5. Our hardware and software contribution to ATLAS

  • Our main obligation on the ATLAS is the proposal, development, testing and maintenance of the on-line calibration system for HEC calorimeters.

    The proposal and the development of the on-line calibration system for HEC calorimeter is one of the crucial tasks.

    In order to achieve an optimal energy resolution of the calorimeters it is necessary to develop a system to calibrate the gain variations in the $ \sim$190000 readout channels. Each individual channel consists of a preamplifier, a shaper and an analog to digital converter. The calibration system has to provide reference currents which are sent into the current sensitive preamplifiers. The shape of the current signal has to resemble the real calorimeter signals as much as possible, with a rise time of $ \sim$1 ns. Since the response of the electronics has to be known over the full dynamic range, the calibration system has to provide currents in a range spanning five decades from 200 nA (lower value in the HEC) up to 20 mA (corresponding current produced by electrons with an energy of 2 TeV in the electromagnetic calorimeter). The linearity of the system is required to be better than 0.1% over the full dynamic range.

    The first prototype of a calibration pulse generator was developed by our institute. We used a specific integrated circuit (ASIC) (CMOS technology with a structure length of 1.2 $ \mu$m). Our hardware was used by Mainz group for the development of the next calibration pulse generator model afterwards. It achieved the required characteristics because one of the main building blocks of this ASIC chip is a differential amplifier with automatic offset compensation. This automatic offset compensation is essential to achieve the required linearity over the full dynamic range. To achieve the linearity of 0.1% an offset value below 20  $ \mu$V has to be reached and has to be stable under operation. This also implies that it is not affeced by the radiation which is induced during the ATLAS operation. Over a period of ten years, for example, the calibration electronics is exposed by an integrated neutron flux of $ \sim$1013 n/cm2 with an equivalent energy of 1 MeV (Si) and a gamma dose of 200 Gy.

    The calibration system was developed and tested several time - in 1999-2002 - combined EMEC/HEC, and in 2003 combined EMEC/HEC/FCAL and HEC/FCAL tests. In Fig. 1 you can see the calibration signal (left) and the prediction for the ionisation signal (right) together with the residuals with the respects to the fit (lower figures).

    Figure 1: The calibration signal (left) and the prediction for the ionisation signal (right) together with the residuals with the respects to the fit (lower figures)

    In 1998 the Cleland method for the noise suppression in the calorimeter readout was successfully applied by our group. The improvement of the signal/noise ratio can be seen in Fig. 2.

    Figure 2: The noise reduction for 100 GeV electron deposits in HEC module 0 a) before and b) after filter applying

    In Fig. 3 is shown the linearity in the four different impact points in the test modules with the points corresponding to uncalibration and calibration data for the various electrons energies. The improvement is markant.

    Figure 3: Response linearity of module 0 in four different impact points. The full points - before and, triangle - after calibration

    The new function describing the response of the whole testing electronics for various signal's types (e..g. calibration one or signal from particle) had been calculated and implemented, which also using less number of free parameters the measured signals describes very well.

  • We participate also on the design and laboratory tests of the Front End Board (FEB) for Atlas experiment. The FEB, see Fig. 4,

    Figure 4: The view on the FEB

    is going to be used for the EM and HEC part of the Atlas calorimeter. As described in the TDR of the liquid argon (LAr) calorimeter, the FEBs contain the electronics for amplifying, shaping, sampling, pipeling, and digitising the LAr calorimeter signals. The FEB electronics e.g. must handle the signal dynamic range of about 16 bits without contributing more than 0.2%.

    On the Fig 5 the FEB test setup at Buld. 180 is shown.

    Figure 5: The FEB test setup

  • We have produced also a set of print boards with the different patterns for the wiring of the final modules of HEC. Parameters of the board divided a calibration signal on 3 pads are crucial for the accuracy calibration signal distribution. A dividing ratio of the front calibration distribution board for all channels is shown in figure Fig. 6.

    Figure 6: Dividing ratio of the Front Module       Calibration Distribution Boards of HEC

  • We finished HV and signal boards production. All these elements are the parts of HEC modules and this year also the first of four large wheels of the HEC calorimeter has been successfully assembled and rotated into its final position on schedule - see Fig. 7. To assemble a wheel have been taken modules that have already been cold tested, do the final electrical testing and locate them onto the HEC wheel assembly table. All mechanical and electrical tests have been successfully passed, it means also Košice elements are satisfied.

    Figure 7: End-Cap Cryostat after the insertion of the second HEC wheel (Dec. 2003), and before the insertion of the forward calorimeter

  • In period 2001-2002 we began start the filter box production, see Fig. 8,

    Figure 8: Filter box

    all needed quality criteria.

  • The filter boxes will be used in all ATLAS calorimeters, e.g. on the Fig. 9

    Figure 9: ATLAS barrel cryostat

    the ATLAS barrel cryostat with our filter boxes (see Fig. 10 in more detail) is shown.

    Figure 10: Detail view: the filter box in the ATLAS barrel cryostat

  • Warm and cold HEC wheels (A, and C) testing. We would like to stress that our team took participation on the all warm and cold tests for both HEC (A, and C) wheels. On the Fig. 11 the example of the test data analyse for End-cap wheel A is shown.

    Figure 11: Delay of calibration signals fro HEC-A

    Now on the detector side, tremendous progress has been done: in building 180 for liquid argon (LAr) HEC (for both two wheels A, and C) the cooling, testing and after than the transport to point 1 in anticipation of its imminent descent into the pit has been done also. See also Fig. 12:

    Figure 12: Transport to point 1 of the End-Cap wheel C.

    On the Fig. 13. the view into ATLAS pit with the Endcap wheels C is shown (November 2005).

    Figure 13: End-Cap wheel C in ATLAS pite (November 2005)