Search

In this article, the quantum phase slip junction (QPSJ) based full adder, multiplexer, and demultiplexer memoryless circuits for superconducting digital applications are designed for the first time. Generally, the memoryless circuits are implemented by interconnection of the basic gates (i.e., AND, OR, NOT). So in this article, the structures of the QPSJ-based AND and OR gates are also modified in order to reduce the complexity of their connections to the other gates by removing or changing the location of the clock pulse from their output branches. When a gate is connected to the next one, the elements of the second gate are loaded to the circuit of the first gate. The output current of the first gate will be affected due to this loading and no Cooper pair may be produced at its output and the next connected gate will not be switched. Therefore, the main tip in designing of these circuits is that each constituent gate must provide the necessary output current to drive the next connected gates, since each gate is driven by one Cooper pair. In this article, providing sufficient output current for each gate has been observed by 1) selecting proper critical voltages of the QPSJs and 2) by adding extra elements to or removing some elements from the connected gates in the proposed memoryless circuits.

Offering extremely high energy-efficiency, the adiabatic quantum-flux-parametron (AQFP) logic family is considered a promising candidate as a future low-power technology for building high-performance computing systems. In this paper, we provide the design of M-word × N-bit dual-port register files as an essential component of an AQFP microprocessor. This design consists of one decoder for data writing, two multiplexers for read out, and an array of registers as storage elements. In order to reduce the design difficulties and improve the design robustness, we employed a top-down AQFP very-large-scale integration (VLSI) design flow to implement this register file. The proposed design flow is composed of high-level synthesis, a specialized AQFP interpreter, a genetic-algorithm-based routing approach, and hardware-description-language-based back-end verification. As a case study, we designed and fabricated an 8-to-1 4-bit multiplexer used as the component circuit of the designed register file. In low-speed test, we confirmed the correct operations across all channels of the multiplexer.

The successful realization and broad deployment of transition edge sensor (TES)-based detector systems has led to significant demand for time-division and code-division superconducting quantum interference device (SQUID) multiplexers time division multiplexing (TDM) and code division multiplexing (CDM) as essential components of the cryogenic readout chain. TDM and CDM circuits are produced by the Boulder Microfabrication Facility in large quantities and in multiple varieties to meet the needs of various bolometric and calorimetric applications. In most cases, the basic functionality of these devices must be verified before they are passed along to internal or external collaborators for integration into scientific instruments. We have developed a test bed that utilizes the NIST TDM/CDM read-out electronics to make automated multiplexed measurements on sixteen devices simultaneously in a 4 K liquid-helium dip probe. The optimization of the measurement process has resulted in vastly improved throughput and enhanced data products. We present a thorough analysis of results from the application of the test process to flux-activated-switch-style multiplexers. The utility of this approach in identifying fabrication trends, yield indicators, and common failure modes will be demonstrated.

Superconducting quantum interference device (SQUID)-based time-division multiplexing (TDM) is a mature and widely implemented technology used to read out transition-edge sensor arrays. As the number of pixels in modern arrays continues to increase, a higher multiplexing factor is required to reduce the number of wires and amplifier channels. However, as the multiplexing factor is increased, the number of row-select wires (used to turn on a row of TDM SQUIDs in a two-dimensional configuration) also increases, limiting the reduction in array wires. We present a more advanced TDM architecture that implements multi-level switching between subgroups of pixels. We show that this technique can dramatically reduce the number of required row-select lines. We also present the design, fabrication, and testing of a TDM multiplexer incorporating a two-level switch, which implements a second switch for each group of ten TDM pixels. In this implementation, a multiplexing factor of 100 can be addressed using ten group-select wiring pairs and ten row-select wiring pairs. We demonstrate multiplexer functionality and present measured operating margins of this new TDM multiplexer.

Accepted: 2019-03-13, 資料番号: SA1180388000

Single flux quantum (SFQ) technology has emerged as a promising beyond-CMOS technology with Josephson junctions (JJs) as active devices in a superconducting circuit. In this paper, we (i) describe the algorithms and the methodology we used to synthesize a synchronous SFQ circuit using a CMOS logic synthesis tool by adding new features to the tool and by modifying the existing features; (ii) introduce the concept of two kinds of complex cells: (a) stand-alone cells and (b) interconnected cells along with the advantage gained by these complex cells through the results of synthesized SFQ netlists. Design of stand-alone complex cells presented here includes and and or gates with more-than-two inputs, high-fanout splitters, and “A+BC” cell. Circuits of decoders, multiplexers, and carrylook-ahead adders are synthesized with complex cells, and the advantage in terms of JJ count and latency is presented. We have also investigated the possibility of SFQ cells supporting multiple fanout drive capability by modifying the input/output interfaces of SFQ standard cells. In this direction, an algorithm to modify the input interface of a cell so that it can be driven along with similar cells by a single splitter output is presented along with the results for a simple clock-distribution line with multiple-fanout capability.

