Manuel Monge

Design and implementation of a fully integrated compressed-sensing signal acquisition system

Compressed sensing (CS) is a topic of tremendous interest because it provides theoretical guarantees and computationally tractable algorithms to fully recover signals sampled at a rate close to its information content. This paper presents the design of the first physically realized fully-integrated CS based Analog-to-Information (A2I) pre-processor known as the Random-Modulation Pre-Integrator (RMPI) [1]. The RMPI achieves 2GHz bandwidth while digitizing samples at a rate 12.5× lower than the Nyquist rate. The success of this implementation is due to a coherent theory/algorithm/hardware co-design approach. This paper addresses key aspects of the design, presents simulation and hardware measurements, and discusses limiting factors in performance.

A 100MHz–2GHz 12.5x sub-Nyquist rate receiver in 90nm CMOS

A fully-integrated, high-speed, wideband receiver called the random modulation pre-integrator is realized in IBM 90nm digital CMOS. It achieves an effective instantaneous bandwidth of 2GHz, with >;54dB dynamic range. Most notably, the aggregate digitization rate is fs =320MSPS, 12.5× below the Nyquist rate. Signal recovery can be accomplished for any signal with a concise representation. The system is validated using radar-pulses and tones as the input and recovering the time-domain waveforms.

A Fully Intraocular High-Density Self-Calibrating Epiretinal Prosthesis

This paper presents a fully intraocular self-calibrating epiretinal prosthesis with 512 independent channels in 65 nm CMOS. A novel digital calibration technique matches the biphasic currents of each channel independently while the calibration circuitry is shared among every 4 channels. Dual-band telemetry for power and data with on-chip rectifier and clock recovery reduces the number of off-chip components. The rectifier utilizes unidirectional switches to prevent reverse conduction loss in the power transistors and achieves an efficiency > 80%. The data telemetry implements a phase-shift keying (PSK) modulation scheme and supports data rates up to 20 Mb/s. The system occupies an area of 4.5 ×3.1 mm2. It features a pixel size of 0.0169 mm2 and arbitrary waveform generation per channel. In vitro measurements performed on a Pt/Ir concentric bipolar electrode in phosphate buffered saline (PBS) are presented. A statistical measurement over 40 channels from 5 different chips shows a current mismatch with μ = 1.12 μA and σ = 0.53 μA. The chip is integrated with flexible MEMS origami coils and parylene substrate to provide a fully intraocular implant.

Packaging study for a 512-channel intraocular epiretinal implant

Much effort has been put into developing multi-channel retinal prosthetic devices. Currently, even the most advanced prostheses do not have enough channels to provide vision to a desirable level. In this paper, we present a system design and a packaging scheme for a 512-channel intraocular epiretinal implant. Both a wireless power coil (with high transfer efficiency) and a data coil are included for this intraocular system. Simulation of the interference between coils is investigated and the results show that the two coils can be put in a co-planar fashion using two notch filters to minimize interference. The complete package is demonstrated with a mechanical model with a parylene-C flexible circuit board, i.e., parylene flex, to show the placement of the IC chips, discrete components, and coils. It also shows the final folded device after surgical insertion into an eye to save space. The feasibility of the proposed structure has been successfully tested in vivo. Experimentally, the maximum allowable pulling force is measured by a dynamic mechanical analysis (DMA) machine to be 8N, which provides a large safety margin for surgery.

A fully intraocular 0.0169mm 2 /pixel 512-channel self-calibrating epiretinal prosthesis in 65nm CMOS

Since their conception and success in human trials, the flexibility and spatial resolution of retinal prostheses have been of major interest. Clinical studies have revealed that hundreds of channels are needed to restore functional visual perception, and more sophisticated waveforms present advantages over biphasic pulses. Initial designs targeted stimulation current levels up to 1mA to ensure functionality. For such designs, an output compliance of >10V was required, and HV technologies were used at the expense of area and power consumption. Human clinical trials have recently shown that implanted electrodes present a stimulus threshold as low as 50μA. In addition, advances in implant technology promise close placement of electrode array and retinal tissue, which can further decrease the required current. Thus, highly scaled LV technologies can provide alternative means to reduce area and power, and to support hundreds of flexible independent channels for fully intraocular implants. In this paper, a self-calibrating 512-channel epiretinal prosthesis in 65nm CMOS is presented. It features dual-band telemetry for power and data, clock recovery, a 2-step calibration technique to match biphasic stimulation currents, and an independent arbitrary output waveform per channel. The implant integrates coils (power and data), IC, external capacitors and electrode array using a biocompatible parylene substrate, providing a fully intraocular solution.

