9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June 24-26, 2008

Transcription

1 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 Advanced front-end electronics to extend the pulse-height spectroscopy range well beyond the ADC analog input range G. PASCOVICI (), A. PULLIA (), F. ZOCCA(), D. BAZZACCO () () Institute of Nuclear Physics, University of Cologne, Germany () University of Milan, Department of Physics and INFN-Sezione di Milano, Italy () INFN-Sezione di Padova, Italy Abstract: - Using innovative front-end electronics developed for the -fold segmented High-Purity Germanium detector of AGATA Array we were able to significantly extend the range of spectroscopic measurements well beyond the fast pipeline ADC limit. To do that above a certain threshold we are automatically switching from a standard pulse height analysis to a Time-over-Threshold [TOT] method (Wilkinson like) and we thus obtain an unprecedented intrinsic dynamic range as large as 00 db. To achieve that performance the structure of the frontend electronics consists of a very low noise and very high dynamic range charge-sensitive preamplifier followed by a passive pole-zero cancellation circuit including a highly accurate Fast-Reset circuit controlled by a fast comparator and zero crossing detector. With a large AGATA detector we could extend the initial dynamic range measured with a standard pulse height spectroscopic method from kev - 0 MeV up to 80 MeV (equivalent gamma energy, measured with large pulser signal). The intrinsic energy resolution (i.e. electronic noise) is pf detector capacity. The energy resolution above the comparator threshold measured with the present TOT method is below MeV (equivalent gamma energy i.e. pulse signals with amplitudes about 0 times higher than the ADC analog input range). The measured energy resolution is in very good agreement with analytical calculation and with intercomparison measurements with normal pulse height mode only and reduced electronic gain. The new time-variant circuit technique, proposed for nuclear pulse spectroscopy, permits a substantial improvement of the energy measurement dynamic range. This technique can be directly used in many other experimental pulse spectroscopic methods where the sensor is in a first approximation an equivalent capacitor. Key-Words: - low noise gamma spectroscopy, high rates electronics, beyond ADC range Introduction A new generation of nuclear physics experiments with radioactive ion beams are foreseen in the near future based on π arrays of high purity, highly segmented HP-Ge detectors based on the novel technique of gamma ray-tracking [, ]. The Advanced Gamma Tracking Array (AGATA) will be a device of major importance for nuclear structure studies at the very limits of nuclear stability, capable of measuring gamma radiation in a large energy range (from a few tenths on kev up to more then 0 MeV) with greatest possible efficiency and very good spectral response. One of the AGATA project s aims is to provide a powerful instrument that can deal with recoil velocities, high gamma-rays background, and low cross-section while preserving high photopeak efficiency, good energy resolution and good peak-to-total ratio, i.e. good spectrum quality also for high-fold coincidences is mandatory. The use of pulse-shape analysis technique raises an additional issue to the preamplifier specification, namely a fast rise time (faster than the fastest collection time of electrons and holes in a detector) and moreover a very clean transfer function in the time domain. The good transfer function has to be preserved in cryostats equipped with multiple highly segmented detectors and real wiring. The crosstalk between channels has to be as small as possible (by design it is in the order of magnitude of ~ 0 - or less). A very large dynamic range at high counting rates is mandatory, to detect gamma-rays alone in the range up to 0-0 MeV and additionally in the presence of pile-up effects induced by some energetic charged particles, e.g. pions, kaons. Some of those charged particles can generate signals in the HP-Ge detector with very high amplitudes, equivalent to MeV gamma energies. Special care has to be taken in the design of the front-end analog electronics to avoid saturation due to pile-up effects of those signals above 0 kcps. We have developed a dedicated front-end electronics device for the central contact (core) of a highly segmented HP-Ge detector consisting of a chargesensitive preamplifier and a programmable spectroscopic pulser circuitry. To reduce substantially the dead time caused by large signals or pileup of large signals and to avoid the saturation of the output buffer, the preamplifier is equipped with a time-variant device which quickly resets the large ISBN: ISSN 90-5