Test point insertion and set/scan techniques for enhanced testability in superconductive single-flux-quantum (SFQ) logic are proposed here. Test point insertion reduces the overhead of a set/scan chain while maintaining most of the functionality. Multiple ways of replacing costly (in terms of the number of Josephson junctions) SFQ multiplexers with mergers and blocking gates are proposed. The multiplexer control signals are replaced with a gated clock signal or separate bias networks for both functional and test paths. Clocked blocking gates or current-controlled Josephson transmission line segments are used to disable undesired data inputs. The clocked blocking gates for test point insertion in a 64-bit register requires 35% fewer Josephson junctions as compared to multiplexers. This advantage further increases for current-controlled blocking gates. Set/scan chain and test point insertion techniques are applied to several SFQ circuits to evaluate error characteristics and to provide built-in self-test of SFQ-compatible memory systems.

We are developing room temperature electronics based upon the ROACH platform to readout microwave multiplexed X-ray TES. ROACH is an open-source hardware and software platform featuring a large Xilinx Field Programmable Gate Array (FPGA), Power PC processor, several 10 GB Ethernet SFP+ interfaces, and a collection of daughter boards for analog signal generation and acquisition. The combination of a ROACH board, ADC/DAC conversion daughter boards, and hardware for RF mixing allows for the generation and capture of multiple RF tones for reading out microwave multiplexed X-ray TES microcalorimeters. The FPGA is used to generate multiple tones in base band, from 10 MHz to 250 MHz, which are subsequently mixed to RF in the multiple GHz range and sent through the microwave multiplexer. The tones are generated in the FPGA by storing a large lookup table in Quad Data Rate SRAM modules and playing out the waveform to a DAC board. Once the signal has been modulated to RF, passed through the microwave multiplexer, and has been modulated back to base band, the signal is digitized by an ADC board. The tones are modulated to 0 Hz by using a FPGA circuit consisting of a polyphase filter bank, several Xilinx FFT blocks, Xilinx CORDIC blocks (for converting to magnitude and phase), and special phase accumulator circuit for mixing to exactly 0 Hz. Upwards of 256 channels can be simultaneously captured and written into a bank of 256 First-In-First-Out (FIFO) memories, with each FIFO corresponding to a channel. Individual channel data can be further processed in the FPGA before being streamed through a 10 GB Ethernet fiber-optic interface to a Linux system. The Linux system runs software written in Python and QT C++ for controlling the ROACH system, capturing data, and processing data.

In order to read out current signals from large-scale arrays of transition edge sensors (TES), microwave SQUID multiplexers (MW-MUX) have been developed by several groups. Reduction of crosstalk induced in the multiplexer is important in order to maintain the energy resolution or the noise equivalent power of detector arrays. In the present study, crosstalk between channels in MW-MUX with five different designs is experimentally evaluated. The resonance frequency separation is ten times as large as the resonator bandwidth. Without flux-ramp modulation, crosstalk between two adjacent channels neighboring in position decreases to ≍1 × 10–3 with increasing resonance frequency separation to ≍100 MHz. Crosstalk between SQUID channels neighboring in resonance frequency also decreases to 4 × 10–3 with increasing the distance to 2.5 mm. We show that nonlinear error can occur due to the crosstalk with flux-ramp modulation. Flux readings from a well-designed multiplexer exhibit crosstalk of less than 1 mΦ0 peak-to-peak and nonlinear errors of less than 2 mΦ0 peak-to-peak, which is sufficient for the readout of the gamma-ray TES arrays being developed. These results can provide guidelines for suppressing both the crosstalk and the nonlinear error to less than the criterion of TES spectrometers with typical energy-resolving ability.

