From: Michal Koziel Date: Thu, 27 Feb 2014 10:13:07 +0000 (+0100) Subject: After M.Deveaux corrections X-Git-Url: https://jspc29.x-matter.uni-frankfurt.de/git/?a=commitdiff_plain;h=34e93dee12e98c33b7bb28db08aee423a61e6b77;p=reports.git After M.Deveaux corrections --- diff --git a/GSI_2014_MK_ProbeTests/MK_ProbeTests.tex b/GSI_2014_MK_ProbeTests/MK_ProbeTests.tex index d75e7c7..68e8e87 100644 --- a/GSI_2014_MK_ProbeTests/MK_ProbeTests.tex +++ b/GSI_2014_MK_ProbeTests/MK_ProbeTests.tex @@ -26,17 +26,19 @@ -The Compressed Baryonic Matter experiment installed at the future FAIR facility will be equipped with a high-precision micro-vertex detector aiming at an outstanding primary and secondary vertex resolution. Highly granular, ultra-low material budget sensors, so-called Monolithic Active Pixel Sensors, manufactured at standard CMOS process, will be employed. Imperfections in CMOS process as well as further dicing and thinning procedures may reduce the final yield of sensors to be mounted in the detector stations to about 60-70$\%$. To select sensors with the best characteristics, probe-testing prior to integration is mandatory. +%The Compressed Baryonic Matter experiment installed at the future FAIR facility will be equipped with a high-precision micro-vertex detector aiming at an outstanding primary and secondary vertex resolution. Highly granular, ultra-low material budget sensors, so-called Monolithic Active Pixel Sensors, manufactured at standard CMOS process, will be employed. Imperfections in CMOS process as well as further dicing and thinning procedures may reduce the final yield of sensors to be mounted in the detector stations to about 60-70$\%$. To select sensors with the best characteristics, probe-testing prior to integration is mandatory. -Such probe-tests were addressed at IKF with the CMOS sensor MIMOSA-26. The motivation to use this sensor is that the readout system \cite{1} for MIMOSA-26 already exists and it can be easily adapted to work with probe-card. Moreover, the thickness of the MIMOSA-26 sensor of 50~$\upmu$m reproduces well expected mechanical properties of final MVD device where mechanical stability is of concern w.r.t. probe-testing. +The future Compressed Baryonic Matter experiment (CBM) will be equipped with a high-precision \mbox{micro-vertex} detector (MVD) aiming at an outstanding primary and secondary vertex resolution. Highly granular and \mbox{ultra-light}, so-called Monolithic Active Pixel Sensors (MAPS), which are manufactured with standard CMOS processes, will be employed. Imperfections in the CMOS production as well as the subsequent dicing and thinning procedures may limit the production yield of the sensors to about \mbox{60-70$\%$}. Probe testing the sensors prior to integrating them to the MVD is a mandatory step of the quality assurance related to the mass production of this detector. This is as it allows to recognize and reject sensors with insufficient performance. -The probe-test bench shown in figure~\ref{fig:setup} was located at the IKF clean room. The probe-card hosts 65 tungsten needles with minimum pitch of 120 $\upmu$m. Signals from sensors are routed through the probe-card PCB to so called adapter-card. The use of this extra PCB was motivated by the fact that the probe-card is fabricated by an external company (HTT-Dresden) and any "customization" is expensive, while an adapter-card we can design ourselves and implemented any kind of useful electronics and logic if needed. In addition, a test board hosting a wire-bonded working MIMOSA-26 sensor was manufactured. This board is used to debug all readout chain including probe-card but not sensors connected with needles, where problem with contact is of concern. To hold an ultra-thin sensor, a probe-station chuck-adapter with micro-vacuum channels was manufactured. +The feasibility of probe testing the only \mbox{$50~\rm \upmu m$} thick sensors was studied with the MIMOSA-26 prototype, which is considered as a realistic precursor of the final sensor of the MVD. Moreover, the existing readout system of the \mbox{MVD-prototype} \cite{1} could be used for