2 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 signals above a threshold. The threshold is put just below the upper limits of the ADC input range and it will be shown how this device reduces the dead time induced by large signals by a large factor of to 5 and extends the pulse-height measurement range well beyond the ADC linear range. Front-End Electronics Structure The structure of the front-end electronics is presented in Fig.. It consists of the following stages: a charge-sensitive loop including a part cooled at about 50 K, a pole-zero cancellation with integrated fast reset circuitry, a balanced differential output buffer and a programmable spectroscopic pulser circuitry. The warm part is built in two versions, one as single core (00mV/MeV) and another one as dual core with two gains, namely 00mV / MeV and 50 mv / MeV, respectively. positive swings, with the unusual solution of two cascaded bipolar transistors in the common-base path (Q; Q). This was a convenient solution with the advantage of increased open-loop gain of the charge-sensitive stage balanced by the small disadvantage of increased noise of the double transistor cascode structure. By properly biasing the circuit structure, the increase in noise due the double transistor cascade structure was rendered negligible. As the closed-loop gain is ~ 5 mv / MeV (for C F = pf) and the output voltage range of the TA stage is about 0 V we obtain an energy range of ~80 MeV with an intrinsic noise of ~ 00 and with a slope of 8 ev / pf (with a total detector bulk capacity of ~ pf) cold part QC D core cold C warm part D core warm Q C R R8 R9 R0 R R R50 R5 C Q +V R Q -V C Q C C C R8 R Fast Reset RC RC RC CC R5 R8 R C C0 C50 FB core cold -V FB core warm C5 R8 R8 R9 Q5 Fig. AGATA Analog front-end electronics for central contact (single core version) and for the segments Pulser Input Fig. Core- preamplifier charge-sensitive loop (left, cold part and right, warm part) -V. Charge-Sensitive Loop The input stage is collecting electrons from the central electrode of a -fold segmented detector. The connection is AC (nf/5 kv) due to a -5 kv bias voltage applied to this electrode. The chargesensitive cooled part comprises an input stage with a very low noise jfet transistor and a passive feedback network. These components are placed in a cryostat near the central electrode, to reduce the noise and the microphonic effects. A temperature between 50 K +/- 5 K was chosen for the input jfet in order to optimize its equivalent noise voltage down to ~ 0. nv/ Hz. The feedback network time constant of the first stage is ms (GOhm and pf, respectively) to optimize both, noise and bandwidth in the first stage. The Transimpedance Amplifier (TA) designed with discrete components is operated at room temperature outside the cryostat. A 'folded cascode' structure was adopted for the TA stage (Fig.). The TA is optimized for a wide unipolar output range of ~ 0 V To get a fast rise time at lowest possible noise, several jfet structures have been tested e.g. IF0 (InterFET), BF8A, -B, -C, BF8 (Philips) and the BF8 was finally selected. We also optimized the double transistors cascode circuitry for a low noise and a wide bandwidth by carefully choosing the collector currents Q and Q to an optimum value for of 50 µa.. Pole-Zero cancellation with integrated Fast Reset circuitry The second stage of the preamplifier section is a passive Pole-Zero cancellation (P/Z Adj.) network and a buffer stage as impedance matcher. The P/Z Adj. stage has the purpose of reducing the decay time of the first charge-sensitive stage (of ~,000 µs) down to ~50 µs. Therefore, the baseline restoration is faster and the event-by-event pile-up is reduced. The P/Z Adj. works properly up to 0-0 kcps and at mean energies around - MeV. Unfortunately, this adjustment is less efficient at much higher counting ISBN: ISSN 90-5

