Full report

Original page 1

NASA / TM- 1999-209689 Temperature Control of Avalanche Photodiode Using Thermoelectric Cooler Tamer F. Refaat Old Dominion University, Norfolk, Virginia William S. Luck, Jr. and Russell ]. DeYoung Langley Research Center, Hampton, Virginia National Aeronautics and Space Administration Langley Research Center Hampton, Virginia 23681-2199 October 1999

Original page 2

Available from: NASA Center for AeroSpace Information (CASI) 7121 Standard Drive Hanover, MD 21076-1320 (301 ) 621-0390 National Technical Information Service (NTIS) 5285 Port Royal Road Springfield, VA 22 !61-2171 (703) 605-6000

Original page 3

1. Introduction The avalanche photodiode (APD) is a solid state, quantum, optical detector which uses the photogeneration of charge carriers in order to detect light. APDs are highly recommended for low-light optical detection applications in the visible and near infrared intensity, compactness, ruggedness, and relatively region due to their linearity with incident light large internal gain. The APD output current is dependent on both the operating temperature and the bias voltage of the device (ref. 1 and 2). The bias voltage can be set using a stable, high-voltage source. To address the temperature dependence of the output current, many APD packages come with a built-in thermoelectric cooler (TEC) for accurately controlling the operating temperature (ref. 3). This memorandum reviews the theory of thermoelectric cooling. Also, the design and performance of a proportional integral (PI) temperature controller and its application to an EG&G APD package (C30649E), which includes a built-in TEC and a thermistor, are discussed. This APD package was chosen, after characterizing a group of APDs, in order to develop an advanced detection system for atmospheric water vapor DIAL measurements (ref. 2, 4, 5 and 6). 2. APD Temperature Dependence An APD is a device similar to a rectifier diode, except its output current contains a term which is dependent on the incident light intensit_¢ on its surface in the operating wavelength range. The APD output current is given by (ref. I, 7 and 8) qV lAp D =-I d + I s e kT -I (I) where IAp D is the APD output current, Id is the detected photo-current, Is is the saturation dark current, q is the electron charge, V is the device bias voltage (negative for reverse bias), k is Boltzmann's constant, and T is the temperature. The second term of equation (1) represents the APD dark current and the first term, Id, represents the photo-current, given by (ref. 1, 7 and 8) Ij = _R. P (2) where _ is the APD responsivity, and P is the incident optical power. The APD responsivity, _ (A/W), is obtained from (ref. I, 7 and 8) = riG. q. hc _. (3) where rl is the wavelength dependent quantum efficiency, G is the APD internal gain, h is Planck's constant, c is the speed of light, and _, is the wavelength of the incident light. At a constant bias voltage, the APD operating temperature affects its output current. The APD gain, and therefore its responsivity, is a strong function of the device temperature. On the other hand, the APD dark current, as well as the dark current noise, is also dependent on the APD temperature.

Original page 4

At a constant bias voltage and wavelength, the APD responsivity increases by decreasing the device temperature. Therefore, cooling the APD is recommended to increase its detected photo-current. Figure 1 clearly indicates this fact by showing the responsivity-versus-temperature relations at three different wavelengths for a 336V bias voltage (ref. 2, 7 and 8). The APD dark current, given by the second term of equation (I), is also dependent on the device operating temperature. Decreasing the APD temperature will increase the dark current. The maximum dark current will be obtained at a temperature of 0 K and will be equal to the saturation current. More importantly, the APD dark current is associated with noise, known as the dark current shot noise, which is given by (ref. 9 and 10) 160 140 120 8O "i 100 60 40 2°s d lb 11 21 3O Temperature in °C Figure 1. EG&G C30649E responsivity variation with temperature at 336 V bias at (o) 720 nm, (+) 820 nm and (x) 940 nm. (See ref. 2) I s. e kT + I + (4) in2 = 2qGFB where in is the shot noise current, F is the excess noise factor, and B is the device bandwidth. The APD temperature affects the dark current shot noise directly in the denominator of the exponential power and indirectly by its effect on the device gain. The first effect is dominant; and for a reverse bias voltage, the dark current shot noise is reduced by reducing the operating temperature. In most applications, the effect of the APD dark current on the output signal can be eliminated by either modulating the input optical signal (ref. 2) or by subtracting the dark current from the device output (ref. 4). Therefore, low temperature operation of APDs is recommended in order to increase the device responsivity and to reduce the dark current shot noise. 2

