US20030033819A1

US20030033819A1 - Current-Mode control of Thermo-Electric cooler

Info

Publication number: US20030033819A1
Application number: US10/032,135
Authority: US
Inventors: Daniel Prescott
Original assignee: Cenix Inc
Current assignee: Cenix Inc
Priority date: 2001-08-10
Filing date: 2001-12-20
Publication date: 2003-02-20

Abstract

According to an exemplary embodiment of the present invention, an apparatus for controlling the temperature of a thermoelectric cooler includes a current sensing circuit which is adapted to control the current through the thermoelectric cooler using proportional, integral, derivative (PID) processing.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS

The present application claims priority from U.S. Provisional Patent Application Serial No. 60/311,496, filed Aug. 10, 2001, entitled “Current-Mode of Thermo-Electric Cooler.” The disclosure of this above-captioned provisional patent application is specifically incorporated by reference herein for all purposes.[0001]

FIELD OF THE INVENTION

The present invention relates generally to optical communications, and specifically to an apparatus for controlling a thermoelectric cooler (TEC) in order to maintain a substantially constant temperature on the surface of the TEC.

BACKGROUND OF THE INVENTION

Digital optical communications have gained widespread acceptance for both telecommunications (telecom) and data communications (datacom) applications. Telecommunication systems typically operate over single-mode fiber at distances from 10 km to over 100 km, and employ lasers emitting light at wavelengths of 1310 nm to 1600 nm. Data communication systems typically cover shorter distances of up to a few kilometers, over multi-mode fiber. Data communication systems can employ laser devices as well, typically having emission wavelengths of 650 nm to 850 nm. As the transmission and reception rates in the telecom and datacom industries continue to increase, there are ever increasing demands placed on the various components of the optical communication system.

One such demand is the continuity of the laser output over time and ambient temperature. As is known, temperature variations can adversely impact the operation of a semiconductor laser. To this end, the wavelength and power output of a semiconductor laser may be influenced by temperature fluctuations. As can be appreciated, in optical communication systems, (e.g., dense wavelength division multiplexing (DWDM) systems) wavelength variation must be maintained within specified boundaries. Output optical power variations must also be minimized in order to maintain certain target operating parameters such as extinction ratio and average output power. Therefore, to facilitate laser operation within close tolerances for both wavelength and optical power, it is useful to closely control the temperature of the active layers of a semiconductor laser.

One method to control the operating temperature of a laser is through the use of a thermo-electric cooler (TEC). A TEC is a device that may be controlled to either add or extract heat to/from a laser. However, in conventional uses of the TEC the amount of power used to cool a laser to a desired temperature often exceeds the amount of power that is used to drive the laser. This can result in an undesirable condition known as thermal runaway, since the additional power supplied to cool the laser may actually increase the temperature of the operating environment of the laser, rather than decreasing it.

One conventional thermoelectric cooler is a Peltier effect thermoelectric cooler (Peltier TEC). A Peltier TEC may consist of pairs of p-type and n-type materials connected in series and sandwiched between two closely spaced ceramic plates. When connected to a DC power source, current flow through the series of p-n junctions transfers heat from one side of the thermo-electric cooler to the other. In a typical application, the “cold” side of the thermoelectric cooler is connected to the device to be cooled, while the “hot” side is connected to a heat sink to disburse the heat to the outside environment. Of course, by changing the current direction, the thermo-electric device can operate as a heater.

Peltier TEC's are often incorporated into laser assemblies to control the temperature of a semiconductor laser over wide ambient temperature ranges. The temperature of the laser may be monitored with a negative temperature coefficient (NTC) thermistor circuit, which provides feedback to the closed-loop temperature control block. The maximum thermo-electric cooler temperature differential may be achieved with the application of a maximum voltage, V _max, which is temperature dependent, and results in the flow of a maximum current I_max, which is substantially temperature independent. Voltages and currents above these levels may result in the opposite effect intended, mainly incrementally heating when cooling is desired. As such, overheating and thermal runaway of the TEC circuit may result if the voltage and/or current applied to the TEC is not bounded or limited in some manner.

One known technique for controlling a TEC is known as voltage-mode control. TEC voltage-mode control results in a closed-loop control circuit that has a temperature dependent closed-loop response. This is because the TEC voltage varies as a function of the temperature differential across the TEC, due to a temperature dependent back-EMF that results from the Peltier effect. The TEC voltage-mode control results in a closed-loop circuit, the performance parameters of which are also subject to a relatively large initial tolerance due to variation in the semiconductor characteristic resistance. Ultimately, these variations in the relationship of heat transfer as a function of TEC applied voltage significantly limit the performance of voltage-mode control of the TEC.

