University of MilanoDepartment of Physics and INFN

HIGH DYNAMIC RANGE LOW-NOISE PREAMPLIFICATION OF NUCLEAR

SIGNALS

A. Pullia, F. Zocca, C. Boiano, R. Bassini, S. Riboldi, D. Maiocchi

Department Conference “Highlights in Physics 2005” October 14, 2005

40 cm

Bea

m

AGATA detector AGATA detector arrayarray

AGATA: an Advanced GAmma-ray Tracking Array

Proposed for high resolution γ-ray spectroscopy with exotic beams

Employing highly segmented HPGe detectors, newly developed pulse-shape analysis and tracking methods

HPGe segmented HPGe segmented detectordetectorγ (1 MeV)

p K

(50 MeV)

Background of energetic particles

Segments

Core

10 cm

Individual highly energetic events or bursts of piled-up

events could easily cause ADC SATURATION and introduce a

significant SYSTEM DEAD TIME

charge loop

charge preamplifier

From detector segmen

t

FR

FCSecond stage

Anti-alias AD

C

The new nuclear experiments with exotic beams pose challenging requirements to the

front-end electronics

Besides having a LOW NOISE, an extremely HIGH

DYNAMIC RANGE is required !

ADC overflow voltage level

New mixed reset technique:

continuous + pulsed

Saturated output without

pulsed-reset

Ideal non-saturated output without pulsed-reset

Preamplifier output with continuous-reset

(50s decay time constant)

Output with pulsed-reset

An ADC overflow condition would saturate the system

for a long while

A pulsed-reset mechanism could permit a fast recovery of the output quiescent value, so minimizing the system dead

time

1st stage 2nd stage

Charge loop

Passive P/Z Amplification

From detector

Cold part of

preamplifier

Warm part of preamplifier

Schmitt trigger comparator

Discharge current From

ADC OVR(optional)

Output

Capacitance to be

discharged to de-

saturate 2nd stageDe-De-

saturation saturation circuitrycircuitry

/Output

1

-1

3rd stage

Implemented mixed reset technique: a time-variant

charge preamplifierCircuit architecture: fast de-saturation of the 2nd stage

Noise is not at risk as no new path is connected to the input node !

Signal acquired at 1st stage output… …and at preamplifier output

1st stage output voltage swing

The realized pulsed-reset technique does not act on the 1st stage and so can’t “protect” it against saturation

The architecture of the 1st stage has been studied to provide a large output voltage swing ( 10 V) and so to a prevent a risk of an overflow condition

Triple AGATA segment preamplifier on alumina substrate

(Mod. “PB-B1 MI” – Milano)

Top view

Bottom view

PZ trimmers

Mechanical dimensions:57x56x5 mm

MDR26 connectors

Segment preamplifiers

Segment preamplifiersCore

preamplifier

0 2 4 6 8 10 120

1

2

3

4

11

109

87

65

43

2

1A

mpl

itude

[V]

Time [µs]

Action of pulsed-reset device

Curve (1)-(10) = from 5 to 50 MeV Curve (11) = 100

MeV

IC

C

qC

I

dT

dE FT

dET / dT = 7.8 MeV/μs

Event energy = 100 MeV : Reset time 13μs !

In a first approximation, a directly proportional relationship exists between the pulsed-reset time T and the event energy ET

I = reset current

= 55 mV/MeV (1st stage conversion gain)

C = 2nd stage capacitance (to be discharged)CF = feedback capacitance

Ψ = 2.92 eV/pair (for HPGe)

Es: CF=1pF, C=4.7nF, I=2mA

Passive P/Z stage: pole

21 RRCP superposition theorem :

1) large signal:

2) tail of previous events:

3) reset current:

Pt

eHtv )(01

Pt

ehtv )(02

P

t

eRRItv 1)( 2103

2121)( RRIeRRIHhtV Pt

PZ

sum of the three contributions: expression of the reset transient

Tt ,0for

by equating to zero at t=T, we derive the relationship between

the total signal amplitude and the reset time :

121

PT

eRRIhH

Detailed analysis of the reset transient

“Reset time-energy” relationship

...TC2

IT

C

IEEE 2

PCST

If we convert the voltage amplitudes H and h in the equivalent energies Es and Ec (by using the conversion gain ), we obtain the

relationship

1eRR

IEEE P

T

21CST

ES = energy of the large signal

EC = equivalent energy of the tail of previous signals

We can expand the exponential term with no loss of accuracy since T<<τP :

large signal energy Es estimated from the reset time T and the tail

contribution Ec

ET = equivalent total energy subjected to reset

T = reset time

C2

PS E...T

C2

IT

C

IE

Energy estimate of a large individual event from the measurement of the

reset time

CP

S ETC

IT

C

IE ...

22

O2112

21S EVVkTbTbE

Contribution of the tail of previous

events

ES = energy of the individual large event

T = reset time

V1 , V2 = pre- and post-transient baselines

b1 , b2 , k1 , E0 = fitting parameters

Tests of the large-signal measurement technique performed with a prototype of

the circuit and a bulky HPGe detector

(Padova, July 2004)

reset device

A spectroscopy-grade pulser injects a large pulse at the preamplifier input

A 60Co source provides a background of lower events which destroys the

large signal resolution if no correction is made

Measurement of large pulses from reset time

Rate of 60Co background

events

Resolution @ 10 MeV in Ge (FWHM)

1 kHz 0.26 %

2 kHz 0.32 %

4 kHz 0.30 %

8 kHz 0.37 %

16 kHz 0.57 %

32 kHz 0.56 %

Rate of 60Co events = 32 kHz

Measurement performed at Padova with HPGe detector (courtesy of D. Bazzacco and R. Isocrate)Measurement performed at Padova with HPGe detector (courtesy of D. Bazzacco and R. Isocrate)

**F. Zocca, ”F. Zocca, ”A new low-noise preamplifier for A new low-noise preamplifier for -ray sensors with smart device for large signal management-ray sensors with smart device for large signal management ”, Laurea ”, Laurea Degree Thesis, University of Milano, October 2004 (in Italian). Degree Thesis, University of Milano, October 2004 (in Italian). See http://topserver.mi.infn.it/mies/labelet_iii/download_file/capitolo6.docSee http://topserver.mi.infn.it/mies/labelet_iii/download_file/capitolo6.doc

OS EVVkTbTbE 2112

21

** EESS = equivalent energy release = equivalent energy release

T = reset timeT = reset time

bb11, b, b22, k, k11, E, E00 = fitting parameters = fitting parameters

VV11, V, V22 = pre- and post-pulse baselines = pre- and post-pulse baselines

Energy range in Energy range in normal mode normal mode ~ 2MeV 2MeV

1408 keV1408 keV

2.02 keV fwhm2.02 keV fwhm

Extending the energy range by reconstruction of the large signals from

reset time

Extended range

+ pulser+ pulser

122 keV122 keV

344 keV344 keV

Future developments

Tests of the pulsed-reset device with a triple AGATA preamplifier coupled to an AGATA HPGe segmented detector

Tests of the large-signal measurement technique when applied to measure the energy of real highly energetic events (photons or energetic particles in the 10-50 MeV range)