A rapid and accurate method of finding light leaks in photomultiplier systems

TL;DR

This paper introduces a quick, reliable method for detecting light leaks in photomultiplier systems using an audible signal, leveraging standard laboratory equipment for practical and efficient leak identification.

Contribution

It presents a novel, simple technique that mimics a Geiger counter to identify light leaks in photomultiplier systems rapidly and accurately.

Findings

Effective detection of light leaks using audible signals

Utilizes common laboratory equipment for practical implementation

Provides a rapid alternative to traditional methods

Abstract

A rapid method of finding light leaks in photomultiplier systems is described, in which an audible signal derived from the light level is produced. It uses equipment commonly available in laboratories. In practice it is like using a geiger counter to detect radioactivity.

Full text

A rapid and accurate method of finding light leaks in photomultiplier

systems

John McMillan

Department of Physics and Astronomy, The University of Sheffield

Sheffield, South Yorkshire, S3 7RH, Great Britain

j.e.mcmillan@sheffield.ac.uk

Keywords: photomultipliers, scintillation detectors, cherenkov detectors

I.   Introduction

When constructing experimental apparatus containing photomultipliers or other photodetectors sensitive at the single photon level, light leaks are often encountered as it is actually quite difficult to fabricate a completely light tight enclosure. These light leaks may be caused by inadequate connectors, misaligned seals, pinholes in materials or even opaque materials being not quite as opaque as had been hoped [1, 2]. All light leaks must be found and rectified before the apparatus will operate correctly.

II.   Standard technique

The standard technique for detecting light leaks is to connect the anode output of the photomultiplier to an oscilloscope and observe the signal with the timebase set so that some hundreds of microseconds are visible. If the signal has a “grassy” look with many pulses visible, then a light leak can be suspected. If this signal changes as the room lights are switched on or off, then a light leak is probable.

This is essentially the technique described by Leo [2], who says that signals from detectors with light leaks appear “unusually intense and noisy”. This advice was valid in the days of analog oscilloscopes, where a change of event rate resulted in a change in the brightness of the trace. With modern digital oscilloscopes, rate changes are not so obvious.

A better approach is to use a photon counting technique. The photomultiplier output is coupled to a discriminator set to trigger at the single photon level. The output from this is then counted over a suitable period (one or ten seconds, possibly more) with a counter-timer module. The counts rate should have a Poisson distribution, so successive counts should differ from each other by typically the square root of the count. If switching the room lights affects this count rate, a light leak is certain.

Identifying the exact point in the apparatus at which the light leak occurs is more of a problem. Many hours can be spent with torches, black tape and blackout cloth, alternately covering and uncovering areas of the detector system in the search for pinhole leaks. At each point, the count period must elapse and the count be visually checked against the previous, with the square root error computed by mental arithmetic. The experimenter must constantly switch from looking at the apparatus and observing the readout of the ratemeter.

III.   Aural technique

A more rapid aural technique is suggested which can easily be implemented with standard units commonly available in laboratories. It is modelled on the use of a sounder in a traditional geiger counter. As before, the output from the photomultiplier is coupled to a discriminator set to trigger at the single photon level. This is passed to a gate-delay generator and the individual discriminator output pulses are widened, to render them audible.

The gate-delay generator should be chosen such that it has a TTL-output. This output is fed to an audio annunciator which can easily be made by soldering a piezoelectric sounder or loudspeaker element to a connector with a short coaxial cable. The arrangement is shown in figure 1. A suitable sounder is the Pro-Signal ABI-003-RC which is widely available, alternatively, piezo sounders scavenged from defunct computers or smoke detectors also work. If a suitable gate-delay generator is not available, a simple circuit using a TTL monostable such as the 74LS221 can easily be constructed.

In use, the speaker produces a rushing sound, the pitch of which is proportional to the light level. As the intensity of the light changes, so the sound alters. The effect is similar to using a geiger counter to detect radioactivity. In order to optimise performance, the photomultiplier should be operated with minimum signal present. This clearly depends on the exact experimental setup.

By working in darkness and illuminating the apparatus with blue or violet LED torches, or even laser pointers, the exact location of leaks can be found in a small fraction of the time that the standard technique takes. The user does not have to wait for time periods to elapse or to continuously observe an oscilloscope or counter-timer display.

The choice of pulse width is determined experimentally. The author has found that with small photomultipliers, $500\mu$s is a good starting point. If too short a value is chosen, then the sound will be beyond human hearing, while if too long a value is chosen, then many pulses are missed.

Note should be taken that with severe light leaks, the illumination may be intense enough to saturate the photomultiplier, or the event rate might become so high that base-line shift occurs and the single-photon discriminator no longer triggers. In both cases the sound stops. Cutting the illumination should restart the sound.

After the light leak has been located and repaired, the photon counting technique using the counter timer should be repeated with a long count period to ensure that the detector system is completely light tight.