Gert van der Horn

Developments in integrated smart sensors

Abstract An integrated smart sensor (ISS) can be defined as a chip that contains the functions of sensing, signal conditioning, AD conversion and a bus output. It may also have on-chip calibration and self-test. This paper concentrates on the standardization of the output interface using on-chip AD conversion, calibration and bus output. First, it is shown how indirect AD-conversion techniques yield simple circuits which can easily coexist with the sensor on one chip. Sigma-delta conversion is a good example. Second, calibration methods are described for analog and digital sensor output signals. Finally, the use of two simple busses is described, the I2C bus and the IS2 bus. The required bus interfaces are relatively simple and can be integrated with sensors in bipolar as well as in CMOS processes.

Integrated smart sensor calibration

In many applications electronic sensors are used toimprove performance and reliability of measurement systems. Suchsensors should provide a correct transfer from the physical signalto be measured to the electrical output signal. One importantstep to achieve this, is to calibrate each sensor by applyingdifferent reference input signals and adjusting the sensor transferaccordingly. Besides expensive reference equipment the calibrationprocess takes much time and attention per individual sensor,which means a considerable increase in sensor production costs.By including at the sensor or sensor interface chip a programmablecalibration facility the calibration of such smart sensors caneasily be automated and can be executed for a batch of sensorsat a time, thus minimizing the calibration time and costs. Thispaper presents a calibration method and options for integrationin the smart sensor concept, in hardware as well as in software.An advantage of the proposed method is that it does not needa large matrix of calibration data, which needs to be storedin a look-up table or converted into a correction formula, butinstead it uses a step-by-step approach to correct the sensortransfer at each calibration measurement until the error is sufficientlysmall.

Calibration and Linearization Techniques

In the first part of this chapter we will identify and distinguish the different types of errors which affect the transfer of the sensor, and explain which errors can be corrected by calibration. In the following section we will explain different linearization techniques which can be used to calibrate the offset, gain, and linearity errors in the sensor transfer. In the last part we will propose and explain a polynomial calibration method which can be used to calibrate and linearize the sensor transfer in a step-by-step approach. It will be shown how the method can be expanded to a two-dimensional polynomial calibration to be used for calibration of a cross-sensitivity error.

Calibration using Digital Signal Processing

It is not the aim of this chapter to investigate or discuss all the possible digital forms of sensor calibration techniques, but rather to demonstrate the feasibility of the progressive polynomial calibration method proposed in chapter 3.

Integrated Smart Sensor Concept

In this chapter we will present a concept for implementing smart sensor interfaces [1]. Within the scope of this work we cannot discuss many types of sensors or sensor interfaces, hence we limit ourselves to two common sensor types, for measuring temperature, and for measuring pressure. This limitation does not mean that the presented matter cannot be extended to other types of sensors.

Calibration using Analog Signal Processing

In this chapter we will focus on calibration techniques based on analog signal processing. Almost all sensors require some kind of analog signal conditioning or amplification before the sensor signal can be converted and processed digitally. Basically, sensors produce an analog output signal only. But even for modern smart sensors with digital output, a standardized analog output signal is often desired as well. Therefore, we will first consider the possibilities of adjusting the transfer of an analog circuit for the purpose of calibration.

Calibration using Sigma-Delta Analog-to-Digital Conversion

In section 4.4 the use of pulse modulation techniques to control the signal transfer of a smart sensor was already suggested. The pulsestream or bitstream modulation can be combined nicely with sigma-delta bitstream techniques. A digital sigma-delta (noise shaper) may be used to generate programmable pulsestreams or bitstreams which modulate the analog signal, and a sigma-delta AD-converter is used to convert the modulated analog sensor signal into a digital bitstream.