FIG. 1 is a schematic diagram of a conventional TEC-

drive circuit

100 used to implement closed-loop temperature control of a TEC. The TEC drive circuit 100 provides voltage-mode control of a TEC 106. A control block 101 consists of a temperature monitor circuit 102, which monitors temperature feedback; a set-point input 103 for target temperature control; and a proportional, integral, derivative (PID) control block 104 ^a. The PID control block 104 may be implemented via a firmware control method driving a digital to analog converter (DAC); or entirely in analog circuitry (e.g., operational amplifiers). The output of the control block 101 drives a summing op-amp 105 which controls the voltage across the thermo-electric cooler 106, as a function of the output of the control block 101. A reference voltage 107 is summed at op-amp 105 to provide an offset to allow positive and negative (bipolar) driving of the thermo-electric cooler (TEC) 106 with a unipolar (PID) signal 104. A unity gain buffer 108 provides the TEC 106 drive current in excess of approximately 1 amp to approximately 2 amps, depending upon the manufacturer's specification for I_maxof the TEC.

In order to suitably protect the

TEC

106 from exceeding the maximum current, I_max, the current is monitored with a current monitoring/protection circuit 112 which includes a sense resistor (R_sns) 109, a first transistor (Q_p) 110 and a second transistor (Q_n) 111. The voltage across the sense resistor 109 turns the first transistor 110 or the second transistor 111 on in an over-current condition, thus shunting the op amp drive signal, protecting the TEC from an over-current condition.

In order for the

sense resistor

109 to turn on first transistor 110 or second transistor 111 in an over-current condition, the voltage drop must be equal to the base-emitter junction turn-on voltage (V_be). However, this turn-on voltage is a function of temperature. As the temperature of the first and second transistors increases, the turn-on voltage is reduced, and the current monitor protection circuit 112 can begin shunting the drive signal prematurely. Thereby, the full range of the TEC is not realized. Further exacerbating this issue is the fact that the sense resistor 109 is relatively large, and dissipates significant power. This can add to the temperature variability of the first and second transistors 110 and 111, as well as make the overall TEC drive circuit 100 very inefficient from a power perspective.

As can be appreciated, using the voltage-mode control technique described above may prohibit the use of the full cooling range of the TEC, resulting in a reduced operational range for laser temperature control. Moreover, the TEC may enter self-heating mode (overheating/runaway) before current limiting occurs. Also, the relatively large voltage drop across the

sense resistor

109 may cause the current buffer 108 to reach voltage saturation in low voltage applications. This may also limit the cooling range of the TEC. Additionally, the power rating of the sense resistor 109 dictates a large package size, which consumes valuable board real-estate.

Accordingly, what is needed, is a technique for controlling a thermo-electric cooler that overcomes at least the drawbacks of the conventional art described above.

SUMMARY OF THE INVENTION

According to an exemplary embodiment of the present invention, an apparatus for controlling the temperature of a thermo-electric cooler includes a current sensing circuit which is adapted to control the current through the thermo-electric cooler using proportional, integral, derivative (PID) processing.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is best understood from the following detailed description when read with the accompanying drawing figures. It is emphasized that the various features are not necessarily drawn to scale. In fact, the dimensions may be arbitrarily increased or decreased for clarity of discussion. [0015]
FIG. 1 is a schematic diagram showing a conventional voltage-mode control circuit for a TEC. [0016]
FIG. 2 is a schematic diagram of a current-mode TEC control circuit in accordance with an exemplary embodiment of the present invention.[0017]

