Fundamentals of Mixed Signal Testing

Fundamentals of Mixed Signal Testing

The main purpose of functional testing is to ensure that the fabricated chip operates in conformance with its specifications. Based on the system specifications, the first step in the test process is to specify the testability features (e.g., scan, test points, etc.) followed by the set of tests defining the test plan. Once the test plan is approved, an initial test program is defined. The latter may undergo continual modifications while the design is still not yet finalized. The process of establishing such test program requires several steps, starting from the design specification to the generation of the test code (Burns & Roberts, 2001). In what follows we will examine these steps as well as the test program functionality in some detail.

Test Specification Process

One of the key elements in the design flow is the specification document also known as datasheet. It should provide explicit information about the functionality requirements that must be satisfied by the design. Moreover, it represents a formal communication channel between the designer and the test engineer. However, it may contain mistakes and ambiguities that must be corrected earlier in the design process. In practice, the specification datasheet has many sections (Burns & Roberts, 2001). Among these, some of the most important to test engineering are: the device description, principles of operation, electrical characteristics, timing diagrams and package/pinout information. A test is derived for the device description and principles of operation to ensure that the device fulfills the requirements. Same goes for the electrical parameters of the design that should be tested in all modes of operation. Typically, a test plan consists of several major components (Burns & Roberts, 2001). Most noteworthy among these are:

• Test Code and Digital Patterns: Test code and digital patterns are the predominant elements of the mixed-signal test program. The former is used to monitor pertinent elements such as the order and timing of instrument settings, signal generation and signal measurements that compose each measurement in the test program. Digital patterns are made up of a sequence of vectors, where a vector defines the required input to the DUT and the expected output value. In mixed signal testing, these patterns must be executed at a precise frequency.

• Simulation Code: Simulation code allows the simulation of the design model with respect to the instructions in the test program. The obtained design responses are compared to the expected test limits.

• Debuggability: Test program debugging process is used to locate hardware problems such as bad Device Interface Board (DIB) layout and broken tester modules. Likewise, it determines measurement correlation errors as well as intermittent bugs. The most prominent test techniques and measurements that make up a mixed signal test program are: continuity test, leakage test, supply current tests and DC measurement.

Mixed-Signal Tester Overview

A typical mixed-signal tester architecture It consists of multiple subsystems such as waveform digitizers, arbitrary waveform generators (AWG) and digital pattern generators. Most mixed-signal testers have several common building blocks, namely DC sources, digital subsystem, AC source and measurement, time measurement system and computing hardware (Burns & Roberts, 2001). In what follows, we provide a brief description of these building blocks.

• DC sources: The DC subsystem is one of the elements that make up most of mixedsignal testers. It consists of several components including multimeters, voltage current sources, precision voltage references, calibration sources and relay matrices. Each of these components serves a different purpose such as providing fast measurements with high-accuracy, or producing the DC voltages or currents that are required to power up the DUT and stimulate its DC inputs. Another important purpose is to maintain flexible interconnections between the tester instruments and the DUT.

• Digital subsytem: Another common feature that make up the bulk of most mixed signal testers is the digital subsystem. It is mainly used to compare the outputs of the DUT with the expected results in order to ensure that the device has been correctly manufactured.

• AC source and measurement: An efficient way to test AC performance is to use DSPBased testing (Burns & Roberts, 2001). The approach involves a stimulus/measurement pair namely the AWG and the waveform digitizer. An AWG consists of a bank of waveform memory, a DAC that converts the waveform data into stepped analog voltages and a programmable low pass filter that smoothes the stepped signal into continuous waveform. It is used to convert digital samples from a waveform memory into continuous time waveforms. Conversely, a waveform digitizer converts continuous-time waveforms into digitized representations. Both the AWG and the waveform digitizer operate from clock sources that are synchronized to each other and to the digital pattern’s frame loop repetition rate. Such synchronization of sample rates between the AWG, digitizers and digital pattern generators is another distinguishing characteristic of mixed signal tester.

