orientation of said transmissive element, incident light impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to said optic axis, regardless of the polarization state of the incident optical beam such that the incident optical beam propagates as an extraordinary beam and and ordinary beam temporally separated dependent upon said predetermined thickness and said predetermined birefringence; and a polarization rotator interposed between said incident optical beam and said optically transmissive element such that the rotational orientation of the optically transmissive element is twice that of the polarization rotator whereby the polarization planes of said propagated extraordinary and ordinary beams remain stationary.

16. An optical apparatus for determining Pulse Mode Dispersion comprising a birefringent optically transmissive element with predetermined thickness and predetermined birefringence and having an incident surface with a surface normal extending therefrom, said incident surface oriented such that an incident optical beam impinges on said incident surface at an angle of 45 degrees to the surface normal, the optically transmissive element having an optic axis oriented at a predetermined angle from said surface normal; wherein said optically transmissive element is rotatable about said surface normal, and for one rotational orientation of said transmissive element, incident optical bean impinging on said optically transmissive element propagates through said optically transmissive element in a direction parallel to said optic axis regardless of the polarization state of the incident optical beam; a stationary quarter wave plate interposed between said incident optical beam and said optically transmissive element such that the incident optical beam propagates as an extraordinary beam and and ordinary beam temporally separated dependent upon said predetermined thickness and said predetermined birefringence.