Thermal and Electrical Properties of Prototype Readout Components for CMB-S4

TL;DR

This work addresses the challenge of scaling the CMB-S4 readout to tens of thousands of detectors by evaluating thermal and electrical performance of prototype cold readout components. The authors perform thermal conductivity measurements on shielded cryogenic cables and conduct electrical tests of SSA based readouts across different configurations and temperatures to assess bandwidth and dynamic impedance. Key findings include that shielding and insulation significantly affect the thermal load on cryogenic stages, and that SSA dynamic impedance and shunt configurations influence bandwidth with complex dependencies. These results inform design trade offs and heat sinking strategies essential to achieving the desired multiplexing factors while maintaining a manageable cryogenic thermal budget.

Abstract

CMB-S4 is the fourth-generation ground-based cosmic microwave background project, designed to probe the early universe and cosmic inflation. CMB-S4 would achieve its science goals in part by dramatically increasing the number of transition edge sensor (TES) bolometer detectors on the sky. The detector readout system for CMB-S4 is time-division multiplexing (TDM) with a two-stage Superconducting Quantum Interference Device (SQUID) system. To accommodate the large increase in detectors, the size of our camera increases, placing physical constraints on the readout, its wiring, and its power dissipation. Therefore, to optimize readout performance, we need to balance competing design considerations such as thermal load and bandwidth. We present results characterizing the thermal and electrical performance of prototype components, including wiring and SQUID arrays for CMB-S4, and discuss the impact on overall system performance.

Figures (4)

• Figure 1: Measurements (crosses) and expectations (lines) based on Equation \ref{['Qin']} with base temperature at $T_{\text{low}}\approx 3.5$ K and $\ell\approx2.6$ cm. Manganin: 8 wires of 32 AWG. The expected $\dot{Q}_\text{in}$ is determined by using thermal conductivity of manganin in duthil2015. Teflon: 4 strips with cross-section $\approx 0.01\times0.0024$ m$^2$. The two lines for expected $\dot{Q}_\text{in}$ shown are derived from the compiled thermal conductivity fit in marquardt2000 (also found in the NIST Cryogenic Material Properties Database) and an extrapolation from thermal conductivity data in reese1965, which had measurements closest to our temperature range. FEP: With no expectation from the literature to compare against, we compared our measurements of one strip of FEP (cross-section $\approx 0.01\times0.0031$ m$^2$) with a measurement of two strips with twice the cross-section, and determined the error based on expecting otherwise identical material properties. The error bars on these measurement data points are derived from comparing these measurements of one and two strips. Shielded Twisted Pair: Dimensions for the cross-section of these wires are shown in Figure \ref{['Thermal']}. The expectation is based on the calculated thermal conductivity contribution from FEP and manganin, and it also considers the theoretically expected contribution from SS 304. The shaded region is the range in expected value which includes the lower and upper bound of the conductance contribution from insulation.
• Figure 2: (a) The lines represent the derivative of Equation \ref{['Qin']}, $\frac{d\dot{Q}_{\text{in}}}{dT_{\text{high}}}$ from data in Figure \ref{['Thermal2']}, with unit length $\ell = 1$ m and $T_\text{low} = 3.49$ K for the different components of four prototype shielded twisted pair wires. The gray highlights the range between the lower and upper bound estimates of FEP/Teflon and the sum of all the wire components for the four prototype wire samples with cross section shown. The green crosses show measurements and error for conductance and stabilizing temperature ($T_\text{high}$) for the same dimensions of wires. (b) Cross section diagram for shielded twisted pair wires. Manganin wires are 36 AWG with .0045 inch FEP coating wrapped in 44 AWG Soft Bare Type 304 Stainless Steel Braid with 14 wires. The outer FEP coating is 0.004 inch thick. The nominal diameter of the entire wire is 0.044 inches.
• Figure 3: (a) $V$-$\Phi$ curves for SA13s with varying shunt resistances. The cross is the selected tuning point on the downward slope (high $R_{\text{dyn}}$), and the dot is the selected tuning points on the upwards slope (low $R_{\text{dyn}}$). (b) The peak-to-peak voltage (top) and bandwidth (bottom) against dynamic impedance for SA13s with varying shunt resistance for both tuning points.
• Figure 4: (a) $V$-$\Phi$ curves for SA23s and SA13s. The dot is the selected tuning point on the downward slope (low $R_{\text{dyn}}$), and the cross is the selected tuning points on the upwards slope (high $R_{\text{dyn}}$). The legend (left) shows the shunt resistor values for each SSA in their corresponding configuration. (b) The peak-to-peak voltage (top) and bandwidth (bottom) against dynamic impedance for SA13s with varying shunt resistance for both tuning points.