--- /dev/null
+\documentclass[a4paper]{JACoW-GSI}
+%\documentclass[12pt,a4paper,oneside,ngerman]{article}
+\usepackage{graphicx}
+\usepackage{url}
+
+%%
+%% VARIABLE HEIGHT FOR THE TITLE BOX (default 35mm)
+%%
+
+\setlength{\titleblockheight}{27mm}
+
+\begin{document}
+\title{Analysis of the CBM MVD prototype CERN SPS beam-time\thanks{supported by BMBF (05P12RFFC7), HIC for FAIR and GSI}}
+
+\author[1]{S Amar-Youcef}
+\author[1]{K. ..}
+
+\affil[1]{Goethe-University, Frankfurt, Germany}
+
+\maketitle
+
+\section{Introduction}
+As part of the development the MVD prototype was commissioned and extensively tested at a beam time at CERN SPS.
+In doing so the performance of the sensors along with the customized frond-end and TRB based DAQ system was investigated
+with regard to the synchronization and long-term stability of the continuous multi-channel readout, the positional stability and the sensor performance (efficiency, resolution, fake hit rate).
+As sensors MIMOSA-26 was used, which is a sizable precursor to the final sensor.
+It comprises the prototype readout concept including on-chip hit discrimination and data sparsification.
+Further, the sensor is based on a high-resistivity epitaxy layer and was thinned down all up to a thickness of 50\ $\mu$m.
+
+Here, the results related to the analysis of the detection performance are discussed.
+A discussion related to the DAQ can be found in \cite{BM, BM2} and related to the mechanics in \cite{TT}.
+It should be shared that related to the synchronization and stability of the DAQ no reservation was diagnosed.
+However, at highest occupancies and/or very low thresholds fluctuations in the supply voltage of the sensors were observed, which led to a bad sensor performance.
+%The stability of the DAQ remained unaffected.
+
+\section{Setup}
+Altogether 12 identical sensors were arranged in four single-sided reference and one double-sided DUT station,
+whereas ``one side'' denotes two sensors side by side.
+According to this the beam particles were tracked by up to 6 hit points.
+Further the setup permitted to vary the relative arrangement of the stations and the relative inclination of the DUT,
+which was positioned between two reference stations upstream and two downstream for best arm of lever.
+In addition the temperature of the DUT could be varied independent of the temperature of the reference system.
+The material budget amounts to 0.05\%\ $X_{0}$ for each of the reference stations and 0.3\%\ $X_{0}$ for the DUT.
+The DUT station corresponds to one fourth of a first MVD station covering an area of up to $\sim$\ 8\ cm$^{2}$.
+
+\section{Results}
+A diversity of measurements at different discriminating thresholds, inclination angles and temperatures were carried out.
+% to determine the performance of the MVD prototype.
+In total 1.2\ TB significant data was recorded.
+The data was analyzed by an internal custom software, which was developed following the development process of the MVD prototype.
+The software performs data consistency checks, unpacking and the high level analysis including quick overview functions and alignment.
+
+In fig. \ref{fig:ShiftSpillTime} and \ref{fig:EffVsFakes} the excellent performance of the MVD prototype should be pointed out.
+Fig. \ref{fig:ShiftSpillTime} shows the temporal course of the alignment of the DUT relative to the reference system.
+A changing of $0.003\ \mu m/s$ could be resolved.
+The single point resolution of the DUT including the resolution of the sensor and the positional stability of the mechanics was found to be about $3.6\ \mu m$.
+This is equivalent to the resolution of the sensor only.
+Fig. \ref{fig:EffVsFakes} shows the efficiency-fake hit properties of the two DUT sensors.
+%For the experiment it is necessary to reduce the number of noisy pixels to a minimum at maximum efficiency in order to reduce the data throughput and ambiguities in the tracking.
+In the figure the two approaches masking of noisy pixels and increasing the threshold are confronted.
+The sensors allow to operate at the desired operating point of above $99\%$ efficiency and below $10^{-5}$ hits per pixel and event fake rate.
+For simplicity reasons only the matrix 1 of both sensors is shown.
+
+\begin{figure}[h]
+ \centering
+ \includegraphics[width=0.335\textwidth]{ShiftSpillTime1.pdf}
+ \caption{Changing of the translational alignment during one run. Each data point corresponds to one spill.}
+ \label{fig:ShiftSpillTime}
+\end{figure}
+
+
+%\begin{figure}[h]
+% \centering
+% \includegraphics[width=0.5\textwidth]{DUTres.pdf}
+% \caption{ Residuals in x and y direction of the two DUT stations referencing each other. (120\,GeV pions, RUN043)}
+% \label{fig:DUTres}
+%\end{figure}
+
+\begin{figure}[h]
+ \centering
+ \includegraphics[width=0.335\textwidth]{EffVsFakes.pdf}
+ \caption{Efficiency as a function of the fake hit rate for two different DUT sensors.
+ %Solid: different threshold settings; dashed: masking of noisy pixels for one threshold setting
+ }
+ \label{fig:EffVsFakes}
+\end{figure}
+
+
+\begin{thebibliography}{9} % Use for 1-9 references
+%\begin{thebibliography}{99} % Use for 10-99 references
+
+\bibitem{BM}
+B. Milanovic: ``Beamtime Results of the MVD Prototype DAQ Network'', GSI Scientific Report 2012
+\bibitem{BM2}
+B. Milanovic: ``...'', this report
+\bibitem{TT}
+T. Tischler: ``...'', this report
+
+\end{thebibliography}
+
+
+
+
+
+\end{document}
jspc29 / reports.git / commitdiff