Tunable laser and laser current source

BACKGROUND OF THE INVENTION

[0001] 1. Field of the Invention

[0002] This invention relates to laser systems and, more particularly, to electronic controllers for controlling and monitoring operations of a tunable laser, such as an external cavity diode laser.

[0003] 2. Description of the Related Art

[0004] Tunable external cavity diode lasers (ECDLs) are widely used in lightwave test-and-measurement equipment and are becoming recognized as essential components for the rapidly expanding fields of wavelength division multiplexed (WDM) optical voice and data communications. The many applications within these fields represent many different sets of performance specifications. However, the following requirements are typical: small size of the optomechanical assembly and control system;, servo control of the wavelength; and controllable frequency modulation (FM) at audio rates (e.g., 100 Hz to 30 kHz) in order to broaden the linewidth.

[0005] To achieve desired control over the operation of external cavity diode lasers, electronic controllers are typically provided that implement various functionality. This functionality may include a current source for providing current to the laser, locked wavelength tuning functionality, a modulation source, and various other functionality to precisely control and monitor operation of the tunable laser. It is typically desirable that the electronic controller allow for versatile control of the tunable laser with reasonable efficiency and a relatively small form factor. It is also typically desirable that electrical noise within the system and its impact upon various measurement functions be minimized.

SUMMARY OF THE INVENTION

[0006] A laser control system for monitoring and controlling various functions of a laser assembly is provided. In one embodiment, the laser assembly comprises a tunable external cavity laser. The laser controller may include a wavelength tuning circuit for adjusting and locking the wavelength of the external cavity. The tuning circuit may include a modulation signal generator for providing a modulation signal to a selected transmission element that causes a corresponding modulation of the optical path of the laser external cavity. Wavelength locking may be achieved by monitoring transmission characteristics that vary due to the slight modulation of the optical path. Such transmission characteristics may be monitored, for example, by detecting variations in the voltage across a gain medium or variations in the intensity of light associated with the laser external cavity. The tuning circuit may include a signal processor such as a microprocessor that performs a Fourier Transform, such as a Fast Fourier Transform, upon data indicative of the transmission characteristics to thereby generate an error signal for adjusting the length of the optical path of the external cavity.

[0007] To perform various monitoring and control functions, the controller may include circuitry for monitoring various parameters associated with operation of the laser, such as temperature indicating signals and/or signals from light detectors such as photodiodes. The controller may additionally detect other parameters, such as a voltage across a gain medium. In one embodiment, the sensing of such parameters is performed synchronously with the generation of various control signals for controlling operation of the external cavity laser. The control signals may include signals for adjusting the external cavity pathlength and for generating a modulation signal. The control signals may be in the form of pulse-width modulated signals, which may be generated by a programmable logic device. In one embodiment, temperature-dependent resistive elements such as thermistors may be used to provide signals indicative of the temperature of various components of the laser assembly. A switching circuit may be employed to couple a selected temperature-dependent resistive element to a common measurement path for detecting a temperature associated with the selected temperature-dependent resistive element. In yet a further embodiment, a control circuit for generating a modulating output signal may include a transformer including a primary coil coupled to an amplifier circuit in a push-pull configuration. A laser current source may be provided that includes a control circuit for controlling a level of current supplied to a laser device through a drive transistor, and a common gate or common base configured transistor coupled between the control circuit and a control terminal of the drive transistor. The laser controller may include a network interface to allow remote control of the laser.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008]FIG. 1 is a block diagram illustrating various aspects of a tunable external cavity laser.

[0009] FIGS. <b>2</b>A-<b>2</b>C and FIGS. <b>3</b>A-<b>3</b>C are diagrams illustrating pass band relationships associated with an external cavity laser.

[0010]FIG. 4 is a functional block diagram illustrating aspects of one embodiment of a laser controller.

[0011]FIG. 5 is a diagram illustrating a relationship of a modulation signal with respect to detected voltage modulation across a gain medium.

[0012]FIG. 6 is a hardware block diagram illustrating various aspects of one embodiment of a laser controller.

[0013]FIG. 6A is a flow diagram illustrating one embodiment of an algorithm for performing wavelength locking.

[0014]FIG. 7 is a circuit diagram illustrating one embodiment of a laser current source.

[0015]FIG. 8 is a circuit diagram illustrating one embodiment of an analog interface.

[0016]FIG. 9 is flow diagram illustrating a method for performing temperature measurements.

[0017]FIG. 10 is a circuit diagram illustrating one embodiment of an amplifier circuit for generating a modulation signal.

[0018] While the invention is susceptible to various modifications and alternative forms, specific embodiments are shown by way of example in the drawings and are herein described in detail. It should be understood, however, that drawings and detailed description thereto are not intended to limit the invention to the particular form disclosed, but on the contrary, the invention is to cover all modifications, equivalents and alternatives falling within the spirit and scope of the present invention as defined by the appended claims.

DETAILED DESCRIPTION

[0019] The term “external cavity laser” as used herein is meant to encompass any laser device wherein at least one external reflective element is used to introduce optical feedback into a gain medium. “External reflective element” means a reflective element that is not actually a part of, or integral to, the gain medium itself.

[0020]FIG. 1 illustrates various aspects of one embodiment of a tunable external cavity laser apparatus <b>10</b>. The apparatus <b>10</b> includes a gain medium <b>12</b> and an end or external reflective element <b>14</b>. Gain medium <b>12</b> may comprise a conventional Fabry-Perot diode emitter chip having an anti-reflection (AR) coated rear facet <b>16</b> and a partially reflective front facet <b>18</b>. Front facet <b>18</b> and end mirror <b>14</b> define an external cavity for the apparatus <b>10</b>. Gain medium <b>12</b> emits a coherent beam from rear facet <b>16</b> that is collimated by lens <b>20</b> to define an optical path <b>22</b> which is co-linear with the optical axis of the external cavity. Rear and front facets <b>16</b>, <b>18</b> of gain medium <b>12</b> are aligned with the optical axis of the external cavity as well. Light reflected from end mirror <b>14</b> is fed back along optical path <b>22</b> into gain medium <b>12</b>. Conventional output coupler optics (not shown) may be associated with front facet <b>18</b> for coupling the output of external cavity laser <b>10</b> into an optical fiber (also not shown).

[0021] Transmission characteristics of the external cavity can be probed or evaluated by monitoring the voltage across gain medium <b>12</b>. In this regard, first and second electrodes <b>24</b>, <b>26</b> may be positioned adjacent to and operatively coupled to gain medium <b>12</b>. First electrode <b>24</b> is operatively coupled to a controller <b>48</b> via a conductor <b>28</b>, where the voltage across gain medium <b>12</b> may be detected. A second electrode <b>26</b> is grounded via conductor <b>30</b>. It is noted that in various other embodiments, transmission characteristics of the external cavity may be alternatively evaluated by monitoring the output of one or more photodiodes <b>15</b> (or any other type of light detector) that may be positioned to receive portions of light associated with the external cavity laser. For example, in one embodiment, a photodiode may be positioned to detect light propagating through the partially reflective end mirror <b>14</b>. In another embodiment, a photodiode may be positioned to detect light propagating through front facet <b>18</b> of gain medium <b>12</b>. Photodiodes may be positioned to detect light associated with the external cavity laser at other particular locations, as desired.

