Transcript Document

FPIX2: A rad-hard pixel readout
chip for BTeV
f
Homestead
David Christian
Fermilab
Vertex 2000
September 14, 2000
• New Fermilab collider experiment (approved in June, 2000)
• Will be installed in a new interaction region at C0
startup ~ 2006-2007
• Primary goals: study of CP violation, mixing
& rare decays in b and c systems
• “Two arm” forward spectrometer: central dipole with pixel planes
in vacuum – as close to the beam as radiation damage will allow
• Level 1 trigger based on tracks & vertices reconstructed using pixel data
Vertex 2000
David Christian
p2
Pixel readout chip for BTeV: requirements
• Radiation hard
• Optimized for 132 ns crossing time
• Able to tolerate large sensor leakage current
• High speed zero-suppressed readout
• We expect the chip we develop for BTeV to be
suitable for use by CDF and D0 also
Vertex 2000
David Christian
p3
FPIX Designers
Abderrezak Mekkaoui:
Lead engineer
(analog design +
overall responsibility)
Vertex 2000
Jim Hoff:
Digital design
David Christian
p4
FPIX Roadmap
• Pixel size = 50m x 400m (matches ATLAS n+ on n test sensors)
• Target rad-hard technology = Honeywell 0.5m CMOS (SOI)
(3 metal, 3.3V) (1 metal layer used for shield between sensor & R/O chip)
• FPIX0 (1997) HP 0.8m CMOS
•Close to final analog front end
•R/O pixel includes a peak sensor – digitized off chip
•Array size = 12 x 64
•Bench tests and beam tests
• FPIX1 (1998) HP 0.5m CMOS
•Optimized front end
•4 comparators per cell (2-bit FADC)
•New fast R/O architecture, allows both self-triggered and externally-triggered operation
•Array size = 18 x 160
•Bench tests and beam tests
• Then (Dec, 1998), a change of plans
•Try to use deep-submicron CMOS
•All subsequent prototypes should be rad-hard.
Vertex 2000
David Christian
p5
FPIX2 Roadmap
• 0.25m CMOS
• (5 metal [6 possible], 2.5V)
• Design for 2 vendors (“lowest common denominator” design rules):
• “CERN” – Very favorable contract, but problems with US Gov. restrictions
• Taiwan Semiconductor Manufacturing Corp (TSMC) – Available through MOSIS
• PreFPIX2-T (1999) TSMC 0.25m CMOS
•New analog front end, with new leakage current compensation strategy
•8 comparators per cell (3-bit FADC); no EOC logic included
•Array size = 2 x 160
•Bench tests (radiation exposure)
• PreFPIX2-I (2000) “CERN” 0.25m CMOS
•Same front end
•Complete “core” – including new, simplified EOC & R/O (self-triggered only)
•Array size = 18 x 32
• PreFPIX2-T2 (2000) TSMC 0.25m CMOS
•New programming interface
•Internal DAC’s – no external currents required; only external voltages are 2.5V & ground.
•Array size = 18 x 64
• FPIX2 (2001) 0.25m CMOS - Final BTeV R/O chip!!??
Vertex 2000
David Christian
p6
FPIX0/FPIX1 front end*
First stage feedback element in box:“Synthetic Resistor” = transistor which
acts as a resistor for small signals and as a constant current source
(discharging the feedback capacitor) for large signals (or large leakage current).
Cinj
Vin
50 W
to grnd
n
p
Chip boundary
Pixel detector
unit cell boundary
-V(detector bias)
* See Blanquart, et al. NIMA 395, p313 (1997)
Vertex 2000
David Christian
p7
FPIX0 feedback & leakage current compensation
FPIX0 Insensitivity to leakage current:
High gain cell can be adjusted
so that rise time & amplitude
with ~20 nA/pixel of leakage
current match the rise time &
amplitude with no leakage current.
High gain cell (9,0)
I_fb = 7 nA
Leakage current ~ 0
Leakage current ~20 nA/pixel
High gain cell (9,0)
I_fb = 0.5 nA
Leakage current ~ 0; I_fb = 7 nA
Leakage current ~20 nA/pixel
Leakage current ~ 100 nA/pixel
Vertex 2000
David Christian
p8
FPIX1 front end layout
Vdda
Iff
Bump bond
Pad (feedback
cap is
underneath)
Sensor
Inject
Test
charge injection
capacitor
Vertex 2000
cascode
1st transistor
feedback and leakage
current compensation
transistors: NOTE
aspect ratio!
