Fiber Test with SiPM

Introduction

Testing the various fibers with a silicon PM at JLAB is done using the same experimental setup as when testing the fibers with an XP2020 photomultiplier. The only difference is that the XP2020 PMT is replaced by a silicon PM from Sensl of type A20HD 3x3 mm. The break down Voltage of this device is at 27.9 Volts and was operated at 2 Volts above break down at 29.9 Volts. The supply voltage for the electroncis board is set at +/- 5 Volts.

Electronics DAQ

Because the signal from the SiPM has a considerable longer decay time then the XP2020 PMT the ADC gate needed to be increased from the initial 60ns to 140ns. This proved to be sufficient for the SiPM signals. At the same time the signal from the SiPM into the ADC had to be delayed by about 30ns because the SiPM has a much shorter transition time as compared to an XP2020 photomultiplier tube. The additional trigger to the LED in the previous setup with the XP2020 PMT is used as an arbitrary trigger to measure the dark rate. This data is needed to correct the SiPM response the extract the number of photo-electroncs from the Sr90 source as will be described below.

SiPM Response

As an example a typical response of the SiPM is shown here on three panels explaining the necessary corrections that need to be taken into account

The top right panel shows the response of the SiPM when triggering the DAQ arbitrary. The shows the effect of the dark rate of the SiPM. The red curve is a fit to the spectrum with three Gauss functions. At the bottom left panel in black the SiPM response to the trigger from the trigger-scintillator is shown and in red the estimated background triggers caused by cosmic rays instead of the electrons from the Sr90 source. The blue historgram is the resulting subtracted spectrum shown once more at the top right panel with a multi-Gaussian fit in red.

Extracting $\mu$

To extract the mean number of photo-electrons $\mu$ from the SiPM ADC spectrum is done as follows:
$\mu ={\frac {<mean>-P_{0}}{P_{1}-P_{0}}}$ $<mean>$ is the mean of the ADC spectrum. Since root does not calculated the mean of a spectrum correctly when subtracting one spectrum form another this number has to be calculated externally. $P_{0}$ is the pedestal peak position and $P_{1}$ is the peak position of the one photo-electron peak which is the second Gaussian peak in the ADC spectrum. Instead of using $P_{1}-P_{0}$ as the denominator the mean distance between neighbouring Gaussina peaks is used.