As the digital data links for superconducting circuits advance and higher data throughput per channel becomes possible, timing margins shrink and data integrity becomes a major challenge. Particular interest for multichannel applications is establishing the high-quality data link to interface with subsequent electronics. In this paper, we focus on integration of an on-chip pseudorandom binary sequence (PRBS) generator into a superconducting analog-to-digital converter (ADC) design to facilitate link stability evaluation and automated interchannel synchronization. PRBS generator and the ADC use a common clock source. An on-chip deserializer/demux, which includes the output drivers, is driven by a set of data sources depending on switch selections on-chip. The outputs are connected to a field-programmable gate array (FPGA) at room temperature, which hosts the developed interface circuitry for data reception, data integrity evaluation, and the synchronization mechanism. The integrated circuit (IC) that combines ADC and PRBS7 generator circuit was designed for the HYPRES 4.5 kA/cm2 four-layer standard fabrication process and features four deserialized outputs. A second similar IC was designed comprising an ADC frontend as well as a PRBS15 generator and was fabricated in the MIT-LL 10 kA/cm2 process. The implemented alignment engine that bonds the individual channels into a single data link was proven up to 10 Gbps while taking 1–2 μ s to complete the alignment. We built chip-to-FPGA data links, comprising the on-chip driver and room-temperature interface amplifier, up to 14 Gbps using FPGA serial-link GTY transceiver. Successful data transport from an ADC using multiple parallel data links to an FPGA upon completion of the channel bonding was demonstrated.

The demonstration of a microwave superconducting quantum interference device (SQUID) multiplexed readout of transition-edge sensor (TES) microcalorimeters has the potential to dramatically expand the scale of arrays of TESs. In this paper, we discuss recent work to develop an instrument for high-resolution high-efficiency gamma-ray spectroscopy that integrates previously demonstrated high-resolution TES microcalorimeters with new lower noise microwave SQUID multiplexers. We will discuss the proposed instrument design and readout noise of the optimized microwave SQUID multiplexed readout. Finally, we will discuss the potential limits on the number of multiplexed TESs imposed by the analog-to-digital converter.

Cryogenic bolometers based on thin silicon nitride membranes show a very high sensitivity, which makes them ideal for ultrasensitive detector applications, particularly in the field of submillimeter-wave imaging. For that, transition-edge sensors (TESs) have been established as a viable approach toward developing multipixel sensor arrays. However, current multiplexer techniques as time- or code-division multiplexing do not scale well to multiplexing levels of several hundred detectors per channel. To achieve such a scalable readout solution, frequency-division multiplexing (FDM) would be an applicable way. For that, the temperature-sensing element of a bolometer has to be resonant or coupled to a resonance circuit, which changes its microwave behavior with the temperature of the absorber.

Recent developments in China for the design and application of high temperature superconducting (HTS) filters are reviewed. Progress in fabrication of HTS thin films with various physical and chemical methods are introduced first, since high quality double-sided thin films provide the groundwork for filters. In the past decade, hundreds of HTS filters have been designed in China with excellent performance, e.g., 0.05 dB minimum insertion loss, -23 dB return loss, -110 dB out-of-band rejection and 220 dB/MHz band-edge slope, etc. Novel design methods, i.e., coupling matrix extracting method, bandpass filters with wide stop band, linear phase filters, multiplexers, multi-band and multi-mode filters as well as tunable filters will be discussed in detail. Applications to mobile communication base stations, meteorological radar, deep space detection and satellite technology will be also reviewed.

For spectroscopic and imaging applications in the submillimeter wavelength band, superconducting radiation sensors are widely used. Thereby, sensor arrays with a rising number of pixels require multiplexing techniques in order to reduce the number of wires leading to the cryogenic stage. A current steering switch (CSS) provides the basis for one kind of promising code division multiplexers. It is composed of two identical superconducting quantum interference devices (SQUIDs) in parallel current paths. Switching one of them from the superconducting into the normal state, which is controlled by the applied magnetic flux, alters the signal path; thus, they can act as a polarity switch for analog signals. In this paper, we describe a method that uses rapid single-flux quantum (RSFQ) electronics for controlling these switches. Therefore, their SQUIDs are inductively coupled each to the storing loop of an RSFQ delay flip-flop (DFF); hence, the state of the analog switch can be controlled by means of digital RSFQ signals. As a first step, we show the change in the critical current of the SQUIDs by the coupled operating digital circuit. For a wider operating range of the CSS, we developed a DFF, which is able to store a larger quantity of magnetic flux to make it possible to apply more flux to the coupled SQUID. For further improvements, we tested asymmetric SQUIDs to reduce the required magnetic flux. The results of the simulations and measurements are discussed.

