The HERA-B Vertex Detector System

The Vertex Detector System (VDS) of the HERA-B detector is designed and built by the
Max-Planck-Institute for Nuclear Physics at Heidelberg and the Max-Planck-Institute for Physics at Munich. The characteristic parameters of the VDS are summarized in a table. While the entire detector environment, including the vacuum vessel, the roman pots for the detector modules and the electronic readout system, is produced in Heidelberg the Munich group does the design and the production of detectors and modules carrying the detector wafer and the frontend electronics.

Introduction to the Vertex Detector System

Diploma and Doctoral Theses
Extracts of Annual Reports
Conference Contributions
Publications & Preprints

Internal Notes


HEPIC | HEP Links | PDG | Preprint Server | Google | DESY | HERA | HERA-B : Runlog / Shiftlog / Run Statistics |

Introduction to the HERA-B Vertex Detector System

Layout | Challenges | Detectors | Electronic Readout | Hardware Setup | Status

The HERA-B experiment - described in more detail in its Proposal and Design Report - is aiming at the detection of CP violation in B meson decays with particular emphasis to the ''gold-plated'' B0 -> J/PsiK0_s decay mode where most selective triggering on the lepton pairs from J/Psi decays is feasible.
The B mesons are produced at a relatively large rate by interactions of the 820 GeV protons stored in the HERA storage ring with internal wire targets. Their decay products are detected, momentum-analyzed and identified in an advanced multiparticle spectrometer.
The purpose of the Vertex Detector System (VDS) is to provide the track coordinates for reconstructing the J/Psi -> e+e-, +- decay vertices and the impact parameters of all tagging particles. The required vertex resolution amounts to about 10% of the average (10 mm) B decay length, resp. 20 to 30 m in transverse direction. Even better resolution is desirable for a measurement of the much faster B_s mixing.
The information of the vertex detector will be used first at the second trigger level in order to reduce data flow by rejecting background. One ultimately requires that the two leptons of the B candidate form a common vertex which is consistent with the B trajectory and displaced from the target wire.



The VDS is located between the target wires (z= -55 mm and 0 mm) and the spectrometer magnet at z = 2.3 m. It comprises 7 superlayers of detector planes, which - typical for a forward microvertex spectrometer - are arranged perpendicularly to the beam axis. Its angular coverage is consistent with the magnetic spectrometer acceptance, 10 mrad to 160 mrad horizontally, and 10 mrad to 250 mrad vertically; in the center of mass system that coverage corresponds to 90% of 4pi. Side and perspective views of the geometrical detector arrangement are shown separately.
A superlayer is segmented into quadrants each of which consists of two double-sided Silicon microstrip detectors providing 4 views in total, 2.5° and 90°2.5°. Each microstrip detector has a sensitive area of 50 x 70 mm^2, such that it can be cut from a 4" wafer. The segmentation allows for two different detector arrangements within one plane which will be exploited to stagger the support structures of subsequent superlayers, and to rotate detector positions in order to distribute the huge radiation load over a larger detector area. During fills of the HERA storage ring the detector elements will be retracted to safe positions. Click here to to see the arrangement of the detector elements in a superlayer.
The performance of the design has been studied by extensive computer model calculations assuming a spatial detector resolution of 12 m which can be achieved with microstrip detectors of 25 m diode strip pitch and 50 m readout pitch. For J/Psi muons the impact parameter distributions are predicted to peak at 20 m and the B vertex resolution is sigma_z = 0.5 mm. The geometrical efficiencies for identified J/Psi leptons and tag leptons are greater than 96%.



The requirements on the VDS imply challenges for the design of detectors, electronic readout, as well as for the mechanical and thermal engineering. Continuous R&D efforts are devoted to approach the optimum solution with respect to performance and cost effectiveness. The various topics include:
  • Design and operation of a VDS that is integral part of a storage ring with detectors to be positioned within the clearance needed for beam injection.
  • Efficient yet low-mass RF shield for the proton beam inside the large vacuum vessel of the VDS.
  • Detectors that tolerate a flux of minimum ionizing particles, mostly pions, of 30 MHz/cm^2 and corresponding 'annual' fluences of 3x10^14 per cm^2.
  • Frontend ASICs with integrated low noise amplifiers, analog pipeline memory and dead-timeless readout.
  • Online processing of the VDS data flow of 8 Gbyte/s or more.
  • Efficient yet low-mass cooling of detectors and readout chips which are positioned in vacuo within the geometrical acceptance region.
  • Rigid yet low-mass support structure for the detectors.
    You can find more in-depth information about these topics in our thesis , annual report , conference contribution , and publication sections.