A Modelling and Nonlinear Equalization Technique for a 20 Gb/s 0.77 pJ/b VCSEL Transmitter in 32 nm SOI CMOS

This paper describes an ultralow-power VCSEL transmitter in 32 nm SOI CMOS. To increase its power efficiency, the VCSEL is driven at a low bias current. Driving the VCSEL in this condition increases its inherent nonlinearity. Conventional pre-emphasis techniques cannot compensate for this effect because they have a linear response. To overcome this limitation, a nonlinear equalization scheme is proposed. A dynamic VCSEL modelling technique is used to generate the time-domain optical responses for “one” and “zero” bits. Based on the asymmetry of the two responses, the rising and falling edges are equalized separately. Additionally, instead of using fixed bit delays, the equalization delay is selected based on the bias current of the VCSEL. The efficiency of the proposed modelling and equalization technique is evaluated through simulations and measurements. The transmitter achieves energy efficiency of 0.77 pJ/b at 20 Gb/s and occupies 100 μm × 60 μm active silicon area.

Design considerations for high-density fully intraocular epiretinal prostheses

Retinal prostheses have successfully proven to be a viable treatment for advanced stages of retinal degenerative diseases such as retinitis pigmetosa. However, current implementations have critical limitations that affect their functionality and resolution. This paper reviews design challenges of the electronics considering the biology of the eye and discusses new approaches for future high-density fully intraocular prostheses. An origami retinal implant that has the potential to alleviate the size, power and cost constraints of such systems is proposed. Measured results of enabling technologies are also discussed.

A 20Gb/s 0.77pJ/b VCSEL transmitter with nonlinear equalization in 32nm SOI CMOS

This paper describes an ultra-low-power VCSEL transmitter in 32nm SOI CMOS. To increase its power efficiency, the VCSEL is driven at a low bias current. The resulting nonlinearity and loss in bandwidth is modelled and compensated by a nonlinear equalization technique. The time domain optical responses for “one” and “zero” bits are used to find the optimum equalization technique. The rising and falling edges were equalized separately and the equalization delay is selected based on the bias current of the VCSEL. The transmitter achieves energy efficiency of 0.77pJ/b at 20Gb/s.

Simulation and measurement of transcranial near infrared light penetration

We are studying the transmission of LED array-emitted near-infrared (NIR) light through human tissues. Herein, we simulated and measured transcranial NIR penetration in highly scattering human head tissues. Using finite element analysis, we simulated photon diffusion in a multilayered 3D human head model that consists of scalp, skull, cerebral spinal fluid, gray matter and white matter. The optical properties of each layer, namely scattering and absorption coefficient, correspond to the 850 nm NIR light. The geometry of the model is minimally modified from the IEEE standard and the multiple LED emitters in an array were evenly distributed on the scalp. Our results show that photon distribution produced by the array exhibits little variation at similar brain depth, suggesting that due to strong scattering effects of the tissues, discrete spatial arrangements of LED emitters in an array has the potential to create a quasi-radially symmetrical illumination field. Measurements on cadaveric human head tissues excised from occipital, parietal, frontal and temporal regions show that illumination with an 850 nm LED emitter rendered a photon flux that closely follows simulation results. In addition, prolonged illumination of LED emitted NIR showed minimal thermal effects on the brain.

A 4μW, ADPLL-based implantable amperometric biosensor in 65nm CMOS

This paper presents a fully implantable, wirelessly powered subcutaneous amperometric biosensor. We propose a novel ultra-low power all-digital phase-locked loop (ADPLL) based potentiostat architecture for electrochemical sensing. The system is wirelessly powered by near-field RF coupling of an on-chip antenna to an external coil at 915 MHz. Bi-directional wireless telemetry supports data transmission from the sensor to the external reader (uplink) via backscattering, and reconfiguration of the sensor chip over the RF downlink. The 1.2×1.2 mm2 prototype is fabricated in TSMC 65nm CMOS process. The potentiostat achieves a 100pA sensitivity over a full scale current range of 0–350nA. The total power consumption of the system is 4μW.