Apparatus and Method for Determination of Wavelength

1. Field of the Invention

The invention is in the field of wavelength measuring devices, particularly those devices utilizing interferometers for measuring the wavelength of monochromatic radiation.

2. Description of the Prior Art

Prior art optical fringe counting techniques are known utilizing interferometers of the Michelson type wherein the optical path length is varied by moving an optical carriage containing a reflecting element within the interferometer system. Interferometers of the Fabry-Perot type have also been widely utilized in studies of coherent radiation as is exemplified, for example, by U.S. Pat. No. 3,824,018 to Crane. In the Crane patent, however, it is necessary to rotate the Fabry-Perot etalon through various angles of incidence inasmuch as no spatial interference pattern is formed. Although wavelength measurements may be made using the resulting frequency pattern as a relative measure, the technique requires movement of the etalon and does not utilize both the spatial frequency (period) and phase of the interferometer pattern to provide an accurate measure of wavelength.

It is an object of the invention to provide a method and apparatus for accurately determining the wavelength of a light beam utilizing a Fizeau interferometer and without requiring the utilization of moving optical components.

It is a further object of the invention to provide a method and apparatus for determining the wavelength of a monochromatic radiation beam by measuring the spatial period and phase of the fringe pattern from a Fizeau-type interferometer.

Yet a further object of the invention is to provide a means for real time monitoring the wavelength of a light source such as a laser to provide a continual readout of the radiation wavelength.

Yet a further object of the invention is to provide a monitor and control apparatus for automatically measuring and controlling the wavelength of a tunable laser.

These and other objects of the invention will become clear in relation to the foregoing description taken in conjunction with the drawings wherein:

FIG. 1 is a diagramatic plan view of the apparatus in accordance with the invention;

FIG. 2 is a front view of the Fizeau interferometer taken along line 2--2 of FIG. 1;

FIG. 3 is a perspective view of a portion of the wavelength meter taken along line 3--3 of FIG. 1; and

FIG. 4 is a flow chart showing the overall program sequence utilized in the calculating means.

FIG. 1 illustrates a block diagramatic view of the wavelength measuring apparatus. The apparatus is utilized to determine the wavelength of a monochromatic source such as, for example, a laser source. The laser may be either pulsed or continuous wave and may be of the tunable type. The wavelength measuring apparatus may be utilized to determine in real time, the wavelength of the light from the laser, and consequently, the apparatus may be utilized as a monitor during the simultaneous use of a laser for other purposes. Consequently, a partially reflecting mirror 4 is positioned to divert a portion 6 of the light beam 5 into the wavelength measuring apparatus through a variable attenuator 7. The wavelength measuring apparatus is contained in a housing 8 which serves to house the various optical elements of the apparatus and isolate the interferometer and photoelectric detector array from extraneous light sources. The portion of the light beam 6 fed into the housing apparatus 8 is directed via a mirror 10 to an objective lens 11, which may be a simple microscope objective, which brings the light beam to a point focus at a small aperture in plate 12 serving as a spatial filter. The aperture may be on the order of 1-10 microns, for example. An achromatic collimating lens 14 is positioned one focal length away from plate 12 and produces a collimated beam directed toward a Fizeau interferometer 16. The Fizeau interferometer comprises two optically-flat plates 18 and 20 which are spaced from one another by means of a plurality of spacers 22-26. Spacers 22-26 provide a wedge angle and gap between the plates 18 and 20. The angle of the wedge is designated .alpha. and may typically be on the order of twenty seconds of arc. The nominal thickness of the gap is 1 millimeter. In FIG. 1, the wedge angle is shown greatly enlarged for purposes of illustration. The plates 18 and 20 may be disk-like as shown in FIG. 2 and fabricated from glass or fused silica. The spacers 22-26 may comprise low expansion glass-ceramic material such as Cer-Vit disk on the order of 1 mm thick and 6 mm in diameter. The gap inbetween the plates 18 and 20 may be simply an air gap or, alternately the interferometer 16 or indeed the entire housing 8 may be evacuated. One may additionally include heating means adjacent the interferometer 16 to reduce temperature variations and resulting dimensional changes.

