Background of the Invention
1. Field of the Invention
The present invention relates generally to linear frequency chirped lasers, and more particularly to a linear frequency chirped laser system that is diffraction grating driven.
2. Description of the Prior Art
The most common system used in the prior art to produce "chirping" uses an electro-optic crystal within the laser resonator. An increasing voltage applied to the crystal causes its index of refraction to change, changing the optical length of the resonator cavity and causing the resonator's longitudinal modes to tune.
The specifics of such a prior art system will be discussed for waveguide CO.sub.2 lasers. The disadvantages and problems of the prior art include, but are not limited to, the following. The electro-optic crystals are typically cadmium telluride (CdTe) having dimensions of 2 mm'2 mm in cross-section and 5 cm long. The crystals must be polished on all surfaces, with the 2 mm.times.2 mm ends being of optical quality and including antireflection coatings. The electro-optic crystals of cadmium telluride are very expensive and very fragile.
The electro-optic effect requires a ramp of 0-1600 volts across the 2 mm crystal width. The ramp must be very accurately linear, and it must be typically swept in 3 to 30 .mu.sec. This is a difficult requirement combining both high voltages and radio frequencies (RF). Furthermore, the crystal has an index of refraction of 2.7 so its optical length is 13.5 cm. This length, plus the length for the laser gain, requires that the laser resonator free spectral range (FSR) is limited to 500 MHz, and the laser cannot reach the chirp amplitude of 1 GHz desired for some applications without adding longitudinal mode suppression to an already very complex device. The present invention adds less than 2 cm to the optical length of the laser, so the laser can easily reach the 1 GHz chirp amplitude.
The electrode connections must be made between two of the 2 mm.times.5 cm faces of the crystal and the high voltage RF power must not arc to other parts of the crystal holder or flash-over from one electrode to the other.
Residual strain in the crystal or strain induced by the crystal holder causes transverse variations in the index of refraction, which distorts the optical beam and leads to optical loss within the resonator and a degraded beam quality of the laser output. The strain also causes birefringence, which converts one polarization of the light into another. This causes further optical distortions because the electro-optic effect is different for the different polarizations. It also increases the optical losses because some of the optical components will be polarization selective (the spectral line-selecting diffraction grating, for example, and possibly the waveguide bore) .
Electro-optic crystals are also piezoelectric, so the applied voltage ramp also causes a dimensional change. Because the voltage ramp occurs rapidly, a spectrum of acoustic waves are generated within the crystal. The acoustic energy will fracture the crystal if it is not removed. Even if attenuated enough so that fracture of the crystal will not occur, the acoustic energy will still induce dynamic strain-optic effects. Therefore, this energy must be removed to a very high degree of completeness.
The crystal also absorbs some optical power, which shows up as heat that must also be removed. The heat input is not uniform across the crystal, but is proportional to the optical intensity, which is most intense on the optical axis, farthest from the cooling walls. The absorbed power sets up a temperature gradient within the crystal, which causes index of refraction gradients and mechanical strains, which in turn cause strain-optic effects. These effects limit the circulating optical power and the achievable laser power and beam quality.
The electro-optic crystal must be aligned with the waveguide bore to within very exacting tolerances in both offset and angle or the insertion losses of the modulator become very large and the optical quality of the laser output drops. This requires a precision adjustable crystal holder and a very difficult alignment procedure. The crystal holder must also absorb the acoustic energy, route the high voltage RF, and provide the cooling, all without straining the crystal. The very best electro-optic modulators should have insertion losses of about 6% round-trip, and more typically they are about 10%. The present invention has an insertion loss of less than 1% round-trip.
The chirp linearity and repeatability limit the resolution of a radar system using a chirped laser transmitter. The linearity of the electro-optic driven chirp depends mainly on the linearity of the voltage ramp with time and the piezo-electric/strain-optic dynamic disturbances in the index of refraction. Other effects such as the thermal and strain-induced index of refraction gradients and birefringence will also contribute to non-linearities in the resulting optical frequency chirp. Similarly, the repeatability of the chirp slope depends not only on the repeatability of the voltage drive, but also on the dynamic strain-optic effects and the temperature dependence of the electro-optic coefficient. The prior art systems can keep deviations from linearity of the chirp ramp to 1% of the chirp amplitude, and the value of the chirp slope can be maintained constant within approximately 1% from chirp-to-chirp. The present invention very significantly improves these numbers by at least a factor of 100? The system of the present invention solves substantially all of the problems of the prior art while avoiding its shortcomings.