• Time measurement system: The time measurement system (TMS) is used to measure various parameters such as frequency, rise and fall times, jitter and propagation delay. Accurate timing measurements necessitate a high-quality signal path between the DUT output and the TMS instruments.

• Computing hardware: A typical mixed-signal tester includes several computers and signal processors serving different purposes such as editing and compiling a test program or performing mathematical operations on the data collected during each test.

Test Techniques for Two-dimensional AMS Integrated Circuits

Several test strategies have been proposed in theory and in practice for testing AMS circuits. In what follows, we present the most common analog test methods and their measurement setups. We also provide a brief description of how analog tests can be performed using digital sampling techniques.

The simplest analog measurement setup is composed of a signal generator to stimulate the DUT and an instrument for output readings on the DUT parameters. The signal generator can produce waves of any shape (e.g., sinusoid, square-wave, etc.) that fit the purpose of the test. Input signals are selected based on the type of measurement to be carried out. The following are the four main measurement categories (Roberts, 1996):

1. DC measurements: used for measuring the static behavior of the design like leakage currents, output resistance, transfer characteristics and offsets.
2. AC measurements: used for measuring the small and large-signal frequency response behavior of the design. Distortion measurements also pertain to this test.
3. Transient or time-domain measurements: used for measuring the behavior of designs prone to signal shapes when used in their intended application.
4. Noise measurements: used for measuring the variations in the signal that usually show up at the design’s output when its input is set to zero.

Table des matières

INTRODUCTION
CHAPTER 1 PRELIMINARIES
1.1 Specifications and Test-related Constraints
1.2 Examples Showing the Impact of Specification and Test-related Constraints on AMS Designs performance
1.3 Analog Behavioral Modeling using PSpice
1.4 Fundamentals of Mixed Signal Testing
1.4.1 Test Specification Process
1.4.2 Mixed-Signal Tester Overview
1.4.2.1 Test Techniques for Two-dimensional AMS Integrated Circuits
1.4.2.2 Test Techniques for Three-dimensional Integrated Circuits
CHAPTER 2 LITERATURE OVERVIEW
2.1 Introduction
2.2 System Verification Techniques for Two-dimensional AMS Integrated Circuits
2.2.1 Simulation-Based Techniques
2.2.2 Formal Verification Techniques
2.2.3 Relevant Work
2.3 System Verification Techniques for Three-dimensional Integrated Circuits
2.3.1 Relevant Work
2.4 Summary
CHAPTER 3 PROPOSED VERIFICATION METHODOLOGY
3.1 Introduction
3.2 Constraint-driven Verification Methodology
3.2.1 Monitoring specifications of AMS designs in the presence of specification and test-related constraints
3.2.1.1 Extraction Phase
3.2.1.2 Modeling Phase
3.2.1.3 Verification Phase
3.2.2 Monitoring specifications of AMS designs in the presence of specification and test related constraints with tester’s components characteristics:
3.2.2.1 Extraction Phase
3.2.2.2 Modeling Phase
3.2.2.3 Verification Phase
3.3 Summary
CHAPTER 4 CASE STUDIES
4.1 Commercial Frequency Synthesizer IC based on a Charge Pump PhaseLocked Loop
4.1.1 Identification of Relevant Specification and Test-related Constraints
4.1.2 CP PLL based frequency synthesizer Behavioral Model
4.1.3 Verification Results
4.2 3D Clock Tree
4.2.1 Identification of Relevant Specification and Test-related Constraints
4.2.2 3D Clock Tree Model
4.2.3 Verification Results
4.3 Sigma-Delta Modulator
4.3.1 Identification of Relevant Specification and Test-related Constraints and Tester Components Characteristics
4.3.2 Sigma-Delta Modulator Behavioral Model
4.3.3 Verification Results
4.4 Complexities and Costs of the Proposed Constraint-based Verification Approach
4.5 Summary
CONCLUSION

Cours gratuitTélécharger le document complet

Télécharger aussi :

Laisser un commentaire

Votre adresse e-mail ne sera pas publiée. Les champs obligatoires sont indiqués avec *