The Sliding Flux Ramp Demodulation Algorithm with High Sampling Rate in Microwave SQUID Multiplexer

GuofuLiao, CongzhanLiu, ZhengweiLi, DaikangYan, XiangxiangRen, YongjieZhang, LaiyuZhang, YuXu, ShiboShu, HeGao, YifeiZhang, XuefengLu, XufangLi, HeXu, DiWu · 2025

Read The Sliding Flux Ramp Demodulation Algorithm with High Sampling Rate in Microwave SQUID Multiplexer by GuofuLiao, CongzhanLiu, ZhengweiLi, DaikangYan, XiangxiangRen, YongjieZhang, LaiyuZhang, YuXu, ShiboShu, HeGao, YifeiZhang, XuefengLu, XufangLi, HeXu, DiWu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-12550.txt.

Introduction

Guofu Liao and co-authors present a faster readout method for a microwave multiplexer used with cold sensors, in a 2025 arXiv paper. The device lets one line read a huge array of heat sensors, a key need for X-ray space work. Readout speed caps how crisp the timing can be.

Their method, called sliding flux ramp demodulation, pulls the sensor signal from a ramped flux with a sliding window. It reaches sample rates near a million per second while still serving about a thousand sensors on the line. That mix of speed and scale suits heat based X-ray sensors, which need fine time stamps and high event counts. The authors say the gain will help X-ray astronomy craft log more events with sharper timing.

Catalog record

CREATORGuofuLiao, CongzhanLiu, ZhengweiLi, DaikangYan, XiangxiangRen, YongjieZhang, LaiyuZhang, YuXu, ShiboShu, HeGao, YifeiZhang, XuefengLu, XufangLi, HeXu, DiWu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE20 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-12550

