MEMS-referenced oscillators

MEMS-based oscillators are important for sensing and timing applications, such as space-constrained mobile devices, and automatic wireless sensor networks (WSNs). This is due to their small form factors (system or node size ∼mm2) and low power consumption (driven by the availability of ultra-high-Q resonators), low cost, programmability, and wide operating temperature range.

This project is a collaboration with Prof. Philip Feng at CWRU. Its goal is to analyze, design, simulate, and experimentally study the behavior of multi-frequency oscillators based on multimode two-dimensional (2D) MEMS/NEMS resonators and programmable CMOS feedback amplifiers. The initial version of the latter contains a differential-difference low-noise amplifier (DD-LNA), Gm-C band-pass filter (BPF), all-pass filter (APF), programmable gain amplifiers (PGAs), and automatic level control (ALC) to control the oscillation amplitude. In addition, the chip contains a compensation circuit to cancel unwanted electrical feedthrough current from the high-Q MEMS device (in this case, a SiC-based comb drive resonator).
Funding: NSF ECCS

Figure: Block diagram of the feedback amplifier (0.5 μm CMOS)

Figure: Simulated phase noise versus open-loop phase shift.

Biologically-inspired spectrum analysis

The biological inner ear, or cochlea, is an amazing sensor that performs auditory frequency analysis over an ultra-broadband frequency range of ~20 Hz to 20 kHz with exquisite sensitivity and high energy efficiency. Electronic cochlear models, which mimic the exponentially-tapered structure of the biological inner ear using transmission lines or filter cascades, have been shown to be fast and extremely efficient spectrum analyzers at both audio and radio frequencies (RF). This project is a collaboration with Prof. Michael Lewicki at CWRU. It aims to develop improved output encoding methods for such cochlea-like analyzers.

We have developed neuron-like asynchronous event-generation circuits to efficiently encode cochlear outputs, including ring-oscillator-based injection-locked frequency dividers (ILFDs) that accurately encode input frequencies and phase-sensitive detectors that encode both amplitude and phase information and thereby improve frequency resolution without reducing temporal resolution. These frequencies are then quantized off-chip by counting edges in the VCO outputs during a sampling period, thus producing multi-bit digital outputs and realizing a set of parallel power-efficient time-domain ADCs.
Funding: NSF CCF

Figure: Die photograph of an RF cochlea chip (65 nm CMOS).

Figure: Phase noise reduction due to injection locking.

Ultra-low-power RF transceivers

In this project we are  developing ultra low power wake-up transceivers for wireless sensor nodes and the internet of things (IoT). These broadband multi-standard receivers will be capable of demodulating on-off-keying and frequency chirped signals. A combination of matched filtering and adaptive notch filtering will be used to cancel both in-band and out-of-band interference. The integrated transmitters are capable of working in unused ISM (Industrial-Scientific-Medical) band from 50 to 2400 MHz.  Moreover, as power consumption is one of the main concerns in IoT applications, we propose electromagnetic  energy harvesting circuits to provide the voltage supply to the system.
Figure: Block diagram of the proposed ultra-low-power RF  transceiver (65 nm CMOS).

Portable bimodal imaging systems

We are pursuing fundamental scientific innovations in sensors, circuits, and computing to address the limitations of portable multimodal imaging, which is of interest for basic materials science and biomedical research and also for various applications.  We propose bimodal imagers for affordable and automated point-of-care (POC) screening and early-detection of skin cancers, bacterial infections, pressure ulcers, and other diseases. These devices are expected to have a significant positive impact on public health by increasing survival rates and reducing treatment costs. In particular, we are integrating low-field magnetic resonance (MR) and ultrasound (US) imaging within portable form factors to take advantage of their complementary contrast mechanisms and information content. Key goals include 1) the development of miniature low-field MR and US sensors to enable portable and potentially wearable devices; and 2) the integration of these sensors with electronics and signal processing within a single device to enable real-time control and optimization of both modalities.
The first figure shows the current MRI sensor and the gradient driver circuit needed to collect images.  The second figure shows some data collected with a commercial ultrasound system and the same MRI sensor.  The data was collected on an NMR tube with an inner structure of 4 smaller tubes.  The gaps in between the inner tubes were filled with a PBS solution (salt water), which is clearly seen in the MRI depth profile (1D image) but invisible to ultrasound.

Figure: Gradient driver circuit (left) and MR sensor (right).