Original page 5

1. Thermoelectric Coolers A thermoelectric cooler module consists of an array of semiconductor (group V-VI, e.g. bismuth telluride) pellets that have been positively (p) or negatively (n) doped. The p-n pellet pairs are connected electrically in series and thermally in parallel (ref. II, 12 and 13). Consider a simple thermoelectric cooling unit, as illustrated in figure 2, in which the electrical resistance between the semiconductor elements and metal links, as well as the resistances of the links themselves, are negligible. Also, the thermal resistance between the semiconductor pair, the heat source, and the heat sink are neglected. For simplicity, all of the material thermal coefficients are assumed constant and independent of temperature. 3.1. TEC Theory of Operation TEC operation is based on the Peltier effect which results in the transport of heat when electric current flows in a conductor (ref. I! and 13). This heat flow is revealed only at a junction between two different materials where the heat transport on either side is different. In the two branches of figure 2, the heat transported (Watts) from the source is given by (ref. 1 I ) Heat Source (Tc) [] Semiconductor [] Ceramic p [] Metal • Heat Sink (TH) n Type L I .... 1 ×=0 | | , I Figure 2. A TEC p-n pair. qp =(ZpIT- KpA dx dT (5) dT qn =-o%IT- _n A- (6) dx where _, and cxn are the Seebeck coefficients (WK-]A -1) of the p and n materials, respectively, 1¢p and _n are their thermal conductivity (Wm-IK-I), I is the TEC current, A is the semiconductor pellet crosssectional area, and T is the temperature, ct n is a negative quantity, and the thermoelectric heat flow from the source through both branches is positive and is opposed by the effect of thermal conduction. Within

Original page 6

the branches,the rateof generationof heatperunit length(js-lm -1) from the Joule effect is given by (ref. I 1) (7) - KnA d2T - I29n (8) dx 2 A where 9p and Pn are the electrical conductivity (f_.m) of the p and n materials, respectively. Now, if the branches are of length L, then the boundary conditions are T=T C at x=0, and T=T H at x=L. Solving differential equations (7) and (8) and substituting (ref. ! 1 ) KpA(TH- TC)I2ppL qp = plT C K'nA(T H -T C) I29n L qn =-CZn ITc - in differential equations (5) and (6) at x--0, we get (9) L 2A (10) L 2A The cooling power Qc at the sink is the sum of the above two equations and is given by Qc =°tlTc -K(TH / -Tc)---I2R (11) 2 where ct.=Otp- n is the differential Seebeck coefficient of the unit. The thermal conductance, K, of the two branches in parallel is given by lCpA lenA K - L and the electrical resistance R of the two branches + (12) L in series is given by R = ppL - (13) A 9n L A From equation (I 1), it is interesting to note that the reversible cooling process (represented by the first term) is opposed by the sum of two irreversible processes, which are the heat conduction (represented by the second term) and half the joule heating (represented by the power loss in the third term). The electrical power consumed in the branches is given by (ref. 1 I ) I2ppL pp =RpI(T H -T C)- (14) A

Original page 7

Pn=-O_nI(TH-Tc) I2pnL +-- (15) A Thus,thetotalpowerinputis givenbythesumof equations(14)and(15) P= otI(TH -T C)+ 12R (16) Applyingthefirstlawof thermodynamics,theheatingpower,QH,atthesinkis thesumof equations(1I ) and(16)andisgivenby QH =ctlTH- K(TH- Tc ) + lI2R (17) 2 Finally,thesystemefficiencyis examinedby thecoefficientof performance,¢p,whichis the ratioof therateatwhichheatis extractedfromthesourceto therateof expenditureof electricalenergy(ref. I 1 and13).Usingequations(I I) and(16),thecoefficientof performanceis givenby 1i2R Qc °dTc- K(TH- TC)- 2 q3- - (18) P I(TH -Tc)+ I2R Fora certaintemperaturedifference,thecurrent,Im,formaximumcoefficientof performance,¢Pm,canbe evaluatedby settingdcp/dI=0. In thiscase,¢Pmisgivenby TCI(1 + ZTm) I/: - (TH/Tc) 1 ¢Pm - 1t2 (19) (T H -Tc I(l + ZTm) +1] where Z is the TEC figure of merit, which is given by (ref. I 1 and 13) o- Z - KR and T m is the average temperature which is given by T H +T C T m - 2 3.2. TEC Cooling Cycle (20) (21) The TEC cooling cycle can be further understood in a qualitative manner by studying the energy band diagram of its materials, as shown in figure 3. An electron at the cold metal absorbs a heat quanta, which leads to an increase in the electron's potential energy. The applied electric field from the supply causes the electron to drift to the n-material. A fraction of the electron energy is converted to kinetic energy as 5

