Magic Numbers
There are several places in the Cait code, where so-called “magic” number appear, e.g. a hard-coded integer as index to an array, with no or only short explanation why this number is taken here. Most of the time, this is due to the numbering of labels, main parameters or fit parameters. These numbers are written as attributes in the HDF5 files, however we also provide lists here.
Labels
A usual pulse, originating in a particle recoil within the absorber crystal.
A testpulse, induced by a temperature pulse from the detector heater. These pulses are intentionally induced every 5 seconds, to measure the variations in the detector response. Their pulse form is different than that of event pulses, with typically a longer rise time.
An empty noise baseline, or a noise trigger.
A jump in the offset level of the baseline, coming from a reset of the SQUID or the loss of a flux quant.
A thin positive or negative spike from the detector electronics.
An absorber pulse with maximum time significantly different than zero (>/< 20 ms).
Two or more events in one record window.
A thin, low energetic pulse, originating in a particle recoil within the carrier crystal of the detector.
Saturated Event Pulse. Saturation happens, when the TES is driven out of its linear operation range by a high energetic particle recoil within the absorber crystal.
Same as Label 9, but for a Test Pulse.
Trigger of noise or a sub-threshold event, that gets elevated above threshold by a decaying baseline.
A sudden and continuous rise of the detector temperature causes a strong positive slope in the noise baseline, that eventually surpasses the threshold.
A pulse with long rise- and decay time, caused by a particle recoil within the sticks that hold the crystal. The sticks themselves are made from e.g. CaWO4.
Small, discrete jumps of the baseline level. A upward jump, followed right away by a downward jump, is also called a jump event and might look like a pulse on first sight.
An oscillating signal, caused by vibration.
Multiple squid resets, cause by fast rising temperature - e.g. due to warm up of the cryostat or very large heat deposition to a sensible TES.
An event with especially long tail, typical for the several Cosinus modules.
A direct hit in the light detector, without corresponding phonon signal.
A recoil in the ring od the detector, this component is e.g. included in Gode/beaker modules.
A very fast rising light event, typically caused by a direct hit in the light detector, with a hit in the detector following after.
In some cases, we have not enough insight to explain an event. If this concerns only a single or very few event, these are together put in an unknown class, which is then excluded from the analysis.
Main Parameters
The Maximum Value of the Event, after application of a 50 sample moving average, in Volt.
This value is calculated backwards in time: Starting at the position of the sample with the maximum value, the last sample that subceeds 20% of the pulse height.
This value is calculated backwards in time: Starting at the position of the sample with the maximum value, the last sample that subceeds 80% of the pulse height.
The index of the maximal sample.
The first sample after t_max, that falls below 90% of the maximal height.
The first sample after t_max, that falls below 73% of the maximal height.
The first sample after t_max, that falls below 36% of the maximal height.
The average of the first 500 samples of the event.
The difference between the average of the first and last 500 samples of the event, divided by the record length.
Usually this is set to zero! The quadratic component of the baseline.
Parametric Fit Parameters
The fit parameters of the parametric pulse shape fit.
Standard Event Fit Parameters
The fit parameters of the standard event fit:
The array fit has the same paramters!
Additional Main Parameters
The maximum of the array.
The minimum of the array.
The variance of the first eight of the record window. This is typically the variance of the baseline noise.
The mean value of the first eight of the record window. The first eight of the record window typically shows only the noise baseline, i.e. this is the mean value of the noise.
The variance of the last eight of the record window. If this value differs strongly from the baseline variance, this indicates a strongly saturated pulse, a Pile-Up Event or an early trigger.
The mean value of the last eight of the record window. A difference of this value from the mean of the first eight indicates a strong baseline tilt.
The variance of the whole array.
The mean value of the whole array.
The skewness of the whole array.
The maximal value of the derivative of the array.
The index of the maximal value of the derivative of the array.
The minimal value of the derivative of the array.
The index of the minimal value of the derivative of the array.
The maximum of the array, after applying the optimum filter.
The index of the maximum of the array, after applying the optimum filter.
The skewness of the array around its peak, after applying the optimum filter. Typically this value is higher, when the array deviates from the standard event.