deeper sampling cmos transient waveform recording asics matt andrew, kurtis nishimura, gary s....
TRANSCRIPT
![Page 1: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/1.jpg)
Deeper Sampling CMOS Transient Waveform Recording ASICs
Matt Andrew, Kurtis Nishimura, Gary S. VarnerUniversity of Hawai’i, and the
Large Area Picosecond PhotoDetector CollaborationTIPP 2011, June 10, 2011
The IRS and BLAB3 Deep Storage ASICs for UHE Radio Neutrino and Next Generation Collider Particle Identification
1
![Page 2: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/2.jpg)
2
History• Traditional “crate based” electronics
– Gated Analog-to-Digital Converters
– Referenced “triggered” Time-to-Digital Converters
• High-rate applications– “pipelined operation”– Low-speed, low-resolution sampling
• High channel counts– Motivation to reduce cabling– Integrate electronics onto detector elements
Det chan
Q-ADC
Disc.TDC
Trigger
Issues: cost, power, resolution, data volume
Det chan FADC
Why Waveform Sampling?
![Page 3: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/3.jpg)
3
Switched Capacitor Array Sampling
Input
Channel 1
Channel 2Few 100ps delay
• Write pointer is ~few switches closed @ once
20fF
Tiny charge: 1mV ~ 100e-
![Page 4: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/4.jpg)
4
“Oscilloscope on a chip”
• 2 GSa/s, 1GHz ABW Tektronics Scope• 2.56 GSa/s LAB
![Page 5: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/5.jpg)
5
Easy access to Waveform sampling
WFS ASICCommercial
Sampling speed 0.1-6 GSa/s 3 GSa/s
Bits/ENOBs 16/9-13+ 8/7.4
Power/Chan. <= 0.05W Few W
Cost/Ch. < $10 (vol)
> 100$
![Page 6: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/6.jpg)
6
An Intrinsic Limitation
We aren’t going to put Analog Devices out of business
No power (performance savings) for continuous digitization
“analog down conversion” For most “triggered” ‘event’ applications, not a serious drawback
![Page 7: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/7.jpg)
7
Gigasamples/s, but Nyquist?Difficult to couple in Large BW (input C is deadly)
f3dB = 1/2pZC
Input Coupling versus total input Capacitance
0
0.5
1
1.5
2
2.5
3
3.5
0 500 1000 1500 2000 2500 3000
Total input Capacitance [fF]
An
alo
g B
and
wid
th [
-3d
B f
req
uen
cy]
R_S = 50Ohm
• More than 3pF input (ESD protection alone often more) limits ABW < 1GHz
• As a result, first generation WFS ASICs a few hundred to 1k samples (ATWD, SAM, DRS, LABRADOR…)
• Is this an fundamental limitation?
![Page 8: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/8.jpg)
8
Why Deeper sampling? ~ km
1018 eV nm
![Page 9: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/9.jpg)
9
Deeper storage: Buffered LABRADOR (BLAB1) ASIC
• Single channel• 64k samples deep,
same SCA technique as LAB, no ripple pointer
• Multi-MSa/s to Multi-GSa/s
• 12-64us to form Global trigger
3mm x 2.8mm, TSMC 0.25um
Arranged as 128 x 512 samplesSimultaneous Write/Read
![Page 10: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/10.jpg)
10
BLAB1 High speed Waveform sampling
• Comparable performance to best CFD + HPTDC
• MUCH lower power, no need for huge cable plant!
