This chapter presents the integrated circuit test basics. Computer-aided design (CAD) computer software is used to design and test ICs. This software is sophisticated and highly specialized. Previously designed cells are stored in the CAD design library. Each cell is specifically tailored in relation to the other cells to perform the intended function. The designer uses these cells to design each layer of the circuit. In deposition, the conducting layers, metal, epitaxial silicon (epi), metal silicides, and insulating material such as silicon dioxide and related materials, are deposited on the wafer. This process is critical because lack of uniformity in the thickness of deposited layers creates reliability problems. In diffusion, the doping agent is introduced to the semiconductor substrate. A precondition for diffusion is the existence of a concentration gradient that facilitates the uniform distribution of atoms in the solid. Testing dies on a wafer or on a packaged die is done by sophisticated application-specific computers known as “automated test equipment” (ATE). ATEs may be different in physical appearance or software design, but they all have the same purpose. It has been found that the most common tests are propagation delay and rise and fall times.

This chapter elaborates the detail of diode and operation transistor. Any semiconductor integrated circuit (IC), regardless of size or technology, consists of hundreds to millions of diodes and transistors packed into a small space, usually less than 0.5 cm2 in size, all operating together. Semiconductors are made from either silicon or germanium. It has been found that the characteristics of these materials fall somewhere between those of conductors and insulators. The resistance of these materials to electron conduction makes them ideal materials for transistors and diodes. In P-type semiconductor creation, some electrons are removed from the semiconductor by the doping process to create more holes. In the N-type creation, electrons are added to carry current. There are many varieties of diodes and their use depends on their application. These devices comprise two terminals with characteristic behavior that is nonlinear with temperature. Transistors are called nonlinear because the relationship between the current and the voltage is not linear. They are current-controlled, meaning that a small increase in the input current causes a large change at the output. It has been found that the common collector configuration functions over wide frequency ranges and is used in impedance stepping.

This chapter presents the fundamentals of semiconductor digital logic network testing. Digital circuits, unlike analogs, deal with discrete pulses known as digits. In these types of networks, discrete signals in the form of either pulse or voltage levels are manipulated as discrete quantities. These discrete quantities are the representations of logical high or logical low. The internal circuitry between input and output can be very complex. The diode structures protect the internal device circuitry from damage by static charges during product handling. The main device circuitry is located in the box named the logic circuit, which can be any type of complex structure. Functional testing is the application of a train of pulses to the input pins of the DUT and checking the outputs for their response. The input stimuli and the output sensing are determined by the logic function of the device given by the manufacturer. Hysteresis in small amounts helps the rapid switching action of internal clocks and covers wide temperature variations. Voltage hysteresis is the difference between the positive-going input voltage when the output switches, and the negative-going input when the output switches again. It has been found that the binary search method reduces the test loop to not more than eight, which, in traditional methods, is in the order of tens for a fine measurement.

This chapter describes the various aspects of digital signal processing. Devices that convert analog signals to digital (A/D) and digital signals to analog (D/A) are the backbone of digital signal processing (DSP). DSP in automated test equipment (ATE) is a microprocessor-controlled operation that processes signals at high rates. DSP is broadly used in telecommunications, especially integrated service digital networks (ISDN), filtering and speech recognition, and comparison and medical imaging such as magnetic resonance imaging (MRI). Fourier analysis (Fourier series and transform) is the fundamental tool used to characterize linear or mixed-signal microchips. A signal sampled by this method can subsequently be analyzed for a variety of complex parameters such as signal-to-noise ratio (SNR), total harmonic distortion (THD), phase and gain discrepancies in the frequency domain. An audio digitizer is a module with a high-speed, high-resolution A/D converter. It digitizes an analog input signal and stores it in accessible memory as a sequence of digits. It has been found that the Fourier transform is used to determine the frequencies of nonperiodic waveforms with finite energy in each period.

