High-accuracy low-noise electrical measurements in a closed-cycle pulse-tube cryostat

TL;DR

The paper demonstrates that high-accuracy, low-noise electrical measurements, including quantum Hall resistance standards, are feasible in a cryogen-free pulse-tube cryostat when perturbations are carefully characterized. It provides a thorough thermal, vibrational, and electromagnetic assessment of a commercial cryomagnet system and a custom coaxial insert, achieving DC measurement uncertainties around $10^{-9}$ and showing comparable performance to liquid-helium systems for long measurements. A key finding is that cryocooler-induced EM noise increases with magnetic field and can couple to DC-SQUID readouts at high frequencies near the SQUID modulation, though this can be mitigated with appropriate damping and filtering. The work offers practical guidance for laboratories to quantify and manage vibration and EM pollution in cryogen-free systems, enabling robust metrological applications in quantum technologies and fundamental constants research.

Abstract

A shift of paradigm to obtain (sub-)Kelvin environment is currently on-going with the democratization of cryogen-free cryocoolers, boosted by their easy-to-use and continuous operation without the need of liquid helium whose cost and scarcity globally increase. Thanks to their large sample space and cooling power, they can host a superconducting magnet and are an adapted platform for quantum technologies, material science, low temperature detectors and even medical fields. The drawback is that this type of system is inherently based on gas compression that induces a certain level of vibrations and electromagnetic perturbations, which can potentially prevent the determination of low amplitude signals or spoil their stability. In this paper we demonstrate that pulse-tube based cryocoolers can be used for electrical precision measurements, using a commercial cryomagnetic system combined with our home-made a coaxial cryoprobe. In particular, parts-per-billion level of measurement uncertainties in resistance determination, based on quantum Hall resistance standards, is achievable at the level of state-of-the-art measurements involving conventional cryostats based on liquid helium. We performed an extensive characterization of the cryomagnetic system to determine the level of vibrations and electromagnetic perturbations, and revealed that although the magnetic field has a drastic effect on the noise level, only marginal interplays on the measurement are observed as long as the working frequencies of the instrumentation are not in the vicinity of the ones of the perturbations. The set of characterization measurements presented here are easily implementable in laboratories, which can help to determine the vibrations and electromagnetic pollution generated by any cryocooler.

Figures (11)

• Figure 1: (a) Schematic view of the facility layout: On the right hand side is the laboratory, regulated in temperature and relative humidity (RH), with an electromagnetic shield. The cryostat, in its own pit, has a separated valve unit. It is connected to the helium tank, pump station and compressor, situated in the technical corridor (left hand side) via wall feedthroughs. (b) Temperature stability of the cold finger when regulated at 4 K with the cold finger heater (black curve on top) with in insert a close-up, when regulated at 4 K using the VTI heater (in blue, middle curve) and at base temperature (at 14 T, in green, lower curve). The typical temperature stability of the second stage of the coldhead is shown in the lower insert (red curve).
• Figure 2: Vibration measurements. (a) Cryomagnetic system and its immediate surrounding, with the position of the accelerometers and the Hall probe attached to the insertion column. (b) Resulting vibration level at the three different positions. Note the break on the y-axis. The arrow shows the signal at 3.4 Hz. (c) Spectral noise density of the Hall probe signal with a DC excitation current of 1 mA, at a magnetic field of 14 T and 0 T.
• Figure 3: The insert assembled. (a) The LEMO connectors for a practical use, which are solely an extension of the SMA connectors on the head of the insert (b). (c) Insert installed on the cryostat, open to have access to the connector, or close as shown in (e). The red circle shows the position where the insert lies on the cryostat. Four connectors A, B, C, D placed on the back of the head are used to the split lines for careful grounding, and the connector for the technical lines in the middle. (f) The assembled cold finger with two TO-8 sample holders installed as shown in (g).
• Figure 4: (a) Picture of the setup: A Hall sensor is tied to an arm itself strongly fixed to the steel plate supporting the cryostat. Directly facing the Hall sensor is placed a magnet on the head of the insert. The oscillations of the insert have been measured along the three space directions, and when the magnet is away (background). (b) Spectral noise density of the Hall sensor signal (biased under a DC current of 1 mA) along the three directions with the background (in green), when the compressor and remote motor are running (in black) or shut down (in red).
• Figure 5: (a) Screenshot of the videos taken of the head of the insert. (b) Frequency spectrum of the luminance of the pixel at the positions 1 and 2. (c) Mapping of the vibrations of the entire head. The color corresponds to the frequency at which the amplitude is maximum. The panels (d), (e) and (f) are the mapping of the luminance respectively at 1.7 Hz, 7.5 Hz and 40 Hz.