Brief Summary of the Invention
A frequency modulated radar transmitter with a repetitive linear increase or decrease in frequency with time (called a "frequency chirp") is required for many radar applications. This invention provides a very simple way of obtaining such a "chirped" frequency modulation for laser radar transmitters. The system of the present invention can generate wider, faster, more linear, and more repeatable chirps than the prior art, and therefore radars made utilizing the concept of the present invention will have higher resolution. The present invention can do this with lower laser internal losses and therefore significantly higher laser output power and better beam quality. The present system is also significantly cheaper to build, easier to maintain, and less prone to failures.
A diffraction grating moving in its own plane imparts a Doppler shift to the diffracted beam. If the diffraction grating is used as the end mirror or end-reflector of a resonator, the Doppler shift is indistinguishable from that due to a mirror changing the resonator's length. In the direction of the resulting diffracted beam, light diffracted off each facet of a diffraction grating has a one wavelength optical path difference from the light diffracted off of adjacent facets. When the grating has moved exactly one groove period in its own plane and its state is indistinguishable from its original state, each facet has moved forward one-half wavelength, and the resonator has tuned through one free spectral range (FSR).
More generally, the moving diffraction grating is not restricted to use as an end-reflector but can be used as any turning or folding optic in the resonator. At these non-end-reflector positions, the grating of course is aligned such that the diffracted beam is in the desired new direction rather than back upon the incident beam. This moving-grating laser chirper is therefore not restricted to ordinary (non-ring) resonators but can be used in ring resonators as well. When used in non-ring resonators in other than the end-reflector position, it interacts with the light on both the forward and reverse passes, which doubles the induced chirp rate.
A simple plane diffraction grating (either transmissive or reflective) moving in its own plane will cause the laser containing it to chirp through successive FSRS. As this continues, the available width of the grating will have passed through the beam path, and if the process is to continue, the grating will have to be stopped and brought back. If the laser is used during this return time, the laser will frequency chirp in the opposite sense from that during the forward grating stroke. This type of action may be useful in some applications, but the reciprocating motion makes it very difficult to drive the grating at a uniform speed to obtain chirps all of the same slope.
Because the grating is moving in its own plane, the reciprocal action can be eliminated by ruling the grating on the rim of a wheel and rotating the wheel to provide the required Doppler shift continuously, without changing the actual location of the grating. Similarly, the diffraction grating can be ruled radially on the planar face of a rotating wheel to provide the required Doppler shift continuously without changing the actual location of the grating.
In one embodiment, the system of the present invention utilizes a diffraction grating on the outer peripheral rim of a rotating wheel disposed within a laser cavity to produce the equivalent of the tuning of the longitudinal modes of the resonator with a moving mirror but without having to eventually stop and reverse direction.
A second embodiment uses a diffraction grating ruled radially on the planar face surface of a rotating wheel to provide the tuning of the longitudinal modes of the resonator.
In one example, a high performance chirp can be made using a 1 GHz frequency ramp in 3 .mu.sec. Some applications can use lesser requirements of 100 MHz in 30 .mu.sec. The present invention can far exceed even the high performance requirement set forth above.
These and other objects and advantages of the present invention will be more fully understood after reading the detailed description of the preferred embodiments, the claims, and the drawings, which are briefly described hereinbelow.
Brief Description of the Drawings
In the accompanying drawings:
FIGS. 1A, 1B, 1C are graphic illustrations representing the tuning of the longitudinal modes to repetitively sweep out one FSR.
FIG. 2 is a schematic illustration of a generalized embodiment of the present invention.
FIG. 3 is a schematic illustration representing another embodiment of the present invention.
FIG. 4 is a partial end view showing the diffraction grating on the peripheral rim of the wheel of FIG. 3 taken along view lines 4--4 thereof.
FIG. 5 is a partial schematic illustration of the diffraction grating on the outer peripheral rim surface of the wheel, as blown up from within the circle designated by reference number 5 of FIG. 3.
FIG. 6 illustrates how the cylindrical corrector lens eliminates optical aberrations by causing all rays to impinge on the diffraction grating on the outer peripheral rim of the wheel at the same angle of incidence.