[0022] Error signals may be derived from the voltage measured across gain medium <b>12</b> to correct or otherwise adjust the transmission characteristics associated with the external cavity. Details regarding specific implementations of this functionality are provided further below.

[0023] Other transmission elements associated with the external cavity may include a grid generator element and a channel selector element, which are respectively shown in FIG. 1 as a grid etalon <b>34</b> and a wedge etalon <b>36</b> positioned in optical path <b>22</b> between gain medium <b>12</b> and end mirror <b>14</b>. Grid etalon <b>34</b> typically is positioned in optical path <b>22</b> before wedge etalon <b>26</b>. Grid etalon <b>34</b> operates as an interference filter, and the refractive index and optical thickness of grid etalon <b>34</b> give rise to a multiplicity of minima within the communication band at wavelengths which coincide with the center wavelengths of a selected wavelength grid which may comprise, for example, the ITU (International Telecommunications Union) grid. Other wavelength grids may alternatively be selected. Grid etalon <b>34</b> thus has a free spectral range (FSR) which corresponds to the spacing between the gridlines of the ITU grid or other selected grid, and the grid etalon <b>34</b> thus operates to provide a plurality of pass bands centered on each of the gridlines of the wavelength grid. Grid etalon <b>34</b> has a finesse (free spectral range divided by full width half maximum or FWHM) which suppresses neighboring modes of the external cavity laser between each channel of the wavelength grid.

[0024] Grid etalon <b>34</b> may be a parallel plate solid, liquid or gas spaced etalon, and may be tuned by precise dimensioning of the optical thickness between its faces by thermal expansion and contraction via temperature control. The grid etalon <b>34</b> may alternatively be tuned by tilting to vary the optical thickness between faces <b>38</b>, <b>40</b>, or by application of an electric field to an electro-optic etalon material. Various other grid generating elements are known to those skilled in the art and may be used place of grid etalon <b>34</b>. Grid etalon <b>34</b> may be thermally controlled using a temperature controller (TEC) <b>66</b> to prevent variation in the selected grid which may arise due to thermal fluctuation during operation of external cavity laser <b>10</b>. Grid etalon <b>34</b> alternatively may be actively tuned during laser operation.

[0025] Wedge etalon <b>36</b>, like grid etalon <b>34</b>, acts as an interference filter, but with nonparallel reflective faces <b>42</b>, <b>44</b> providing a tapered shape. Wedge etalon <b>36</b> may comprise a tapered transparent substrate, a tapered air gap between the reflective surfaces of adjacent transparent substrates, or a thin film wedge interference filter as described further below.

[0026] The relative size, shape and distances between the various optical components of external cavity laser <b>10</b> are in some instances exaggerated for clarity and are not necessarily shown to scale. External cavity laser <b>10</b> may include additional transmission elements (not shown), such as focusing and collimating components, and polarizing optics configured to remove spurious feedback associated with the various components of external cavity laser <b>10</b>. The location of grid generator <b>34</b> and channel selector <b>36</b> may vary from that shown in FIG. 1.

[0027] Wedge etalon <b>36</b> defines a plurality of pass bands which are substantially broader than the pass bands of the grid etalon <b>34</b>, with the broader pass bands of the wedge etalon <b>36</b> having a periodicity substantially corresponding to the separation between the shortest and longest wavelength channels defined by the grid etalon <b>34</b>. In other words, the free spectral range of the wedge etalon <b>36</b> corresponds to the full wavelength range of the wavelength grid defined by grid etalon <b>34</b>. The wedge etalon <b>36</b> has a finesse which suppresses channels adjacent to a particular selected channel.

[0028] The wedge etalon <b>36</b> may be used to select between multiple communication channels by changing the optical thickness between faces <b>42</b>, <b>44</b> of wedge etalon <b>36</b>. This is achieved by translating or driving wedge etalon <b>36</b> in a direction parallel to the taper of wedge etalon <b>36</b> and perpendicular to optical path <b>22</b> and the optical axis of external cavity laser <b>10</b>. Each of the pass bands defined by the wedge etalon <b>36</b> supports a selectable channel, and as the wedge is advanced or translated into optical path <b>22</b>, the beam traveling along optical path <b>22</b> passes through increasingly thicker portions of wedge etalon <b>36</b> which support constructive interference between opposing faces <b>42</b>, <b>44</b> at longer wavelength channels. As wedge etalon <b>36</b> is withdrawn from optical path <b>22</b>, the beam will experience increasingly thinner portions of wedge etalon <b>36</b> and expose pass bands to the optical path <b>22</b> which support correspondingly shorter wavelength channels. The free spectral range of wedge etalon <b>36</b> corresponds to the complete wavelength range of grid etalon <b>34</b> as noted above, so that a single loss minimum within the communications band can be tuned across the wavelength grid. The combined feedback to gain medium <b>12</b> from the grid etalon <b>34</b> and wedge etalon <b>36</b> support lasing at the center wavelength of a selected channel. Across the tuning range, the free spectral range of the wedge etalon <b>36</b> is broader than that of grid etalon <b>34</b>.

[0029] Wedge etalon <b>36</b> is positionally tuned via a tuning assembly which comprises a drive element <b>46</b> structured and configured to adjustably position wedge etalon <b>36</b> according to selected channels. Drive element <b>46</b> may comprise, for example, a stepper motor together with suitable hardware for precision translation of wedge etalon <b>36</b>. Drive element <b>46</b> may alternatively comprise various types of actuators, including, but not limited to, DC servomotors, solenoids, voice coil actuators, piezoelectric actuators, ultrasonic drivers, shape memory devices, and like linear actuators.

[0030] Drive element <b>46</b> is operatively coupled to controller <b>48</b> which provides signals to control the positioning of wedge etalon <b>36</b> by way of drive element <b>46</b>. Controller <b>48</b> may include a data processor and memory (not shown in FIG. 1) having lookup tables of positional information for wedge etalon <b>36</b> which correspond to selectable channel wavelengths.

[0031] When external cavity laser <b>10</b> is tuned to change from one communication channel to another, controller <b>48</b> may convey signals to drive element <b>46</b> according to positional data in a lookup table, and drive element <b>46</b> translates or drives wedge etalon <b>36</b> to a position wherein the optical thickness of the portion of the wedge etalon <b>36</b> positioned in optical path <b>22</b> provides constructive interference which supports the selected channel. A position detector <b>50</b> such as a linear encoder may be used in association with wedge etalon <b>36</b> and drive element <b>46</b> to ensure correct positioning of wedge etalon <b>36</b> by driver <b>46</b>. Alternatively, a single point position electro-optic detector may be provided to locate a “home” position associated with wedge etalon <b>36</b> during initialization of the system.