the test. -After debugging the probe-test bench with the adapter card, the first tests with 300~$\upmu$m MIMOSA-26 sensors were addressed. The sensor response to JTAG programming sequence as well as to various threshold setting resulted in the expected data output. Next, the 50~$\upmu$m sensors were probe-tested. Here, we noticed that an overdrive allowing for proper connection between probe-card needles and sensor pads was higher than in a case of tests with thicker sensors. This was attributed to the fact that 50~$\upmu$m sensors become soft enough to adapt the surface to imperfection in the support material. Nevertheless, it was also possible to probe-test successfully the thinned sensors. +The probe-tests were carried in the IKF technology lab. As shown in Figure~\ref{fig:setup}, the probe tester was equipped with a \mbox{probe-card} hosting 65 tungsten needles with a minimum pitch of 120 $\upmu$m. The MIMOSA-26 sensors were held by a chuck adapter with micro-vacuum channels, contacted with the needles and their signals were routed through the probe-card to a so-called adapter-card. The latter was introduced as building it came out to be easier and cheaper than adapting the probe-card itself to our readout system. In addition, a test board hosting a wire-bonded working MIMOSA-26 sensor was manufactured. It is to test the readout chain including the probe-card independently of the delicate issue of contacting of sensors with needles. -Currently, we are working on accessing dedicated test modes of the MIMOSA-26 sensor and implementing routines allowing for more adequate and precise sensor characterization. The currently used DAQ system (TRBv2 based) will be migrated to a TRBv3 platform and some of readout chain components will be upgraded to the newest version \cite{4}. +After commissioning the system, first tests with 300~$\upmu$m thick MIMOSA-26 sensors were carried out. The \mbox{response} of the sensors to JTAG programming and to various \mbox{threshold} settings matched our expectations. Next, the 50~$\upmu$m sensors were probe-tested. The distance between the probe-card and the sensor, which is needed to establish a save contact between sensors and needles, was found smaller than expected from the tests with thicker sensors. This is attributed to the fact that the 50~$\upmu$m sensors become soft enough to follow the imperfections of the surface of their support. Nevertheless, we succeeded to probe-test the thinned sensors. -The 50~$\upmu$m-thin precursors of a final CBM-MVD sensors were successfully tested with a probe-card. The activity reported here is also of interest for a much larger community being at the stage of employing thinned MIMOSA-26 sensors into experimental setups, i.e. PLUME project and an upgrade of the NA-61 experiment at CERN. +In a next step, we are working on implementing a full test protocol, which is suited for testing the sizable number of sensors foreseen in the future mass production. To do so, we intend to update the software of our readout system and to migrate this system from the current TRBv2 to a TRBv3 platform \cite{4}. + +Concluding, a 50~$\upmu$m-thin precursor of a final \mbox{CBM-MVD} sensors was successfully tested with a probe-card. This activity and the related know-how found meanwhile some interest of a larger community, which is employing thinned MIMOSA-26 sensors into various experimental setups. This includes the PLUME project and the vertex detector project of the NA-61 experiment at CERN. @@ -54,7 +56,7 @@ The 50~$\upmu$m-thin precursors of a final CBM-MVD sensors were successfully tes \begin{figure}[htb] \centering \includegraphics*[width=75mm]{setup.eps} -\caption{Probe-station setup: (1) microscope lens, (2) adapter-card, (3) connectivity to DAQ, (4) probe-card, (5) test-board hosting the reference sensor and (6) chuck-adapter with one Mimosa-26 sensor.} +\caption{Probe-station setup: (1) microscope lens, (2) adapter-card, (3) connectivity to DAQ, (4) \mbox{probe-card}, (5) test-board hosting the reference sensor and (6) \mbox{chuck-adapter} with one Mimosa-26 sensor.} \label{fig:setup} \end{figure}