FIG. 7 is a schematic illustration of a laser system for simultaneously generating both an up-chirp and a down-chirp signal using the embodiment of FIG. 3.
FIG. 8 is a schematic illustration of yet another embodiment of the present invention.
FIG. 9 is a schematic illustration of the embodiment of FIG. 8 viewed from the side angle to show the impingement on the face of the rotating wheel.
FIG. 10 is a schematic illustration of an alternative to the system of FIG. 7 using the embodiment of FIGS. 8 and 9 therein.
Detailed Description of the Preferred Embodiment
This application is closely related to my copending application entitled "Phase Plate or Spiral Phase Wheel Driven Linear Frequency Chirped Laser," application Ser. No. 07/871,888, filed Apr. 2, 1992, the same date as this application and assigned to the same assignee, the disclosure of which is incorporated herein.
Referring first to the diagrams of FIG. 1A, 1B, and 1C, a laser resonator can operate at any of its longitudinal mode frequencies. Which mode will lase is determined by the gain of the medium--the longitudinal mode at the frequency with the highest gain will lase (for homogeneously broadened gain media as in the CO.sub.2 waveguide laser). By moving an end mirror of the resonator or otherwise changing the optical length path of the resonator, the longitudinal modes will tune, and so will the operating frequency of the laser. The laser frequency will follow the longitudinal mode until another longitudinal mode is closer to the gain line center and therefore has a higher gain. The laser will then change to this higher gain longitudinal mode. The tuning range available, called the free spectral range (FSR), is the frequency separation between longitudinal modes and is given by the formula FSR=c/2L, where "c" is the velocity of light and "L" is the optical length of the resonator. The FSR is 1 GHz for a 15 cm resonator optical length. The FSR will be swept every time the resonator optical length changes by one-half wavelength. The tuning can be done at essentially any rate because the Doppler shift off of the moving mirror exactly matches the frequency shift of the tuning longitudinal mode, or the element changing the optical length changes the frequency of the light within itself so that emerging light exactly matches the frequency shift of the tuning longitudinal mode.
FIG. 2 shows a first embodiment of the present invention wherein a diffraction grating 19 moving in its own plane imparts a Doppler shift to the diffracted beam. FIG. 2 shows a laser 11 which emits an optical beam 13 within the resonant cavity and along the optical path of the laser. The light beam 13 passes through an output mirror 15 of the laser 11 as output beam 20. At the other end of the laser resonator, beam 13 is designated as beam 17 and is shown as impinging on the face 23 of the moving diffraction grating 19. The diffraction grating 19 is shown as having a base 21 and a face 23, which contains the actual diffraction grating. The arrow 25 shows the direction of movement of the diffraction grating in its own plane.
It will, of course, be recognized by those skilled in the art, that the designation of laser 11 is used herein as a shorthand method for designating anything not actually shown in the laser. Actually, the reference numeral 11 designates the gain media and the term "laser" actually includes the gain media, the resonant cavity, the output mirror, any corrective optics in the optical path, the end-reflector (for non-ring resonators), and any optical devices inserted into the optical path. With this clarified, the gain media will continue to be referred to as the laser, and it is defined to encompass the gain media and any other laser component not expressly shown.
If the diffraction grating is used as the end optic occupying the end-reflector position of a resonator, as in FIG. 2, the Doppler shift caused by the moving grating is indistinguishable from that due to a mirror changing the resonator's length. The light diffracted off of each facet of the diffraction grating 23 in the direction of the resonator axis 17, 13 has one wavelength optical path difference from the light diffracted off of adjacent facets, since that is the manner in which a diffraction grating functions. When the grating has moved exactly one groove period in its own plane, and its state is indistinguishable from its original state, each facet has moved forward one-half wavelength, and the resonator has tuned through one FSR. Because the grating is moving in its own plane, it can be put on the face or rim of a rotating wheel, for example, to provide the required Doppler shift continuously, without changing its actual location, as shown in FIG. 3.
FIG. 3 shows a waveguide laser 27 having a central waveguide bore 29 wherein the laser 27 emits a laser output beam 13 through output mirror 15. The optical path of the resonator further includes the optical beam 17, which passes through the corrective lens 31 and impinges at a point 43 on the outer peripheral rim surface 37 of a rotating wheel 33 having a pair of opposed, generally planar, circular faces 35. The wheel 33 rotates counterclockwise in a direction shown by the arrow 41 about its axis of rotation 39. As shown in FIGS. 3, 4, and 5, the outer peripheral rim surface 37 includes a diffraction grating thereon. Observe that whatever power is absorbed by the grating is spread out over the "entire peripheral rim 37 of the wheel 33 and is not concentrated on the beam spot area on the grating.