This chapter describes different aspects of operational amplifier. An operational amplifier (op amp) amplifies or enlarges the input signal. At the transistor level, application of an AC voltage signal at the emitter of a transistor that is correctly biased in its active region causes a fluctuation in the collector current. An op amp is tested in a closed-loop configuration. Some form of feedback is required to set the gain and control of all the device functions. The feedback components also determine the frequency response of the device. In the fabrication of this type of op amp, only the resistors and transistors that are easily manufacturable on a chip are needed. It has been found that if the transistors are exactly identical, an increase in the collector current will not cause the output voltage to vary with temperature. The high-frequency op amp produces high output current that can drive a capacitive load or any side-effect of transmission line capacitance at the device output. In the chapter, the relationship between gain and the common input/output voltage of an op amp in terms of common mode rejection ratio has also been presented. It has been found that slew rate is the most important specification affecting the dynamic operation of an op amp and poses a severe limitation on the large-signal operation of the device.

This chapter deals with the different aspects of data acquisition devices. Data acquisition devices—analog-to-digital (A/D) and digital-to-analog (D/A) converters—are explained from a test engineer's point of view. Data acquisition devices that are used extensively in communication networks contain circuitry that deals with analog and digital quantities at the same time. Technologies such as MOS, bipolar, and BiCMOS are used in the fabrication of analog-to-digital converter (ADC) devices. The selection of any data converter must be based on the most important function one expects the device to perform satisfactorily. The main factors are accuracy, speed of conversion, and resolution. Among the secondary factors in choosing a converter are the output data format, temperature stability, required clock rating, power supply requirement, and output coding. The performance and parameter measurement of D/A or A/D devices can be severely affected by temperature changes. The only parameters that are not affected by temperature changes are quantization and resolution. It has been found that the input stimulus and the predefined output response are determined by the logical function of the device based on the truth table or timing diagram.

This chapter elaborates the noise identification and elimination methods. Noise identification and elimination methods are extremely valuable for design, product, or test engineers at any level. Discontinuity of electrical charges is the primary problem related to noise. These charges constitute the noise, or unwanted signal, when carried as discrete pulses in a conductor. They become more consequential as they are multiplied by the amplification stages. It has been found that grounding, if done properly, can be the primary answer to noise problems. Single-point grounding has the advantage of being the simplest method of grounding, but it is not very effective for controlling noise. The amount of noise in a resistor depends mainly on the shape of the component and the type of material in its structure. The resistor lead is the source of inductance and makes the element vulnerable to picking up surrounding magnetic fields, and the inherent capacitance of any resistor is about 0.5 pF. Most devices are capable of generating flicker noise, which is directly dependent on current flow. The main source of this noise is contamination at the emitter/base layer of the crystal in transistors and diodes. In this chapter the popcorn noise and contact noise have also been described.

This chapter describes important parameters in CODEC (coder/decoder) testing. The CODEC device is the backbone of telephone communications. The important parameters in CODEC testing are the total harmonic distortion (THD) and signal-to-noise ratio (SNR), which require the DSP process as well as the gain error. The details of an on-bench test connection to test the above parameters in single-tone methodology along with the explanation of the circuit operation are presented. The CODEC also coordinates the timing between itself and the networks with which it interfaces. Data transmission is synchronized and multiplexed. It has been found that when the encoding section receives the analog signal, the signal is passed through filtering stages to suppress low-frequency noise and the anti-aliasing process is then applied to it. A process called “companding” is essential to the CODEC's performance. Digitization of the analog pulse is not uniform, because input pulses themselves are not uniform. Companding is used to upgrade the quality of the pulse, which increases the signal-to-noise ratio and reduces the peak power to prevent overloading. The most important noise affecting CODEC design is the idle channel noise (ICN), which exists even when there is no signal traffic. It has been found that ICN occurs as a result of the quantization of a 0 V input analog signal as the digital output bounces around the assigned binary value for zero along with any nonlinearity existing along the transmission line.