DETAILED DESCRIPTION

In the following detailed description, for purposes of explanation and not limitation, exemplary embodiments disclosing specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one having ordinary skill in the art having had the benefit of the present disclosure, that the present invention may be practiced in other embodiments that depart from the specific details disclosed herein. Moreover, descriptions of well-known devices, methods and materials may be omitted so as to not obscure the description of the present invention. [0018]
Briefly, the present invention relates to a method and apparatus for controlling the current through a thermal electric cooler during operation using a current-mode control circuit configuration. Advantageously, the TEC current is a substantially linear function of the TEC heat transfer. As will become more clear as the description proceeds, the current-mode control in accordance with an exemplary embodiment of the present invention enables TEC current control, the flow of which is directly related to the temperature differential across the TEC, given a fixed thermal load. To this end, the thermal energy of the TEC is transferred by the majority carriers in the semiconductor material of the TEC, and is therefore substantially proportional to the current, neglecting I[0019] ²*R losses. Moreover, current-mode control of the TEC in accordance with an exemplary embodiment of the present invention is substantially independent of the semiconductor resistance initial tolerance, to a first order approximation. This enables the various performance parameters of the current mode TEC control circuit of an exemplary embodiment of the present invention to be tightly controlled, and optimized. These and other advantages will become more apparent to one of ordinary skill in the art upon a review of the present invention as described in connection with exemplary embodiments thereof.
As described, conventional voltage-mode control methods of a thermo-electric cooler are inherently dynamically inaccurate because of the variation in voltage across the TEC from device to device as well as over-temperature. This inaccuracy translates into inaccuracies in the closed-loop performance parameters (e.g., the transient response) of the TEC, and the attendant shortcomings of conventional voltage-mode TEC controllers described in more detail above. In clear contrast, the current-mode TEC controller in accordance with an exemplary embodiment of the present invention accurately controls the current through the thermoelectric cooler. This accuracy results from the proportional relationship between change in temperature across a thermo-electric cooler and the change in the current through the device. [0020]
Turning to FIG. 2, a current-[0021] mode TEC controller 200 according to an exemplary embodiment of the present invention is shown. The current-mode TEC controller 200 includes a control block 201. The control block 201 further includes a temperature monitor circuit 202 which provides feedback to a PID controller 204. The PID controller 204 includes a second input for the set-point 203 of the TEC temperature. As will become more clear as the description of the present exemplary embodiment proceeds, the PID controller 204 changes the current of a TEC 209, as is needed to maintain the sensed temperature at a level equal to the set-point temperature. The TEC 209 is illustratively a Peltier-effect TEC.
A reference voltage (V[0022] _ref) 205 provides the offset necessary for bipolar control of the TEC current (to enable both heating and cooling) by the adjustment of the unipolar output of the PID controller 204. The output from the PID controller 204 is a voltage between 0 and V_refvolts, which is calculated by the PID function to produce a current through the TEC necessary to achieve a TEC temperature equal to the set-point input. The difference amplifier, comprising of the opamp 207 and resistors R (208), R (208), R_ref(211) and R_cntl(212), controls the amount of current through the TEC.
When the value of R[0023] _ref(211) is set equal to the quantity (2*R_cntl+R), the TEC current (sensed by R_sns) will vary from (−R*V_ref)/(R_ref*R_sns) to (+R*V_ref)/(R_ref*R_sns) as the voltage output of the PID control block is varied from 0 Volts to V_ref, where V_refcan be any standard value reference, typically provided for use by the DAC output of the PID block. The resistors R, R_refand R_cntlare selected such that the TEC current is bounded within the TEC manufacturer's specified limits by the output range of the PID block (V_ref). This inherently provides the current limiting function. The output from the op amp 207 drives the high-current buffer 210 such that the current that is sensed, or converted to a voltage, by R _sns 206, equals that determined by the PID control function. It is noted that the current-mode controller of the exemplary embodiments of the present invention may be applied to both linear power as well as switched-mode power designs. The current buffer block 210 shown in FIG. 2 is a generic block that may utilize either method, depending upon the requirements of the application for efficiency, noise, etc.
The [0024] control block 201 is one portion of the closed-loop feedback of the present invention. The input to the temperature monitor circuit 202 is typically provided by a NTC thermistor 213. The temperature monitor circuit 202 provides an input voltage signal to the PID control block 204 that is representative of the temperature of the TEC 209. The PID control block 204 also has a set-point 203 which is the target voltage level [representing the target TEC temperature] for operation of the TEC 209.
The [0025] PID control block 204 calculates the required change in current through the TEC 209 to effect the desired temperature change of the TEC to heat or cool the laser (not shown). The calculation is illustratively carried out using a closed-loop temperature control method. An exemplary of closed-loop temperature control method incorporates a combination of Proportional, Integral, and Derivative (PID) feedback in an attempt to achieve the optimal compromise between response time, accuracy, and stability. In a temperature control application, the PID equation can be represented as: $\begin{matrix} W = P \cdot (T_{s} - T_{0}) + I \int (T_{s} - T_{0}) \partial t + D \cdot [\frac{\partial}{\partial t} (T_{s} - T_{0})] & (1) \end{matrix}$
