Prsentation PowerPoint

1
Electronic Design for a proton trigger and a
Time of Flight measurement with the Recoil
Proton Detector for the hadron program
Nicole dHose and CEA-Saclay (SPhNSEDI)
and Bonn, Mainz, Warsaw
(groups involved in the GPD recoil
prototype)
CERN, 30 May, 2007
2
Status of the present RPD
Electronics solution - readout time and
amplitude measurements
with standard COMPASS equipement -
proton trigger (coincidence A-B selection of a
proton small time window)
3
Time of Flight concept
zB tB
tdoB
tupB
LB 106 cm thick.1cm
65cm
zA tA
tupA
tdoA
LA 50 cm thick.0.5cm Ltarget 40 cm
?3.5cm
beam
12cm
target
zB (tupB - tdownB) VB/2 LB/2 Coruptw
Cordowntw Offup-Offdown
tB (tupB tdownB)/2 LB/2VB Coruptw
Cordowntw OffupOffdown
To be precisely determined (tw time walk
correction)
4
Criteria for a ToF measurement
Required resolution 300ps
Time walk corrections
t


Thr0


Thr1


Thr2
15ns
5ns
Methods 1- Sampling of the
signal each ns (with MATACQ board) -
necessity if more than 1 hit - not yet
available 2- Time over 3 thresholds (Reiner G.s
idea) Amplitude measurement
Vmax
V
5
Expected counting rates in each scintillator
For the recoil foreseen in the GPD program
target 250cm acceptance from 15 to 90
in 5?108 ? / spill of 5s (100 MHz)
- counting 100 MHz in the total
acceptance (in 24 elts!) - (4
MHz?150ns) ½ of all the scintillators are
touched - the scintillators
have less than 4 hits in 150 ns
For the RPD
in the hadron program target 40cm with
same thickness acceptance from 42.5 to
90 in 5?107 ? / spill of 10s (5 MHz)

Reduction by a factor 200 for the counting rates
We can expect really less than 1 hit in one
scintillator in 50 ns Simulation in progress
with low threshold (300 keV)
6
Objectifs in 2007 ? Objectifs in 2010
Recoil detector size 1m
size 4m
A 5mm B 10mm A 4mm B 50mm
Identification of protons reasonable
perfect
Proton Range in momentum 290 ? 800 MeV/c
270 ? 800 MeV/c in
angle 45 ? 90 36 ? 90
Resolution in ToF 300 ps
lt 300ps
Obtained results with the prototype in 2006 with
the MATACQ at CERN (with position reference)
at Saclay (with time reference)

• ?(tupB - tdownB) 200 ? 6 ps ?(tupB
  tdownB) 145 ps ? 10 ps
• ?(tupA - tdownA) 270 ? 6 ps
• ToF ? (tupB tdownB) - (tupA tdownA)
• 315 ? 12 ps
• to be still improved but

7
Results on the Ring B prototype

Saclay Protvino Number of ?e
( _at_ 1MIP) 60 (cosmics) 120 (hadron
beam) Attenuation length 70-80
cm 85 cm Speed of light in scint.
6.1 cm/ns 6.3 cm/ns
?(tupB tdownB) 250 ps is achieved at Saclay
(with a sampler and a time
reference) it was 430 ps at Protvino by a
diff. method
?ToF 350 ps can be expected
Production of Ring B scintillators is started
8
Requirement for the hadron program a smallest
timing window for the trigger to reduce the rates
Pulse shape and timing for the Ring A
The A downstream signal is nearly independent of
the vertex position The A
downstream signal is considered as time leading
in all the following coincidences
9
Pulse shape and timing
all the signals within 50ns and relatively
time correlated
10
Energy lost in the A and B scintillators
A thickness 5mm B thickness 10mm
Tests in 2001
protons

