UST 031

Renesas and Hitachi have developed a rugged analogue-to-digital converter (ADC) for autonomous systems such as driverless cars (writes Nick Flaherty). Engineers from the two companies developed a technique for continuous-time digital calibration of a delta-sigma ( ΔΣ ) modulator in the ADC. The ΔΣ approach uses oversampling and a finite impulse response (FIR) filter to reduce the silicon area, and so leads to less expensive ADC chips. However, it struggles with accuracy and reliability in harsh environments such as automotive electronic control units. With many more sensors required in autonomous systems, more high-speed ADCs are needed to convert data from Lidar and millimetre-wave radar. The new technique provides more accuracy by using a least mean square (LMS) algorithm to measure and calibrate the transfer function of a continuous-time ΔΣ modulator. To do that, the team also developed the world’s first multi-rate LMS search algorithm. This reduces the operating frequency of the process of searching for the best coefficients for the FIR digital filter, cutting power consumption as a result. The new circuit reduces the operating frequency of the chip by 75%. A test chip built on a 28 nm process achieved a signal bandwidth of 15 MHz and dynamic range of 74.3 dB when operating at an oversampling frequency of 480 MHz. By reducing the digital calibration circuit operating frequency to 120 MHz, the chip consumed 37 mW – 19 mW for the analogue section and 18 mW for the digital block. A vital factor is that the chip is stable over a wide temperature range, of -20 to +125 C, with the signal-to-noise ratio varying by no more than ±1 dB. This is achieved by digitally calibrating the transfer function, which enables the digital circuits in the background to be calibrated even if the analogue blocks are affected by environmental variations such as temperature changes. Renesas says it plans to use the technology in its next generation of chips for automotive sensors for self-driving technology. The new design of analogue-to-digital converter allows sensor data to be captured in harsh environments Rugged ADC aimed at harsh conditions Driverless cars