[0032] An electro-optically activated modulation element <b>58</b> is also shown positioned in optical path <b>22</b> before end mirror <b>14</b>. In the embodiment of FIG. 1, end mirror <b>14</b> is formed as a reflective coating directly on the electro-optic material of modulation element <b>58</b>. Thus, the end mirror <b>14</b> and modulation element <b>58</b> are combined into a single component. In other embodiments, end mirror <b>14</b> may be formed on an element that is separate from modulation element <b>58</b>. Details regarding the function of modulation element <b>58</b> will be provided further below.

[0033] The pass band relationship of the grid etalon <b>34</b>, wedge etalon <b>36</b> and the external cavity defined by front facet <b>18</b> and end mirror <b>14</b> are illustrated graphically in FIG. 2A through FIG. 2C, which show external cavity pass bands PB<b>1</b>, grid etalon pass bands PB<b>2</b>, and wedge etalon pass bands PB<b>3</b>. Relative gain is shown on the vertical axis and wavelength on the horizontal axis. As can be seen, free spectral range of the wedge etalon <b>36</b> (FSR_{Channel Sel}) is greater than the free spectral range of the grid etalon <b>34</b> (FSR_{Grid Gen}), which in turn is greater than the free spectral range of the external cavity (FSR_cavity). The band pass peaks PB<b>1</b> of the external cavity periodically align with the center wavelengths of pass bands PB<b>2</b> defined by the wavelength grid of grid etalon <b>34</b>. There is one pass band peak PB<b>3</b> from the wedge etalon <b>36</b> which extends over all of the pass bands PB<b>2</b> of the wavelength grid. In the specific example shown in FIGS. <b>2</b>A-<b>2</b>C, the wavelength grid extends over sixty four channels spaced apart by one half nanometer (nm) or 62 GHz, with the shortest wavelength channel at 1532 nm, and the longest wavelength channel at 1563.5 nm.

[0034] The finesse of grid etalon <b>34</b> and wedge etalon <b>36</b> determine the attenuation of neighboring modes or channels. As noted above, finesse is equal to the free spectral range over the full width half maximum, or finesse=FSR/FWM. The width for a grid etalon pass band PB<b>2</b> at half maximum is shown in FIG. 2B, and the width for a wedge etalon pass band PB<b>3</b> at half maximum is shown in FIG. 2C. The positioning of grid etalon <b>34</b> and wedge etalon <b>36</b> within the external cavity improves side mode suppression.

[0035] The tuning of the band pass PB<b>3</b> of wedge etalon <b>36</b> between a channel centered at 1549.5 nm and an adjacent channel at 1550 nm is illustrated graphically in FIGS. <b>3</b>A-<b>3</b>C, wherein the selection of a channel generated by grid etalon <b>24</b> and the attenuation of adjacent channels or modes is shown. The external cavity pass bands PB<b>1</b> shown in FIGS. <b>2</b>A-<b>2</b>C are omitted from FIGS. <b>3</b>A-<b>3</b>C for clarity. The grid etalon <b>34</b> selects periodic longitudinal modes of the external cavity corresponding to the grid channel spacing while rejecting neighboring modes. The wedge etalon <b>36</b> selects a particular channel in the wavelength grid and rejects all other channels. The selected channel or lasing mode is stationary at one particular channel for filter offsets in the range of approximately plus or minus one half channel spacing. For larger channel offsets the lasing mode jumps to the next adjacent channel.

[0036] In FIG. 3A, the wedge etalon pass band PB<b>3</b> is centered with respect to the grid channel at 1549.5 nm. The relative gain associated with pass band PB<b>2</b> at 1549.5 nm is high, while the relative gain levels associated with adjacent pass bands PB<b>2</b> at 1549.0 nm and 1550.0 nm are suppressed relative to the selected 1549.5 nm channel. The gain associated with pass bands PB<b>2</b> at 1550.5 nm and 1548.5 nm is further suppressed. The dashed line indicates the relative gain for pass bands PB<b>2</b> without suppression by wedge etalon <b>36</b>.

[0037]FIG. 3B shows the wedge etalon pass band PB at a position in between the channels at 1549.5 nm and 1550.0 nm, as occurs during channel switching. The relative gain associated with pass bands PB<b>2</b> at 1549.5 nm and 1550.0 are both high, with neither channel suppressed. The relative gain levels associated with pass bands PB<b>2</b> at 1549.0 nm and 1550.5 nm are suppressed relative to the 1549.5 nm and 1550.0 nm channels. The dashed line indicates the relative gain for pass bands PB<b>2</b> without suppression by wedge etalon <b>36</b>.

[0038]FIG. 3C shows the wedge etalon pass band PB<b>3</b> centered with respect to the grid channel at 1550.0 nm, with the relative gain associated with the pass band PB<b>2</b> at 1550.0 nm being high, while the relative gain levels associated with adjacent pass bands PB<b>2</b> at 1549.5 nm and 1550.5 nm are suppressed relative to the selected 1550.0 nm channel, and the gain associated with pass bands PB<b>2</b> at 1551.0 nm and 1549.0 nm is further suppressed. Again, the dashed line indicates the relative gain for pass bands PB<b>2</b> without suppression by wedge etalon <b>36</b>.

[0039] As can be seen from FIG. 2 and FIG. 3, non-optimal positioning or tuning of channel selector <b>36</b>, end mirror <b>14</b> and/or grid generator <b>34</b> will result in mis-alignment of pass bands PB<b>1</b>, PB<b>2</b> and PB<b>3</b> and give rise to attenuation in optical output power from the external cavity laser <b>10</b>. Monitoring of voltage across gain medium <b>12</b> via voltage sensor <b>32</b> allows such external cavity transmission characteristics to be probed or evaluated during laser operation. Error signals derived from the monitored voltage can then be used to adjust or servo the transmission profile of the external cavity during operation such that pass bands PB<b>1</b>, PB<b>2</b> and PB<b>3</b> are optimally aligned with respect to each other by suitable repositioning or adjusting of end mirror <b>14</b>, wedge etalon <b>36</b> and/or grid etalon <b>34</b>, thus providing accurate wavelength tuning and stability. As stated previously, in various other embodiments, such error signals may be alternatively derived using one or more photodiodes <b>15</b>.

[0040] The electro-optic modulation element <b>58</b> as shown in FIG. 1 provides a signal modulation in the form of a frequency dither, which may be introduced into the optical path length of the external cavity laser <b>56</b> by the presence of element <b>58</b> in optical path <b>22</b>. Modulation element <b>58</b> may comprise an etalon of electro-optic material such as lithium niobate, and has a voltage-adjustable refractive index. The signal modulation may comprise, for example, a frequency modulation of about 20 KHz. Adjustment of voltage across the electro-optic material of tuning element changes the effective optical thickness of modulation element <b>58</b>, and hence the overall optical path length l across the external cavity (between diode facet <b>18</b> and end mirror <b>14</b>) of external cavity laser <b>56</b>. Thus, electro-optic modulation element <b>58</b> may provide both (i) a frequency modulation signal or dither to the external cavity, and (ii) a mechanism to tune or adjust the external cavity optical path length by voltage applied across modulation element <b>58</b>. Electro-optic modulation element <b>58</b> may alternatively comprise an acousto-optic device, mechanical device, or other device capable of introducing a detectable frequency dither or modulation signal into the output of the external cavity.