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The Sliding Flux Ramp Demodulation Algorithm with High Sampling Rate in Microwave SQUID Multiplexer GuofuLiao, CongzhanLiu, ZhengweiLi, DaikangYan, XiangxiangRen, YongjieZhang, LaiyuZhang, YuXu, ShiboShu, HeGao, YifeiZhang, XuefengLu, XufangLi, HeXu, DiWu Abstract Microwave SQUID Multiplexing ( μ \mu MUX) is a widely used technique in the low-temperature detectors community as it offers a high capacity for reading large-scale Transition-Edge Sensor (TES) arrays. This paper proposes a Sliding Flux Ramp Demodulation (SFRD) algorithm for μ \mu MUX readout system. It can achieve a sampling rate in the order of MHz while maintaining a multiplexing ratio of about one thousand. Advancing of this large array readout technique makes it possible to observe scientific objects with improved time resolution and event count rate. This will be highly helpful for TES calorimeters in X-ray applications, such as X-ray astrophysics missions. keywords Microwave SQUID Multiplexer, Flux Ramp Demodulation, Transition-Edge Sensor 1 Introduction The \mu MUX is invented to enable the readout of large arrays of TESs. As a superconducting thermal equilibrium detector, the latter has an excellent signal-to-noise ratio and shows great potential in science applications that require high energy resolution or low noise equivalent power (NEP). Compared with other TES multiplexing techniques, known as Time-Division Multiplexer, Code-Division Multiplexer, and Frequency-Division Multiplexer, \mu MUX requires fewer wires that run from the room-temperature stage to the cryogenic stage, which makes it attractive to TES involved experiments that need large-scale TES arrays or compact, low cooling-power detector chambers. One field that relies on large-scale TES arrays is cosmic microwave background (CMB) research. The demand for CMB studies involves scaling up to tens of thousands of pixels or more. The \mu MUX with a high multiplexing factor of about 1000 to 2000 has been used in several recently deployed and under-construction CMB telescopes, such as SO-SAT Galitzki et al. (2024) and AliCPT-1 Salatino et al. (2020) . Another scenario is space-based X-ray astrophysics telescopes with high-energy resolution. Several space-based telescopes for X-ray astrophysics have been proposed to use thousands of pixels from TES calorimeters as focal-plane detectors in recent years Bandler et al. (2023) ; Barret et al. (2023) ; Gottardi and Nagayashi (2021) ; Sato et al. (2020) ; Cui et al. (2020) ; Bandler et al. (2019) . Driven by the need to measure finer energy spectra, structures, and temporal variability of astrophysical X-ray sources, large-scale TES arrays and high sampling rates are required. Additionally, a compact detector chamber is preferred in space. Meeting these requirements poses a significant challenge, as a high sampling rate necessitates a high flux ramp frequency, leading to high resonator bandwidth and frequency spacing requirements. Increasing the sampling rate often results in a decrease in the multiplexing ratio. In this study, we propose a Sliding Flux Ramp Demodulation (SFRD) method to enhance the effective sampling rate of the \mu MUX system while maintaining a high multiplexing ratio. 2 Algorithm 2.1 The \mu MUX Signal Modulation and Demodulation With \mu MUX, the current signal from a TES is inductively coupled to an RF-SQUID (Radio Frequency Superconducting Quantum Interference Device) and then inductively coupled to a microwave resonator. By accommodating multiple such microwave resonators of different resonant frequencies into a common feedline, a number of TESs can be read out through one coaxial cable simultaneously. Each TES channel is measured by the S21 response of a microwave resonator: Jonas Zmuidzinas (2012) : S_{21}=1-\frac{Q_{r}}{Q_{c}}\frac{1}{1+2jQ_{r}\frac{f_{exc}-f_{r}}{f_{r}}} (1) where Q_{r} is the total quality factor and Q_{c} is the coupling quality factor of the resonator, f_{exc} is the fixed excitation frequency on the microwave feedline, and f_{r} is the resonant frequency that changes with RF-SQUID inductance L(\phi) John Mates (2011) : \displaystyle f_{r}=\frac{f_{1}}{1+4f_{1}C_{c}Z_{1}+4f_{1}L(\phi)/Z_{1}}, (2) \displaystyle L(\phi)=L_{c}-\frac{M_{c}^{2}}{L_{s}}\frac{\lambda\cos(\phi)}{1+\lambda\cos(\phi)}, (3) \displaystyle\lambda\equiv\frac{L_{S}}{L_{J}},\quad\phi\equiv 2\pi\frac{\Phi}{\Phi_{0}}. (4) In Eq. ( 2 ), f_{1} is the intrinsic resonant frequency, C_{c} is