FIG. 4 shows a partial sectional side view of the outer peripheral rim portion 37 of the wheel 33 of FIG. 2 and includes the raised portions or peaks 45 and the depressions or grooves 47 making up the diffraction grating on the surface 37 of the wheel 33. FIG. 5 shows a blown-up partial section of the outer peripheral rim 37 of the wheel 33, as designated within the circle 5 of FIG. 3, and shows a side view of the raised portions 45 and depressions or grooves 47 comprising the diffraction grating on the outer peripheral rim surface 37 of the wheel 33.
Putting the diffraction grating on the outer peripheral rim 37 of the wheel 33 makes the diffraction grating act as a divergent cylindrical mirror, with spherical aberration, instead of as a flat mirror. The spherical aberration is exactly corrected by the cylindrical converging lens 31, which makes all rays arrive at the same angle of incidence at the diffraction grating as shown in FIG. 6.
FIG. 6 shows the center 7 and two extreme 6, 8 rays passing through lens 31 to arrive at the grating on the rim surface 37 of wheel 33. The central ray 7 impinges at point 43, while the extreme rays 6, 8 impinge at points 42 and 44, respectively. Lens 31 bends the rays such that they all arrive at the grating surface 37 at the same angle of incidence. Each ray is then retrodiffracted to return along its own path. The lens is mainly cylindrical, with some higher order corrections. The required cylinder and corrections do not change as the wheel is rotated, so they can be completely corrected with an appropriately figured lens.
With or without the cylindrical lens, all of the diffracted beam gets the same Doppler shift. After the grating has moved in its own plane one groove spacing, its state is identical to that before the move, and so the state of the diffracted beam must also be identical to that before the grating motion. The phase of the diffracted light at any location must therefore have been changed by exactly one wavelength (or n wavelengths if the diffraction was into the n-th order). Since each grating motion of one groove causes a phase shift of one wavelength, then a grating motion of m grooves per second causes a phase change in the diffracted light of m cycles per second (for a first order diffraction). The Doppler shift due to a diffraction grating moving in its own plane is therefore equal to the velocity of the grating measured in grooves per second.
A specific example is described hereinbelow. For a conventional CO.sub.2 laser operating at 10.6 .mu.M wavelength, a diffraction grating of 150 lines/mm is used at an angle of incidence of 52.7 degrees. To tune through one FSR in 3 .mu.sec, the wheel 33 must have a tangential velocity of (1/150)mm/3 .mu.sec=2.2 m/sec. If the wheel is 5 cm in diameter, it will rotate at 14 Hz, or 850 rpm. A system like the one shown in FIG. 2 would add less than two centimeters to the overall resonator optical length, compared to 14 cm of optical length required for the electro-modulator, so the "chirp" could be a full 1 GHz and could actually be swept in less than 3 .mu.sec, depending on the rotation speed and the diameter of the wheel.
For very fast chirps, there might be some problem with the laser's having time to establish a new mode when it switches longitudinal modes as indicated in FIG. 1C. The new mode must build up from random noise and if the chirp is very fast, the chirp might be over before the mode has had time to come up to significant power. In these cases, a seed laser should be used to establish the new mode above the noise level. The seed beam can be injected into the chirped laser through, for example, the zeroth or other unused order diffraction from the moving grating. The seed frequency should be set slightly higher (for an up-chirp system; slightly lower for a down-chirp system) than the beginning frequency of the operating free spectral range of the chirp. (The moving grating will impart a small Doppler shift, but not a chirp, to the seed beam.) When the chirped laser longitudinal mode reaches the seed frequency, the chirped laser will pick up from the seed beam and chirp from there.
The chirp linearity of this invention depends on the uniformity of the speed of rotation of the wheel during one chirp. It would be very difficult to make a rotating device that was not highly uniform in rotation speed over such a short time. For example, consider a 5 cm diameter wheel that has a mass of only 10 g and is rotating at 14 Hz. To change the rotation speed by 1% during a 3 msec chirp time would require a torque of 0.3 nt-m (0.2 ft-lb). Considering that the motor necessary to drive this device has a torque of only 1/100 of this amount, rotational uniformity during a chirp would probably be better than one part in one million.