Original page 8

COLD Dev_e (-) Energy Heat • _ .i• Abs0rpti°:t _ Valance Band Electron _ HOT COLD Metal (+) Heat Form,Lov,! ! Rejecti°n • Figure 3. Energy band diagram for a thermoelectric cooling unit. the electron proceeds through the n-material. At the end of the n-material, the electron rejects its energy gain to the hot metal in the form of heat. In order to complete the cycle, the electron must reject more energy to pass through the p-material using the lower band. On the other hand, at the p-material coldmetal junction, the electron gains more energy, through heat absorption, in order to complete its cycle. Similarly, the holes do the same thing in the opposite direction, which leads to a heat transfer from the cold metal to the hot metal by both charge carriers. The power supply must provide the energy for the carriers to continue the cycle and also provide the system with the energy that is lost due to the electrical resistance of the cooling unit. 4. APD Temperature Controller The EG&G APD C30649E package, shown schematically in figure 4, is supplied with a built-in TEC cooler to control the detector temperature. The temperature status of the APD is sensed by the thermistor, which is located as close as possible to the APD in order to ensure a minimal temperature gradient between the two devices. For this particular thermistor, the relation between its resistance (f) and the temperature (K) is given by (ref. 3) RT=104'exp[3940( TEC I I TIA Feedback _j.....- Resistance = '/ Thermistor - ! IT 2981/] (22) I I Figure 4. APD package physical diagram. 6

Original page 9

The APD temperaturecontrollercircuit is shownin the circuit diagramof figure 5. The APD temperaturefeedbackis obtained by placing the thermistor in an arm of a Wheatstone bridge. The remaining arms are formed by three equal-valued resistors, R s. According to equation (22), the value of R s determines the balance condition of the bridge and sets the detector temperature to a value given by [,°( 3940 2-_ (23) 't-' The bridge is supplied by the zener diode, D, with a zener voltage V Z. R Z is a current-limiting resistor which stabilizes the zener diode voltage and prevents its break down. V CC Rz Thermistor V Z Rs v, D_ v2 I Rs Rs V CC C Rc _r I a L [ I Vo _.Q TEC VTM 7 Op Amp 1 Figure 5. APD temperature controller circuit diagram. The bridge balance error, V2-V I, is sensed by a PI controller formed by two operational amplifiers (op amp's). Op amp 1 acts as a voltage follower with an output voltage VTM. This voltage is used to monitor the APD operating temperature. The output of op amp 2 is applied to a potentiometer, Rp, to set the TEC current by controlling the emitter current of transistor Q. The setting of Rp changes the controller overall gain KpI. According to the analysis of the circuit shown in figure 5, the controller input-output relation is given by Vo = K H V2 +(v2-v,)]sCRc 7 (24)

Original page 10

In termsof theerrorsignal,Err,definedasthedifferencebetweenV2 andV1,the sameequationcanbe rewrittenas (25) Err Resistor R L is used to limit the TEC current that corresponds to its maximum value (See Appendix A). The temperature monitor voltage reading, VTM, is given in terms of the zener voltage as VT M - R s + R T RS Vz (26) If the APD temperature reaches its steady state value, the monitor voltage is given by VTM = 2 Vz (27) I Appendix B gives the value of the controller circuit components used to construct this controller. 5. Heat Load Estimation The heat loads at the cold side of the TEC cooler are due to active and passive heat sources. In the EG&G C30649E APD package, the detector output photo-current variation is converted into a voltage variation by a trans-impedance amplifier (TIA) (ref. 1,2 and 3). The feedback resistor R F of this amplifier is cooled by the same TEC cooler in order to reduce the Johnson noise. The APD, thermistor, and TIA feedback resistor define the three active heat loads QAPD, QTHR, and QTIA, respectively. The power consumed in these elements is directly transformed to heat and is given, respectively, by. VBIAS "IAPD (28) QAPD = QTHR - (RT+Rs)2 RT V_ (29) QTIA = I2pD "R F (30) TEC passive heat loads come from the temperature gradient between the cooled components and the ambient environment. The first passive load, QRAD, is the heat radiation from the ambient environment to the TEC cold side and is given by (ref. 12) QRAD = F.e .s. ATECt amb-T 4 ) (31) T 4 where F is the surface shape factor, e is the surface emissivity, s is the Stefan-Boltzmann constant, ATE C is the area of the TEC cold surface, and Tam b is the ambient temperature. The second passive load, QCON,