[0041] Modulation of the optical path length l via frequency dither introduced by element <b>58</b> produces intensity variations in the output power of external cavity laser <b>56</b> which are detectable in the monitored voltage across gain medium <b>12</b>, due to optical feedback thereinto from the external cavity. These intensity variations will decrease in magnitude and phase error as a laser cavity mode is aligned with the center wavelength of the pass bands defined by grid generator <b>34</b> and channel selector <b>36</b>. In other words, the intensity variations and phase error in the modulation signal are minimal or nominally zero when pass bands PB<b>1</b>, PB<b>2</b> and PB<b>3</b> are optimally aligned as shown in FIGS. <b>2</b>A-<b>2</b>C. The use of intensity variation and phase error in the modulated signal with respect to error signal determination is described further below with reference to FIG. 5.

[0042] A tuning arm <b>74</b> may further be employed to positionally adjust end mirror according to input from controller <b>48</b>. Tuning arm <b>74</b> may be made from a material having a high coefficient of thermal expansion, such as aluminum or other metal or metal alloy. Controller <b>48</b> is operatively coupled to a thermoelectric controller <b>78</b> via line <b>80</b>. Thermoelectric controller <b>78</b> is coupled to tuning arm and is configured to adjust the temperature of arm <b>74</b>. Thermal control (heating or cooling) of tuning arm <b>74</b>, according to signals from controller <b>76</b>, may be used in this embodiment to control the position of end mirror <b>14</b> and the length of optical path l of the external cavity defined by end mirror and front facet <b>18</b> of gain medium <b>12</b> in an optimal position.

[0043] The frequency modulation introduced by modulation element <b>58</b> is detectable by controller <b>48</b> by monitoring the voltage across the gain medium <b>12</b> or a signal from one or more photodiodes <b>15</b>, and the frequency modulation includes variations in magnitude and phase error indicative of laser cavity mode alignment with the center wavelength of the pass bands defined by grid generator <b>34</b> and channel selector <b>36</b>, as noted above. Controller <b>48</b> maybe configured to derive an error signal from the modulation introduced: by the frequency dither, and to communicate a compensation signal to thermoelectric controller <b>78</b>, which accordingly heats or cools tuning arm <b>74</b> to position end mirror <b>14</b> and adjust the optical path length l of external cavity laser to null out the error signal.

[0044]FIG. 4 is a functional block diagram illustrating aspects of one embodiment of controller <b>48</b>. The controller of FIG. 4 includes a tuning circuit <b>84</b>, a current source <b>86</b> operatively coupled to gain medium <b>12</b> via line <b>28</b>, a grid controller <b>88</b> operatively coupled to thermoelectric controller <b>66</b> via line <b>68</b>, and a channel controller <b>90</b> operatively coupled to drive element <b>46</b> via line <b>70</b>. The current source <b>86</b> controls the power delivered to gain medium <b>12</b>. The grid controller <b>88</b> maintains the referential integrity of grid etalon <b>34</b> by thermal control thereof using thermoelectric controller <b>66</b> to heat or cool grid etalon <b>34</b> as required. Channel controller <b>90</b> directs drive element <b>46</b> to position or otherwise adjust channel selector <b>36</b> for selection of desired transmission bands in the grid defined by grid etalon <b>34</b>.

[0045] Tuning circuit <b>84</b> comprises a signal processor <b>94</b>, a voltage detector <b>96</b>, a path length adjuster <b>100</b>, and a modulation signal generator <b>102</b>. Modulation signal generator <b>102</b> provides a frequency dither or modulation signal to a selected loss element (e.g., modulation element <b>58</b>) that causes a corresponding modulation of the optical path l of the laser external cavity. The modulation frequency and amplitude may be selected, for example, to increase effective coupling efficiency. The voltage across gain medium <b>12</b> (or a signal derived from a photodiode <b>15</b>, as discussed previously) may be detected by voltage detector <b>96</b> and communicated to signal processing circuit <b>94</b>. The signal processing circuit <b>94</b> may be configured to determine the alignment of passbands PB<b>1</b> (FIG. 2 and FIG. 3) of the external cavity with passbands PB<b>2</b> of grid etalon <b>34</b> and passbands PB<b>3</b> of channel selector <b>36</b>, and to generate corresponding error information.

[0046] Pathlength adjuster <b>100</b> generates an error correction or compensation signal, from the error information provided by signal processing <b>94</b>, that is used to adjust the optical path length l of the external cavity in order to optimize the relationship between the modulation signal and the intensity signal. When an external cavity mode or pass band PB<b>1</b> is aligned with bands PB<b>2</b> and PB<b>3</b> generated by grid generator <b>34</b> and channel selector <b>36</b>, intensity variations at the modulation frequency (and odd multiples thereof) in the coherent beam traveling optical path <b>22</b> are substantially minimized, as discussed further below with reference to FIG. 5. Concurrently, the voltage signal intensity will vary at twice the modulation frequency. Either or both of these detectable effects are usable to evaluate external cavity loss associated with loss characteristics associated with the positioning or inter-relationship of end mirror <b>14</b>, grid generator <b>34</b> and channel selector <b>36</b>, and to generate error signals usable for adjustment of cavity loss characteristics such that the modulation signal and intensity signal are optimized. As discussed previously, in one embodiment, adjustment of the optical path length l may be carried out via thermal positioning of end mirror <b>14</b> in conjunction with tuning arm <b>74</b> and temperature controller <b>78</b>. In other embodiments, pathlength adjuster <b>100</b> may control other elements that adjust the optical pathlength l of the external cavity laser.

[0047] Referring now to FIG. 5, the relationship of the dither modulation signal introduced to an external cavity with respect to the detected voltage modulation across gain medium <b>12</b> is illustrated graphically as wavelength versus relative intensity. FIG. 2 shows a grid etalon pass band PB<b>2</b>, together with frequency or dither modulation signals <b>104</b>A, <b>104</b>B, <b>104</b>C corresponding to external cavity laser modes <b>106</b>A, <b>106</b>D and <b>106</b>C respectively. Frequency modulation signals <b>104</b>A-C are introduced to the laser external cavity by voltage modulation of electro-optic element <b>58</b> in the manner described above. As shown in FIG. 6, laser mode <b>106</b>A is off-center with respect to the center of pass band PB<b>2</b> towards the shorter wavelength side of pass band PB<b>2</b>, while laser mode <b>106</b>B is located at about the center wavelength of pass band PB<b>2</b>, and laser mode <b>106</b>C is located on the longer wavelength side of pass band PB<b>2</b>. Laser mode wavelength <b>106</b>B corresponds to a wavelength lock position and represents an optimal loss profile for the external cavity. Laser modes <b>106</b>A and <b>106</b>B are off-center with respect to pass band PB<b>2</b> and result in non-optimal cavity loss profiles which will require adjustment of the external cavity length l, either by adjusting the effective optical thickness of electro-optic element <b>58</b> or by positioning end mirror <b>14</b> as described above.