Another non-linearity mechanism is mode pulling. The laser does not lase exactly at the frequency of the cavity longitudinal mode but is pulled slightly towards the center of the gain line. How much it is pulled depends on the "Q" of the laser cavity, the frequency of the gain transition, the linewidth of the gain transition, and the tuning of the cavity resonance off the center of the gain line. The majority of the mode pulling is proportional to the tuning of the cavity mode and so does not give rise to a chirp non-linearity. A calculation of the mode pulling for a CO.sub.2 laser at 10.6 .mu.m with a 20 cm laser cavity, a 20% combined loss and output coupling (which gives a "Q" of 100,000), a gain linewidth of 1/2 GHz, and a tuning over a 1 GHz FSR gives a deviation from linearity of one part in 200,000.
The chirp slope repeatability requirement depends on the signal direction and processing techniques--some schemes require long term repeatability, while other schemes require repeatability only over times of up to 10 .mu.sec, or a few chirps.
Chirp repeatability will depend on the uniformity of the average groove spacing over the illuminated spot size. This can be held better than 0.01% around the entire circumference of the wheel, and several orders of magnitude better than that over a few chirps. Repeatability also depends on the uniformity of the speed of rotation of the wheel during a revolution and over many revolutions, which can be held below 0.01% long term and several orders of magnitude better than that of a few chirps. Thermal expansion of the groove spacing will effect long term repeatability, but this is less than 0.002%/.degree. C., and can be lowered further with low thermal expansion materials.
Some frequency modulated radar systems require unidirectional chirps while others require up-chirps and down-chirps. The prior art electro-optic devices favor the bidirectional chirps because the electro-optic voltage must be brought back to the starting value, and unidirectional chirps require very fast reset times on the order of 10 nsec to bring the voltage back down from 1600 volts to 0. This invention favors unidirectional chirps.
Some radar schemes require chirp-to-chirp coherence; that is, the optical phase is synchronized from one chirp to the next. The seed injection described above starts each chirp up in phase with the seed laser, so it provides chirp-to-chirp coherence, even if the seed is not needed for rapid initiation of the chirp for short chirp durations.
In situations that require bidirectional chirps, it is often not necessary that the up-chirps and down-chirps alternate on a chirp-by-chirp basis. It is acceptable to alternate a sequence of up-chirps with a sequence of down-chirps, as long as the alternation frequency is fast enough, such as 4 msec of up-chirps followed by 4 msec of down-chirps, with a 1 or 2 msec in between for switching. Such alternation could easily be accomplished with two counter-rotating concentric wheels. The outer wheel would be slotted so that half of the time the inner would be visible and drive one type of chirp, and half of the time the outer would be controlling, driving the other type of chirp. This is the simplest chirp alternation scheme, but there are others that would be obvious in view of the present invention if the chirping and switching times given above are too long.
FIG. 7 shows a two laser system for generating both an up-chirp and a down-chirp signal simultaneously. FIG. 7 shows the first laser 49 having a waveguide bore 53 and a second laser 51 having a waveguide bore 55. A first laser output 57 passes through an output mirror 61 while the second emitted laser output 59 passes through an output mirror 63. The second beam 59, which represents the down-chirp signal, is then impinged upon and reflected off of the face of a mirror 65 at point 66 as reflected beam 71. Beam 71 is then transmitted through a one-half waveplate optical device 67 as beam 73 and impinged off of a point 74 on the surface of a polarizing beamsplitter device 69. The beamsplitter 69 also passes the output beam 57 of the first laser 49, so that the beams are combined and both the up-chirp signal (represented by the first laser output 57) and the down-chirp signal (represented by the second output laser 59) are combined as the optical output 75 containing both the up-chirp and the down-chirp signals, simultaneously.
A cylindrical corrective lens 81 is disposed within the optical path of the resonator cavity of the first laser 49, and the optical beam 77 passing therethrough is impinged off of a point 93 on the outer peripheral rim 87 of a rotating wheel 85. The wheel 85 rotates counterclockwise as indicated by the arrow 91 about its axle or shaft 89 comprising its axis of rotation.
