Day 2: High Performance Sensor Interfaces
"Sensor Applications in Advanced Lithography and Metrology"
The minimum feature size or Critical Dimension (CD) achievable by optical lithography continues to decrease. Today, devices with 5 nm feature sizes are possible. High production yields require lithography tools that maintain good CD uniformity (CDU) and overlay (OV) in the presence of unavoidable process variations. This is achieved by combining precision metrology to monitor OV and CD with advanced control strategies to minimize CDU and OV. Lithography and metrology tools rely heavily on a wide variety of sensors to measure many different parameters: temperature, pressure, distance, light level and intensity distribution (images). In this invited talk a concise overview will be given of some of the main sensors applications in lithography and metrology and the main sensor developments that have helped the semiconductor device industry to maintain its aggressive shrink roadmap.
"Tailor-Made Precision Position Sensors"
Eddy current sensor systems consist of a sensing coil located near a conducting target. A power-efficient way to excite the coil is to combine it with a fixed capacitor to form a resonant circuit. The current required to drive this circuit is then a function of the distance between the coil and the target. The inherently non-linear sensor characteristic can be linearized in a post processing step. Over time, there has been a shift away from the use of complex analog signal conditioning circuits towards the combination of smart analog front ends with advanced digital signal processing. This presentation will discuss the design of a modular electronics platform concept that can be scaled in terms of size, accuracy, resolution, power consumption and other specifications to meet the needs of particular customers and applications. As an example of customer adaption we will present the development of a precision position sensing system developed for the European Extra Large Telescope. This system achieves a resolution of 1 nm over a range of +/- 200um.
"A Stray-field-Immune Magnetic Displacement Sensor with 1% Accuracy"
We present a novel magnetic angle sensor for the accurate and robust measurement of small angular displacements. Implemented in CMOS, the sensor is based on a novel gradient measurement concept made possible by combining Hall sensors with integrated magnetic concentrators. In typical applications, the peak output voltage of the Hall sensors will only be 1.5 mV at the maximum operating temperature (160°C), and thus requires high-performance low-offset readout electronics. Over its 14mm linear displacement range, the sensor’s total error is less than 1% including manufacturing tolerances, trimming accuracy, temperature and ageing effects, which meets automotive requirements. The realization of the sensor involved several novel design considerations, e.g. related to the on-chip signal processing and to the sensor’s calibration, which will be discussed during the presentation.
"High-Resolution Temperature Sensors for State-of-the-Art MEMS Frequency Reference"
In temperature-compensated MEMS frequency references, stringent requirements on phase-noise and jitter, require the use of high-resolution temperature sensors. Such sensors must also be fast enough to maintain frequency stability during fast thermal transients. In this paper, two different temperature sensor designs will be presented. For battery-powered systems, a BJT-based sensor is presented which achieves sub-mK resolution at a conversion rate of 2kS/s. For more demanding telecommunications and networking applications, a sensor based on the use of dual-MEMS oscillators is presented. At a conversion rate of 200S/s, it achieves an extremely high resolution of 20uK, which is essential to meet the stringent phase noise and integrated phase jitter (IPJ) of such applications. Both sensors achieve state-of-the-art energy efficiency.
"101 ways to readout a photodiode"
The design of a pixel of an image sensor involves several contradictory trade-offs: between silicon area and complexity, noise and uniformity, quantum efficiency and fill factor, dynamic range and signal to noise ratio, speed, conversion gain, response speed and power dissipation, amongst others. In consumer imaging, pixel pitch is a dominant specification. However, in higher-end applications, there is often a need and opportunity to do “more” with and in the pixel. We will present several cases where more complex in-pixel electronics enables higher performance compared to “plain vanilla” types of pixels, or even completely new applications. Examples of these are: pixels that can handle wide dynamic range scenes, that can operate in “IWR” global shutter. Pixels that can measure distance by “time of flight” or “LIDAR” pixels. Pixels with sense amplifiers that maintain a constant bias across the photodetector. Pixels that separate AC and DC optical information, pixels that demodulate the optical information. Pixels may even record short movies in local memory. Pixels that count photons or particles.
"Smart Ultrasound Probes: Going Digital in the Probe Tip"
While medical ultrasound imaging is currently predominantly done using hand-held probes connected to relatively bulky imaging machines, various new application areas are emerging that call for advanced miniaturized imaging devices. Examples include catheters capable of providing real-time 3D images to guide minimally-invasive interventions, and wearable devices for new monitoring and diagnostic applications. In contrast with conventional probes, which tend to contain little or no electronics, these new devices need to become “smart”: electronics need to be integrated into the probe to interface with the transducer elements. This is especially critical for 3D imaging devices, which readily contain more than 1000 elements. Each of these elements receives echo signals that need to be communicated via a limited number of cables, or even wirelessly. In-probe integrated circuits are needed to locally reduce the number of channels. Several analog approaches have been reported to do so, including multiplexing and sub-array beamforming. These approaches still rely on sensitive analog connections to the imaging system, where the received echo signals are digitized and further processed. This talk explores the possibility of in-probe digitization, to improve signal quality and leverage the advantages of in-probe digital signal processing and high-speed digital data links. Due to the stringent size and power-consumption constraints, this requires state-of-the-art application-specific ADCs. Several design examples will be presented, including a beamforming SAR ADC and an element-matched delta-sigma ADC.