3 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 rates of much higher gamma energies up to MeV equivalent gamma energies. To speed up the baseline restoration in such cases, we implemented a fast time-invariant reset circuitry. It consists of a fast accurate trigger comparator which switches on/off a temperature-compensated current sink placed behind the passive P/Z Adj. circuit. When a large signal arrives, the current sink is switched on, which discharges the capacitance C and quickly restores the output voltage [y (t)] to 0 V. As soon as the voltage y (t) reaches 0 V, the current sink is switched off and normal operation is restored. comparator threshold. It turned out that after measuring T = t t (Fig.5) we can estimate the energy of the large signal as: E w = b T + b T - E c () Where b and b are parameters independent of T and the E c is the exponential tail, namely: E c = E E + E o () +V -V Charge Sens. Loop R +V + U R - C9 R Pole Zero Adj. R5 + U5-5 8 UB -V R R8 R9 I -V R0 R R5 +V R R9 Fast Comparator +VA R5 Active Reset R Q8 Q9 A_ U A_ A_ +VA -V Vee - U9 + +Vs Vcc SHDN R R R R5 R R SHDN_Core +VA INH_Core Fig. Fast Reset current through Pole-Zero network As shown in Fig. an ultra fast comparator is continuously sensing the preamplifier output signal. The LT9 with less than ns output/input delay at only 5 mv input overdrive has been chosen for this purpose. Fig.5 Measurement technique of time over-threshold technique Where E and E are the averaged pre- and postpulse baseline estimates expressed as energy values and E o is an offset term. The final expression for the large signal energy as function of the reset time and pre- and post-pulse voltage baseline V and V is given by: E w = b T + b T -(V -V )/G +E 0 () were I CF b =ψ q C and b b = () τ P Fig. Time diagram of the time-variant fast reset circuit (left above, SHDN OFF, right above and below, SHDN ON) and Fast Recovery < µs after INH signal (left below, expanded baseline). F. Zocca [] calculated exactly the relation between the width of the INHIBIT signal (T) and the amplitude signal of a very large signal, over the Unfortunately, while C F and C are not calibrated we can hardly calculate parameters b and b, but we can take advantage of an existing spectroscopic pulser on board and calculate b and b providing large precise pulser signals of different precalibrated and over-threshold equivalent energies. Additionally, if we simultaneously irradiate the detector with gamma-rays, with energies well below the threshold, then we can also generate a statistical distribution of the disturbing tails as shown in Fig.. The equation () establishes a linear relation between the measured data and the parameters b, b, E o, thus we can extract those parameters from three ISBN: ISSN 90-5

4 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 or more experimental points. We need to generate different values for T, each with different pulser amplitudes above the threshold. Following this procedure energy resolution better as ~0.0% can be obtained at ~00MeV equivalent gamma energies.. The differential balanced output buffer A differential signal transmission mode was selected to enhance the rejection to common-mode noise and disturbances picked up along the output cable. As output stage, a balanced differential output stage has been designed around LM dual operational amplifier which features low noise, low power and wide bandwidth (Fig.). Only +/- V power supply have been chosen due to the overall power consumption limitation of the triple cryostat, holding spectroscopic channels of three AGATA detectors (x segment channels and x core channels). -V Programmable Spectrometric Pulser (PSP) has been developed. It is used for both, detector electrical characterization and for self calibration of the TOT method. The output signal of the PSP is DC coupled to the source pin of the core jfet through a resistor divider consisting of a 8.5 Ohm resistor and a grounded.8 Ohm resistor. Thereafter, the signal reaches each of the detector segments via the source-gate capacitance of the jfet, the HV coupling capacitor and the core-segment bulk capacitance (Fig.). A block diagram of the PSP is showing in Fig.. The PSP consists of three stages: a programmable reference voltage (V Ref ), a chopper triggered by an external signal (Pulser In) and a programmable attenuator which comprises also an output buffer stage (Programmable Attenuator-Buffer). R UA R8 /Out_C +V from P/Z R5 + U5-8 R +V R -V /OUT_C R -V R0 5 R5 8 UB +V R R to Comparator OUT_C (/OUT_C+OUT_C) Out_C Fig. Balanced differential buffer and twisted cable driver circuit The signals are transmitted to the remote ADCs modules through 0 m individually shielded twisted pairs cables preserving the original quality of the signals, which makes the preamplifier assembly suitable for noisy experimental environments. A high quality -wire LVDS cable assembly with individually shielded twisted pairs ( camera-link like format) and MDR (Mini Delta Ribbon) male connectors on either side have been chosen.. Programmable spectroscopic pulser One of the important issues in the framework of AGATA detector and triple cryostat development is the characterization of its response function, including the response function of the preamplifiers (both in time and frequency domain), e.g. stability, linearity, treatment of saturated pulses, dependency of the energy resolution on counting rate, dead time estimation, microphonics effects of the detector and cryostat, intrinsic detector characterization (bulk capacities) etc. Many of these parameters can be measured with the help of a high precision spectroscopic pulser and this is the main reason why a Fig. Block diagram of the Programmable Spectroscopic Pulser (PSP) The programmable reference is built around an ultra high precision voltage reference of.09 V (MAXESA) with a very low temperature coefficient ( ppm/ C), very low noise (.5 µv pkpk) and long-term stability of 0 ppm/000 hr and a bit serial input, voltage DAC (+/- bit), type AD55CR. Data is written to it in a -bit word format, via a -wire industry standard serial interface with CS, Clock, Data In signals. The output of the programmable reference stage is a low noise and very fast rail-to-rail operational amplifier (OPA5) with relatively large slew rates in both directions, being able to source and sink fast chopped currents. The Chopper is built around two ADG5 analog switches. It is designed in a T-switch configuration, which results in excellent Off Isolation while maintaining good frequency response in the ON condition. The chopper stage is triggered by an external signal. A balanced differential transmission (LVDS type) has been adopted for this signal to reduce its crosstalk with the analog signals (Fig. 8). The chopper mode of operation is programmable; a rectangular or an exponential-decay mode can be selected. In exponential mode the decay time is set ISBN: ISSN 90-5

5 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 by default to 50 µs but it is adjustable in the range of 50 µs up to 000 µs. If the exponential-decay mode was selected, a linear gate circuitry selects only the positive polarity. Vref Trigger S In C5 -V +V +V _D 8 VDD D R9 V Shown for In = _D _D _D C0 R90 Mode R0 U -V R0 C0 C9 8 C59 +V VDD V5 _D _D R5 S In D +V 8 VDD V D Shown for In = Fig.8 Spectroscopic pulser, chopper and linear gate _D R80 C8 R9 _D -V R8 R +V U -V Chopper Out The last stage is a programmable attenuator, which consists of four stages, each of 0dB; thus the coarse attenuation factor can be selected between 0 and 0 db, with a granularity of 0 db. The buffer stage (LT85) can drive a maximum DC current of 0 ma onto a tiny load resistor with a nominal value of.8 Ohm, placed in the cryostat between the source pin of the core jfet and internal (Fig.). Table Specification for spectroscopic pulser Property Value Dynamic range core [MeV] ditto segments [MeV] Resolution core [kev] <.5 Resolution segments [kev] < 0.09 Rise time (default) [ns] 50 (opt.0-80 ) Fall time ( default] [µs] 50 ( opt. 000) Stability: long term ~..0 - / h Stability: temperature ~.5 ppm / o C.5 Dual Gain Core Structure From the discussion of the TOT method, it is evident that a small open gap in signal amplitude range remains between the classical amplitude spectroscopy method and the Time-Over-Threshold method. Analogous to the subranging ADC methods, where some extra bits are used in the extra stages to correct the errors made in earlier stages, it was decided to build a new version of the AGATA Core Preamplifier, namely one board with two conversion gains. The previous AGATA Single Core board had only one conversion range of ~00mV/MeV, switching from the classical amplitude mode to the TOT method at about ~ 0MeV range. The Dual Core preamplifier has one channel with larger conversion gain range of ~ 00 mv / MeV and