Sketch of a CUORE bolometer. The crystal is held by teflon supports, the thermistor is glued to the crystal, and its wires are glued to copper frames. The supports and the thermistor wires thermally couple the crystal to the copper frames, which act as heat bath.
Biasing circuit of the thermistor. A voltage generator biases the thermistor resistance in series with a load resistance . The bolometer signal is the voltage across . The wires used to read have a non-negligible capacitance .
Pulse generated by a 2615 keV particle. The baseline is related to the temperature of the thermistor before the particle interaction, the amplitude carries information on the amount of energy released.
Energy spectrum. The line at 5407 keV is generated by contamination of the crystal, the rest of the spectrum is due to ’s and ’s generated by a source. The most prominent peaks are labeled with their energy (in keV).
Pulse shape parameters vs energy. The rise time (a) and the decay time (b) are anticorrelated and correlated with energy, respectively.
Amplitude of 5407 keV pulses vs baseline. A change in the bolometer temperature also changes its response, worsening the resolution.
Residuals obtained using a linear calibration function. The difference between the estimated energy and the nominal one is not compatible with zero, considering that the energy resolution is FWHM. The error bars refer to the uncertainty on the estimated peak position, that depends on the FWHM and on the number of events in the peak as .
Comparison between the original signal generated by a 2615 keV particle, and the same signal transformed into and . The signal is faster than the one, because the low pass filter of the biasing circuit has been deconvoluted. The signal is even faster since the thermistor slowing down effect has been removed.
Pulse shape parameters vs energy. The data (light gray) are less dependent on energy than data (dark gray), the correction is particularly effective on the decay time (b). The rise time (a) of data is faster, because the low pass filter of the biasing circuit has been deconvoluted.
Amplitude of 5407 keV pulses vs baseline. The data (light gray) are less correlated with the baseline than data (dark gray). The ratio of the slopes is about 20.
Residuals of the linear calibration function. The residuals of data (light gray) are very close to zero and have no trend compared to data (dark gray).
Article metrics loading...
Full text loading...