Original page 11

is dueto theconductionbetweentheTECcoldsurfaceandtheambientenvironmentthroughtheair space belowthepackagewindow,whichhasa lengthLp(seeAppendixA). Thisis givenby (ref. 12) QCON - KAirATEc (Tamb - Tc ) (32) Lp where Kai r is the air thermal conductivity. At the TEC hot side, heat accumulation will lead to device failure. Therefore, a heat sink is used for heat dissipation to the ambient environment. Considering the worst-case situation in which the heat transfer from the heat sink to the ambient environment is only through radiation, the amount of heat radiated from the heat sink, QHS, is given by T4 -T4mb) (33) QHS = FeSAHs( where AHS is the heat sink surface area. The heat transfer problem is summarized in the power flow diagram of figure 6. Assuming the worst case in which the surface shape factor and emissivity are both equal to one, the heat loads at the source are evaluated using equations (28) through (32) by assuming an ambient temperature of 27 "C and maximum possible values for the active heat loads. The sum of the heat loads is equated with the required cooling power given by equation (IlL Knowing the TEC parameters (given in Appendix A) and setting the cold temperature equal to 0.3 "C, the hot temperature, T H, at the sink is evaluated, using equation ( 1 I ), to be 30.7 "C. Substituting these temperatures into equation (17), the heating power at the sink is obtained. Since, in the worst case, this heating power is dissipated only by radiation, substituting its value in equation (33) defines the heat-sink surface area required (104.5 cm2). In reality, a considerable amount of the heating power is dissipated by conduction rather than only through radiation. Therefore, for mechanical assembly considerations, the heat sink used is only half of the calculated area given above. However, extensive testing has proven that the reduced heat sink area is sufficient under all expected operating conditions. QAPD QTHR QTIA QRAD QCON Heat ourc P Figure 6. Power flow diagram for the APD cooking system. 9

Original page 12

1. Controller Performance Referring to equation (23), the APD temperature is set to 0.3 "C by setting R S equal to 33 k_. For an initial APD temperature of 23.6 "C, figure 7 shows the transient response of the APD temperature monitor voltage VTM, the thermistor resistance, the APD temperature, and the TEC current. The rise time is approximately 5 s, and the 5 percent settling time is approximately 50 s with about 2 "C of overshoot. The steady-state temperature was 0.3_+0.3 "C. The temperature measurement is obtained by converting the temperature monitor voltage to a thermistor resistance according to equation (26), and then this is further converted into temperature according to equation (22). From figure 7(a) and (c), it is interesting to note a linear relationship between the APD temperature, 30 "C operating range, which can be approximated T, and the temperature monitor voltage, VTM, over a by = 17.2-VTM -43.6 (34) T Figure 7(d) shows a steady-state TEC current of approximately 0.35 A. From equation (20), the figure of merit of the TEC is calculated to be 1.6x10 -3. From equations (16) and (18), the TEC input power is 174.2 mW, which results in a coefficient of performance equal to 38.4 percent. The maximum theoretical coefficient of performance is 39.1 percent. For successful temperature controller operation, the maximum ambient temperature is 27 "C; and since the controller was designed to cool the detector, the minimum ambient temperature is theoretically equal to the controller temperature setting. 7. Conclusion For this memorandum, an APD temperature controller was designed, and its performance was evaluated. The APD considered is an EG&G APD package C30649E with a built-in thermoelectric cooler, thermistor, and trans-impedance amplifier. The theory of thermoelectric cooling is discussed, and the importance of sufficient heat sinking, in order to stabilize controller operation, is shown. The temperature controller performance is acceptable. The final steady-state temperature of 0.3 "C and power consumption of 174.2 mW, corresponding to a 38.4 percent TEC coefficient of performance, is close to the theoretical maximum value. For successful operation, the maximum allowable ambient temperature variation is between 27 "C maximum and a minimum equal to the temperature setting. This temperature controller design has adequate accuracy and is suitable for any APD which has a built-in or external thermoelectric cooler and a temperature feedback element. The design can be simply changed to meet any temperature setting, assuming that it is below the ambient temperature. Other advantages of this controller circuit are its simplicity, low-cost, and compactness. 10

