Behaviour of Belle II ARICH Hybrid Avalanche Photo-Detector in magnetic field

The proximity-focusing Aerogel Ring-Imaging Cherenkov detector (ARICH) has been designed to separate kaons from pions in the forward end-cap of the Belle II spectrometer. The detector will be placed in 1.5T magnetic field and must have immunity to it. In ARICH R&D, we solve the problem with new equipment called Hybrid Avalanche Photo-Detector (HAPD) which developed by Hamamatsu Photonics. Recently the production of about 500 HAPDs was completed. We test HAPDs in magnetic field in KEK. We found some HAPDs have significant amount of dead time, which reaches up to 30% in the worst case. The dead time is caused by very large (more than 10,000times larger than a single photon signal) and frequent (∼5Hz) signals, which make electronics paralysed. The huge signals are observed in about 30% of HAPDs. To identify the origin and understand the mechanism, we perform some extra test of HAPDs. We find a strange dependence of the huge signals to the APD bias voltage. If we reduce the bias voltage applied to one of the 4 APDs by 10V, the frequency of the huge signals is much reduced. On the other hand, if we reduce the voltage of all the 4 HAPDs, huge signals do not decrease, or even increase in some case. We also find the huge signals seems to be related to the vacuum inside HAPD. We present about the observation of the huge signals of HAPDs in the magnetic field, and our strategy to manage it.