Документ взят из кэша поисковой машины. Адрес оригинального документа : http://rtm-cs.sinp.msu.ru/docs/borexino/sn-daq.pdf Дата изменения: Thu Jan 24 13:54:13 2008 Дата индексирования: Mon Oct 1 20:14:18 2012 Кодировка:

Supernova Acquisition in Borexino
Didier Kryn, Michel Obolensky (APC) October 22, 2007

1 Introduction
We present in this document a Data Acquisition system which was rst set up in july 2007 with the help of Marco Pallavicini and Massimo Orsini, and then consolidated in october 2007. The main goal was to be continuously up and running, independantly of the operation of the main acquisition of Borexino, so as to have a 100% eciency on Supernova detection. It appeared at the same time that it was dicult to the main DAQ to eciently detect neutrons and, following the suggestion of Marco, we decided to acquire both neutrons and Supernova events in this new system. The Supernova signal in Borexino would be easy to identify, consisting essentially in a high rate of high energy events. The requirement for a dedicated acquisition system is therefore much more the high eciency than a good precision energy measurement and position reconstruction. The system described here features actually only one ash-ADC channel properly dedicated to the Supernova signal and fourteen dedicated to neutron detection, by their dierent energy ranges. This acquisition system sits in one single VME crate. It also needs a few devices in a NIM crate to build its trigger. Below we describe the system and show some preliminary analysis of the data.

2 The Acquisition System
2.1 Analog signals
Each of the Marathon Adders located on the top of each Front-End rack features two pairs of outputs providing the analog sum of one half of the PMTs connected to this rack. For each half rack there are two outputs, one of which is used to build the total sum of the detector, by the mean of a passive cirtuitry, and the other to provide partial sums for the super-nova/neutron DAQ. These last signals are summed two by two by the means of resistor networks so as to obtain one signal per Front-End rack. This last summation is done on a patch pannel located in the VME crate of the Supernova DAQ.

1


The total sum is distributed to several systems, including the Sums DAQ, the Muon/Neutron acquisition system of Princeton and the Super-Nova DAQ. It is available from a fan-out in the NIM crate at the bottom of the leftmost FADC rack. The total sum, attenuated by 10db and the fourteen partial sums are connected to the 15 waveform digitizer channels of the Supernova DAQ.

2.2 Clock distribution and trigger
The ve waveform digitizer modules can be operated with their own internal clock or with an external clock. In order to keep them synchronous, one of them works with its own internal clock which is then sent to the others with cables of apropriate length to give them the same phase (with a precision of 1ns). To be certain that waveform digitizers detect the trigger always all during the same clock cycle, the trigger (originaly asynchronous) is synchronized with the clock. The trigger is rst formed by a simple discriminator with the total sum as input. Its threshold is 35mv (350mV at the test point). Then it is vetoed by the pulser signal, the laser signal and the busy from the rst waveform digitizer board. Then it is delayed by a long delay line of more than one micro-second, so that the signal is positionned near the beginning of the 1280ns time window of the waveform digitizers. Finally it is synchronized with the clock.

2.3 Waveform digitizer conguration and read out
The waveform digitizers used for the Supernova DAQ are the same model as used for the Sums DAQ; however they have a dierent conguration. In the Sums, they have a buer of 64 events of 10240ns, while for the Supernovae they have a buer of 512 events of 1280 ns. This means that they can absorb a burst of 511 triggers without incurring any dead time. They can actually absorb more since the DAQ is reading the buer on the y at the rate of 300 events/s. The DAQ is running in a PowerPC of the same model as the ones used in Borexino, however this one boots and runs from an SCSI disk sitting in the same VME crate. The acquisition program is the very one used during years to take CTF data. It complies to the same protocol as the Borexino PowerPCs for what concerns the run control. Two client run control interfaces are available to interact with the DAQ, one of which has a simple graphical interface.

2.4 Summary of operation
The Supernova DAQ is currently operated from Paris and the data transfered to Lyon. Given the necessary space, the raw data and/or the root les will be stored in the Borexino storage area where they will be publicly available. The operation can easily be controlled from LNGS and made part of the shifters' duty.

2


Figure 1: Trigger

3


The system has been almost continuously up and running as from october 1st 2007. The periods of inactivity are essentially due to triggering problems caused by the instability of the Front-End electronics. We don't know for sure where the source of the noise is but, periodically, a low frequency signal with an amplitude of 50 to 100mV appears in some or all partial sums and on the total sum. Under these conditions, the trigger rate increases dramatically and the data ow as well, which can ll the disk in a few hours.