Original page 13

4 3.8 3.6 3.4 (/) 3.2 "5 > 3 2.8 2.6 2.4 2.2 | | 40 35 30 25 2O 15 -20 0 2'0 40 & 80 1°20 0 2;_ 4b eb 80 Seconds (a) Temperature Monitor Output 25 20 15 oo 10 -5 ! I Seconds (b) Thermistor Variation 0.4 0.35 0.3 0.25 0.2 E < 0.15 0.1 0.05 0 -0.05 I I -20 8o 0 2; 4o 6o 8; 100 0 20 4; 6; Seconds (c) APD Temperature Variation Seconds (d) TECCurrent Figure 7. APD temperature and temperature monitor transienls, 11

Original page 14

References I. B. Salah & M. Teich, Fundamentals of Photonics, Wiley Intcrscience Publication, 1991 2. T. Refaat, G. Halama & R. DeYoung, "Characterization of Advanced Avalanche Photodiodes lbr Lidar Receivers", NASA/TP, under review. 3. EG&G SLIK APD C30649E data sheet. 4. T. Refaat, W. Luck & R. DeYoung, "'Design of advanced atmospheric water vapor differential absorption lidar (DIAL) detection system", NASA/TP- 1999-209348, 5. T. Refaat, W. Luck & R. DeYoung, "Advanced July 1999. Detector and Waveform Digitizer for Water Vapor DIAL Systems", 19" Intetvtational Laser Radar Conference, Annapolis, 1998. 6. T. Refaat, W. Luck & R. DeYoung, "Advanced Water Vapor DIAL Detection System", Conference on Lasers and Electro Optics, 1999. • 7. C. Crowell & S. Sze, "Temperature Dependence Physics Letters, Vol. 9, No 6, pp 242-244, 1966. of Avalanche Multiplication in Semiconductors", Applied 8. J. Conradi, "'Temperature Effects in Silicon Avalanche Dicxles", Solid State Electtvnics, Vol. 17, pp. 99-106, 1974. 9. R. Boyd, Radiometry attd the Detection of Optical Radiation, Wiley, New York, 1983. 10. G. Agrawal, Fiber-Optic Communication Systems, Wiley, New York, 1997. 1 I. H. Goldsmid, Electronic Refrigeration, Pion Limited, 1986. 12. Marlow Industries, Inc. "Thermoelectric Cooling Systems Design Guide", Publication No 017-7939, Rev 1, 1994. 13. A. Kapitulnik, "Thermoelectric cooling at very low temperature", Applied Physics Letters, 60, 1992. 12

Original page 15

Appendix A: EG&G C30649E Detector A1 Manufacturer Test Data Sheet DESCRIPTION CONDITION Positive Bias Current Vr, dark, 25 "C Negative Bias Current Vr, dark. 25 "C Output Offset Voltag.c Vr, dark, 25 "C APD Breakdown Voltag.e lhv=2uA, dark, 25 "C APD Operating Voltaic Resp=50 MV/W, 25 "C APD Dark Current Vr, dark, 25 "C Spectral Noise Peak Vr, clark, 25 "C Noise Equivalent Power Vr, dark, 25 "C APD Operating. Voltage Resp=50 MVAV, 0 "C APD Dark Current Vr, dark, (1"C Spectral Noise Peak Vr, dark, 0 "C Noise Equivalent Power Vr, dark, 0 "C Bandwidth lhv=2uA, dark, 25 "C Test are done at + 5V on the ampllifier with 5(g2 AC coupled. Test wavelength is 820 nm. Serial Number 147 A2 Parameters AAPD Detector Sensitive Area Resistance 560 k_ RI= Trans-lmpedance Amplifier Feedback QAPD Detector Maximum Power Loss Ilnax TEC Maximum Current Io oc TEC Rated Current @ 0 "C ATEC TEC Area R TEC Electrical Resistance K TEC Thermal Conductance Differential Seebeck Coefficient Lp APD Surface - Window Distance RT Thermistor Resistance 13 Package MIN. DATA MAX. UNITS II 18 22 mA 7.1 15 mA -I.5 0.19 0.5 V 325 356 500 V 300 349 48O V 35 nA 150 150 0.003 0.(X)6 pW/a_z 290 336 470 V 4 15 nA 140 15O 0.003 0.006 pw/,/h7 10 II MHz 0.24 mm 2 (measured) 50 mW 1.8A 0.5 A 3.96x3.96 mm 2 (I.4833_+(t. l f2 0.0102222 W/"(2 3.2222x 10-3 W/AK 5 mm R T =10 4 .e L -