the coupling capacitance from the resonator to the common feedline, and Z_{1} is the intrinsic impedance of the resonator. In Eq. ( 3 ), L_{c} is the coupling inductance on the resonator to the RF-SQUID, L_{s} is the self-inductance of the RF-SQUID that originates from its loop structure, and M_{c} is the mutual inductance between the inductor and the RF-SQUID. In Eq. ( 4 ), \Phi is the magnetic flux in the RF-SQUID loop, and \Phi_{0} is the magnetic flux quantum. By simply transforming Eq 1 , we can obtain : S_{21}=1-\frac{Q_{r}}{2Q_{c}}-\frac{Q_{r}}{2Q_{c}}e^{-2j\arctan(2Q_{r}\frac{f_{exc}-f_{r}}{f_{r}})} (5) Here, when the excitation frequency f_{exc} is constant, it can be seen that the trajectory of S_{21} forms an arc of a circle on the I-Q plane. The center of the circle is 1-\frac{Q_{r}}{2Q_{c}} , the radius is \frac{Q_{r}}{2Q_{c}} , and the angle of the arc is related to f_{r} . Assuming f_{exc}=\frac{f_{1}}{1+4f_{1}C_{c}Z_{1}+4f_{1}L_{c}/Z_{1}} , then we can get 2Q_{r}\frac{f_{exc}-f_{r}}{f_{r}}=-2Q_{r}\frac{4f_{1}\frac{M_{c}^{2}}{L_{s}Z_{1}}\frac{\lambda\cos\phi}{1+\lambda\cos\phi}}{1+4f_{1}C_{c}Z_{1}+4f_{1}L_{c}/Z_{1}}. (6) Here, considering 1+4f_{1}C_{c}Z_{1}+4f_{1}L_{c}/Z_{1}\approx 1 , and assuming 8Q_{r}f_{1}\frac{M_{c}^{2}}{L_{s}Z_{1}}\ll 1 . Therefore, \arctan(2Q_{r}\frac{f_{exc}-f_{r}}{f_{r}})\approx 2Q_{r}\frac{f_{exc}-f_{r}}{f_{r}}=-k\frac{\lambda\cos\phi}{1+\lambda\cos\phi} (7) here, 0<k=\frac{8Q_{r}f_{1}\frac{M_{c}^{2}}{L_{s}Z_{1}}}{1+4f_{1}C_{c}Z_{1}+4f_{1}L_{c}/Z_{1}}\ll 1 . Considering a small value of \lambda<0.5 John Mates (2011) ; Mates et al. (2012) , then Eq. 5 can be further simplified to S_{21}\approx 1-\frac{Q_{r}}{2Q_{c}}-\frac{Q_{r}}{2Q_{c}}e^{2jk\lambda\cos\phi} (8) Therefore, when we obtain the S_{21} trajectory at the excitation frequency, we can remove the center of the circle by translation. Then, we use the arctangent to obtain 2k\lambda\cos\phi , which is exactly the SQUID signal. In practice, the measured S_{21} contains cable delay components, which will cause the S_{21} arc trajectory to rotate by a certain angle on the complex plane. So, the practical process includes translation to move the center to zero (or the origin) and rotation to remove the cable delay before performing arctangent to obtain the SQUID signal 2k\lambda\cos\phi . The \mu MUX system adopts a sawtooth signal to linearly ramp the magnetic flux in the RF-SQUID, so called “flux-ramp modulation” Mates et al. (2008) . Based on Eq. ( 8 ), the flux ramp modulation leads to a sinusoidal response in the finally measured SQUID signal. The TES current signal, superimposed to the flux-ramp signal, will cause an extra phase shift in the sinusoidal SQUID response. Therefore, the measured signal has the form: \theta(t)=2k\lambda\cos(2\pi f_{ramp}N_{\Phi_{0}}t+\phi_{TES}(t)) (9) where f_{ramp} is the frequency of the sawtooth flux-ramp signal. N_{\Phi_{0}} , usually an integer, is the number of \Phi_{0} the ramp signal generates in the RF-SQUID, and is directly controlled by the ramp signal amplitude. \phi_{TES}(t)=2\pi\frac{\Phi_{TES}(t)}{\Phi_{0}} represents the phase shift caused by the TES signal. \Phi_{TES} represents the magnetic flux caused by the TES signal in the SQUID loop. According to the Eq 9 , it can be seen that the phase shift of the SQUID signal is proportional to the TES signal. When 2\pi f_{ramp}N_{\Phi_{0}} is much greater than the slope of \phi_{TES}(t) , this phase shift can be solved as \phi_{TES}=\arctan(-\frac{\int\theta(t)\sin(\omega_{mod}t)}{\int\theta(t)\cos(\omega_{mod}t)}) (10) Here, \omega_{mod}=2\pi f_{ramp}N_{\Phi_{0}} represents the modulation angular frequency. The integration interval spans an integer number of modulation cycles. This process is referred to as Flux Ramp Demodulation (FRD). Considering the discrete nature of the digitizing system, with a sampling frequency of f_{s} , the actual demodulated TES signal is: \phi_{TES}^{\prime}[m]=\arctan\big(-\frac{\sum_{n=Nm}^{N(m+1)-1}\theta[n]\sin(\Omega_{mod}n)}{\sum_{n=Nm}^{N(m+1)-1}\theta[n]\cos(\Omega_{mod}n)}\big) (11) Here, \Omega_{mod}=2\pi\frac{f_{ramp}N_{\Phi_{0}}}{f_{s}} represents the digital modulation angular frequency and N=f_{s}/f_{ramp} is the number of data samples in one flux-ramp cycle. The effective sampling rate of the \mu MUX system, therefore, is the flux-ramp frequency f_{ramp} . 2.2 Sliding Flux Ramp Demodulation Performing a Discrete Fourier Transform (DFT) to the time series \theta[n]

Read the complete text