MANTIS

MANTIS

The Monopix-bAsed Next generation Tracking Instrumentation System (MANTIS) provides a graphical user interface (GUI) for the slow control of the telescope in a test beam environment. The graphical user interface is designed to be user friendly, so that inexperienced students can operate the telescope. The figure shows the components of the slow control. The TJ-Monopix2 data acquisition software is presented in yellow. The slow control uses the existing software to control the hardware components. The hardware components are the detector chips, the BDAQ53-boards and the power supply devices, shown in light green. Power supply devices can be controlled using the basil framework, which eases expansion to support a wide range of different device types. The MANTIS slow control is shown in teal. It is the link between all components and allows full remote communication with the detectors including scan control and live visualization of the accumulated hitmap as well as live correlation.

The graphical user interface has been implemented using PyQt6. The different functionalities are separated into designated classes to improve readability and modifiability. For full functionality the BDAQ53-boards have to be connected to a computer via an RJ45-cable, the power supplies are connected via USB. The connection details must be specified in the user_configuration.toml. The GUI can then be started from a terminal using the shortcut command mantisgui. The software consists of multiple pages that are accessible via tabs. The first page provides control of the power supplies. The configuration can be edited in GUI. The scans are started from the scan control. A visualization of the scan data, including hitmaps and live correlation of two selected detectors, is provided through the online monitor.

The state machine presents the flexibility provided by the GUI. For example it is possible to skip the configuration step and just start one scan after another and it is possible to edit the configuration between scans but skip powering, as long as the powering setup does not change. To handle multiple BDAQ53-boards multiple instances of the GUI can be started, the BDAQ53- board connection has to be specified in the configuration.

Power Manager

The power manager controls all necessary power supplies used to power BDAQ53-boards and TJ-Monopix2 chips in remote control. It allows full control over the settings of the power supplies while running, for example when access is restricted while measuring in a test beam. It displays all live values (current and voltage) at the sidebar visible from all pages of the GUI. The power supplies can be added and edited in the user_configuration.toml. Implemented power supply types are Keithley2400 SourceMeter, Keithley2410 SourceMeter and TTi QL355TP power supply LXI. The base for this is the basil framework developed by the Silicon Lab Bonn (SiLab). The framework minimizes the necessary code modifications for the addition of more power supply types to the setup and GUI.

Configuration

The configuration is split in two .toml files to improve the readability for the user. The fixed_configuration.toml includes usually fixed settings that are needed for the trigger logic unit, the time-to-digital-converter, conversions, calibrations and analysis. The second file user_configuration.toml includes all the user specific settings like detector information, power devices and scan parameters. The configuration can be opened and modified in the graphical user interface. After saving it, the last modified user_configuration.toml will automatically be applied at the configuration step in scan control. This updates not only the scan parameters but also the power manager and the online monitor. Both configuration files are combined automatically when they are being applied.

Scan Control

The scan control combines the most relevant scans of a TJ-Monopix2 chip for test beam usage in one user interface:

• Analog Scan: The analog scan injects a fixed charge into each pixel of the chip multiple times and measures the response to determine the functionality of each pixel. Noisy pixels will measure significantly more hits than the number of injections, broken or disabled pixels detect significantly less hits than the number of injections. The left figure shows an occupancy map created from an analog scan of a TJ-Monopix2 chip. The dark line in the middle indicates that multiple pixels at the border between the two main flavors are defective. The white pixels are disabled pixels. In this case the two AC-coupled flavors are disabled.

• Source Scan: The source scan is a simple measurement scan. All hits are being recorded with an individual timestamp defined by the chip and the BDAQ53-board. The hitmap of a strontium measurement using only the DC-coupled standard front-end flavor is shown in the figure on the right. Since the number of detected hits is higher towards the center of the chip the position of the strontium source can easily be extrapolated to a point near the center of the chip.

• External Trigger Scan: The external trigger scan is a measurement scan with additional synchronisation possibilities using a trigger signal from a TLU. The TLU trigger signal defines a common timestamp over multiple devices. The scan is very similar to a source scan and produces the same output plots.

Threshold scans or tunings are not included in the GUI, as the GUI mainly aims for test beam usage. Optimally tunings have already been prepared and are provided to the user. The scan type can be selected using the QComboBox provided by PyQT6. The scan procedure has been separated into three steps: Initialize, Configure and Start Scan. This allows more flexibility and spares the user from repeating the lengthy process of configuration if the scan configuration does not change. A Source Scan or an External Trigger Scan can be stopped at any time. The Analog Scan does not implement a stop signal as a fixed readout sequence is completed automatically. MANTIS provides scan information using different log levels. The output is reduced to the minimum necessary information as well as errors and warnings. At debug level all available terminal outputs will be written into the GUI. A progress bar displays the progress of the scan.

Monitoring

The Monitoring is derived from the Online Monitor developed by the SiLab. It supplies live cumulative hitmaps of each detector. This has been enhanced by adding a functionality, which allows the live visualization of the correlation of two selected detectors while scanning. This allows quick checks of the functionality of the setup including synchronization and geometry of the setup without extensive analysis of the data after a measurement. The code for the correlator has originally been inspired by the pymosa correlator. An illustration of the correlation procedure at the example of a 2D row/row correlation plot is shown in . It displays the hit positions of one detector against those of another detector. The correlator goes through the detector A hit data linearly. The detector B hit timestamps are compared to every timestamp of the detector A hits in a linear search through the data of detector B. Whenever a detector B hit timestamp falls within the timestamp window defined by the detector A timestamp: $w = (ts_A − dt, ts_A + dt)$ the 2D correlation histogram in row- (and column-) direction increases the number of correlated hits of this combination of row(/column) of detector A and row(/column) of detector B by 1 until the detector B timestamps are higher than $ts_A + dt$. Then the next detector A hit is being investigated and compared to detector B data.

That way all potentially correlated hits can be found but double correlations may be possible as shown at the example of detector B hit $B2(r_{B2}, c_{B2})$. It falls within the time window of two detector A hits and therefore increments the 2D correlation histogram at two positions. For both online monitor and online correlator masking can be applied. This allows to suppress displaying noisy pixels to enhance the contrast of the remaining hitmap. The threshold of the masking is defined by the mean hit occupancy $\mu$ and standard deviation σ of all pixels: $\text{threshold} = \mu + 3 \cdot \sigma$. The threshold is calculated at the moment masking is applied.