David Christian
p9
Radiation damage to CMOS transistors
Positive charge trapped in the oxide layer effectively biases the transistors.
Gate oxide
Gate
n+
Source
(normally connected to gnd)
n+
Drain
p bulk
Conductive channel is induced by
positive voltage applied to the gate
“Threshold voltage” shifts with exposure to radiation
BUT, the effect gets smaller as the oxide gets thinner (with smaller feature size)
… by 0.25m the threshold shifts are small enough to be “benign.”
Vertex 2000
David Christian
p10
Radiation induced leakage current
Trapped charge in the field oxide also causes leakage current
in nmos devices by inducing an n-channel in the p-bulk.
source
drain
pmos leakage current does not increase (glass charges +; doesn’t
induce a p-channel).
Vertex 2000
David Christian
p11
Rad-hard nfet layout (very schematic!)
“gate all around” layout (guard rings to prevent latchup not shown)
Large W/L is “easy”
Small W/L is hard
Or, impossible!
Vertex 2000
David Christian
p12
FPIX1 front end layout
Vdda
Iff
Bump bond
Pad (feedback
cap is
underneath)
Sensor
Inject
Test
charge injection
capacitor
Vertex 2000
cascode
1st transistor
feedback and leakage
current compensation
transistors: NOTE
aspect ratio!
David Christian
p13
FPIX2 feedback solution
•
•
•
•
•
One NMOS feedback transistor
biased by a global voltage VFF.
VFF generated such as to track
(to the 1st order) the preamp
DC level shifts due to global
changes (process,
temperature…)
Feedback is current controlled
as before. This current can be
much higher than in the
previous scheme.
It is more reliable to work with
higher currents.
Leakage current compensation
assured by a separate scheme
(next slide).
Vertex 2000
Vff
Vbp
One bias c ell per c hip
Sens or
Vbp
Iff
Vff
Injec t
Vbn
Tes t
David Christian
p14
FPIX2 leakage current compensation
Vdda
gmc
• Compensates one polarity
only (n+ on n sensor)
+
Vff
• Differential amplifier in feedback must
have VERY low bandwidth!
Rf
Cf
Sensor
-A
Iin
Vo1
Ileak
Inject
Test
Vertex 2000
David Christian
p15
(pre)FPIX2 front end layout
charge
Injection
capacitor
Capacitors used to limit frequency response of op amp in
Leakage current compensation circuit
Vdda
+
Vff
Bump bond
Pad (feedback
cap is
underneath)
+
Sensor
Inject
Test
Vref
leakage current
compensation
op amp
Vertex 2000
cascode
1st transistor
feedback
resistor
(transistor)
David Christian
p16
(pre)FPIX2 pulse shapes
Qin=3260e- channel R. 3 different feedback currents.
Vertex 2000
David Christian
p17
(pre)FPIX2 leakage current compensation
After the first nA no change in the response is observed !
Vertex 2000
David Christian
p18
(pre)FPIX2 Irradiation Tests
• Just starting irradiation tests!
•1st test used 60Co source at Argonne
• After ~33 MRad:
•Circuits are fully functional
•No degradation in speed
(inferred from kill/inject shift register operation)
•Less than 10% change in analog power dissipation;
less after irradiation.
(as expected; due to small VT change in PMOS)
Vertex 2000
David Christian
p19
(pre)FPIX2: front end response, before and after irradiation
No change in settings (bias voltages, feedback current,…) before/after irradiation
Vertex 2000
David Christian
p20
(pre)FPIX2 Noise and discriminator threshold distributions
=> Practically no change in noise and threshold dispersion.
=> 200 e- change in the threshold voltage.
Vertex 2000
David Christian
p21
Next Steps
• Irradiation of preFPIX2 chips bump bonded to sensors using protons
(Indiana University cyclotron – 200 MeV):
• Total dose effects
• Single event effects
• Latchup
• Single event upset (single bit errors)
• Tests of preFPIX2-T2
• DAC’s
• VLDS I/O
• Final Specification of FPIX2
• Array size (18 x 160?)
• Output format
• serialized, high speed VLDS?
• point to point?
• drive signals out of high-radiation environment?
Vertex 2000
David Christian
p22