[0048] The voltage detected across gain medium <b>12</b> by voltage detector <b>96</b> for dither signals <b>104</b>A, <b>104</b>B and <b>104</b>C are shown respectively as voltage modulation signals <b>108</b>A, <b>108</b>B and <b>108</b>C on the right side of FIG. 6, which correspond respectively to the laser mode wavelengths <b>106</b>A, <b>106</b>B and <b>106</b>C. The location of laser mode <b>106</b>A at a wavelength shorter than that of the center wavelength of pass band PB<b>2</b> results in voltage signal <b>108</b>A having a modulation that is in phase with the dither modulation signal; <b>104</b>A. The location of laser mode <b>106</b>C at a greater wavelength than the center wavelength of pass band PB<b>2</b> results in a modulation of voltage signal <b>108</b>C that is out of phase with respect to the modulation of dither signal <b>104</b>C.

[0049] The location of each laser mode wavelength with respect to the slope of pass band PB<b>2</b> affects the amplitude of the corresponding voltage signal. Thus, voltage signal <b>108</b>A, which corresponds to laser mode <b>106</b>A wavelength on a relatively steep slope of pass band PB<b>2</b>, has a relatively large modulation amplitude, while voltage signal <b>108</b>C, which corresponds to laser mode <b>106</b>C associated with a portion of pass band PB<b>2</b> having a less steep slope, has a correspondingly smaller modulation amplitude. Voltage signal <b>108</b>B, which corresponds to centered laser mode <b>106</b>B, has a minimal modulation amplitude since the period of the dither modulation signal <b>104</b>B occurs symmetrically about the center wavelength of pass band PB<b>2</b>. The frequency of the dominant intensity in the case of voltage signal <b>108</b>B in this instance is twice the frequency of dither modulation signal <b>104</b>B.

[0050] From FIG. 5 it can be seen that the amplitude of the modulation detected in the voltage across gain medium <b>12</b> indicates the magnitude of correction or adjustment required for the laser external cavity, while the phase of voltage signal modulation indicates the direction of the adjustment. The amplitude of dither modulation signals <b>104</b>A-C is selected so that, during wavelength lock, the variation in the intensity of voltage signal modulation is held to acceptable levels for the particular use of the external cavity laser. The frequency of the dither modulation is chosen to be high enough to provide coherence control, but low enough to prevent interference with information modulated onto the carrier signal provided by the external cavity laser during transmission.

[0051]FIG. 6 is a hardware block diagram illustrating various aspects of one embodiment of a laser controller that may be configured to implement the functionality of the control system as depicted in FIG. 4. Various features of a laser assembly such as the assembly <b>10</b> described previously in conjunction with FIG. 1 are also illustrated in FIG. 6. Features that correspond to those of FIG. 1 and FIG. 4 are numbered identically for simplicity and clarity. It is noted that in other embodiments, various features of the laser controller of FIG. 6 as discussed below may be used in conjunction with other configurations of laser assemblies. Furthermore, such controllers and laser assemblies may omit various functionality as discussed above in conjunction with FIGS. <b>1</b>-<b>5</b>.

[0052] The laser controller of FIG. 6 includes a microprocessor (CPU) <b>602</b> coupled through an interconnect bus <b>610</b> to a read-only memory (ROM) <b>604</b>, a random access memory (RAM) <b>606</b> and a field programmable gate array (FPGA) <b>608</b>. FPGA <b>608</b> is coupled to a stepper motor driver <b>612</b>, amplifiers <b>614</b>-<b>616</b>, and a low pass filter <b>618</b>. FPGA <b>608</b> is further shown coupled to a digital-to-analog converter <b>620</b>, an analog interface unit <b>622</b>, and an analog-to-digital converter <b>624</b>. A laser current source <b>86</b> is shown coupled to an output of digital-to-analog converter <b>620</b>.

[0053] Power to the components of the laser controller illustrated in FIG. 6 is provided by a power source <b>630</b>. In one embodiment, power source <b>630</b> receives 5 volt input power and generates output power of varying voltage levels to appropriately supply power to the components of controller <b>600</b>. Power source <b>630</b> may be implemented using a high efficiency switching regulator circuit.

[0054] Microprocessor <b>602</b> and FPGA <b>608</b> operate concurrently and in cooperation with each other to perform various functionality as depicted in FIG. 4 and described hereinbelow. It is noted that operations performed by microprocessor <b>602</b> may be conducted in accordance with the execution of software code stored within ROM <b>604</b>. In one embodiment, microprocessor <b>602</b> is implemented using a general purpose microprocessor, such as a Motorola MCF5206e microprocessor. It is noted that in other embodiments, a digital signal processor or other specialized hardware may be employed in place of microprocessor <b>602</b>. It is further noted that in other embodiments, other programmable logic devices, such as a CPLD (Complex Programmable Logic Device) may be employed in the place of FPGA <b>608</b>. Alternatively, one or more ASICs (Application Specific Integrated Circuits) could be employed. Still additional embodiments are contemplated that combine various functionality of microprocessor <b>602</b> and FPGA <b>608</b> as described herein within a single device.

[0055] Generally speaking, microprocessor <b>602</b> and FPGA <b>608</b> collectively operate to measure and process various parameters associated with the operation of laser assembly <b>10</b> and to perform various control functions. In one particular implementation, microprocessor <b>602</b> and FPGA <b>608</b> are clocked at 40 MHz.

[0056] As illustrated in FIG. 6, laser assembly <b>10</b> may include a laser temperature sensor <b>631</b> located in proximity to gain medium <b>12</b>, a grid generator temperature sensor <b>632</b> located in proximity to grid etalon <b>34</b>, a cavity length actuator temperature sensor <b>633</b> located in proximity to tuning arm <b>74</b>, and an ambient temperature sensor <b>634</b>. Each of the sensors <b>631</b>-<b>634</b> may be implemented using a thermistor, although other temperature dependent devices may be employed in other embodiments. Laser assembly <b>10</b> may further include one or more photodiodes <b>15</b> positioned at selected locations of the laser assembly to receive light associated with the operation of the external cavity laser. In the illustrated embodiment, FPGA <b>608</b> may be programmed to periodically detect signals associated with each of sensors <b>631</b>-<b>634</b>, photodiodes <b>15</b>, and/or gain medium <b>12</b> through analog interface <b>622</b> and analog-to-digital converter <b>624</b>. For this purpose, analog interface <b>622</b> includes multiplexers <b>650</b>-<b>652</b> and an anti-alias filter <b>653</b>. Multiplexers <b>650</b>-<b>652</b> operate under the control of FPGA <b>608</b> to periodically couple a signal associated with a selected one of sensors <b>631</b>-<b>634</b>, split detector <b>658</b>, or gain medium <b>12</b> for signal detection. These operations will be described in further detail below.

