GRID-02

GRID-02 is part of GRID, a constellation of transient gamma-ray detectors. It was launched on 06/Nov/2020 at an orbit of 457 × 467 km 97.2$^{\circ}$ LEO.

• GRID-02 Paper (Ground)
• GRID-02 Paper (In-Orbit)
• Presentation

In addition to the optimization of the GRID detector, several functions are specifically designed for performance studies of SiPMs. The current, bias voltage and temperature of the SiPMs are monitored. A charge injection module is used to analyze the noise contributions.

The detector is subdivided into 4 channels, later referred to as CH0 - CH3. Each channel is comprised of of a scintillator crystal and a 4 × 4 array of MicroFJ-60035-TSV1 SiPMs. Furthermore, each channel has its own bias voltage power supply, readout electronics and dedicated SiPM characterization circuits.

Every second, during scientific observations, the bias voltage, current and temperature are recorded as housekeeping data. They are later used to analyze the SiPM dark current. A special trigger initiates IV measurements of the SiPM at 40 different bias voltages (above and below V$_{bd}$).

The temperature dependance of the breakdown voltage (V$_{bd}$) is fitted with a linear function. It's gradient (or temperature coefficient) $k_b$ is found to be 21.5mV/K. As the figure to the right shows, after applying the temperature correction, no significant features are observed. V$_{bd}$ therefore is minimally affected by radiation damage (more time in space = more radiation damage).

The SiPM dark current is expressed through $I_{dark}=DCR*C_{pix}*(V_{bias} − V_{bd}) · ECF$ (see paper for more information), where the DCR is expressed by the Field-Enhanced Shockley-Read-Hall (FE-SRH) model. For a constant bias voltage, the effective electric field strength is thought to remain constant. The dark current is therefore fitted with respect to temperature (on a daily basis) to the aforementioned model where the electric field strength is a free parameter. After the fit, the expected dark current at a reference temperature of $5^{^\circ}C$ is shown with respect to time. This is shown in the figure (to the right) where the dark current increases by $\approx 93/96/98/110 \frac{\mu A}{year}$.

Charge injections are used to look into the energy resolutions and noise contributions within the detector. Results in ADC channels are subsequently converted to equivalent energy with a scale factor ~0.027 keV/channel. When the bias voltage is switched off the width of the charge injection peak is heavily dominated by the electronics noise (irrespective of radiation damage). However, when the bias voltage is switched on the dark count noise becomes significant. Through Campbell’s theorem, the total noise can be correlated and fitted with respect to the dark current. This is shown in the figure below. The results show a noise increase of ~7.5 keV/year for each channel.