another one with only ~ 50 mv / MeV. The advantage of a board with dual channel gains is at least two fold; firstly the channel with higher gain will be able to analyze the smaller range of spectra with higher accuracy (and in normal applications this part of the spectra has higher incidence rate) and it switches to TOT method at about ~ 5 MeV, while the second channel will switch to TOT method at about ~ 0 MeV. In this way, the second channel will cover the narrow subranging open-gap of the first channel, being at about ~ 5 MeV in its full linear range and therefore can be used to correct the first channel. Measurements with dual core preamplifier and spectroscopic pulser A full set of core-pulser and triple preamplifier boards has been installed in the warm part and in the cold part of AGATA test cryostat operated in conjunction with an AGATA -fold segmented detector, as biased at kv. Table Full specifications of the AGATA Single and Dual Core Preamplifiers Property Value Tolerance Conversion gain for seg- 00 mv / MeV ±0 mv ments and single core Conversion gain for dual core Noise (terminated) 00 mv/mev (Ch ) 50 mv/mev (Ch ) 0. kev fwhm (C d =0 50K) ±0 mv ±5 mv Noise slope + 8 ev / pf ± ev Rise time ~ ns (0 pf) ± ns Rise-time slope ~0. ns / pf Decay time 50 µs ± µs Integral non linearity < 0.05% (Dyn.~.5V) Output polarity Differential, Z o =00Ω Fast reset speed ~0 MeV / µs Inhibit output LVDS (EIA/TIA-) Power supply ±.5 V, ±.5 V ±0.5V Power consumption jfet < 0 mw Power consumption (except diff. buffer) < 80 mw (80 mw for a triple) Supplementary power diff. buffer very high ~00 mw counting rates Mechanical dimension (0 x 50 x )mm segm. ( x 5 x 8 mm core As mentioned all jfets ( segments + core) and all feedback components (GOhm and pf) are placed in the cryostat and cooled down with liquid nitrogen to about 50 K, while the detector is cooled down very close to the liquid nitrogen temperature. To preserve the good transfer function of all preamplifiers (Fig.9 left) special attention has been paid to the cryostat wiring, especially to the return ground paths. At least for the cold core and pulser parts, individual return ground paths have been provided. ISBN: ISSN 90-5

6 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 Fig.9 Block diagram of the dual gain core and programmable spectroscopic pulser kev FWHM as mean value for the all segment channels (Fig.), i.e. the equivalent intrinsic energy resolution of the spectroscopic pulser is less then ~0. kev FWHM. With a real symmetric AGATA detector placed in an AGATA test cryostat, the mean energy resolution observed in the core spectra on the. MeV line of 0 Co was.9 kev FWHM and about.-. kev FWHM on the kev/ kev lines of 5 Co source. Two of the acquired spectra are shown in Fig. for the high energies part (below) and low energies part (above, left), respectively. The structure of the dual core preamplifier is shown in Fig. 9 (almost ideal transfer functions in time domain for the two core channels Ch and Ch at pulser amplitude of equivalent ~9 MeV are presented) and the layout is displayed in Fig.0. Fig.0 Dual gain core preamplifier and programmable spectroscopic pulser (top view) The amplitude resolution was tested in a setup with a test cryostat (with pf as segmented dummy detector) an Ortec Filter Amplifier (Model No.A) with µs shaping time and a k channels IKP MCA [0]. In this setup we obtained Fig. Energy resolution of AGATA Core-PSP vs. PB- (BNC) spectroscopic pulser for the spectroscopic pulser an equivalent energy resolution of ~ kev FWHM for the core Ch., about ~. kev FWHM for core Ch. and ~0.9 Fig. Acquired spectra with gamma-ray lines and intercalated pulser lines. Detector electrical characterization Taking advantage of the built-in pulse generator in the AGATA Core Preamplifiers one can measure directly some intrinsic detector electrical parameters, not specified by detector manufacturer but very useful in data treatment and interpretation. One of these parameters is the bulk capacity between the central electrode (core) and individual segments. As seen in (Fig.) the pulser signal injected in the source of the core jfet is distributed over all detector segments by its central contact as presented in the subsection.. The pulser signal amplitude measured in each segment will be, therefore, inversely proportional to the core-to-segment bulk capacity. Since the preamplifiers are not calibrated in absolute sense, the core-to-segment bulk capacity measurements only provides relative data but these values can be calibrated with the total detector capacity, i.e. their sum should be equal to the whole detector bulk capacity. For example, in the case of the first symmetric -fold segmented AGATA detector, the detector manufacturer guaranteed. pf for detector bulk capacity. From the pulser core-segment amplitude ratios the bulk capacities of the segments could be measured, normalizing them ISBN: ISSN 90-5

