H.J.M. ter Brake

Fabrication of a micro cryogenic cold stage using MEMS-technology

This paper describes the design and production process of a variety of reliable micro cryogenic coolers. The different cold stages are based on an optimized design found during a study which was done to maximize the cold-stage effectiveness. Typical cold-stage dimensions are 30 × 2 × 0.5 mm with an expected net cooling power varying from 10 mW to 20 mW at a tip temperature of 96 K. A cold stage consists of a stack of three fusion bonded D263T glass wafers. The production process has 7 lithography steps and roughly 100 process steps. In order to determine the maximum bend, shear and bond stresses inside a 175 µm thick D263T glass wafer, several pressure tests were performed.

Elimination of flux-transformer crosstalk in multichannel SQUID magnetometers

Multichannel SQUID magnetometers are being developed for signal-field mapping in biomagnetic experiments. A problem that becomes more serious as the number of channels is increased is the crosstalk caused by the mutual inductances between the individual sensing coils. A simple and effective method for eliminating this crosstalk is presented in this Paper. The method is based on a rearrangement of the feedback loops which causes the flux-transformer circuits to become currentless. The feasibility of the method is verified experimentally.

Characterization of micromachined cryogenic coolers

Micro cryogenic coolers can be used to cool small circuitry and improve their performance. The authors present a variety of micro coolers which are fabricated using MEMS technology production processes only. The typical dimension of a micro cold stage is 30 × 2.2 × 0.5 mm. It cools down to 96 K, applying Joule–Thomson expansion in a 300 nm high flow restriction and has a cooling power ranging from 10 mW to 25 mW. This paper discusses the operation of the micro cold stage and the characterization measurements done.

Long-life vibration-free 4.5 K sorption cooler for space applications.

A breadboard 4.5 K helium sorption cooler for use in vibration-sensitive space missions was developed and successfully tested. This type of cooler has no moving parts and is, therefore, essentially vibration-free. The absence of moving parts also simplifies scaling down of the cooler to small sizes, and it contributes to achieving a very long lifetime. In addition, the cooler operates with limited dcs so that hardly any electromagnetic interference is generated. This cooler is a favorite option for future missions such as ESAs Darwin mission, a space interferometer in which the sensitive optics and detectors can hardly accept any vibration. The system design consists of a hydrogen stage cooling from 80 to 14.5 K and a helium stage establishing 5 mW at 4.5 K. Both stages use microporous activated carbon as the adsorption material. The two cooler stages need about 3.5 W of total input power and are heat sunk at two passive radiators at temperatures of about 50 and 80 K-radiators which are constructed at the cold side of the spacecraft. We developed, built, and tested a demonstrator of the helium cooler. This demonstrator has four sorption compressor cells in two compressor stages. Test experiments on this cooler showed that it performs within all specifications imposed by ESA. The cooler delivered 4.5 mW at 4.5 K with a long-term temperature stability of 1 mK and an input power of 1.96 W. So far, the cooler has operated continuously for a period of 2.5 months and has not shown any sign of performance degradation.

165 k microcooler operating with a sorption compressor and a micromachined cold stage

This paper presents the fabrication of a microcooler consisting of small stainless steel sorption compressor cells, micromachined silicon check valves, and a micromachined cold stage that incorporates glass-tube heat exchangers. The design, fabrication and experiments on the different elements are described. Two compressor cells were thermally cycled to investigate the dynamic behaviour. The cold stage could reach a stable temperature of 169 K with a cooling power of about 200 mW.

The UT 19-channel DC SQUID based neuromagnetometer

A 19-channel DC SQUID based neuromagnetometer is under construction at the University of Twente (UT). Except for the cryostat all elements of the system are developed at the UT. It comprises 19 wire-wound first-order gradiometers in a hexagonal configuration. The gradiometers are connected to planar DC SQUIDs fabricated with a Nb/Al, AlO kappa/Nb technology. For this connection we developed a method to bond a Nb wire to a Nb thin-film. The SQUIDs are placed in compartmentalised Nb modules. Further, external feedback is incorporated in order to eliminate cross talk between the gradiometers. The electronics basically consist of a phase-locked loop operating with a modulation frequency of 100 kHz. Between SQUID and preamplifier a small transformer is used to limit the noise contribution of the preamplifier. In the paper the overall system is described, and special attention is paid to the SQUID module (bonding, compartments, external-feedback setup, output transformer).

On the fetal magnetocardiogram

Fetal magnetocardiography is a non-invasive technique for studying the electrical activity of the fetal heart. Fetal magnetocardiograms (fMCG) can be used to diagnose and classify fetal cardiac arrhythmias reliably. An averaged fMCG shows a QRS-complex, a P-wave, and a T-wave. However, it is still unknown if the currents in the tissues surrounding the fetal heart disturb these features. Furthermore, the measuring technique is not yet optimised for fMCG registrations. Simulation studies may provide guidelines for the design of an appropriate magnetometer system. Therefore, finite-element and boundary-element models were constructed in order to study the possible influence of the volume conductor. Especially, the influence of the layer of vernix caseosa, a fatty layer that covers the fetus, was investigated. The computations showed that the layer of vernix caseosa will affect the waveform of the fMCG. The signal processing procedure used is also discussed. It turned out to be difficult to deduce the onset and offset of the T-wave from the resulting averaged signals. Possibly, the QRS-complex does not provide a correct trigger to obtain a distinguishable T-wave in the averaged signal, because the RT-interval may be variable.

Cryogenic Systems for Superconducting Devices

It may happen, sometime, that the reader spends his or her well deserved holidays in the beautiful country of Greece and, sitting on a typically unstable chair in one of the nice taverns, one may ask for “kryo nero”. It may take a while, but you will get a nice jug filled to the top with icy cold water. Indeed, “cryo” means “cold”, and cooling was already known as a means to improve the quality of life in the early times The Egyptians put wet cloths over their foods, and placed them in the sun, so that the heat associated with the evaporation of the water cooled their food and drinks The Romans cooled their foods with ice blocks that were taken from Alpine regions and were stored underground in vaults insulated with straw. Our forefathers in Europe cut ice blocks out of rivers or shipped ice from Norway or Canada. In special ice factories rods of ice were made to be used for instance in butcheries and breweries.

Multichannel heart scanner based on high-T/sub c/ SQUIDs

A 7-channel magnetometer for magnetocardiography based on high-T/sub c/ SQUIDs has been realized. This magnetometer is used for test experiments in the development of a multichannel high-T/sub c/ SQUID based heart-scanner for clinical applications. The intrinsic noise level of the channels in the 7-channel system is typically 120 fT//spl radic/(Hz) down to 1 Hz. Magnetocardiograms were recorded inside a magnetically shielded room. Introductory experiments were performed on the suppression of noise by combining magnetometers to form planar gradiometers. The noise suppression that can be established appeared to be limited by the imbalance of the gradiometric configuration, which is roughly 2%. This relatively poor balance of the system is caused by inaccuracies in the transfer functions of the individual SQUID magnetometers, and by deviations from the planar geometry.

Magnetic Noise of Small Stirling Coolers

We measured the magnetic noise generated by three small Stirling coolers (cooling powers 0.5 to 1.5 W at 80 K). Such a cooler will be used for cooling a high-Tc SQUID magnetometer. The measurements were performed with a fluxgate magnetometer and a 3-axis low-Tc SQUID magnetometer, the latter in a magnetically shielded room. The measuring setup is shortly described, and results on the coolers are given and compared to a simple dipole model. Consequences for SQUID-cooling are shortly discussed.