AD590 thermal connection. Power source P represents the power ated on the chip The rise of the junction temperature T), ADE8 uation Table I gives the sum of Bjc and ecA for several common thermal media for both the"H"and"F"packages. The heatsink used was a common clip-on. Using Equation 1, the temperature rise of an aD590“H” package in a stirred bath at+25°C,when driven with a 5 V supply, will be 0.06 C. However, for the same conditions in still air the temperature rise is 0.72C. For a given Figure 7A. Two Temperature Trim supply voltage, the temperature rise varies with the current and PTAT. Therefore, if an application circuit is trimmed with the sensor in the same thermal environment in which it will be +2c used, the scale factor trim compensates for this effect over the entire temperature range Table l thermal resistances Medium OJc+0cAC'C/Watt) t(sec)(Note 3) Aluminum block 0.6 0.1 Stirred Oill 0.6 Figure 7B. Typical TwO-Trim Accuracy Moving air2 With Heat Sink VOLTAGE AND THERMAL ENVIRONMENT EFFECTS Without Heat Sink115190 13.510.0 The power supply rejection specifications show the maximum Still Air expected change in output current versus input voltage changes With Heat sink The insensitivity of the output to input voltage allows the use of Without Heat Sink 480 650 unregulated supplies. It also means that hundreds of ohms of resistance(such as a CMOS multiplexer) can be tolerated in INote: t is dependent upon velocity of oil average of several velocities listed above series with the device 2 Air velocity≡9t/se It is important to note that using a supply voltage other than 5v The time constant is defined as the time required to reach 63. 2% of an words, this change is equivalent to a calibration error and can be The time response of the AD590 to a step change in tempera- removed by the scale factor trim(see previous page) The AD590 specifications are guaranteed for use in a low thermal 0.04 watt-secFC for the AD590. Cc varies with the measured resistance environment with 5 V across the sensor. Large changes in the thermal resistance of the sensors environment medium since it includes anything that is in direct thermal contact with the case. In most cases, the single time constant ill change the amount of self-heating and result in changes in the output which are predictable but not necessarily desirable exponential curve of Figure 9 is sufficient to describe the time response, T(t). Table I shows the effective time constant, t, for The thermal environment in which the ad590 is used deter- several media mines two important characteristics: the effect of self heating and the response of the sensor with time FINAL T(t]= TINITIAL +ITFINAL-TINITIAL HC Figure 8. Thermal Circuit Model Figure 8 is a model of the AD590 which demonstrates these characteristics. As an example, for the To-52 package, Ajc is he thermal resistance between the chip and the case about 26 C/watt. ecA is the thermal resistance between the case and the surroundings and is determined by the characteristics of the Figure 9. Time Response Curve V.BAD590 –6– REV. B Figure 7A. Two Temperature Trim Figure 7B. Typical Two-Trim Accuracy VOLTAGE AND THERMAL ENVIRONMENT EFFECTS The power supply rejection specifications show the maximum expected change in output current versus input voltage changes. The insensitivity of the output to input voltage allows the use of unregulated supplies. It also means that hundreds of ohms of resistance (such as a CMOS multiplexer) can be tolerated in series with the device. It is important to note that using a supply voltage other than 5 V does not change the PTAT nature of the AD590. In other words, this change is equivalent to a calibration error and can be removed by the scale factor trim (see previous page). The AD590 specifications are guaranteed for use in a low thermal resistance environment with 5 V across the sensor. Large changes in the thermal resistance of the sensor’s environment will change the amount of self-heating and result in changes in the output which are predictable but not necessarily desirable. The thermal environment in which the AD590 is used deter￾mines two important characteristics: the effect of self heating and the response of the sensor with time. Figure 8. Thermal Circuit Model Figure 8 is a model of the AD590 which demonstrates these characteristics. As an example, for the TO-52 package, θJC is the thermal resistance between the chip and the case, about 26°C/watt. θCA is the thermal resistance between the case and the surroundings and is determined by the characteristics of the thermal connection. Power source P represents the power dissipated on the chip. The rise of the junction temperature, TJ, above the ambient temperature TA is: TJ −T A = P (θJC + θCA ) Equation 1 Table I gives the sum of θJC and θCA for several common thermal media for both the “H” and “F” packages. The heatsink used was a common clip-on. Using Equation 1, the temperature rise of an AD590 “H” package in a stirred bath at +25°C, when driven with a 5 V supply, will be 0.06°C. However, for the same conditions in still air the temperature rise is 0.72°C. For a given supply voltage, the temperature rise varies with the current and is PTAT. Therefore, if an application circuit is trimmed with the sensor in the same thermal environment in which it will be used, the scale factor trim compensates for this effect over the entire temperature range. Table I. Thermal Resistances Medium θJC + θCA (8C/Watt) τ (sec)(Note 3) H F HF Aluminum Block 30 10 0.6 0.1 Stirred Oil1 42 60 1.4 0.6 Moving Air2 With Heat Sink 45 – 5.0 – Without Heat Sink 115 190 13.5 10.0 Still Air With Heat Sink 191 – 108 – Without Heat Sink 480 650 60 30 1 Note: τ is dependent upon velocity of oil; average of several velocities listed above. 2 Air velocity ≅ 9 ft./sec. 3 The time constant is defined as the time required to reach 63.2% of an instantaneous temperature change. The time response of the AD590 to a step change in tempera￾ture is determined by the thermal resistances and the thermal capacities of the chip, CCH, and the case, CC. CCH is about 0.04 watt-sec/°C for the AD590. CC varies with the measured medium since it includes anything that is in direct thermal contact with the case. In most cases, the single time constant exponential curve of Figure 9 is sufficient to describe the time response, T (t). Table I shows the effective time constant, τ, for several media. Figure 9. Time Response Curve