System noise cancellation by digital signal processing for SQUID measurement (original) (raw)

Superconductor Science and Technology, 2006

Abstract

It is important to suppress both environmental noise and system noise as much as possible when a weak magnetic signal is measured by a superconducting quantum interference device (SQUID). Environmental noise can be suppressed by using magnetic shielding or a gradiometer. However, we still have system noise produced by the electronics such as coils and amplifiers, even if we use perfect magnetic shielding or a gradiometer. This research has been aimed at reducing this system noise signal in measured data using digital signal processing. Two SQUIDs are placed close together, and the same magnetic field is detected with these two SQUIDs simultaneously. The outputs of these SQUID magnetometers, however, are different from each other, because the system noise included in the signal has random phase, amplitude and frequency for each respective SQUID. By extracting the in-phase components from these two SQUID output signals, the system noise signal can be reduced, and as a result, the signal from the measuring object in which we are interested can be obtained. An adaptive digital filter (ADF) algorithm was used for this extraction of the in-phase components. When the signal-to-noise ratio was 0.5, the noise signal was decreased by about 10 dB by this processing. In addition, the frequency division by the wavelet transform was used to raise the performance of the in-phase component extraction. The noise signal is reduced at each frequency band, and each of the band elements is reconstructed by an inverse wavelet transform to obtain the signal of the object. The noise removal performance was improved to about −20 dB when this method was used. In addition, the waveform distortion became lower than that processed without wavelet transform.

hideo itozaki hasn't uploaded this paper.

Let hideo know you want this paper to be uploaded.

Ask for this paper to be uploaded.