The second laser 51 has a cylindrical corrective lens 83 within its optical path, and beam 79 impinges off of point 95 on the outer peripheral surface 87 of the rotating wheel 85. This configuration illustrates how two separate laser systems, or lasers 49 and 51, can use a common rotating wheel 85 having a diffraction grating on its outer peripheral rim surface 87 to generate both an up-chirp signal and a down-chirp signal, simultaneously. The beams are of orthogonal polarizations so they do not interfere with one another, and they are combined at the polarizing beamsplitter 69 and simultaneously outputted as the common output 75.
FIGS. 8 and 9 represent an alternate embodiment of the system of FIG. 3. Again, a diffraction grating moving in its own plane, as shown in FIG. 1, imparts a Doppler shift to the diffracted beam. If the diffraction grating is used as or in place of the end-reflector or mirror of a resonator, the Doppler shift is indistinguishable from that due to a mirror changing the resonator's length.
In this case, the diffraction grating is ruled radially on the face of a rotating wheel to provide the required Doppler shift continuously, as shown in FIGS. 7 and 8. The diffraction grating on the face of a wheel, however, adds some distortion to the diffracted beam. The grooves closer to the center of the wheel are closer together than those farther out, so there is a small variation in diffracted angle over the horizontal extent of a beam spot size. Also, the angle between the incident beam and the grating grooves changes slightly over the vertical extent of the beam, such that light at the top of the spot is diffracted to the left and light at the bottom of the spot is diffracted to the right. These are all static effects, and so they can be corrected with a corrector plate. One might also expect that since the grooves farther from the center of the wheel are moving faster than those nearer the center, that they might impart a larger Doppler shift to the diffracted beam. However, this is compensated by the change in the diffracted angle caused by variation in groove spacing, so that there is no difference in frequency shift. The frequency shift is simply equal to the number of grooves per second passing any point (for a grating used in first order), which is the same for all radial distances. Note also that whatever power is absorbed by the grating is spread out over the entire path of the beam on the wheel and is not concentrated on the beam spot area of the grating.
FIG. 8 shows a laser 11 emitting a laser output beam 13 which passes through an output mirror 15 as the laser output beam 14. Within the optical path of the laser's resonator, depicted as optical beam 17, is a corrector plate 97, and the beam 17 is shown as impinging at a point 107 on the diffraction grating 105 disposed on the face 101 of a rotating wheel 99. The wheel 99 includes a front surface or face 101, an opposite surface or face 103, an outer peripheral rim portion 109, and an axle or shaft 89 through which the axis of rotation 123 passes. The diffraction grating 105 on the front face 101 of the wheel 99 includes a plurality of radially-oriented, spaced diffraction grooves as previously described. FIG. 9 shows the system of FIG. 8 from a side angle so that the point of impingement 107 of the beam 17 on the face 101 of the wheel 99 is more clearly shown. A specific example will be described herein below.
For a CO.sub.2 laser operating at 10.6 .mu.M wavelength, a grating of 150 lines/mm is used at an angle of incidence of 52.7 degrees. To tune through one FSR in 3 .mu.sec, the wheel must have a tangential velocity of (1/150) mm/3 .mu.sec=2.2 m/sec. If the wheel is 5 cm in diameter, it must rotate at 14 Hz, or 850 rpm. A system like that shown in FIGS. 8 and 9 would add less than 2 cm to the resonator's optical length, compared to 14 cm of added optical length for the electro-optic modulator, so the chirp would be a full 1 GHz and could actually be swept in much less than 3 .mu.sec, depending on the rotational speed and the diameter of the wheel. The distortion due to the variation in groove spacing in the radial direction amounts to 1/10 wavelength over a 2 mm spot size. The distortion due to the upper part of the spot being diffracted to the left and the lower portion of the spot being diffracted to the right is 3 wavelengths. These distortions can be easily removed with a single corrector plate.
The chirp linearity of this invention depends on the uniformity of the speed of rotation of the wheel during one chirp. It would be very difficult to make a rotating device that was not highly uniform for the reasons given previously. Another non-linearity mechanism is mode-pulling, and this is identical to that previously described with respect to FIG. 3. The chirp slope repeatability requirement again depends on the radar signal detection and processing techniques used. Some schemes requiring long-term repeatability while others only require repeatability over a few times up to 10 .mu.sec, or over a few chirps.