275 410 760 MeV/c
For protons
50mm thickness for B would have been necessary
for a better p/? separation
11
Energy lost in the A and B scintillators
Dynamical range In A From 1 to 20 MeV Max
attenuation factor 0.8 (?2m) ? detection from
0.4 to 20 MeV
ratio 50 In B From 2 to 40 MeV Max attenuation
factor 0.28 (?0.8m) ? detection from 0.28 to 40
MeV ratio 150 PMT
Output signal for A from 100mV to 5V
for B from 30 mV to 5V !!!
Any improvement in the B attenuation length would
be welcome
12
Energy lost in the A and B scintillators
Tests in 2001
protons
due to attenuation length EB ? (EupB
EdownB) or rather less rigorously, but easier
for electronics EB EupB EdownB
double threshold coincidence
EA gt SAHigh and EB gt SBLow or
EA gt SALow and EB gt SBHigh
13
Possibility for a trigger scheme and ToF
measurement
14
Possibility for a trigger scheme and ToF
measurement
The time of all the coincidences is determined
by Aido and the resulting jitter has to be lt 3
ns
Splitter discriminator with CFD with
adjustable threshold adjustable
threshold
ADC?2
The time dispersion due to the different
channels has to be lt 3 ns
TDC
Aido
Coincidence OR
FPGA
ALow
AiL B2iH
ADC?2
Ai B2i
trigger
TDC
Aiup
AiH B2i L
AHigh
12?3 OR
12?3?2 COINC
ADC?2
TDC
B2ido
B2ido B2iup
Many adjustable delay
TDC
B2iup
CFD with adjustable threshold
analogic sum
ADC?2
Proton Trigger  in yellow 
A_down  has the smallest timing dispersion and is
time leading in all the coincidences. CFD will
still improve the size of the time window
Threshold on the Sum (BdownBup) will avoid the
huge pb of attenuation lenght The double
threshold (Low and High) coincidence between A
and B will eliminate electrons, pions 500ns
time to deliver the proton trigger (70m cables
to the trigger barrack) gt 50ns before FPGA
10ns for FPGA 30ns for coinc beamproton 280ns
cables

15
FPGA
AiL B2i-1 H
Ai B2i-1
AiH B2i-1L
AiL B2i H
Ai B2i
AiH B2i L
12 OR for the TRIG GER
AiL B2i1H
ADC
Ai B2i1
TDC
Ai1do
AiH B2i1 L
SAi1L
ADC
TDC
Ai1up
Ai1L B2i1H
Ai1 B2i1
SAi1H
Ai1H B2i1L
(12-2) ?2 lignes
Ai1L B2i2H
Ai1 B2i2
B2i-1
..
Ai1H B2i2 L
B2i-1do B2i-1up
ADC
TDC
Ai1L B2i3H
B2ido
Ai1 B2i3
Ai1H B2i3 L
10 other groups
B2iup
TDC
FPGA very compact solution

ADC
Control of these individual coincidences
..
B2i1do B2i1up
B2i1
(24-3) ?2 lignes .
16
Tests on FPGA
Transit time and dispersion of the signal through
the FPGA
Test done at Saclay by Denis Calvet (DAPNIA-SEDI)
with a FPGA build from an evaluation kit using
Xilink Virtex 4 (FX60-10)
ns
Transit time 10ns -------
Dispersion 3ns without optimisation
Question what is the requirement for a maximum
dispersion of the trigger ?
Channel number
Test done also by Krzysztof Zarembas group in
Warsaw Same transit time 8.4 ns with Xilink
XC95144XL-5-TQ144
17
CAEN module V1495 general purpose VME board
Delay input-output 15ns Programmable with
Quartus Timing of individual channels can be
checked and optimized In the design phase The
dispersion can be larger due to the use of
several boards (for the total inputs)
The realisation at Saclay (DAPNIA-SEDI)
has started
18

• On 23 February 2007 Andrea
  Ferrero mentioned
• The gates of the A and B signals in coincidence
  are set to 4 and 20ns
• respectively
• A time jitter of 2.5ns can be achieved with a
  careful tuning of
• delays and coincidences

