Temperature Monitoring
Temperature monitoring is mainly used as a safety feature to switch off the unit should the temperature rise above a predefined threshold. Temperature monitoring can be implemented at the Level 1 Control. The cause of an over-temperature event can be:
- Insufficient airflow in the system, perhaps caused by a failure of the cooling fan
- Operation of the system at a high ambient temperature
It's a good practice for the power converter designers to implement temperature monitoring of the power switches due to protection and performance (current consumption, Direct Current (DC), and Alternate Current (AC) characteristics) reasons.
Temperature Monitoring: Analog Solution
A temperature sensor, such as the MCP9700A, outputs an analog voltage proportional to the measured temperature. This temperature knowledge may be used to change the loop performance in order to compensate for it. To prevent damage due to excessive temperature, a comparator may also be used to rapidly detect the threshold and immediately shut down the unit.
The output of the temperature sensor can be connected to an analog input of the dsPIC® Digital Signal Controllers (DSCs). The board temperature is measured in the Analog-to-Digital Converter (ADC) Interrupt Service Routine (ISR) and checked for over-temperature protection in the fault loop. Figure 1 shows an example of an MCP9700A temperature sensor interfacing to a PIC® microcontroller.
The maximum temperature set point is configured in reference to the maximum operating temperature, which can be found in the datasheet. If the measured temperature exceeds the maximum setpoint, a fault is generated and the Pulse Width Modulation (PWM) outputs are turned OFF.
Temperature Monitoring: Digital Solution
On the dsPIC DSCs side, the low and high-temperature thresholds can be changed. In the case of over-temperature, the voltage can be driven smoothly to zero. Upon detection of an over-temperature event, the dsPIC DSC can wait for a certain amount of time (or check the continuous presence of high temperature) before concluding that a real over-temperature event has occurred. Variable hysteresis can be also added. The system can easily make a predetermined number of attempts to restart from an over-temperature status. If these attempts fail, the system will shut down permanently. Redundant power supplies can balance the temperature by adjusting their output currents, and optimizing the current sharing.
The dsPIC DSC can easily accommodate any of them. Because temperature is simply a number inside the dsPIC DSC, it is very easy to set the temperature threshold and also set up a hysteresis cycle in software. The temperature information can be used to change the loop performance. In an over-temperature event, the information can be used to smoothly shut down the unit (soft start). Moreover, there is a much higher versatility, since, for instance, the designer could define that an over-temperature event only occurs when the temperature remains higher than the threshold for a predetermined period of time. Even in an over-temperature event, the system could make a predetermined number of attempts to start the unit again. If the system is always in the over-temperature state, after the predefined number of attempts, the system remains shut down.
Temperature Sensor: Diode
A diode can be used as a temperature sensor. The MCP9700A provides an accurate temperature measurement by monitoring the voltage of a diode located inside the Integrated Circuit (IC). Figure 3 shows the cross-sectional diagram of the temperature sensing diode based upon a Positive Negative Positive (PNP) transistor. The details of the voltage to temperature transfer function can be found in the MCP9700A datasheet.