If up-chirps and down-chirps must be within one chirp period of each other, the two-laser system of FIG. 10 could be used. FIG. 10 shows a first laser 111 emitting the laser output signal 57, representing the down-chirp signals, which passes through the output mirror 61 and then passes through a polarizing beamsplitter 69 as one portion of the combined output beam 75. Within the optical path of the resonator cavity of laser 111 is a corrector plate 119 through which the beam 115 passes and impinges upon the diffraction grating 105 which is radially disposed on the face surface 101 of the wheel 99 as a series of radial lines radiating from the center to the outer peripheral edge of the face 101. The wheel 99 rotates in a clockwise direction, as illustrated by the arrow 91, about the rotational axis 123.
A second laser 113 emits a laser output beam 59 representing an up-chirp signal. This signal is passed through an output mirror 63 and reflected off of the face of a mirror 65 as reflected beam 71. Beam 71 passes through a one-half wavelength plate 67 and emerges as beam 73, which impinges on the polarizing beamsplitter 69 to combine with the down-chirp 57 at the combined output 75. Since the up-chirp signal 59 and the down-chirp signal 57 are of orthogonal polarization, they are combined as the output beam 75 without interfering with one another, as conventionally known. The optical path of the resonant cavity of laser 113 also includes a corrector plate 121 through which the optical beam 117 passes to impinge upon an opposite and adjacent portion of the diffraction grating 105 of the face plate 101 of the wheel 99 opposite the point at which the beam 115 of the first laser 111 impinges.
The system of FIG. 10 will, therefore, transmit both up-chirps and down-chirps simultaneously. In practice, both lasers 111 and 113 could not run off of the same diffraction grating because it would be extremely difficult to shape the grooves for high efficiency for both directions of illumination. One laser could run off of one ring of radial grooves blazed for its direction of illumination, and the other laser could run off of a concentric ring of grooves blazed for the other direction. Alternatively, one grating could be on top of the wheel and the other on the bottom; or for a pedestrian solution, two separate wheels could be used. The two outputs with orthogonal polarizations are combined by the polarizing beamsplitter 69. The half waveplate could be eliminated by folding the path out of plane to rotate the plane of polarization, or one could start with the lasers initially in orthogonal polarizations.
Another method of transmitting up- and down-chirps simultaneously is to run the two lasers of FIG. 7 or FIG. 10 on different spectral lines and combine the two output beams with a diffraction grating or a prism.
The present invention does not teach simply the first and second embodiments utilizing a diffraction grating placed on the outer peripheral rim of a wheel or a grating radially disposed on one of the planar faces of a wheel, but more generally teaches the use of a diffraction grating moving in its own plane to cause a laser to "chirp". Any ordinary diffraction grating moving in its own plane will cause chirping, but eventually, one reaches the end of the grating and has to stop and pull it back. Putting the diffraction grating on a wheel is simply one means to continuously generate chirps without the necessity of having to bring the grating back. The wheel also makes it easier to maintain a uniform grating velocity, the velocity determining the chirp rate of change of frequency. An endless loop or belt could also be used to carry a diffraction grating moving in its own plane and other possibilities will occur to those of ordinary skill in the art given Applicant's disclosure herein.
Even the ordinary diffraction grating, with its limited duration of continuous chirping, may have applications as a radar chirper in some areas. If 10,000 chirps are required, they could be obtained with systems of the prior art by shortening the laser cavity by moving the end mirror or grating along the optical axis for 10,000 half wavelengths, or 5 cm for a 10.6 .mu.n wavelength CO.sub.2 laser. That is, the laser cavity is shortened (or lengthened for down-chirps) by 5 cm. It may not be possible to change the laser cavity by this much for many lasers, particularly waveguide lasers, where such length change would increase the optical losses and the laser would extinguish. But a diffraction grating slid along its own plane a distance such that the component of the motion along the laser axis is the required "shortening" will cause the same number of chirps without any real change in the cavity length. For a 10.6 .mu.M wavelength CO.sub.2 laser used with a 150 line/nun diffraction grating, the actual grating motion in its own plane would be 6.7 cm for the grating to be moved 10,000 of its own groove spacings.
It will be understood by those skilled in the art that various modifications, changes, variations, substitutions, and alterations can be made in the systems, methods, and apparatus of the present invention without departing from the spirit and scope thereof, which is limited only by the appended claims.