• Using full samples reduces the impact of noise
• Photodetector limited 6.4 psRMS
CH1
CH2
NIM A602 (2009) 438
![Page 11: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/11.jpg)
Ice Radio Sampler (IRS) / Buffered LABRADOR 3 (BLAB3) Specifications
32768 samples/chan (8-32us trig latency)8 channels/IRS ASIC8 Trigger channels
~9 bits resolution (12-bits logging)64 samples convert window (~16-64ns)
1-4 GSa/s1 word (RAM) chan, sample readout
16 us to read all samples100's Hz sustained readout (multibuffer)
• Difference between IRS/BLAB– BLAB has input amplifier
– IRS doesn’t really use internal trigger capability
11
![Page 12: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/12.jpg)
12
IRS/BLAB3 Single Channel
• Storage: 64 x 512 (512 = 8 * 64)
• Sampling: 128 (2x 64) separate transfer lanesRecording in one set 64, transferring other
(“ping-pong” “2 stage sampling”)
• Wilkinson (32x2): 64 conv/channel
![Page 13: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/13.jpg)
13
IRS/BLAB Die photo
8x RF inputs(die upside down)
5.82mm
7.62mm
32k storage cells per channel(512 groups of 64)
![Page 14: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/14.jpg)
Analog bandwidth
Analog BW~1GHz
14
![Page 15: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/15.jpg)
ARA
Trig
/Dig
Ele
ctro
nics
- 17
-AU
G-2
010
Input coupling simulation (35fF sample)
~1 GHz stored
Onto chip (flip chip)
Magnitude [dB]
15
![Page 16: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/16.jpg)
16
Wilkinson ADC – easy to integrate on-chip
• Excellent linearity• Basically as good as can make current source/comparator
12-bit ADC
• No missing codes• Linearity as good
as can make ramp• Can bracket
range of interest
Run count during ramp
Modified! (on-chip clock/counter)[~0.7 GHz]
![Page 17: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/17.jpg)
17
Storage Cell - compact
• Diff. Pair as comparator– Density ~ 25k storage cells/mm^2 (0.25um)
![Page 18: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/18.jpg)
18
On chip Wilkinson Clock
~1us to 9-10 bits~4us to 12 bits512 ADC in parallel
![Page 19: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/19.jpg)
Linearity Calibration
19
CAM and/or LUT in FPGA
![Page 20: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/20.jpg)
• Need dT calibrations – but only 128 per channel • Any way to automate? (see K. Nishimura talk)
20
Example: 100 MHz sine wave input
![Page 21: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/21.jpg)
21
More like 10% effect at 3.2GSa/s
J. Davies –UC London Not a small effect
![Page 22: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/22.jpg)
ASIC Amplification? # chan Depth/chan Sampling [GSa/s] Vendor Size [nm] Ext ADC?
DRS4 no. 8 1024 1-5 IBM 250 yes.
SAM no. 2 1024 1-3 AMS 350 yes.
IRS2 no. 8 32536 1-4 TSMC 250 no.
BLAB3A yes. 8 32536 1-4 TSMC 250 no.
TARGET no. 16 4192 1-2.5 TSMC 250 no.
TARGET2 yes. 16 16384 1-2.5 TSMC 250 no.
TARGET3 no. 16 16384 1-2.5 TSMC 250 no.
PSEC3 no. 4 256 1-16 IBM 130 no.
PSEC4 no. 6 256 1-16 IBM 130 no.
Success of PSEC3: proof-of-concept of moving toward smaller feature sizes.• Next DRS plans to use 110nm; next SAM plans to use 180 nm.
Now a variety of options…
22
![Page 23: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/23.jpg)
23
Future Prospects• Expect many other designs in
future– Barriers to entry are low– Many different reference designs out there
• Challenges (R&D continues):– Fine timing (~ 1ps)– Larger dynamic range– Deeper (continuous) sampling– Faster, sparsified readout– Calibration
• Key enabling technology– Large telescope arrays (CTA)– > 100km3 neutrino detectors– PID, Fast x-ray detectors
![Page 24: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/24.jpg)
24
Back-up slides
![Page 25: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/25.jpg)
25
Underlying Technology• Track and Hold (T/H)
• Pipelined storage = array of T/H elements, with output buffering
2v V1=VQ=Cs.V1
N capacitorsWrite Bus
Return Bus
1
N caps
Vout=A / (1+A) * Q/Cs=V1 * A/(1+A)
3
Bottom Read BUS
4
Top Read Bus
Cs
CAnalog Input
SampledData
T/H
![Page 26: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/26.jpg)
26
G. V
arne
r -- D
eepe
r Fas
t Wav
efor
m S
ampl
ing
-- p
icoS
econ
d W
S in
Kra
kow
Real MCP-PMT Signals (with BLAB2)
7-8 psRMS
Residual Time Walk
Rather robust for amplitude invariant signals,TOF still hard, but can shape extract