We report here a real-time pulse processing algorithm for superconducting transition-edge sensor (TES) based x-ray microcalorimeters. TES-based microcalorimeters offer ultra-high energy resolutions, but the small volume of each pixel requires that large arrays of identical microcalorimeter pixels be built to achieve sufficient detection efficiency. That in turn requires as much pulse processing as possible must be performed at the front end of readout electronics to avoid transferring large amounts of data to a host computer for post-processing. Therefore, a real-time pulse processing algorithm that not only can be implemented in the readout electronics but also achieve satisfactory energy resolutions is desired. We have developed an algorithm that can be easily implemented in hardware. We then tested the algorithm offline using several data sets acquired with an 8 × 8 Goddard TES x-ray calorimeter array and 2 ×16 NIST time-division SQUID multiplexer. We obtained an average energy resolution of close to 3.0 eV at 6 keV for the multiplexed pixels while preserving over 99% of the events in the data sets.

We have been studying single flux quantum (SFQ) digital signal processing (DSP) circuits for superconducting tunnel junction (STJ) photon detector arrays. The proposed DSP circuit consists of a binary tree of multiplexers, a time counter, a binary counter, and a readout memory. Each multiplexer was designed such that the first arriving SFQ pulse train connects to the binary counter until the time counter reaches a set value. The time counter is set to a sufficiently long time compared with the response time of STJ photon detector. After processing the multiplexer, SFQ pulses are integrated by the binary counter. The number of integrated SFQ pulses is proportional to the number of quasi-particles generated by photon irradiation in an STJ. After the time counter reaches the set value, the time counter generates a set SFQ pulse to set the multiplexer in the ready state, the integrated value is sent to the readout memory and then the binary counter is reset. The binary tree of multiplexers could be extended to have address bits for photon imaging using an STJ array. One key component of the proposed DSP circuit was fabricated and its operation using an on-chip clock generator was successfully demonstrated.

International audience; To further increase the sensitivity of future telescope projects dedicated to photometric astronomy observation at millimeter and sub-millimeter wavelengths, large number of bolometer is currently developed. In this context, The DCMB (Dévelopement Concerté de Matrice de Bolomètre) French collaboration makes an R&D effort to develop large bolometer arrays. This paper concentrates on a first demonstration of NbSi TES (Transition-Edge Sensors) array development: a 23 NbSi superconducting thermometer array. Firstly, the NbSi thin film alloy is described then the 23 TES array topology is presented. The readout of large TES arrays requires ultra low noise amplification and multiplexing electronics. The use of a first stage transducer such as a SQUID (Superconducting QUantum Interference Device) allows ultimate performances in terms of noise. However, the linearization of the SQUID characteristic requires a Low Noise Amplifier (LNA) to generate a Flux Lock Loop (FLL). We implement this component in a cryogenic SiGe Integrated Circuit (IC) that could also contains the control of the multiplexing. Using this readout chain, a one-pixel operation using NbSi thermometer readout by SQUID and a cryogenicLNAis demonstrated. Finally, the development of a specific cryogenic IC including amplifiers, addressing and switching current sources needed for a 24 to 1 time domain SQUID multiplexer is presented

We propose a single-flux-quantum (SFQ) based readout circuit for a transition edge sensor (TES) array for X-ray radiation detection. Utilization of SFQ circuits for this purpose enables large-scale integration of TESs due to very high speed processing ability of the received signals and using an already established integrated circuit design environment. We utilize the cooling time dependence of the TES on the incident X-ray energy. Time to digital conversion is made by using an SFQ based quasi-one-junction SQUID (QOS), which works as a 1-bit comparator with an adjustable current threshold level, and an SFQ based high speed counter. The readout system is composed of two separate chips that are connected to each other with flexible superconducting wiring. The QOS-multiplexer chip is directly connected to the TES array and it performs the digital conversion of the TES output. The demultiplexer-counter chip, which will be placed at elevated temperatures, receives the multiplexed SFQ pulses and determines the duration of the TES output which is above the predetermined threshold level. Final goal of this work is to readout more than 10 k TES pixels and in this paper, feasibility of the system and current status of the development process is demonstrated.

We have developed a single-stage SQUID multiplexer (SQMUX) for time-domain multiplexing of transition edge sensors (TESs). The individual channels in our SQMUX circuit are SQUID series arrays (SSAs). Several of these SSAs form one column, of which exactly one SSA is active at a time. The channel to be read out is selected by setting a superconducting-to-normalconducting (SN) switch into its normalconducting state. All other channels are short-circuited by SN-switches in their superconducting state. This SQMUX topology ensures that there are no noise contributions from off-channels. The SN-switches employed are SSAs, as well. The addressing is performed inductively by means of coupling magnetic flux into the switch. Our SQMUX can be operated at the TES operating temperature. It is directly read out by a room temperature preamplifier. The first realizations of this single- stage SQMUX are 4:1 and 2:2 multiplexers. We present the application of the new SQMUX to read out TES X-ray microcalorimeters. Furthermore, we discuss the application of SQMUXs to the readout of distributed power TES Detectors.