[0057] FPGA <b>608</b> is additionally configured to generate control signals for controlling various functionality of laser assembly <b>10</b>. More particularly, in the embodiment of FIG. 6, FPGA <b>608</b> is configured to generate a control signal for controlling the position of a channel selector stepper motor <b>46</b><i>a </i>through a stepper motor driver <b>612</b> (which are collectively representative of the drive element <b>46</b> of FIG. 1). As discussed previously, a position indicator <b>50</b> coupled to FPGA <b>608</b> may further provide an indication of the position of drive element <b>46</b> (or to indicate when the drive element is at a home position). The control signal generated by FPGA <b>608</b> for controlling the position of stepper motor <b>46</b>A may be driven in accordance with a control value stored within a storage location of FPGA <b>608</b>. This storage location may be periodically updated with new values through the execution of instructions executed by microprocessor <b>602</b>.

[0058] FPGA <b>608</b> may also be configured to generate a control signal for controlling grid generator temperature controller (TEC) <b>66</b>, which regulates the temperature of grid etalon <b>34</b>. FPGA <b>608</b> may similarly generate control signals for controlling a cavity length actuator temperature controller <b>78</b>, which regulates the temperature of tuning arm <b>74</b>, and a laser <b>79</b>, which regulates the temperature of gain medium <b>12</b>. In one embodiment, each of the temperature controllers <b>66</b>, <b>78</b> and <b>79</b> are controlled by pulse-width modulated (PWM) signals generated by FPGA <b>608</b>. Each of the temperature controllers may be implemented using a peltier device. In one specific implementation, the pulse-width modulated signals are generated at a repetition rate of 200 kHz. Amplifiers <b>614</b>-<b>616</b> are provided to amplify the PWM signals generated by FPGA <b>608</b>. It is noted that in alternative embodiments, other forms of control signals may be generated to control selected functions of laser assembly <b>10</b>.

[0059] FPGA <b>608</b> may further be configured to generate a modulation signal for driving modulation element <b>58</b>. For this purpose, FPGA <b>608</b> may be configured to generate a pulse-width modulated signal which is input to a low-pass filter <b>618</b> which correspondingly provides an analog modulation signal that is passed to an amplifier <b>619</b>. In one particular implementation, the modulation signal provided from the output of low-pass filter <b>618</b> is in the form of a sinusoidal wave at 20 kHz. The PWM signal generated by FPGA <b>608</b> may have a frequency consistent with that of the other PWM signals generated by FPGA <b>608</b>. For example, in one embodiment, the PWM signal has a frequency of 200 KHz. Further details regarding generation of a modulation signal for driving modulation element <b>58</b> will be provided further below.

[0060] The sampling of signals associated with sensors <b>631</b>-<b>634</b>, photodiodes <b>15</b>, and/or gain medium <b>12</b> may be performed synchronously with the generation of the PWM control signals that drive temperature controllers <b>66</b>, <b>78</b>, and <b>79</b>, as well as the PWM signal provided to low pass filter <b>618</b>. The precise timing and synchronization of the control signals with the detected signals reduces the potential noise sources to a DC offset by mixing the fundamental component down to 0 or DC. The DC offsets can be subtracted from the signal in interest.

[0061] As described previously in conjunction with FIG. 5, the amplitude of the modulation detected in the voltage cross gain medium <b>12</b> indicates the magnitude of correction or adjustment required for the laser external cavity, while the phase of voltage signal modulation indicates the direction of the adjustment. Accordingly, in one embodiment the voltage across laser <b>12</b> is periodically measured by FPGA <b>608</b> through analog interface <b>622</b> and analog-to-digital converter <b>624</b>. The voltage signal may be amplified with a single stage pre-amp within analog interface <b>622</b> and then multiplexed through multiplexer <b>650</b> into a common anti-alias filter <b>653</b>. Multiplexer <b>652</b> is set to provide the output of anti-alias filter <b>653</b> to analog-to-digital converter <b>624</b>.

[0062] In one particular implementation, following a predetermined settling time after FPGA <b>608</b> sets multiplexers <b>650</b> and <b>652</b> in a manner to convey a signal corresponding to the voltage across gain medium <b>12</b> to analog-to-digital converter <b>624</b>, FPGA <b>608</b> performs a burst of, for example, <b>50</b> separate and consecutive voltage readings associated with the voltage across gain medium <b>12</b>. Each of the voltage readings (in the form of digital data generated by analog-to-digital converter <b>624</b>) may be temporarily stored within FPGA <b>608</b>, and is subsequently transferred into RAM <b>606</b>. Upon receipt of data from analog-to-digital converter <b>624</b> by FPGA <b>608</b>, FPGA <b>608</b> may signal microprocessor <b>602</b> which may responsively invoke an internal direct memory access control mechanism to carry out the transfer of the data from FPGA <b>608</b> to RAM <b>606</b>.

[0063] Upon storing a set of data indicative of the voltage across gain medium <b>12</b> within RAM <b>606</b>, microprocessor <b>602</b> performs a Fourier Transform to transform the temporal data to a frequency domain to separate the DC, fundamental and/or harmonic terms. In one embodiment, microprocessor <b>602</b> executes a Fast Fourier Transform (FFT) routine. The FFT routine may be optimized for integer input data as supplied from analog-to-digital converter <b>624</b>, and may be configured to compute only the output terms of particular interest, such as the fundamental component. As discussed previously, by calculating, for example, the magnitude and phase of the fundamental component, an error signal may be generated to adjust the cavity length. Thus, upon calculation of the error signal, microprocessor <b>602</b> writes a value derived from the error signal to a location within FPGA <b>608</b> which controls the pulse width of the PWM signal provided to amplifier <b>616</b> to drive cavity length actuator temperature controller <b>78</b>. It is noted that in other embodiments, the error signal may be used to control other mechanisms within a laser assembly to adjust cavity length. It is also noted that in other embodiments, similar measurements may alternatively be taken from one or more photodiodes <b>15</b> (or other light detectors) to derive the error signal. In various embodiments and depending upon the signals of interest, multiplexer <b>650</b> and/or anti-alias filter <b>653</b> of analog interface <b>622</b> may be omitted.