![Page 27: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/27.jpg)
27
G. V
arne
r -- D
eepe
r Fas
t Wav
efor
m S
ampl
ing
-- p
icoS
econ
d W
S in
Kra
kow
Time Difference Dependence on Signal-Noise Ratio (SNR)
0
2
4
6
8
10
12
14
16
18
20
10 100 1000
Signal Noise Ratio
Tim
e D
iffe
ren
ce R
eso
luti
on
[p
s]Simulated Performance vs. SNR
kTZ
300MHz ABW, 5.9GSa/s
![Page 28: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/28.jpg)
28
IRS Input Coupling
• Input bandwidth depends on 2x terms– f3dB[input] = [2*p*Z*Ctot]-1
– f3dB[storage] = [2*p*Ron*Cstore]-1
Input Coupling versus total input Capacitance
0
0.5
1
1.5
2
2.5
3
3.5
0 500 1000 1500 2000 2500 3000
Total input Capacitance [fF]
An
alo
g B
and
wid
th [
-3d
B f
req
uen
cy]
R_S = 50Ohm
Input coupling versus frequency
-10
-9
-8
-7
-6
-5
-4
-3
-2
-1
0
0.1 1 10 100
Frequency [GHz]R
elat
ive
amp
litu
de
[dB
]
C=15fF,Ron=1k
C=15fF,Ron=5k
C=25fF,Ron=1k
C=25fF,Ron=5k
![Page 29: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/29.jpg)
29
Wilkinson Clock Generation
• Strictly only 5 channels necessary– 4x antenna, 1x reference channels
– Could interleave for twice depth, or multiple reference channels
![Page 30: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/30.jpg)
30
Wilkinson Recording
Start = start 0.5-1.5GHz Clock Ripple counter (run as fast as can)
![Page 31: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/31.jpg)
31
Triggering
• Need 9th channel for monitoring
![Page 32: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/32.jpg)
32
Triggering – same as previous resultsTrigger 1-shot Width Adjust
1
10
100
0 20 40 60 80 100 120
Discharge Current [uA]
Ou
tpu
t W
idth
[n
s]
T_1_TRG
Power (T_1_TRG)
• Monitor 9th channel (uses Ch.1 threshold) to compensate for temperature dependence
![Page 33: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/33.jpg)
33
Hit Processing numbers Assume: 100kHz charged track hits on each bar
~32 p.e./track (1% of 100ns windows)30kHz trigger rateEach PMT pair sees <8> hits240k hits/sEach BLAB3 has an average occupancy <1 hit (assume 1)400ns to convert 256 samples16ns/sample to transferAt least 16 deep buffering(Markov overflow probability est. < 10-38)
Each hit = 64samples * 8bits = 512bits~125Mbits/s (link is 3.0 Gb/s ~ x30 margin)
BLAB3 ASIC
8
Trans-Imp Amps 512 x 64 samples
Per channel
Fast conversionMatrix (x256)
BLAB3 sampling
Improvements based uponLessons learned from BLAB2
Plan to model in standard queuing simulator, but looks like no problem(CF have done same exercise with Jerry Va’vra for 150kHz L1 of SuperB and can handle rate)
![Page 34: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/34.jpg)
34
Front-end Electronics studies
J-F Genat, G. Varner, F. Tang, H. FrischNIM A607 (2009) 387-393.
G. Varner and L. RuckmanNIM A602 (2009) 438-445.
1GHz analog bandwidth, 5GSa/s Simulation includes detector response
![Page 35: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/35.jpg)
35
Ice Radio Sampler (IRS)RF input coupling (S11)
P. Gorham -- measurement
![Page 36: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/36.jpg)
36
Ice Radio Sampler (IRS)
P. Gorham -- measurement
![Page 37: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/37.jpg)
37
WFS Evolution and Philosophy• SAM
1. Maximize dynamic range and minimize signal distorsion.2. Minimize need for calibrations and off-chip data corrections.3. Minimize costs (both for development & production)
• DRS family– Get a solid, general design working– Something that can work, in volume, for many apps
• LAB and siblings (68 generations)– Continue to explore parameter space– Concentrate on applications where more general solutions above many not
be best choice
Approaches very complementary
![Page 38: Deeper Sampling CMOS Transient Waveform Recording ASICs Matt Andrew, Kurtis Nishimura, Gary S. Varner University of Hawai’i, and the Large Area Picosecond](https://reader035.vdocuments.mx/reader035/viewer/2022070412/56649f155503460f94c2aea3/html5/thumbnails/38.jpg)
38
References• PSI activities (DRS)
– IEEE/NSS 2008, TIPP09– http://midas.psi.ch/drs
• DAPNIA activities– MATDAQ: IEEE TNS 52-6:2853-2860,2005 / Patent WO022315 – SAM; NIM A567 (2006) 21-26.
• Hawaii activities– STRAW: Proc. SPIE 4858-31, 2003.– PRO: JINST, Vol. 3, P12003 (2008). – LABRADOR: NIM A583 (2007) 447-460.– BLAB: NIM A591 (2008) 534-545; NIM A602 (2009) 438-445.– STURM: EPAC08-TUOCM02, June, 2008.