Automotive test solutions risk overlooking the majority of field failures in automotive ICs.
The amount of electronic content in passenger cars continues to grow rapidly, driven mainly by the integration of various advanced safety features. The industry’s move towards fully autonomous vehicles promises to even further increase the number of these safety features and consequentially, the electronic content required in each vehicle.
Recent reports indicate that hundreds of semiconductor devices are now being integrated into higher-end vehicles. What’s more, these components are becoming increasingly complex. The frontier for sophisticated semiconductors is in fact being led by the chips needed to execute artificial intelligence algorithms that govern emerging self-driving capabilities. It is of course critical that these safety-related devices adhere to the highest possible quality and reliability requirements. These requirements are formalized in the ISO 26262 standard that is being rapidly adopted by automotive manufacturers and suppliers worldwide. The standard is comprehensive and covers all aspects of the hardware and software lifecycle from design through testing and in-field operation.
There has been a lot of material published related to automotive semiconductor test solutions. The bulk, if not all, of this material has exclusively addressed the test and diagnosis of digital logic and memories. However, it turns out that the majority of field failures in automotive ICs now occur within the mixed-signal portion of the chip (see figure 1). This is not so surprising when one considers the great effort and cost applied to digital testing. The successful elimination of most digital defects means that any remaining defects will likely be mixed-signal in nature. These defects, albeit often small in number, cannot be tolerated in safety-critical automotive applications. Solutions are therefore needed to address this testing shortfall.
Figure 1: Source of electronic breakdown in automotive ICs.
Analog designers and test engineers unfortunately do not have DFT tools comparable to those used by their digital counterparts. A basic prerequisite to automating the generation of analog tests is an automated means of measuring the fault coverage achieved by any test. Although fault simulation for digital circuitry has been commercially available for almost 30 years, analog fault simulation has only been discussed in academic papers and, recently, in a few industrial papers that describe proprietary software. Fortunately, commercial fault simulators for analog circuits are beginning to appear.
The basic approach is to measure the coverage of opens and shorts and related parametric variations within a transistor-level netlist. Coverage of a given defect is determined by evaluating a change in the circuit response in the presence of the defect through analog simulation. What makes this approach practical now is the use of a number of significant speed-up techniques to reduce the analog simulation time by several orders of magnitude over simulating each defect one at a time on the flat netlist. These techniques include such things as likelihood-weighted random sampling to ensure most time is spent simulating the most likely defects and mixed-model simulation where the highest-level model or netlist is used for each sub-circuit instance that does not contain the defect being simulated.
This new automation opens up a new world of analog test related capabilities. First, existing analog tests can be evaluated for their effectiveness. Ineffective tests can be eliminated to reduce test time and cost. The new defect coverage metric can also be used to guide the generation of new efficient tests. This metric also allows measuring a circuit’s tolerance to defects. Defect tolerance is a measure of circuit’s ability to continue to work within acceptable operational parameters in the presence of certain defects. This metric is very important in automotive applications as it directly relates to long-term reliability. Both defect coverage and defect tolerance metrics are becoming hard requirements in the delivery of automotive semiconductor parts.
The expanding automotive semiconductor market continues to drive the need for new test solutions. Both the digital and analog portions of the chip must be fully addressed to meet chip-wide quality and reliability requirements. New analog fault simulation solutions are the first step in the development of a comprehensive set of analog test solutions.
Further reading: For more on new approaches to transistor-level fault simulation for analog, mixed-signal, and non-scan digital circuits, download the Mentor whitepaper Analog Fault Simulation Challenges and Solutions.