Matías Senger
July 14th, 2022
Contents
Introduction
A number of times when "playing" with our new Photonis PMT I noticed that the signals on the LGAD get smaller if I require coincidence with the PMT, which is below. So I decided to perform a beta scan to see more precisely what is going on.
In there is a graphical description of the setup arrangement. The LGAD, an HPK device whose IV curve is shown in , was mounted on a Chubut board . Both the LGAD and the PMT were connected to the oscilloscope through one of our EBay amplifiers. As shown in the trigger was set only on the LGAD. In these conditions several beta scans were performed at different bias voltages with "low statistics" and only a single one, at 140 V, was performed at high statistics.
Setup arrangement description.
Configuration in the oscilloscope. Channel 2 is the PMT while channel 3 is the LGAD.
IV curve of the LGAD used for this test.
Results
99999 triggers were acquired with the oscilloscope in the conditions described above and the LGAD at a fixed voltage of 140 V. The events were filtered according to 3 categories:
- LGAD yes PMT don't care: Events that produced a signal on the LGAD, independently of the fact that there was signal in the PMT.
- LGAD yes PMT no: Events that produced a signal on the LGAD and no signal on the PMT.
- coincidence: Events that produced a signal in both devices.
Events with no signal in any of the two devices were discarded as background. The charge distribution for each of these three conditions is shown in . Three Langauss fits where performed, once for each distribution. The distributions seem to qualitatively match the fits. Looking at the parameters of each fit, it can be seen that the Landau's most probable value (xMPV) is practically the same in each case, but the Landau's width parameter ξ is the one changing when the events are required to produce a signal not only in the LGAD but also in the PMT.
Distribution of the collected charge in the LGAD for 3 different conditions.
The same behavior was observed at different bias voltages, though with less statistics.
Conclusions
The effect of the PMT filtering out the higher charge events was measured. The most probable value of the Landau does not seem to change but the width of it. I am still not sure what the reason for this behavior is, some colleagues suggested that probably the non-MIP part of the 90SR spectrum, i.e. low energy electrons, are probably not reaching the PMT because they deposit much more energy on the LGAD (and maybe the glass of the PMT) and they are stopped before. If this is the case, this would actually be beneficial as we are interested only on MIPs.