where T[0026] _sis the temperature setpoint or target, T₀is the current sensed temperature, W is the calculated power necessary to be applied to the system in order to minimize the difference (T_s−T₀), which optimally would approach zero. The factors P, I and D are Proportional, Integral, and Derivative constants, respectively. The combination of values of these parameters dictates the overall system response, accuracy and stability. Many implementations of this method set I and D both to zero and merely utilize proportional control because of the simplicity of stabilizing the closed-loop system.
The output from the [0027] PID controller 204 is the calculated control voltage which will provide the desired TEC current needed to effect the desired temperature change of the TEC 209. Quantitatively, the TEC current may be expressed as (assuming a large open-loop OpAmp gain and a unity gain current buffer 210): $\begin{matrix} I_{TEC} = \frac{\begin{matrix} [(R \cdot R_{ref} + R^{2}) \cdot V_{cntl} + (- R \cdot R_{ref} - R^{2} - R_{cntl} \cdot R_{ref} - R_{cntl} \cdot R) \cdot \\ V_{ios} + (- R^{2} - R_{cntl} \cdot R) \cdot V_{ref}] \end{matrix}}{(R \cdot R_{sns} \cdot R_{ref} + R_{cntl} \cdot R_{sns} \cdot R_{ref})} & (2) \end{matrix}$
where V[0028] _iosis the input-offset voltage of the OpAmp 207; and V_cntlis the output voltage of the PID portion of the control block 201.
In the event that symmetric heating and cooling is desired, R[0029] _ref(shown at 211 in FIG. 2) can be shown to be:
R _ref=2·R _cntl +R (3)
where R[0030] _cntlis the control resistance 212 and R is the resistance of monitor resistors 208 shown in FIG. 2.
Moreover, when V[0031] _ios=0 in eqn. (2), and R_refas given by eqn. (3) is substituted into eqn. (2), I_TECbecomes: $\begin{matrix} I_{TEC} = (2 \cdot V_{cntl} - V_{ref}) \cdot \frac{R}{R_{ref} \cdot R_{sns}} & (3a) \end{matrix}$
From eqn. (3a), it can be seen that the TEC current limit is bounded substantially by the range of the output voltage V[0032] _cntlof control block PID 204, which is from 0 volts to V_ref.
To this end, from eqn. (3a), at the maximum V[0033] _cntloutput voltage (V_ref), the maximum thermo-electric cooler current is bounded by: $\begin{matrix} I_{TEC_\max} = \frac{R}{R_{ref}} \cdot \frac{V_{ref}}{R_{sns}} & (4) \end{matrix}$
One particularly useful aspect of the present exemplary embodiment is the accuracy of the control of I[0034] _TEC. To wit, using standard 1% resistors, the TEC current may be controlled to approximately 2% or better. This allows the TEC 209 to be used for cooling up to within a couple percent of its maximum capacity, without entering into the self-heating range, thereby substantially preventing thermal runaway. Of course, this provides clear benefits to both the performance and the life of the laser device being cooled by the TEC 209.
From equations (2)-(4), it can be appreciated that the control of the TEC cooling is independent of the resistivity of the TEC device, to a first-order approximation, and independent of the associated voltage non-linearities over temperature. Again, this independence of the voltage across the TEC fosters accurate linear control of the TEC temperature independently of the intrinsic inaccuracies associated with voltage-mode control. Also, the tolerance of all parameters may be selected to allow the precision of the design to be dictated by the application. [0035]
A more exact expression, which takes into account the finite op-amp open-loop gain (A[0036] _v) the current buffer voltage-gain (A_buffer), as well as other second-order effects may be obtained by summing currents at each current node in FIG. 2 and simultaneously solving the resultant equations. From this analysis, the current through the TEC may be expressed as: $\begin{matrix} I_{TEC} (V_{cntl}) = - A_{buffer} \cdot A_{v} \cdot \frac{(V_{ref} \cdot R + V_{ios} \cdot R_{ref} - 2 \cdot V_{cntl} \cdot R + V_{ios} \cdot R)}{\begin{matrix} (R_{sns} \cdot R_{ref} + R_{TEC} \cdot R_{ref} + A_{v} \cdot A_{buffer} \cdot \\ R_{sns} \cdot R_{ref} + R \cdot R_{sns} + R_{TEC} \cdot R) \end{matrix}} & (5) \end{matrix}$
where R[0037] _TECis the intrinsic resistance of the TEC 209.
Certain aspects and advantages are worthy of specific mention at this point. For example, because the current-mode TEC controller of the exemplary embodiment of the present invention directly senses the current through the TEC via the [0038] sense resistor 206, and since the voltage across this resistor is not needed for current limiting as in conventional voltage-mode control techniques, the current sense resistor (R_sns) 206 may be made arbitrarily small The value of resistor 206 is substantially limited only by the input offset voltage of OpAmp 207 (e.g.: 0.1 ohms will result in an approximate error of 10 mA for an op amp with an input offset voltage (V_ios) of 1 mV) and the tolerance of monitoring resistors 208 as can be shown by partial differentiation of eqn. (2) (or even more precisely eqn. (5), depending on the desired accuracy). This ultimately enables the sense resistor 206 to be physically much smaller than sense resistors used in conventional voltage-mode control circuits, for example that shown in FIG. 1.
This reduction in the size of the [0039] sense resistor 206 helps conserve valuable board space in a deployed TEC device. Moreover, because the resistance value of the sense resistor 206 is relatively low when compared to conventional voltage-mode sense resistors, a lower supply voltage may be used to power the TEC. In contrast, the sense resistor of a conventional voltage-mode control circuit can consume on the order of approximately 20% of the supply voltage, thereby limiting the heating/cooling range of the TEC by limiting the available supply voltage. Furthermore, because of the relatively low resistance of sense resistor 206, a relatively insignificant amount of heat is dissipated by this device at maximum TEC current. Of course, this reduces the contribution of the sense resistor 206 to thermal run-away, and is, therefore, another significant advantage when compared to the conventional voltage-mode TEC control circuit.
Finally, it is noted that limiting the TEC current in accordance with the present exemplary embodiment does not require transistor-based sensing circuitry. Among other advantages, this avoids the limiting of the TEC operational range as may occur in conventional voltage-mode TEC controllers as a result of the temperature dependent V[0040] _beassociated with Q_pand Q_nin FIG. 1.
The invention having been described in detail in connection through a discussion of exemplary embodiments, it is clear that various modifications of the invention will be apparent to one having ordinary skill in the art having had the benefit of the present disclosure. Such modifications and variations are included within the scope of the appended claims. [0041]

