Lower Threshold Study:
Setup Threshold vs. Measured Threshold

The relation between the setup threshold and the actual threshold is still not well known. Actually the problem is that the most of the block calibraion runs were acquired with 12 mV of "setup threshold", and little data was acquired with other voltages. But here, we found good data sets available for Blcok 10.

Data sets examined:

Co-57 block calibration runs:

Data Number	File Name	Input Threshold Value*
1	blk_10_co57_200_0_0_0_021216_1	8 mV
2	blk_10_co57_200_0_0_0_021212_2	12 mV
3	blk_10_co57_150_0_0_0_021212_2**	20 mV

*: The values are from the log notebooks and binders stored in the BAT lab.

**: This file has a name "..._150_...", but the actual bias voltage was 200 V.

I) Measured lower threshold

For each of the above data sets:

1. corrected for the gain, offset and non-linearity for the electronics.
2. rebinned into 4096 ch for the range of 3000 mV in pulser voltage.

(The 11 testpulse peak and DNL run data were used to convert the raw ADU channel to pulser voltage unit.)

3. found the first five continuous non-zero bins for each detector channel.

The lowest bin in the five bins was assumed as the lower threshold.

Measured lower threshold:

The measured thresholds are plotted below in pulser voltage unit as well as in energy.

The distributions are evaluated by Gaussian function.

1. The setup threshold is 8 mV.

The distributions of measured thresholds are shown on the left-hand side. The fitted parameters (peak height, sigma and mean values) are shown in the same panel.

On the right-hand side, the measured thresholds are plotted against the detector ID. The guidelines are drawn to separate the 16 sandwiches.



2. The setup threshold is 12 mV



3. The setup threshold is 20 mV

II) Setup Threshold vs. Measured Threshold

The mean value of the measured thresholds are plotted against the setup threshold, shown in pulser voltage on the left side and in energy on the right side.

The error bars indicate variation between detector channels (1 sigma obtained with the above Gaussian fitting).

The three data points are fitted with a linear function very well.

When we increase the setup threshold from 12 mV to 20 mV, the mean value increases by ~6 keV.

III) Conclusions:

1. The mean value of the measured thresholds is linearly depnedent on the setup threshold value while the variation between detector channels is not dependent on the setup threshold very much.
2. When we increase the setup threshold from 12 mV to 20 mV, the mean value increases by ~6 keV.
3. If this linear relation between the setup and measured threshold holds in the other blocks, the distribution of the thresholds will shifts upward so that the yield at 15 keV will decrease to 50-60 %.
4. In order to analyze spectra to be observed by BAT accurately around threshold, it is necessary to use wellknown setup value (12 mV?) or to develop a tool to search the lower threshold through the 32K detectors, and to implement the batdrmgen to handle the measured threshold information.