The fabrication of the Fizeau interferometer 16 may be done using conventional techniques. Preferably, the plates 18 and 20 are made of fused silica having an anti-reflective coating on the non-gap forming surfaces. The plate surfaces which define the gap are uncoated to eliminate optical phase dispersion. The spacers 22-26 are glass soldered to one gap forming plate surface and the assembly is lapped to obtain the desired wedge angle. The angle may be tested using a standard autocollimator. The spacers are then polished to obtain an optically flat surface and are secured by optical contact with the other gap forming plate surface to produce the interferometer 16.

The Fizeau interferometer 16 produces a fringe pattern 27 such as shown in FIG. 2 which is seen to comprise a parallel array of light fringes 28 (maxima) and dark fringes 30 (minima). The fringe pattern 27 is produced by reflections from a back surface 32 of plate 18 and front surface 34 of plate 20. Light reflected from these two surfaces 32 and 34 produce the fringe pattern 27 which is subsequently reflected from the Fizeau interferometer 16 onto a photoelectric detector array 40 via a mirror surface 42. The mirror surface 42 is positioned below the laser beam which is incident on the Fizeau interferometer as best illustrated in FIG. 3. The Fizeau interferometer 16 is tilted such that the fringe pattern is reflected onto the mirror surface 42 and subsequently to the photoelectric detector array 40. The photoelectric detector array 40 serves to provide electrical signals corresponding to the intensity of light at discrete points along the fringe pattern 27. These signals are utilized to calculate the period between the fringes as well as the position of the fringes (phase) relative to an origin or reference point by electronic equipment which is typically housed outside the housing 8 and comprises a signal processing circuit 44, analog-to-digital converter 46 and calculating means 48. The output of the calculating means 48 is fed to a display 50 for readout of the calculated wavelength. Calculating means 48 may also provide the wavelength signals as control signals to a comparator circuit 60 which also receives a reference or desired wavelength signal. The output of the comparator circuit is fed to tuning means 64 for controlling the laser frequency so that an effective feedback tuning capability is provided. The comparator circuit 60 may alternately be implemented by the computing means 48 wherein the reference wavelength is an adjustable input parameter. The tuning means 64 includes D/A conversion means for providing an analog output signal for tuning the laser.

The photoelectric detector 40 may for example comprise a Reticon RL-1024C/17 photodiode array manufactured by Reticon Corporation, Sunnyvale, California. The diode array is essentially a monolithic self-scanning linear photodiode array which comprises 1024 separate elements, each 25.4 micrometers wide and spaced 25.4 micrometers on centers. The array elements may have an aperture length of 0.43 mm although other size apertures may also be employed. Each element or cell consists of a photosensor in parallel with a storage capacitor and is connected through a field effect transistor switch to a common video output line. The plurality of photosensor cells are scanned utilizing shift registers driven by two-phase clocks. The signal processing circuit 44 supplies the two-phase clock signals to the diode array 40 and receives the common video output signal which is amplified using an op amp connected as a current amplifier. The signal processing circuit 44 may comprise for example Model RC-100 and 103 array boards also manufactured by Reticon Corporation. The video output signal from the signal processing circuit 44 is fed to an A/D converter such as, for example, Model ADC-EH8Bl manufactured by Datel Systems, Inc., Canton, Mass. An eight bit digital output signal is provided from the analog-to-digital converter as an input to the computing means 48. A blanking signal which marks the end of a photodiode readout scan is also fed from the signal processing circuit 44 to the computing means 48. The fringe pattern is typically sampled and digitized every 20 milliseconds. The computing means 48 stores the fringe pattern for processing in accordance with the algorithm set forth in FIG. 4. Any number of commonly available minicomputers or microcomputers may be utilized as the computing means 48, and in preferred embodiment, an Interdata 7/16 computer is employed, programmed in accordance with the instructions set forth in Appendix A.