As large arrays of ultrasensitive cryogenic detectors of radiant energy are developed, multiplexing schemes are required to manage system complexity. Any such scheme must also meet stringent sensitivity, bandwidth, dynamic range and power dissipation specifications. We describe preliminary tests of a microwave frequency-division multiplexed readout of SQUID amplifiers. From these initial tests we estimated the number of SQUIDs that can be multiplexed with this scheme and the sensitivity and bandwidth of each SQUID amplifier.

We have developed a data-driven self-timed (DDST) rapid-single-flux-quantum (RSFQ) demultiplexer (demux) for the interface between on-chip high-speed RSFQ circuits and off-chip low-speed circuits. In order to eliminate the timing issue in a synchronous clocking system we employed the DDST architecture, where a clock signal is localized within a 2-bit basic demux module and dual rail lines are used to transfer the timing information between the modules. A larger demux can be produced simply by connecting the 2-bit modules in a tree structure. The DDST demux was designed for 10 Gb/s operation with sufficient dc bias margin using HYPRES 1 kA/cm/sup 2/ Nb process. We have successfully tested operation of the 2-bit demux up to 18 GHz using the DDST on-chip high-speed test system which was developed in our group.

A single-chip superconducting quantum interference device (SQUID) array interfaced to a Josephson multiplexer has been developed. A single-chip SQUID consists of a SQUID sensor and a superconducting feedback circuit. Its output is digital and can be processed with a Josephson digital circuit. The parallel output signal from the SQUID array was time-division multiplexed and summed at a single output cable by a Josephson multiplexer in a cryogenic environment. This method reduces the number of cables required for connection between the SQUID chips and room-temperature electronics. The eight-channel SQUID was tested with a 160-gate Josephson multiplexer. The authors verified the operation of the SQUID array at bias frequencies up to 5 MHz, while the multiplexer was operated at 120 MHz. The measured sensitivity was 1.13*10/sup -5/ Phi /sub 0//(Hz)/sup 1/2/ and the cut-off frequency was 350 Hz. >

Magnetically coupled calorimeters (MCC) have extremely high potential for X-ray applications. Although very high energy-resolution has been demonstrated, until now there has been no multiplexed read-out of MCCs. We report on the first realization of a time domain multiplexed read-out of MCCs. Although this has many similarities with time domain multiplexing of transition-edge sensors, for MCCs the energy resolution is limited by the SQUID read-out noise, and requires the well established scheme to be altered in order to minimize degradation due to noise aliasing effects. In our approach, each pixel is read out by a single first-stage SQUID that is operated in open loop. The outputs of the SQ1s are low-pass filtered with an array of low cross-talk inductors, then fed into a single-stage SQUID multiplexer. The multiplexer is addressed from room temperature and read out through a single amplifier channel. We present the noise performance and compare to expectations. We have demonstrated multiplexed X-ray spectroscopy at 5.9 keV and for improved readout noise achieved an energy resolution ΔEFWHM

In this paper, we present a novel design for a signal processor based on rapid-single-flux-quantum (RSFQ) circuits for a compact neutron diffraction system with an array of 10B-enriched MgB2 nanowire neutron detectors. By introducing an RSFQ signal processor, the system is able to accommodate a large number of detectors and determine the kinetic energy of each incident neutron. The processor consists of quasi-one-junction SQUIDs (QOSs), time-to-digital converters, and a multiplexer. A QOS is connected to each detector that operates at 27 K. The QOS picks up a voltage impulse with a width smaller than 2 ns generated across a detector and creates an SFQ pulse for RSFQ signal processing. The diffraction patterns are determined by the position and the kinetic energy of the incident neutrons. The time-to-digital converter is used to measure these energies by the time-of-flight method. Time resolution down to tens of picoseconds is possible, even with an array of detectors, and it is easy to multiplex signals in the time domain. We successfully demonstrated the RSFQ signal processor in liquid helium, and also tested a prototype, consisting of the processor connected with four MgB2 nanowire detectors in a Gifford-McMahon cryocooler, by irradiating a focused pulsed laser instead of neutrons, and obtained preliminary experimental results.