INNOVATORS IN ANALOG AND MIXED SIGNAL TESTING
Aiming high: ECpE Professors Randall L. Geiger and Degang Chen recently discovered a way to solve one of the most challenging problems in analog and mixed signal testing, even though critics and industry professionals told them their goal—to accurately test high-precision devices with imprecise instruments, and ultimately reduce test costs—was impossible. “Industry professionals told us our idea and initial results were too good to be true,” Chen says.
Achieving the impossible: Geiger and Chen are leading the search for new ways to integrate more analog and mixed signal and RF functions onto a signal chip, and provide consumers with high-performance, low-cost products. The researchers made breakthroughs in three key areas that will enable on-chip testing. They began by greatly reducing test-signal linearity requirements for analog to digital converter (ADC) testing. “Researchers have been looking at ADC testing for decades, so we were told it was impossible to reduce test-signal linearity requirements to the level we did,” Chen says.
Continuing to innovate: Next, the researchers created an algorithm for reducing data acquisition time. “Time equals money and the time a circuit has to stay on a test is determined by the data acquisition time,” Chen says. “We introduced a method that worked four to eight times faster and achieved the same accuracy.”
Additionally, Geiger and Chen introduced a method for digital to analog converter (DAC) testing. This method involves using imprecise measurement devices. Geiger says conventional wisdom required using measurement devices that performed better than the device you’re testing, but now engineers can conduct tests with a device that is up to 100 times poorer in linearity metrics.
Raising the bar: This technology could improve functionality and lower prices to consumers of everyday electronics such as PCs, cell phones, and MP3 players. Geiger and Chen’s advancements also will urge engineers to find new methods of designing high-performance circuits. The researchers have earned a U.S. patent for their work and have a second patent pending.