Figure: Images collected from the same phantom with a commercial ultrasound unit and our custom MR sensor.

Ultra-broadband magnetic resonance (MR) front-ends

The front-end electronics in magnetic resonance (MR) systems commonly uses resonant circuits for efficient RF transmission and low-noise reception. These circuits are narrow-band analog devices that are inflexible for broadband and multi-frequency operation at low operating frequencies. We are addressing this issue by developing an ultra-broadband MR probe that operates in the 0.1-3 MHz frequency range without using conventional resonant circuits for either transmission or reception. This “non-resonant” approach significantly simplifies the probe circuit and allows robust operation without probe tuning while retaining efficient power transmission and low-noise reception. We will demonstrate the utility of the technique through a variety of nuclear magnetic resonance (NMR) and nuclear quadrupole resonance (NQR) experiments in this frequency range.

Funding: Schlumberger

Figure: Photograph of a custom ultra-broadband MR front-end  board used for low-field MR experiments.

Authentication of medicines and food products

Access to safe medicines and food is a serious concern around the world. The goal of this project is to develop a low-cost, portable device designed to authenticate medicines and nutritional supplements. It is a collaboration with Prof. Swarup Bhunia at the University of Florida.

The proposed device is based on Nuclear Quadrupole Resonance (NQR), a non-invasive and non-destructive analytical technique suitable for characterizing a variety of solid materials. NQR is also known as zero-field magnetic resonance. It is mediated by the interaction of electric field gradients within atoms with the quadrupole moment of nuclear charge distributions. Almost every medicine has active ingredients with NQR-active nuclei. The ultimate goal of the project is to use NQR measurements to create a chemometric passport for quality assurance, where the information of packaged medicines can be derived from spectroscopic analysis. The contents then will be authenticated by matching to reference spectra stored in a secure database.

Funding: NSF CCF

Figure: The proposed chemometric passport approach.

Figure: Measured NQR signal versus number of Tylenol caplets.

Active structural health monitoring (SHM)

Applications of structural health monitoring (SHM) in the aerospace industry are currently of great interest due to its ability to ensure structural safety and reduce high maintenance costs. This project is a collaboration with Virtual EM, Inc. It proposes to develop a highly expandable, lightweight and flexible sensor network for SHM of aerospace structures by directly addressing the roadblocks of high size, weight, and power requirements and high communication bandwidth. It is based on a custom front-end IC developed by CWRU using commercial 0.5 μm CMOS technology that contains both a programmable transmitter (waveform generator) and a receiver suitable for a range of active SHM applications using arrays of piezoelectric transducers.

Funding: Air Force

Figure: Block diagram of the proposed integrated SHM transceiver (0.5 μm CMOS).

Power-efficient biopotential amplifiers

An integrated low-power analog front-end (AFE) for simultaneously detecting electrocardiogram (ECG) and respiration rate (RR) has been  designed and tested. This chip could be used in wearable biopotential measurement systems for wireless monitoring of heart rate, heart rate variability, ventilation rate, and ventilation pattern variability. We use a three-lead ECG and impedance pneumography (IP) for this purpose. Both signals are measured between standard RA (right arm) and LA (left arm) chest electrodes, with a driven ground reference electrode (REF) for rejecting common-mode interference.
Figure: Die photograph of the ECG/RR AFE (0.5 μm CMOS).

Figure: ECG measured from the upper arm using the low-power AFE.

Power-efficient wireless telemetry

The photograph shows the test boards used to test a low-power transcutaneous wireless telemetry link for biomedical implants. This broadband bidirectional link was an important part of a larger project to develop implanted neural prosthesis systems (brain-machine interfaces). It demonstrated excellent power-efficiency (measured to be < 1pJ/bit), and was later used for successful wireless recording of action potentials from the brain of an awake, behaving primate (rhesus macaque).

Low-Power CMOS Rectifiers for RFID

During this project we investigated efficient methods for extracting DC power from electromagnetic radiation, which is important for a number of applications, such as Radio Frequency Identification (RFID) tags and bionic implants. Theoretical bounds on system performance were first derived, and then circuit, antenna and impedance matching network designs were designed that efficiently approached these bounds. An experimental power extraction system was then built to collect RF energy (around 900 MHz) at low electromagnetic field strengths, thus enabling a substantial increase in the operating range of remotely-powered devices.