Transcript Diapositive 1 - indico in2p3

R&D status of the very front end ASIC
Tilecal week (7 October 2011)
François Vazeille
Jacques Lecoq, Nicolas Pillet, Laurent Royer and Irakli Minashvili
 Status by June 2011
 New results
 Conclusion and next steps
1
2
Status by June 2011
 The first chips in IBM 130 nm technology were made (CERN order)
then tested at Clermont-Ferrand
- First one: only a current conveyor.
- Second one: current conveyor + shaping stage.
With 3 different gains (1, 8, 64) for the two chips.
0.8 cm
1.7 cm
3
Chip 1
Chip 2
 Main results
-
They work.
A good correlation between the 3 gains.
Dynamics fully effective.
Good agreement with simulations.
BUT:
- Large offset for the 3 gains: solved (due to a lay out problem).
- Oscillation problems  length of wire bonding to the package:
cured by adding an external R-C.
- Linearity not well tested because the test bench generator is not enough precise.
- Pulse width not optimized with respect to Tilecal needs,
because of the cure of the oscillation problems (pulse tail).
 New studies of integrator
- Needs of a very high gain and of a very small offset.
- A compromise choice could associate 2.7 V “analog” transistors
and 1.6 V transistors (low power) .
4
 Next steps foreseen by June
□ Chip 2:
- Tests behind a PMT at Clermont-Ferrand using Laser pulses (Scheduled in July)
 More realistic situation.
 A first approach of the absolute scale/Chip design.
-Tests with cosmics at CERN inside a Tilecal module (Scheduled in September)
The most realistic detector using real particles.
 A better approach of the absolute scale/Chip design.
 A good learning exercise around a Tilecal module.
□ Integrator
- Simulations of the new design.
- Goal:
 To reach the performances of the Chicago amplifier (discrete solution):
Gain of 50 000 and max offset < 1 mV.
5
New results
Chips 1 and 2
 Identity names
Chip1  FATALIC 1
Chip 2  FATALIC 2
Front end for Atlas Tile cAL Integrated Circuit (IC for every Clermont chips)
 Laser pulses at Clermont-Ferrand
(July)
- The first time behind a PMT and with light.
- Large uncertainty on the Laser amplitude
(10 to 20%).
But first check of the dynamical range
(in equivalent charge) for the 3 gains.
6
 Cosmics at CERN (September): building 175.
7
D cell
Long cable
Super-drawer LBA65 #27
PMT # 1 Hvout = 678.5 V (setting: 678.0 V)
calibrated with Cesium.
LED signal: Oscillation
due to the 4 m long cable,
with a circuitry not suited to it.
8
Muon signals: first particle events seen by FATALIC2
~ One event every 2 minutes.
Choice of the high gain (64)  saturation of many events above 500 mV
It was not the best choice !
9
- D cell   1 GeV muons (Claudio information)  50 pC
- Gain 64: 625 µA max or 12 pC
 Saturation of many pulses.
Nevertheless,
these results do confirm the first approach of the requested scale:
- Gain 64: 625 µA or  12 pC max.
- Gain 8: 6.25 mA or  120 pC max.
- Gain 1: 62.5 mA or  1200 pC max.
10
 New Integrator simulations of amplifier
Bode diagram:
Gain > 50 000
Output offset (gain = 1 000):
86 µV converted to input
(Monte carlo with 500 events)
 Results fitting the goals (Chicago reference).
11
Conclusion and next steps
The chips have “seen” light: laser (Clermont-Ferrand)
and cosmic particles (CERN).
 These tests put the chips in realistic environments,
besides, it was a good learning for everybody.
The absolute choices of the gains and ranges
are not far away the physical needs.
 New steps are scheduled.
12
 New foundry order on November 7th  FATALIC 3
Circuit corrected + Integrator amplifier.
+ inside: capacitor and resistor used to reduce parasitic self effects
due to cable length.
Expensive:  10000 $
 Parallel studies
- Improvement of the test bench
(In particular by buying an “expensive” generator for precise linearity tests).
- Pulse width, peaking time.
- ADC studies.
- CIS part.
- Analog output for trigger.
- Environment (3in1 card, LV regulators, MB-2, data format, DCS, connections…)
with contributions of both micro-electronics and electronics people.
3-in-1
FE-ASIC
QIE
MB-1
MB-2
MB-3
DB
STD-DRAWER
13
.
MINI-DRAWER
.
… towards a very aggressive planning !!!
FE-ASIC alternative
Feasibility tests
Done
Design ASIC FATALIC 1 & 2 in IBM 130 nm
Done
Test FATALIC1 & 2 @ C-F
Done
Test FATALIC2 in Bld 175
Q3-11
Design FATALIC3
Q4-11
Test FATALIC3 @ C-F
Q2-12
3in1 prototype design
Q4-11
3in1 prototype tests at C-F
Q1-12
Prototype design of MB-2 (similarto3-in-1 DB) Q4-11
MB-2 prototype tests at C-F
Q2-12
System tests in Bld 175
Q3/Q4-12
Design of FATALIC4 (ADC)
Q2-12
FATALIC4 tests at C-F
Q4-12
Radiation tests
Q1/Q2-13
Final designs for production
Q3-13
Tests in test beam
Q3-14
Comments:
Depending from the true upgrade schedule, the above “final” design
could be reviewed a new time if more time is available.
Question: is it judicious to speed up the studies
while Phase 2 is starting … many years later ?
14