3 Upgrade plan
In the next months we plan the following stagged upgrade of the DAQ 1. During this fall, we will replace the current trigger system with one entirely contained in the same VME crate as the FADC boards and processor. This new trigger should be safer. In addition, it might be possible to record Muon Veto tags and BTB event numbers. 2. During this fall we will replace the MVME2300 processor by an MVME3100, state of the art for VME and communications, with a 40GB hard disk (instead of 8GB currently) and the latest stable version of Debian Linux. As a side eect, the MVME2300 will be available as a spare for the normal Borexino DAQ. 3. During next winter we will replace the V896 Waveform Digitizers with V1721, a model codevelopped with CAEN. These sample the signals at 500MHz and have 8 analog channels per board. Therefore, two boards can replace the current ve V896 and provide a second channel for Supernovae proper. These new boards show a much lower input noise than the V896 (0.3 ADC counts RMS instead of 1.5); this might be usefull if the quality of the input signals could be improved.

4 Preliminary Data Analysis
The data acquisition is running continuously, when the high voltage is on, except when the trigger rate is very high. The usual rate is of the order of 0.2 trigger/s, except when a noise show up in one or in all the channels. The simple discriminator we use for trigger does not allow to obtain a sharp threshold. So we do not use the 210 Po peak to dene the energy scale, but the peak of the neutron capture on hydrogen. But at the same time this simple trigger, and the dead time free digitizers, give the opportunity to detect all the spallation neutrons, as it has been shown in our analysis of CTF3 data1 . The statistics of this very preliminary analysis represents 9.28 days obtained after some tuning of the threshold, and the implementation of a veto on the laser
1 See the internal note "CTF data analysis with the FWFD system" M.Ob olensky, S.Sukhotin. August 22, 2006.

4


and prepulser signals. But as mentionned above, no information is recorded, for the moment, from the Outer Muon Veto. Following the method used for the analysis of CTF data, separate samples of pure single, double and multiple clusters have been dened.

4.1 Neutrons
Most of the multiple clusters are dened by a muon (rst pulse in the cluster) followed by spallation neutrons. The energy spectrum of those neutrons is presented on gure 2, with a time selection [20,1500]µs, and an energy cut (1MeV) to avoid the PMT afterpulses after the muons. The main peak corresponds to the neutron capture on Hydrogen, and the tiny one may be due to the capture on Carbon. Figure 3 shows the time distribution of the neutron capture, after a muon. The tted value may be a little higher than expected, but the accidentals events have not been yet taken into account. The gure 4 shows the distribution of the neutron multiplicity. As expected, clusters with huge number of neutrons are detected, even during this short period of data taking. Figure 2: Energy distribution for events more than 20µs and less than 1.5ms after a crossing muon. Energies below 1MeV have been cut.
Q(i)-DSum Final_Mult>1 Q(i)>1MeV
350

QD(i)_FM>1_1
Entries 1712 Mean 2.17 RMS 0.5145 Underflow 0 Overflow 1 Integral 1700 2 / ndf 22.54 / 4 Constant 314.6 ± 12.0 Mean 2.18 ± 0.01 Sigma 0.1815 ± 0.0067

H
300 250 200 150 100 50 0 0

C

1

2

3

4

5

6

7

8

9

MeV

10

5


Figure 3: Time netween Muon and neutron candidate

Neutron Time µsec
2 50 const 40 / ndf norm

Time
98.84 / 124 50.65 ± 2.40 245.4 ± 11.7 0.6673 ± 0.2071

30

20

10

0 0

200

400

600

800 1000 1200 1400 1600 1800 2000 2200 2400

6


Figure 4: Multiplicity of neutron-like events between 20µs and 1.5ms after a muon

Multiplicity without first 20 µ sec && Q(i)>1MeV

Mult_20_1
Entries 22128 1.568 0.8051 0 1 2.213e+04 Mean RMS Underflow

10

4

10

3

Overflow Integral

10

2

10

1 0 5 10 15 20 25 30 35 40 45 50

7


Figure 5: Events with multiplicity 2: Energy deposited by rst event vs energy deposited by second event
Q1(Y) vs Q2(X) FinalMult=2
50 45 40 35 30 25 20 15 10 5 0 0 2 4 6 8 10 12 2 1 0 0 0.5 1 1.5 2 2.5
MeV
Entries 871

Q1(Y) vs Q2(X) FinalMult=2
MeV

Entries

871

8 7 6 5

T(astr) = 12.1 days

4 3

4.2 Double Clusters
The gure 5 shows the distribution of the energy of the rst pulse (Q1 ) versus the energy of the second one (Q2 ), within a time window of [20,1500]µs. The accumulation of events with a high Q1 (> 30M eV ) and a low Q2 (< 1M eV ), corresponds to after-pulses following a through going muon probably in the time interval between 20 and 40µs after the muon. Most of the events with Q1 < 20M eV and 1M eV < Q2 < 5M eV are neutrons induced by muons. A zoom of the scatter plot for Q1 < 8M eV , and Q2 < 2.5M eV is shown on the right; this zone of energies should contain the 214 Bi-Po, the e (reactor or Ї Geoneutrinos) and the alpha-n background. But a muon ag is mandatory to make a conclusion about the real nature of those events.