7 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, 008 to a total bulk capacity of. pf. The mean values for segment capacities are presented in Tab. (there are rings each segmented in sectors). Table Mean values for ring bulk capacities Ring number 5 Mean segment capacity [pf] Due to the symmetry of the AGATA_S00 detector, these capacities are only ring dependent showing a very small deviation for further -fold segmentation inside each ring. The deviations of the measured values with the listed averages of the further six segments in each ring are within a standard deviation of.5%.. Crosstalk inside one segmented detector Due to the relatively high density of electronic channels in the -fold segmented AGATA detector, special care has been taken at the level of preamplifier schematics and its PCB design, interconnecting motherboards, up to the shielding of the individual channels against cross-talk. In such a complex system where the sensor itself is highly segmented, crosstalk between segments is observed, as well as between core and segments and vice versa. Special care has been taken to minimize the crosstalk between segments, between core and segments and between detectors at the level of triple cryostat. The cold jfet PCBs have been carefully designed to minimize the inductive coupling between neighboring segments channels [], also additional number of return ground wires for the segments wiring in the cryostat have been added. In the case of triple cryostat separated grounding wires, for the three central detectors contacts and related core-pulser signals, have been provided. The crosstalk measurements were routinely performed with digital electronics, using a set of 0 DGF-C CAMAC modules developed by XIA LLC for the MINIBALL array of segmented detectors []. Each DGF-C module has four spectroscopic channels, each channel has a similar analog input stage consisting of a programmable gain, offset circuitry and a Nyquist filter followed by a fast pipe-line ADC with bit resolution operated at 0 MHz. Nine of DGF- C modules are used for the segment channels and one for the central core contact. The experimental data are stored on disk. After off-line sorting of -fold hits, energy shifts are observed as a function of hit pattern (as presented in Fig. for all x combinations). One can distinguish the same structure ( fingerprint ) due to the segment ring bulk capacities that was measured using the pulser (Fig., above). Fig. Core-Segments pulser ratio and crosstalk measurement, energy shift for core and energy shifts for segments in all -fold hits ( x combinations) Indeed, as documented by E. Gatti et al. [], Pullia et al. [8] and B. Bruyneel et al. [9] the AC equivalent model of a depleted detector can be described by a network of capacities. In such models each electrode i is coupled to other electrode j by a capacity C ij (e.g. as those presented in Table for core-segments). The current induced on electrode i by movement of free charge carriers inside the detector is given by the Shockley-Ramo theorem [9] and is incorporated in the model as an AC current source i i generated between ground and electrode i. The relation between electrode potentials v i and their currents is given by Kirchhoff s law and can be expressed in a matrix form as: i i =Y ij.v i () with Z ij = (s-c ij ) (5) These non-diagonal Z - couple the potentials to currents induced on other electrodes and is, therefore, responsible for this kind of crosstalk.. TOT technique from MeV to 80MeV The spectroscopic pulser resolution measured in normal spectroscopic mode is between.05 kev and.5 kev fwhm in relative wide range of equivalent energies, from 50keV up to MeV. Then tested with TOT technique and in the presence of a background of regular statistical signals from a calibration radioactive source (such as 0 Co) the resolution is better than 0.% over the full range and better as 0.0% between MeV range, as shown in Table. Table Pulser resolution at low counting rates: Pulser Amplitude, Equiv. energy [MeV] Pulser resolution [%] Increasing the counting rates from ~ 800 cps up to 5 kcps the pulser resolution is still very good, i.e. better than 0. % for all tested counting rates of, ISBN: ISSN 90-5