Have these requirements to be achieved
necessarily? Until which values can these
conditions be relaxed? ? jitter
smaller than 5ns timing
window smaller than 30ns
19
Final selection and costs

20
RPD read out done by Damien Neyret
in k
1 9U VME crate (needed ???), near MM rack or
Silicon rack 10 or 0? 1 VME CPU
(needed ???)
5? 7 translator ECL/LVDS
7 4
Catches into new VME crate or into RICH-MaPMT VME
crate 24 16 TDC-CMC
? 4
TCS receivers they are no more built! 1
multiplexer
2? 3 SADC modules

6 1 Gesica
4 2 Odin
Slinks to be exchanged with HOLA 2
fiber pair links, in MM rack or Silicon rack

Total lt60 k
21
planning All the readout system has to be
ready for the commissioning of the RPD in the
muon beam halo which will begin this year at the
end of September.
22
(No Transcript)
23

• Comments on the
  calibration methods
• with cosmics,
• with laser light distributed to each
  scintillator,
• with muon tracks of determined direction in the
  muon beam halo,
• with elastic events ? p ? ? p

24
Rough calibration with a green Laser
illuminated a totaly diffusive sphere
n optical fibers connected to the centre
in each scintillator
to control easily the electronics and to get a
quick determination of timing offsets.
Necessity to put the connectors before wrapping
the scintillators
25
Study with cosmics at Saclay
µ
Ref1 Scint to study Ref2
PM
t ( tL ref1 tR ref1 tL ref2 tR ref2) / 4
- (tL sc tR sc) / 2
The resolution on t is s2 s ref 2 s sc 2

with s ref ½ s TOFref1/ref2
z ( tL ref1 - tR ref1 tL ref2 - tR ref2) / 4
- (tL sc - tR sc) / 2
26
Test bench at Saclay for the cosmics
calibration
We can test Protvino and Mainz scintillators
light guides PMTs
27
Our ultimate goal for recoil detector studies
study in details the shape of the analog signals
collected on a few inner and outer elements to
continue our study with a new generation of
MATACQ boards to sample the signal and quantify
the background (to be taken into account in the
splitter with an extra ouput) with the
hydrogen target, the hadron recoil detector and
the positive and negative muon beams
during one or two shifts

28
Possibility for a trigger scheme and ToF
measurement
Splitter discriminator with CFD with
adjustable threshold adjustable
threshold
The time of all the coincidences is determined
by Aido and the resulting jitter has to be lt 3
ns
ADC
The time dispersion due to the different
channels has to be lt 3 ns
ADC
TDC
Aido
FPGA
Coincidence OR
test
ALow
AiL B2iH
ADC
ADC
Ai B2i
trigger
TDC
Aiup
AiH B2i L
AHigh
test
12?3 OR
12?3?2 COINC
ADC
ADC
TDC
B2ido
test
B2ido B2iup
Many adjustable delay
TDC
B2iup
CFD with adjustable threshold
analogic sum
ADC
test
ADC
In blue 3 time measurements after 3
thresholds 2 energy loss
determinations in 2 dynamical ranges In yellow
Proton trigger performed with a FPGA
29
Final selection for the RPD electronics

Discriminators 3 ? 72 14 modules Lecroy 4416
for example
CERNs pool 12 months 4
k 7 translator ECL/LVDS
7 k F1
TDC multihit 4 catches with mezannine
24 k
SADC 5 SADC modules
pre-shapers 1 gesica 64 k
this
solution has to be tested (help of Warsaw?)
Splitters analog sum CFD discriminator
Mainz/Bonn


?? FPGA 1rst tests of the CAEN module V1495
Saclay


10 k number of crates, translators, cables,
connectors Saclay
? to be determined before end of june
??



Saclay lt 60 k
Total costs between 60 and 100 k
The readout system has to be ready for the
commissioning of the RPD in the muon beam halo
which will begin at the end of 09-2007.