Original page 16

A3 TEC Characteristics Curves 300 290 ._ 280 .= _ 270 I1 E _ 260 I-- 25O 24O 0 0.5 1 1.5 2 TEe Current in A (a) 1 > .Q. 0.8 0.6 & m _ .4 I,- 0.2 0 I 285 280 275 .1= 270 265 _ 260 E i 255 250 245 240 2.5 -0.1 (_ 011 012 013 014 015 016 Cooling Power in W (b) I I I I 0 0.5 1 1.5 2 TEe Current in A (c) (a) Cold side Temperature, Tc, versus TEC current, I, at cooling power, Qc=0 with still air and TH=300K. (h) Cold side Temperature, T C, versus cooling power, Qc, at TEC current, I=2A with still air and TH=300K. (c) TEC current, 1, versus TEC voltage drop at TH=Tc=300K. 14

Original page 17

Appendix B: Applied Temperature Controller System B1 Heat Sink Design I I I_ m_, t I '; f 15.00 ! i I.ql-_ m = 1.5 + Screw Head 15 50.00 _ I 31.5O _1 I I 1 I I i 24.50 Dimension in mm Material: Aluminum

Original page 18

B2 Dimension of the APD Package 8.75 _ ._ ., -41,-,_ 10O0 B3 Temperature Controller Parameters Vcc Power Supply Voltage Vz Zener Diode Voltage Rz Zener Diode Current Limiting Resistor Rs Temperature Setting Resistance Rc Integration Time Constant Resistance C Integration Time Constant Capacitance Rp Transistor Gain Control Resistance RL TEC Current Limiting Resistor 23._ .. ! o° i ." o' f i2X2.00 .... _ TM I ! ! Dimension in mm 12V 5.1V 2 k_ 33 k_ 200 k_ 2 t.tF I0 k_ 5f2 16

Original page 19

Original page 20

REPORT DOCUMENTATION PAGE I OMBFormNo.Approve_0704-0188 = Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information, Send comments regarding this burden estimate or any other aspect o| this collection ot information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Inlormation Operations and Reports. 1215 Jefferson Davis Highway. Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington, DC 20503 1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE October 1999 4. TITLE AND SUBTITLE 3. REPORT TYPE AND DATES COVERED Technical Memorandum 5. FUNDING NUMBERS Temperature Control of Avalanche Photodiode Using Thermoelectric Cooler 6. AUTHOR(S) WU 622-63-13-70 Tamer F. Refaat, William S. Luck, Jr., and Russell J. DeYoung 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) NASA Langley Research Center Hampton, VA 2368 I-2199 8. PERFORMING ORGANIZATION REPORT NUMBER L- 17906 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/MONITORING National Aeronautics and Space Administration Washington, DC 20546-0001 11.SUPPLEMENTARY NOTES 12a. DISTRIBUTION/AVAILABILITY STATEMENT Unclassified-Unlimited Subject Category 33 Distribution: Nonstandard Availability: NASA CASI (301) 621-0390 13. ABSTRACT (Maximum 200 words) AGENCY REPORT NUMBER NASA/TM- 1999-209689 12b. DISTRIBUTION CODE Avalanche photodiodes (APDs) are quantum optical detectors that are used for visible and near infrared optical detection applications. Although APDs are compact, rugged, and have an internal gain mechanism that is suitable for low light intensity; their responsivity, and therefore their output, is strongly dependent on the device temperature. Thermoelectric coolers (TEC) offers a suitable solution to this problem. A TEC is a solid state cooling device, which can be controlled by changing its current. TECs are compact and rugged, and they can precisely control the temperature to within 0. I "C with more than a 150 °C temperature gradient between its surfaces. In this memorandum, a proportional integral (PI) temperature controller for APDs using a TEe is discussed. The controller is compact and can successfully cool the APD to almost 0 "C in an ambient temperature environment of up to 27 °C. 14. SUBJECT TERMS APD; TEC; PI controllers; Temperature controllers 17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION OF REPORT OF THIS PAGE Unclassified Unclassified NSN 7540-01-280-5500 15. NUMBER OF PAGES 21 16. PRICE CODE A03 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT OF ABSTRACT Unclassified UL Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z-39-18 298-102

Original page 21

Original page 22