8 9th WSEAS International Conference on AUTOMATION and INFORMATION (ICAI'08), Bucharest, Romania, June -, or 0 MeV. The TOT technique has been tested in a setup with real high energetic γ -rays produced by Am+Be radioactive source with a Ni target generating high-energy photons by neutron capture reactions []. In this experimental setup the reset threshold was intentionally set to a lower value of ~MeV, in order to reserve a large part of the whole spectrum of Am+Be γ -ray source for the reset mode (i.e. TOT technique). The reconstructed energy spectrum from the fast reset pulse width, putting ADC limit to MeV is shown in Fig.. Fig. Demonstration of TOT method withγ -rays Conclusion A very low noise, very wide dynamic range with a very accurate transfer function charge-sensitive preamplifier has been developed and tested to be used with a highly segmented and encapsulated HP-Ge detector in the frame of the AGATA project. Furthermore its wide spectroscopic range has been successfully extended by more than one order of magnitude, by switching (below the maximum of the ADC range) from standard amplitude spectroscopic mode to TOT method. The disadvantage of a small gap in the conversion range has been eliminated by implementing a dual range AGATA core preamplifier. An excellent energy resolution over the whole range was achieved, about.kev@59.5 kev,. kev@. MeV (pulse height analysis mode) and 0. %@8.99 MeV and less than 0.0% above 50 MeV (in Time over Threshold mode). There are already some very interesting applications of these techniques developed initially mainly for the large array of segmented detectors. Due to mentioned advances in the development and manufacturing of the large and highly segmented HP-Ge detectors and related fast and highprecision front-end electronics and digital electronics, it is now possible to build efficient and high-resolution Compton cameras []. By employing gamma-ray tracking procedures it is possible to get position resolution of 0.5 mm at kev without the use of an attenuating collimator. The authors thank the AGATA detector group for strongly supporting this work and C. Görgen, G. Richardt N. Warr, for valuable suggestions and help. References: [] D. Bazzacco, B. Cederwall, J. Cresswell, G. Duchene et al., AGATA Technical Proposal, available online, [] J. Simpson, The AGATA Project, EPS Euro Conference XIX Nuclear Physics Divisional Conference, Journal of Physics: Conference Series, 00, -80 [] G. Pascovici, AGATA Core Pulser, AGATA Week Orsay, 5-9 Jan. 00, available online, a/agata-week/monday-session//pascovici.ppt [] J. Eberth, G. Pascovici, H.G. Thomas, N. Warr et al., Miniball: A Gamma-Ray Spectrometer with Position-Sensitive Ge Detectors for Nuclear Structure Studies at REX_ISOLDE, AIP Conf. Proc., 00, Vol. 5, pp 9-5 [5] A. Pullia, G. Pascovici, B. Cahan, D. Weisshaar, C. Boiano, R. Bassini, M. Petcu, F. Zocca, The AGATA charge sensitive preamplifiers with buitin active reset and pulser, Nucl. Sci. Symp. Conf. Record, 00 IEEE Vol, Issue, Oct. 00, pp - [] F. Zocca, A. Pullia, D. Bazzacco, G. Pascovici, Wide Dynamic Range Front-End Electronics for Gamma Spectroscopy with a HP-Ge crystal of AGATA, Nucl. Sci. Symp. Conf. Record, 00 IEEE, Vol., pp - [] E. Gatti, G. Padovini, Signal evaluation in multi electrode radiations detectors by means of a time depending weighting vector, Nucl. Instr. Meth. Vol. 9, 98, pp 5-5 [8] A. Pullia, R. Isocrate, R. Venturelli, D. Bazzacco, R. Bassini, C. Boiano, Characterization of HPGe-segmented detectors from noise measurements, IEEE Trans. Nucl. Sci., Vol 5, 00, pp [9] B. Bruyneel et al., Crosstalk properties of - fold segmented symmetric hexagonal HP-Ge Detectors, Nucl. Instr. Meth., 008, to be published [0] G. Pascovici et al., Programmable stretcher and k channels portable MCA, private communication [] K. Vetter, M. Bursk, L. Mihailescu, Gammaray imaging with position sensitive HPGe detectors, Nucl. Instr. and Meth. in Physics Research, Section A, 00, vol.55, pp. -. ISBN: ISSN 90-5