    The baseline design assumes AC coupled double-sided silicon detectors of 50x70 mm^2 active area, 280 m thickness, and strip resp. readout pitches of typically 25 m and 50 m. Single-sided detectors are considered to be a backup solution if radiation damage issues would justify their use. The detectors have been designed by the semiconductor laboratory of the MPI Munich and they are produced both there as well as by commercial companies. The special design of those detectors allows operation at reverse bias voltages exceeding 300 V.
    Since radiation damage is expected to limit the lifetime of silicon detectors to one year of HERA-B operation, alternative detectors of higher radiation hardness are desirable. Recognizing the potential of diamond detectors, members of the VDS groups have joined the R&D efforts of the RD42 diamond tracker group.


    Electronic Readout

    The requirements on the VDS detector readout closely ressemble those of the LHC tracking detectors, and it is no coincidence that the HERA-B VDS is using the architecture developed by the RD20 group that has been also adopted by the CMS detector. It is based on a relatively slow low noise amplifier of 50 to 75 ns peaking time. Detector signals are sampled at the bunch crossing frequency of 10.6 MHz and stored in a 128 cell deep analog pipeline to await the Level-1 trigger decision. In the event of a Level-1 trigger, the appropriate analog samples of 2x128 detector channels are multiplexed to one serial output line. Occuring concurrently with data sampling, readout at up to 40 MHz is practically dead-timeless. Prototype chips, HELIX and its support chip SUFIX, processing 128 input channels have been implemented at the Heidelberg ASIC laboratory for a standard CMOS process. The time-multiplexed output signals of the frontend chips are transmitted via analog optical fiber links to the counting room for digitization in VDS specific front-end driver modules. The interface to the detector wide data acquisition system is common to all subdetectors and represented by the 'SHARC boards' each of which is equipped with six ADSP 21060 digital signal processors including memory for the second level buffers. In case of the VDS, the task of these digital signal processors will be to do on-line the pedestal subtraction, common baseline shift correction, cluster finding, data sparsification formatting for a total of about 150.000 detector channels.


    Hardware Setup

    The detector system including the target wire assemblies are contained in a vacuum vessel with an exit window and an integrated tapering beam pipe for the rest of the HERA-B experiment at the one end and a connection to the standard beam line system at the other. The overall length of the vessel is about 2.5 m and its maximum radius is 58 cm. The exit window is kept as thin as possible, i.e. about 3 mm if fabricated out of aluminium. The silicon wafers are maintained at a secondary vacuum of 10^{-6} mbar and the main stainless steel vessel at 10^{-8} mbar.
    Each quadrant of a superlayer is contained in a removable pot assembly that can be displaced individually in radial and lateral sense to the beam by an external motorized mobile bearing unit. The wafers of the three superlayers next to the target wires are contained in a single pot. A 125 m thick aluminium shielding cap separates secondary and primary vacua and serves as protection against RF interference from the beam. Detectors and readout chips are cooled via separate cooling paths connected to a cooling block located outside the acceptance cone. Heat drains are carbon composite materials of high thermal conductivity (~1000 W/m/K) and large radiation lengths (~30 cm). Positions of pots along each quadrant row will be continously monitored by using collimated laser beams as alignment references and semi-transparent optical position sensors being attached to the pots. At an active area of 20x20 mm^2, these sensors are found to deliver both the x and y coordinates with a precision on the order of 1 m.



    The HERA-B experiment has been fully approved by DESY in February 1995, and the detector is expected to be completed and commissioned in 1998. To gain early experience, some detector components were already installed during the 1995/96 shutdown period including the spectrometer magnet, parts of the muon system and prototypes of the electromagnetic calorimeter counters, of the transition radiation detectors and the tracking chambers (see A.Spengler, NIM A384 (1996) 106 for more details) . As to the VDS, the 1995/96 installation included
  • the 2.6 m long stainless steel vacuum vessel equipped with manipulators
  • a preliminary version of the RF shield ('96 & '97 version)
  • prototype primary and secondary vacuum systems, and
  • a protoype laser reference system.
    With three prototype silicon detector modules installed in the vessel extensive test measurements have been performed throughout the 1996 run. The results obtained from this running demonstrate both the full functionality of the present setup as well as the power of the VDS as a diagnostic tool for the HERA-B wire targets.
    More details about VDS components and status are found in our annual reports and in the transparency collections of the recent VDS plenary talks.


    June 1997, K.T.Knpfle (