[0064]FIG. 6A illustrates one embodiment of an algorithm for performing wavelength-locking. The wavelength-locking algorithm as depicted in FIG. 6A may be implemented by code executed within microprocessor <b>602</b>, and in conjunction with the control of FPGA <b>608</b> as described herein. When initiated, the algorithm begins by computing offsets, initializing variables, and placing the cavity length actuator <b>78</b> in an initial starting position (steps <b>670</b> and <b>671</b>). Next, the algorithm enters a locking loop where the cavity length actuator sensor <b>633</b> is measured and the quality of the locking is determined. The quality of locking may be determined by computing a decaying integral of the error signal. If the cavity length actuator sensor indicates a temperature within a predetermined range and if the lock quality is sufficient as determined during step <b>672</b>, modulation data is acquired during step <b>673</b>. As discussed previously, the modulation data may be in the form of a set of readings associated with the voltage across gain medium <b>12</b>, or may be associated with a set of readings taken from one or more photodiodes <b>15</b>. The fundamental modulation component of the most recent gain medium potential measurement may be used to compute cavity length errors (step <b>674</b>) and is applied to a compensator in order to minimize the fundamental component. As stated previously, the fundamental modulation component may be computed by an FFT routine executed by microprocessor <b>602</b>. In other embodiments, other harmonics of the gain medium voltage or photodiode currents may alternatively or additionally be determined and used to compute the error signal. The slew rate associated with the error signal may be limited during step <b>675</b>. During step <b>676</b>, microprocessor <b>602</b> may write a value in a corresponding storage location of FPGA <b>608</b> that controls the generation of the PWM signal to cavity length actuator temperature controller <b>78</b> to thereby cause corrections to the cavity length to be made. The locking algorithm repeats these steps unless the cavity length actuator sensor <b>633</b> indicates a temperature that is out of a predetermined range or if the lock quality is poor (step <b>672</b>). An integral error term of the error signal may be reset during step <b>676</b>, and the cavity length actuator (e.g., tuning arm <b>74</b>) may be returned to the initial starting position during step <b>677</b>. The locking loop is subsequently reentered and modulation data is acquired during step <b>673</b>.

[0065] Returning to FIG. 6, laser assembly <b>10</b> may further include an EEPROM (electrically erasable programmable read-only memory) <b>83</b> or other non-volatile storage device for storing information particular to laser assembly <b>10</b>. EEPROM <b>83</b> may be embodied upon the same base or within the same housing that includes elements forming the external cavity laser assembly (e.g., including gain medium <b>12</b>), and separate from, for example, a printed circuit board upon which the hardware associated with controller <b>600</b> is mounted. Data may be stored within EEPROM <b>83</b> that contains information relevant to wavelength calibration, tuning hints such as temperatures or positions, power or temperature calibration factors, identifying numbers, and operating data. The operating data may contain, for example, information relevant to laser lifetime, such as time-current profiles. By storing this information within EEPROM <b>83</b> that may be provided as an integral part of laser assembly <b>10</b>, interchangeability between laser optic assemblies and controller boards may be possible while retaining device-specific data.

[0066] In one particular embodiment, calibration coefficients associated with sensors <b>631</b>-<b>634</b> are stored within EEPROM <b>83</b>. The calibration coefficients may represent deviations from nominal values of sensors <b>631</b>-<b>634</b>. For example, each of the sensors <b>631</b>-<b>634</b> may nominally have the same value of a resistance at ambient temperature. However, due to specific device variations, the actual values associated with sensors <b>631</b>-<b>634</b> may deviate from the nominal value. The calibration coefficients may represent the relative differences between the resistances of sensors <b>631</b>-<b>634</b> when each is measured at an equal ambient temperature. These calibration coefficients may be stored in EEPROM <b>83</b> following manufacture of the laser assembly, and may be used to scale temperature measurements taken from sensors <b>631</b>-<b>634</b>, as described further below.

[0067] The laser controller may further include a network interface such as an Ethernet interface to allow control of the laser functionality by a remotely connected device. In one embodiment, the Ethernet functionality may be used to support an HTTP interface. Additionally, code for controlling operations of microprocessor <b>602</b> may be upgraded by downloading through an interface such as, for example, an RS-232 or Ethernet interface. This functionality allows for on-the-fly upgrades. Similarly, the logic configuration of FPGA <b>608</b> (or any other programmable logic device) may be modified through an interface such as an RS-232 or Ethernet interface.

[0068]FIG. 7 illustrates one embodiment of laser current source <b>626</b>. The laser current source of FIG. 7 is preferably configured to provide a low noise current to gain medium <b>12</b> with reasonable efficiency. As will be described further below, laser current source <b>626</b> may also include a mechanism to provide a shutdown of the laser output in the case of a fault condition.

[0069] Current flowing through gain medium <b>12</b> is passed through a transistor <b>702</b> and resistor <b>704</b>. Transistor <b>702</b> may be implemented using a MOSFET (metal oxide semiconductor field effect transistor device). A filter <b>706</b>, which may be implemented as an LCR filter, is provided to filter high frequency noise at a power supply VCC. Digital-to-analog converter <b>620</b> is provided to receive a programmed value from microprocessor <b>602</b> which sets the current flowing through gain medium <b>12</b>.

[0070] An operational amplifier <b>716</b> regulates the current flowing through transistor <b>702</b> by comparing an output of digital analog converter <b>620</b>, which may be passed through an RC filter formed by resistor <b>717</b> and capacitor <b>719</b>, to a signal at node <b>721</b> which is dependent upon current sensed through transistor <b>702</b>. The RC filter formed by resistor <b>717</b> and capacitor <b>719</b> may provide residual noise attenuation at mid and upper frequencies. Current flowing through transistor <b>702</b> is sensed in accordance with sense resistor <b>704</b> and an operational amplifier <b>722</b>. More particularly, operational amplifier <b>722</b> is configured to sense the current flowing through transistor <b>702</b> by measuring the voltage across resistor <b>704</b>. The output of operational amplifier <b>722</b> is reflected down to a ground-based voltage using transistor <b>725</b>, which regulates current flow through a resistor <b>723</b> depending upon the voltage across resistor <b>704</b>. Thus, the voltage at node <b>721</b> is a ground referenced voltage indicative of the current flowing through transistor <b>702</b>. It is noted that the circuit configuration of FIG. 7 allows one of the nodes (e.g., the cathode) of gain medium <b>12</b> to be grounded.

[0071] The current source of FIG. 7 further includes a transistor <b>708</b> connected in a common gate configuration between a node <b>710</b> and the gate of transistor <b>702</b>. In the embodiment shown, transistor <b>708</b> is implemented using a FET (field effect transistor). In other embodiments, transistor <b>708</b> may be implemented using a bi-polar transistor coupled in a common base configuration. Due to the high output impedance looking into the drain of transistor <b>708</b>, lower frequency noise on the power supply at node <b>712</b> is reflected onto the gate of transistor <b>702</b> thus causing VGs to remain constant. Transistor <b>708</b> provides level translation up to the gate of transistor <b>702</b> without introducing a significant power supply voltage dependence. Thus, while the current flowing through transistor <b>702</b> will be dependent upon the voltage at node <b>710</b> which is controlled by the output of operational amplifier <b>716</b>, the output current of transistor <b>702</b> is largely unaffected by low frequency noise on the power supply at node <b>712</b>. Operational amplifier <b>716</b> maintains the DC current at the programmed level.

