Active-pixel-sensor (APS) circuits and perhaps other dense complementary metal oxide/semiconductor (CMOS) very-large-scale integrated (VLSI) circuits would be powered by infrared beams transmitted by laser diodes and received by photodetectors, according to a proposal. Clock signals for synchronizing the operations of such a circuit would be transmitted as modulation on the infrared power-supply beam. Command data signals could be received via other, low-power infrared beams. Digital APS output signals would likewise be sent to external circuits via modulation of infrared beams transmitted by low-power laser diodes incorporated into the VLSI APS chips.
The power-supply part of the proposal has been made feasible by advances that have reduced the power demands of CMOS VLSI circuits. The power demand of a typical CMOS VLSI APS chip is now low enough that a single, sufficiently illuminated infrared photodetector could serve as the source of a galvanically isolated power supply on the chip. With a sufficiently high duty factor, the clock modulation on the infrared power-supply beam should exert little effect on power-coupling efficiency.
The data-communication part of the proposal has been made feasible by the evolution of sensitive infrared detectors and low-power, frequency-tunable laser diodes. The infrared beams for input and output of data would have wavelengths different from that of the power-input beam. By use of tuned laser diodes in the transmitters and narrow-band dielectric filters in the receivers, it would be possible to communicate simultaneously over multiple infrared bands; thus, it would be possible to use a wavelength-multiplexing scheme to achieve a high data rate.
Multiple CMOS VLSI APS chips could be operated under common control and readout by use of a combination of wavelength and time multiplexing. The multiplexing scheme could be simplified, at the cost of some increase in structural complexity, by using a dedicated optical fiber for data communication between each APS and the common readout and control circuitry.
As APS and other VLSI circuits become denser and more complex, design problems pertaining to reliability of, and power dissipation in, electrical interconnections, become increasingly difficult. The problems are further intensified in cases in which VLSI circuits are required to be connected together in many-to-one networks. In general, the complexity of, and power dissipation in, electrical interconnections increase approximately exponentially with the number of nodes, while reliability decreases approximately exponentially with the number of nodes. The use of all-optical input and output connections according to the proposal described above could reduce overall complexity and increase reliability. In particular, if full-duplex communication with frequency multiplexing of data signals were used, then the complexity of a network with all optical interconnections would increase only linearly with the number of nodes.
This work was done by Frank Hartley and Bedabrata Pain of Caltech for NASA's Jet Propulsion Laboratory. NPO-20438
This Brief includes a Technical Support Package (TSP).

Optical power supply and data communication for APS circuits
(reference NPO-20438) is currently available for download from the TSP library.
Don't have an account?
Overview
The document discusses a proposal from NASA's Jet Propulsion Laboratory (JPL) that explores the use of optical power supply and data communication for active-pixel-sensor (APS) circuits, leveraging advancements in infrared (IR) technology. The primary innovation is the elimination of electrical connections between APS circuits and external systems, which is made feasible by the low power requirements of contemporary complementary metal oxide semiconductor (CMOS) very-large-scale integrated (VLSI) circuits.
The proposal outlines how infrared beams, transmitted by low-power, frequency-tunable laser diodes, can serve as both a power source and a communication medium for APS chips. A single, well-illuminated IR photodetector can provide a galvanically isolated power supply for the APS, while the modulation of the IR power beam can also transmit clock signals for synchronization. Command data signals can be sent via separate low-power IR beams, allowing for digital output signals to be communicated back to external circuits.
The document highlights the advantages of using optical communication over traditional electrical interconnections, particularly in terms of reliability and complexity. As the number of interconnected nodes increases, the complexity and power dissipation of electrical connections grow exponentially, while reliability decreases. In contrast, optical communication systems can maintain linear complexity and improve reliability, making them suitable for large networks of APS devices.
The proposal also discusses the potential for wavelength and time multiplexing to achieve high data rates, enabling multiple APS chips to operate under common control and readout. This multiplexing can be further simplified by using dedicated optical fibers for data communication, although this may increase structural complexity.
Overall, the document emphasizes the transformative potential of optical powering and communication in enhancing the performance of APS circuits, particularly for applications requiring ultra-definition imaging and high-speed data transfer. The integration of these technologies could lead to smaller, lighter, and more reliable systems, addressing the challenges posed by dense interconnects in modern electronic designs. The work is credited to Frank Hartley and Bedabrata Pain of Caltech, conducted for NASA's JPL, and represents a significant step forward in the field of optical communication and power supply for advanced electronic systems.

