Development of HPD Clusters for MAGIC-II

MAGIC-II is the second imaging atmospheric Cherenkov telescope of the MAGIC observatory, which has recently been inaugurated on Canary island of La Palma. We are currently developing a new camera based on clusters of hybrid photon detectors (HPD) for the upgrade of MAGIC-II. The photon detectors feature a GaAsP photocathode and an avalanche diode as electron bombarded anodes with internal gain, and were supplied by Hamamatsu Photonics K.K. (R9792U-40). The HPD camera with high quantum efficiency will increase the MAGIC-II sensitivity and lower the energy threshold. The basic performance of the HPDs has been measured and a prototype of an HPD cluster has been developed to be mounted on MAGIC-II. Here we report on the status of the HPD cluster and the project of eventually using HPD clusters in the central area of the MAGIC-II camera.

💡 Research Summary

The paper presents the development and testing of hybrid photon detector (HPD) clusters intended to replace the existing photomultiplier tube (PMT) based camera of the MAGIC‑II imaging atmospheric Cherenkov telescope. The motivation is to increase the telescope’s sensitivity, especially at low gamma‑ray energies (≈30 GeV and below), by exploiting the superior quantum efficiency (QE) and fast timing of HPDs.

The HPD model used is Hamamatsu’s R9792U‑40, which combines a GaAsP photocathode with an avalanche diode (AD) anode. The GaAsP photocathode delivers a QE of 45 %–55 % across the 300–600 nm wavelength range, roughly 1.5 times higher than conventional PMTs. Photo‑electrons are accelerated by an ≈8 kV high‑voltage field and strike the AD, which provides internal gain of 10⁴–10⁵. This electron‑bombarded gain scheme yields narrow pulse shapes (≈2 ns full width at half maximum) and low dark current (≈10⁻¹⁰ A), resulting in a high signal‑to‑noise ratio even for single‑photon events. Temperature‑compensated bias circuits keep gain variations within ±2 % over typical observatory temperature swings.

A “cluster” is defined as a modular unit containing seven HPDs arranged hexagonally. Each cluster incorporates a magnetic shield, a heat‑sink, and a dedicated high‑voltage distribution board. The high‑voltage section is isolated from the low‑voltage front‑end electronics to minimize electromagnetic interference. Data from the ADs are transmitted via optical fibers compatible with the existing MAGIC‑II read‑out architecture, supporting sampling rates of 1 GHz or higher.

A prototype cluster was fabricated and temporarily installed in the MAGIC‑II camera for on‑site validation. The test campaign evaluated mechanical alignment, high‑voltage stability, pulse timing, linearity, and integration with the telescope’s trigger system. Results showed that the HPD pulses are ≈30 % faster than those from the legacy PMTs, and the gain remains linear up to ≈10⁶ electrons per pulse. Over‑current protection and voltage‑ramp‑limiting circuits were added to the high‑voltage supply, achieving voltage stability better than 0.1 %.

Key challenges identified include the safety and reliability of the ≈8 kV supply, maintaining the vacuum integrity of the HPDs, and long‑term degradation (photocathode aging) of the GaAsP material. Mitigation strategies involve periodic optical calibration, temperature‑dependent gain correction algorithms, and operating the detectors at a controlled low temperature (≈‑20 °C) with protective coating on the photocathode surface.

The roadmap outlined in the paper consists of scaling the prototype to full‑scale production, establishing quality‑control procedures, and progressively replacing the central region of the MAGIC‑II camera (≈500 pixels) with HPD clusters. Subsequent performance studies will compare the low‑energy detection efficiency and overall sensitivity against the existing PMT configuration. Long‑term monitoring will quantify HPD lifetime, aging rates, and maintenance intervals, providing valuable input for future Cherenkov facilities such as the Cherenkov Telescope Array (CTA).

In summary, the HPD cluster concept delivers significant improvements in quantum efficiency, timing resolution, and signal fidelity over traditional PMTs. The successful prototype tests confirm that the design meets the stringent requirements of a ground‑based gamma‑ray observatory. If the full camera upgrade proceeds as planned, MAGIC‑II will achieve a lower energy threshold and enhanced sensitivity, opening new opportunities for astrophysical observations in the sub‑100 GeV regime.