4.3 Menagerie of noises
From time to time the trigger rate begins to grow due the noise induced in one or several channels. Some examples are show in the follwoing gures(noise(i).eps). Each of the 3 gures show one trigger as seen by the 14 channels, the analog sum (1st blue one), and the digital (the 2nd blue one). Some of those periodic signals may be rejected by oine lters, but the eciency of such cuts must be tuned.

8


Figure 6: Various kinds of noises recorded
FADC_CHANNEL_1 Ev_44
15

CH_1 Ev_44 Entries 512

FADC_CHANNEL_2 Ev_44
14 12

CH_2 Ev_44 Entries 512

FADC_CHANNEL_3 Ev_44
15

CH_3 Ev_44 Entries 512

FADC_CHANNEL_4 Ev_44
15

CH_4 Ev_44 Entries 512

10

10 8

10

10

5

6 4

5

5

0

2 0

0

0

-5

-2 -5 -4

-5

-10

-6 100 200 300 400 500 -8 0 -10 100 200 300 400 500 0 100 200 300 400 500

-10 0 100 200 300 400 500

0

FADC_CHANNEL_5 Ev_44
15

CH_5 Ev_44 Entries 512

FADC_CHANNEL_6 Ev_44
15

CH_6 Ev_44 Entries 512

FADC_CHANNEL_7 Ev_44

CH_7 Ev_44 Entries 512

FADC_CHANNEL_8 Ev_44

CH_8 Ev_44 Entries 512

15 10 10 10 10 5 5 5 5 0

0

0

0

-5

-5

-5

-5

-10 0 100 200 300 400 500

-10 -10 0 100 200 300 400 500 0 100 200 300 400 500 -10 0 100 200 300 400 500

FADC_CHANNEL_9 Ev_44
15

CH_9 Ev_44 Entries 512

FADC_CHANNEL_10 Ev_44
15

CH_10 Ev_44 Entries 512

FADC_CHANNEL_11 Ev_44
15

CH_11 Ev_44 Entries 512

FADC_CHANNEL_12 Ev_44

CH_12 Ev_44 Entries 512

15 10 10 10 5 5 0

10

5

5

0 -5 -5 -10

0

0 -5

-5 -10

-10 0 100 200 300 400 500 -15 0 100 200 300 400 500

-10 0 100 200 300 400 500 0 100 200 300 400 500

FADC_CHANNEL_13 Ev_44
15

CH_13 Ev_44 Entries 512

FADC_CHANNEL_14 Ev_44
25 20

CH_14 Ev_44 Entries 512

Deconvoluted ASum
6

ASum_Dec Entries

Deconvoluted DSum

512

200 150

DSum_Dec Entries

512

10 15 5 10 5

4 100 2 50 0 0 -2 -50

0

0 -5

-5

-10 -4 -15 -100 -6 100 200 300 400 500 0 100 200 300 400 500 -150 0

-10 0 100 200 300 400 500 -20 0 100 200 300 400 500

FADC_CHANNEL_1 Ev_35
15 10

CH_1 Ev_35 Entries 512

FADC_CHANNEL_2 Ev_35
10

CH_2 Ev_35 Entries 512

FADC_CHANNEL_3 Ev_35
15 10 5

CH_3 Ev_35 Entries 512

FADC_CHANNEL_4 Ev_35
10 5 0 -5

CH_4 Ev_35 Entries 512

5 5 0 -5 -5 -10 0

0 -10 -5 -15 -15 -20 0 100 200 300 400 500 0 100 200 300 400 500 -10 -10 -20 -15 0 100 200 300 400 500 0 100 200 300 400 500

FADC_CHANNEL_5 Ev_35

CH_5 Ev_35 Entries 512

FADC_CHANNEL_6 Ev_35

CH_6 Ev_35 Entries 512

FADC_CHANNEL_7 Ev_35

CH_7 Ev_35 Entries 512

FADC_CHANNEL_8 Ev_35
20 15 10 5 0

CH_8 Ev_35 Entries 512

10

10 5 0

15 10 5

5 0 -5 -10

0 -5 -5 -10 -10 -15 -15 -20 -20 0 100 200 300 400 500 0 100 200 300 400 500 -15 -20 0 100 200 300 400 500 -10 -15 -20 -25 0 100 200 300 400 500 -5

