Temperature sensor versus thermistors
Many methods exist to measure temperature in electronic systems. To measure the local temperature of an object or area of a PCB, the most common methods are to use a thermistor or local temperature sensor. Table 1 summarizes the basic principles of operation for both thermistors and silicon temperature sensors, and briefly summarizes the advantages and disadvantages of each.
Let's look specifically at digital temperature sensors. These integrate additional signal conditioning including an analog-to-digital converter (ADC). The output is a digital word that represents the temperature experienced by the bandgap junction. Here are some of the advantages they provide over thermistors.
A digital interface, such as I2 C, SMBus, or SPI, by definition, has significantly better noise immunity versus an analog interface. As the analog output of a thermistor transits the PCB trace to the signal conditioning circuitry, it is subject to multiple noise sources, whether from electromagnetic interference (EMI) from off-board sources, or potential cross-talk with neighboring traces.
Digital signals are subject to the same noise sources. However, since only a threshold (digital 0 or 1) is important, it takes a significant noise source to create an error in a bit stream. Analog signals, by nature, are continuous. So any noise that adds to or subtracts from the signal level creates additional error in the measured temperature reading.
A thermistor requires a constant current excitation to create a voltage drop across it that is representative of the temperature being measured. This current is usually on the order of 100μA. In addition to this current flow, any signal conditioning circuitry (amplifiers and/or ADCs) add to this value. Digital temperature sensors can vary greatly in power consumption. However, Texas Instruments has focused on minimizing power consumption and have multiple digital temperature sensors that are 50μA and under. Included is the TMP103, which has a maximum power consumption of 3μA, or 3 percent of just the excitation current required by a thermistor. System management resources
A digital temperature sensor can free up system management resources in many ways. The most obvious is that with the ADC integrated on chip, no additional signal conditioning is required to provide temperature to the system management decision making module.
Additionally, the resistance change of a thermistor relative to temperature is non-linear (Figure 1). This requires the system management controller (SMC) to implement a look-up table to accurately understand the temperature. Also, this non-linearity causes the ADC error to be non-uniform across the full input range as each LSB represents a different change in temperature.
Many digital temperature sensors offer an ALERT capability for over/under temperature conditions which can be used as interrupts to the SMC. This allows it to ignore the actual temperature that the system is experiencing until it reaches a threshold at which point the SMC can start monitoring the temperature. Otherwise, you'd have to monitor the thermistor output in real time and make decisions on every reading.
Multiple temperature readings
Finally, if multiple temperature reading locations are required to ensure system management, thermistors require signal conditioning resources for each. This ties up multiple pins and ADCs within the SMC, if not separate external ADCs for each thermistor. Power consumption becomes even more evident as each thermistor requires its own current excitation source of 100μA. Since the industry standard interfaces utilized by digital temperatures all allow for multiple devices to exist on a single bus, SMC resource requirements are minimized.
For example, the TMP103 allows for up to eight devices on a single I2 C master via eight unique address options. The TMP104 utilizes a novel one-wire interface called SMAART wire that allows for up to 16 devices to be daisy chained and managed via a single serial (UART) interface on the SMC (Figure 2).
Please join us next month where we will see how to optimize a SAR ADC driver based on a given precision versus power requirement.
For more information, visit: www.ti.com/tempsensor-ca.
About the author:
Dan Harmon is Sensing Business Development Manager for TI's sensing group. In his 25+ year career at TI, he has supported a wide variety of technologies and products including interface products, imaging analog front-ends (AFEs), and charge-coupled device (CCD) sensors. He also has served as TI's USB-IF Representative and TI's USB 3.0 Promoter's Group Chair. Dan earned a BSEE from the University of Dayton and a MSEE from the University of Texas in Arlington. You can reach Dan at . Related posts:
- Signal Chain Basics #78: How to Avoid Common-Mode Limitations on Instrumentation Amplifiers
- Signal Chain Basics #77: Enjoying the VU: Real-World Implementations
- Signal Chain Basics #76: Design a Low-Cost PWM Circuit for Single IGBT-Drive Applications
- Signal Chain Basics #75: Driving a Low-Impedance Load With a Precision Reference
- Signal Chain Basics #74: Band-Limited Dither Improves SFDR in High-Speed DACs
- Signal Chain Basics #73: Audio metering, enjoying the VU
- Signal Chain Basics #72: Isolating analog signals using a digital isolator
- SIGNAL CHAIN BASICS #71: How supply noise impacts clocking devices
- SIGNAL CHAIN BASICS #70: Digital correction of analog quadrature modulator imbalances