The interference fringes are measured using the photoelectric detector array 40 using, for example, the edge of the first element as the origin or reference point. It is also possible, however, to calculate a reference or origin using, for example, the midpoint between two markings 70 and 72 on the interferometer plate 18. The markings 70 and 72 may be narrow reflective strips placed on the outer surface of plate 18, and the computer software may be used to identify these markings by their narrowness and the saturated output of the associated diode elements. Calculating the origin in this fashion has the advantage of automatically self-calibrating the apparatus in the event of an extraneous mechanical vibration or jarring of the interferometer 16.

It is evident that the lens 14 may be replaced by a mirror and an off-axis optical scheme to eliminate chromatic aberrations. Additionally, the mirror 42 may be simply a beam splitter. Other equivalent optical steering and collimating designs will be evident to those of skill in the art.

The flow chart showing the overall algorithm of the instruction set of Appendix A is shown in FIG. 4. Step 1 in the flow chart of FIG. 4 is the initialization step which serves to reset and read various computer parameters and load the data reading starting address (address of SELCH) into a designated general purpose register (R3). Data is read into the computer (step 2) upon the occurrence of a blanking signal which is generated by the signal processing circuit 44 at the end of each photodiode array scan. The data is first checked for saturation (step 3) which would occur, for example, if extraneous light enters the housing 8. The saturation check is done by looking at every five consecutive data points to determine if they are all at a maximum count value which is 255 for the eight bit input data word. If saturation does occur a display FFFF is generated by the computing means on display 50 and the program is returned to the restart address RSTR of step 1.

If the input data is not saturated, the computer proceeds to step 4 to smooth and differentiate the received data. The smoothing and differentiating routine is simply a convolution of the raw data with a single period of a symmetrical square wave function of unity amplitude. The zero crossings of the convolution correspond to the maxima and minima of the fringe pattern 27. To find the zero crossing, designated u.sub.i ', the computer program interpolates the convolved data to the nearest 1/64 of a photodiode element spacing (step 5). It is recalled that the photodiode array comprises 1024 separate elements. Consequently, the zero crossing points u.sub.i ' are simply numbers between 1 and 65536 (1024.times.64) which represent the locations of the fringe minima and maxima at various points, i, along the photodiode array as measured in units of 1/64 of an element (cell) spacing (25.4 micrometers on center, and 25.4 micrometer wide cells). Nominally, these zero crossings are equally spaced and a plot of u.sub.i ' as a function of i should be linear. In practice the plates 18 and 20 will never be completely optically flat and will have a curvature which may be represented by a polynomial function. Step 6 of the program adjusts the zero crossings u.sub.i ' to correct for the plate curvature using a polynomial function experimentally determined during initial calibration. The resulting corrected zero crossings designated as u.sub.i are then fit with the most probable straight line fit by the method of least squares as indicated in step 7. The fitting line is given by u.sub.i =(LAMBDA/2)i+LL.sub.0. The slope of the straight line fit is half the fringe spacing and is given by: ##EQU1## where m=total number of zero crossings u.sub.i

i=1, 2, . . . m

LAMBDA=fringe spacing

The value of LL.sub.0 is the most probable position of the calculated zero crossing for i=0 and is given by: ##EQU2## For a fringe minimum at the calculated i=0 crossing the optical interference path difference L is given by:

where L.sub.0 is the optical path difference at the origin from which u.sub.i is measured, and is a calibration constant of the instrument previously determined using a laser with an accurately known wavelength. LL.sub.0, properly scaled, is then simply the change in the optical path difference between that at the origin and that at the calculated zero crossing for i=0. If however, there is a fringe maximum at the calculated zero crossing for i=0 then the fringe at the calculated zero crossing for i=1 must be a minimum. In this event, the optical path difference is given by:

LAMBDA is a measure of the spatial period of the fringe pattern, which is proportional to the wavelength of the light beam that produces the fringe pattern. When correctly scaled, LAMBDA is thus a first approximation to the wavelength of the light beam being measured. The true value of the wavelength, .lambda., is given at the first (or any) fringe minimum strictly by the formula:

where

L=the optical interference path difference at the fringe minimum, and

n=an integer (the interference order number)

The algorighm for calculating .lambda. is simply to calculate LAMBDA from equation (1) and L from equation (2) and either equation (3) or (4). The ratio L/LAMBDA will be approximately equal to the order number n, as per equation (5). In practice, however, there will be a remainder after dividing L by LAMBDA and the remainder may be used to correct the first approximation to the wavelength .lambda. as described below.

Steps 8-14 of the flow chart of FIG. 4 illustrate the above-described algorithm. To reduce round-off errors, the formulas used in the program to compute LAMBDA and LL.sub.0 have been modified slightly from equations (1) and (2). As a result, the units for both LAMBDA and LL.sub.0 are 1/128th of the photodiode element spacing. The calculated value of LAMBDA is checked to see if it is within certain limits, namely 66-195 photodiode elements to insure that previous computer steps have resulted in a reasonably correct value of LAMBDA. If LAMBDA is correctly within the acceptable limits, step 9 of the program converts LAMBDA and LL.sub.0 to appropriate units using the calibration factor SCALE. This factor SCALE is proportional to the interferometer wedge angle .alpha. and is experimentally determined using a laser of known wavelength during initial instrument calibration. Step 10 of the program determines whether the calculated zero crossing i=0 corresponds to a minimum or maximum light intensity value. This is done simply by looking at the slope of the zero crossing, i.e. positive slope corresponds to maximum and negative to minimum values. Step 11 of the program thus calculates L using either equation (3) or (4) in combination with equation (2). The value of L is then divided by LAMBDA in step 12. The nearest integral value of the quotient n.sub.0 is equal to the interferometer order number. The remainder is designated n.sub.1 and the relative remainder defined as n.sub.1 /n.sub.0 is the correction factor .delta.LAMBDA which is added to the first approximation LAMBDA to get the wavelength .lambda.. Step 13 of the program calculates .delta.LAMBDA=n.sub.1 /n.sub.0 and step 14 calculates the final value of .lambda.=LAMBDA+.delta.LAMBDA. The resulting output wavelength is displayed after conversion from binary to decimal form (B/D subroutine) on display means 50 as indicated in step 15. A detailed program listing in accordance with the flow chart of FIG. 4 is given in Appendix A.

In operation, the displayed value of .lambda. is updated 10-20 times a second. The instrument has a wavelength range of approximately 400 nm-1.1 um with possible resolution and accuracy of a few parts in 10.sup.8. The resolution and accuracy of the apparatus are determined by the phase measurement, somewhat as in the case of a Fabry-Perot interferometer, but the order number ambiguity of the phase measurement is removed by determining the spatial frequency of the fringes. The resulting instrument and method permits an absolute determination of the measured wavelength to an accuracy of a small fraction of the free spectral range of the interferometer.

It is evident that the apparatus of the invention may be utilized for measuring the wavelength of both CW and pulsed lasers. If a pulsed laser is employed it is necessary to sync the pulsing of the laser with the scanning of the linear photodiode array. The "start pulse" from the signal processing circuit 44 may conveniently be used for this purpose.

The apparatus and method of the invention are applicable for visible as well as non-visible light sources. For example, UV and IR sensitive photodiode arrays may be utilized in the disclosed apparatus. Further, it is not necessary to utilize laser radiation inasmuch as image dissecting photomultipliers will allow wavelength measurements of monochromatic non-laser sources. Additionally, modifications of the computer program will allow analysis and measurements of multi-line sources with any of these systems. ##SPC1## ##SPC2##