Claims

I claim:

1. An apparatus for controlling the operating temperature of a thermo-electric cooler, comprising:

a current-sense circuit which is adapted to control a current through the thermo-electric cooler based on a set-point temperature and a detected temperature of the thermo-electric cooler using proportional, integral, derivative (PID) processing.

2. An apparatus as recited in claim 1, wherein said current-sense circuit limits a maximum current through the thermo-electric cooler.

3. An apparatus as recited in claim 1, wherein said current through the TEC is controlled to approximately 2% or better.

4. An apparatus as recited in claim 1, further comprising a control block.

5. An apparatus as recited in claim 4, wherein the control block further comprises a temperature monitor circuit and a proportional, integral, derivative (PID) controller.

6. An apparatus as recited in claim 1, further comprising a difference amplifier circuit.

7. An apparatus as recited in claim 1, wherein a sense resistor is operatively connected to said thermo-electric cooler.

8. An apparatus as recited in claim 6, wherein said difference amplifier further comprises an operational amplifier and a plurality of resistors including a sense resistor.

9. An apparatus as recited in claim 8, wherein said difference amplifier controls an amount of current through the thermo-electric cooler.

10. An apparatus as recited in claim 8, wherein said plurality of resistors are selected such that a maximum current through the thermoelectric cooler is not exceeded.

11. An apparatus for controlling the operating temperature of a thermo-electric cooler, comprising:

a current-sense circuit which is adapted to control a current through the thermo-electric cooler.

12. An apparatus as recited in claim 11, wherein said current-sense circuit limits a maximum current through the thermo-electric cooler.

13. An apparatus as recited in claim 11, wherein said current through the TEC is controlled to approximately 2% or better.

14. An apparatus as recited in claim 11, further comprising a control block.

15. An apparatus as recited in claim 14, wherein the control block further comprises a temperature monitor circuit and a proportional, integral, derivative (PID) controller.

16. An apparatus as recited in claim 11, further comprising a difference amplifier circuit.

17. An apparatus as recited in claim 11, wherein a sense resistor is operatively connected to said thermo-electric cooler.

18. An apparatus as recited in claim 16, wherein said difference amplifier further comprises an operational amplifier and a plurality of resistors including a sense resistor.

19. An apparatus as recited in claim 18, wherein said difference amplifier controls an amount of current through the thermo-electric cooler.

20. An apparatus as recited in claim 18, wherein said plurality of resistors are selected such that a maximum current through the thermo-electric cooler is not exceeded.