[0072] Switch <b>730</b>, which may be implemented using a transistor such as a FET or bipolar transistor, provides the user with a fast acting laser shutdown. Preferably, switch <b>730</b> may have a low voltage threshold so even in a worst-case, a relatively low voltage may be sufficient to drive the transistor into conduction and divert the current source from the gain medium <b>12</b>. It is noted that control of the switch <b>730</b> may be conducted independent of the operation of microprocessor <b>602</b> (FIG. 5). Thus, the laser may be shut down even if malfunctions associated with the execution of instructions by microprocessor <b>602</b> occur.

[0073] It is noted that in other embodiments, other particular current source circuits may be employed for providing current to gain medium <b>12</b>. Such alternative circuit configurations may employ a drive transistor for supplying current to a laser device, a control circuit for controlling the level of current supplied to the laser device, and a common gate transistor (or common base transistor) coupled between the control circuit and a control terminal of the drive transistor to reduce the effects of noise. Such circuits may additionally employ a switch for diverting current from the laser device.

[0074]FIG. 8 illustrates one embodiment of analog interface <b>622</b> for the measurement of temperatures within laser assembly <b>10</b>. Circuit portions that correspond to those of FIG. 6 are numbered identically. FIG. 9 is a flow diagram depicting a method for temperature measurements.

[0075] Referring collectively to FIGS. 6, 8 and <b>9</b>, FPGA <b>608</b> sets multiplexers <b>651</b> and <b>652</b> in modes to selectively convey a signal generated by one of the temperature sensors <b>631</b>-<b>634</b> or other input to analog-to-digital converter <b>624</b> for data capture within FPGA <b>608</b>. Additional multiplexer <b>651</b> inputs include a ground reference <b>802</b> and a precision reference <b>804</b>. Precision reference input <b>804</b> may be implemented using a precision resistor. Depending upon the mode of multiplexer <b>651</b> as controlled by FPGA <b>608</b>, one input at a time is coupled to the output of multiplexer <b>651</b>, which in turn is coupled to a fixed voltage reference through a fixed resistance <b>806</b>. Thus, one of the temperature sensors <b>631</b>-<b>634</b> or precision reference <b>804</b> may be connected to form the lower leg of voltage divider. For example, when FPGA <b>608</b> sets multiplexer <b>651</b> in a mode that connects temperature sensor <b>631</b> to the output of multiplexer <b>651</b>, current flows from the fixed voltage reference through resistor <b>806</b> and temperature sensor <b>631</b>, and the voltage at node <b>808</b> is measured. The voltage at node <b>808</b> is conveyed through a common measurement path through multiplexer <b>652</b> to analog-to-digital converter <b>624</b>, where the voltage is converted to a digital value which may be sampled by FPGA <b>608</b>, as discussed previously. FPGA <b>608</b> may set multiplexer <b>651</b> and <b>652</b> to select a particular one of sensors <b>631</b>-<b>634</b>, ground reference <b>802</b>, or precision reference <b>804</b> to take a corresponding measurement.

[0076] Measurements associated with ground reference <b>802</b> and precision reference <b>804</b> are performed to allow for the correction of DC offsets and gain associated with the temperature measurement circuitry. As illustrated in FIG. 9, in one embodiment, after FPGA <b>608</b> has acquired voltage readings associated with all inputs of multiplexer <b>651</b> during step <b>902</b>, microprocessor <b>602</b> may execute code stored in memory (within RAM <b>606</b>, for example) to filter the reference value associated with ground reference <b>802</b> and/or precision reference <b>804</b> (step <b>904</b>) to compute corrected sensor values (step <b>906</b>). Subsequently, microprocessor <b>602</b> may execute code to scale the sensor values using calibration coefficients stored within EEPROM <b>83</b> (step <b>908</b>).

[0077] A lookup table may further be provided within memory (e.g., RAM <b>606</b>) which correlates various corrected voltage readings with temperature. Thus, during step <b>910</b>, microprocessor <b>602</b> may access entries within the lookup table to determine a corresponding temperature associated with each of the temperature sensor measurements. In one implementation, microprocessor <b>602</b> may perform linear interpolation to increase the resolution of the lookup table result.

[0078] Turning finally to FIG. 10, a circuit diagram illustrating one embodiment of an amplifier circuit <b>619</b> for generating a modulation signal to drive electro-optic modulation element <b>58</b> is shown. Circuit portions that correspond to those of FIG. 6 are numbered identically for simplicity and clarity.

[0079] Referring collectively to FIGS. 6 and 10, FPGA <b>608</b> may be programmed to generate a pulse-width modulated signal at line <b>617</b> which is provided to low-pass filter <b>618</b>. In one particular implementation, the pulse-width modulated signal is modulated according to variations which approximate a sinusoidal wave at 20 kHz. The pulse-width modulated signal may be generated according to a set of stored values within FPGA <b>608</b>, that are provided by microprocessor <b>602</b>. In one particular implementation, a set of 10 values are stored within FPGA <b>608</b> to control the particular modulation associated with the pulse-width modulated signal at line <b>617</b>.

[0080] Low-pass filter <b>618</b> filters the pulse-width modulated signal at line <b>617</b>. Thus, a sinusoidal wave form at 20 kHz may be output from low-pass filter <b>618</b>. As illustrated in FIG. 10, amplifier circuit <b>619</b> includes a transformer <b>1004</b> having a primary connected in a push-pull configuration (also known as a bridge-tied load). The amplified modulation signal is inverted by a first operational amplifier <b>1006</b> and again by another operational amplifier <b>1008</b>. Thus, the output of operation amplifier <b>1008</b> takes the form of a sinusoidal wave that is <b>180</b> degrees out of phase with respect to a similar sinusoidal signal at the output of operational amplifier <b>1006</b>.

[0081] It is noted that in other embodiments, other signal conversion circuits such as other types of analog filters may be employed in the place of low-pass filter <b>618</b> for converting the digital output of FPGA <b>608</b>. It is further noted that in other embodiments, other forms of amplifier circuits may be coupled in a push-pull configuration to the primary of transformer <b>1004</b>. For example, in one embodiment, a class D amplifier may be employed in the place of low-pass filter <b>618</b> and the amplifier circuitry including operation amplifiers <b>1006</b> and <b>1008</b>. The output of the class D amplifier may be coupled to drive the primary of transformer <b>1004</b> through an LC filter, and may be coupled in a push-pull configuration.

[0082] As a result of the push-pull configuration, a voltage of approximately 2V_CCpeak-to-peak variations (twice the supply voltage) may be generated across the primary of transformer <b>1004</b>. Return current through the primary of transformer <b>1004</b> is passed through operational amplifier <b>1008</b>, rather than running return current through ground. Noise due to the generation of the 20 kHz modulation signal on the ground reference may thereby be reduced. In one embodiment, transformer <b>1004</b> has a coil ratio of 120 to 1 thereby generating a voltage of up to 1000 volts peak to peak at the output of the secondary of the transformer <b>1004</b> to drive the modulation element <b>58</b>.

[0083] Although the embodiments above have been described in considerable detail, numerous variations and modifications will become apparent to those skilled in the art once the above disclosure is filly appreciated. It is intended that the following claims be interpreted to embrace all such variations and modifications.