FADC_CHANNEL_9 Ev_35
15 10 5 0 -5 -10 -15 -20

CH_9 Ev_35 Entries 512

FADC_CHANNEL_10 Ev_35
15 10 5 0 -5

CH_10 Ev_35 Entries 512

FADC_CHANNEL_11 Ev_35

CH_11 Ev_35 Entries 512

FADC_CHANNEL_12 Ev_35
15 10

CH_12 Ev_35 Entries 512

10

5

5 0 -5 -10 -15

0

-5 -10 -15 -20 -10

-15 100 200 300 400 500 0 100 200 300 400 500 0 100 200 300 400 500

-25 0

100

200

300

400

500

0

FADC_CHANNEL_13 Ev_35
10 5 0

CH_13 Ev_35 Entries 512

FADC_CHANNEL_14 Ev_35
20

CH_14 Ev_35 Entries 512

Deconvoluted ASum

ASum_Dec Entries

Deconvoluted DSum

512

200

DSum_Dec Entries

512

6 10 4 2 0 100

0

0

-5 -10 -10 -15 -20 0 100 200 300 400 500 -20 -4 -200 -6 -30 0 -8 0 -2 -100

100

200

300

400

500

100

200

300

400

500

0

100

200

300

400

500

FADC_CHANNEL_1 Ev_592

CH_1 Ev_592 Entries 512

FADC_CHANNEL_2 Ev_592

CH_2 Ev_592 Entries 512

FADC_CHANNEL_3 Ev_592
10 8 6

CH_3 Ev_592 Entries 512

FADC_CHANNEL_4 Ev_592
10

CH_4 Ev_592 Entries 512

4

4

8 6 4 2 0

2 2 0 0 -2

4 2 0

-4 -6 0 100 200 300 400 500

-2 -2 -4 -4 0 100 200 300 400 500 0 100 200 300 400 500

-2 -4 -6 0 100 200 300 400 500

FADC_CHANNEL_5 Ev_592
10

CH_5 Ev_592 Entries 512

FADC_CHANNEL_6 Ev_592

CH_6 Ev_592 Entries 512

FADC_CHANNEL_7 Ev_592
80 60 40 20

CH_7 Ev_592 Entries 512

FADC_CHANNEL_8 Ev_592
10 8 6 4 2

CH_8 Ev_592 Entries 512

8 6

5 4

0 0 2 0 -5 -2 -10 -4 -100 -6 0 100 200 300 400 500 0 100 200 300 400 500 -120 0 100 200 300 400 500 -20 -40 -60 -80

0 -2 -4 0 100 200 300 400 500

FADC_CHANNEL_9 Ev_592
8 6 4 2 0 -2

CH_9 Ev_592 Entries 512

FADC_CHANNEL_10 Ev_592
10 8

CH_10 Ev_592 Entries 512

FADC_CHANNEL_11 Ev_592
10 8

CH_11 Ev_592 Entries 512

FADC_CHANNEL_12 Ev_592
12 10 8

CH_12 Ev_592 Entries 512

6 6 4 4 2 2 0 0 -2 -2 -4 -4 -6 100 200 300 400 500 0 100 200 300 400 500 0 100 200 300 400 500

6 4 2 0 -2 -4 0 100 200 300 400 500

-4 0

FADC_CHANNEL_13 Ev_592
8 6

CH_13 Ev_592 Entries 512

FADC_CHANNEL_14 Ev_592
15 10 5

CH_14 Ev_592 Entries 512

Deconvoluted ASum
6

ASum_Dec Entries

Deconvoluted DSum

512

200

DSum_Dec Entries

512

4 2

100

4 0 2 0 -5 -2 -10 -4 -2 -4 0 -15 -20 100 200 300 400 500 0 100 200 300 400 500 -6 0 100 200 300 400 500 0

0

-100

-200

-300 0

100

200

300

400

500

9


5 Conclusion
There is a Data Acquisition system recording fourteen partial sums and the total sum of the signals from the PMTs of Borexino. This system is able to register the event of a Supernova and to record the neutrons with a high eciency. It can run 24 hours per day, given good trigger conditions. Future improvements will allow it to record Muon Veto tags and to synchronize with the main Borexino data.

6 Acknowledgement
We are very grateful to Sergei Sukhotin (Kurtchatov Institute) for his help in analysing the data.

10