for Nuclear Physics Applications Digital Signal Processing ...nucalf.physics.fsu.edu/~mriley/GRETINA_ElecWG_July04/Skulski.pdf · Digital Signal Processing Electronics for Nuclear

W.Skulski GRETINA Electronics Workshop, Argonne, 24-25 July 2004

Wojtek Skulski

SkuTek Instrumentation

and

University of Rochester

Digital Signal Processing Electronics

for Nuclear Physics Applications

Small Business Innovation ResearchDepartment Of Energy Grant DE-FG02-03ER83778


Outline

• Outline of grant work.

• Accomplishments.

• Hardware, firmware, software.

• Good resolution, low noise.

• Research and student projects @ UofR.

• Plans.

• Acknowledgements.


Outline of the work

• The company: 75% of the grant work

• Development of waveform digitizers:• Single-channel, 12-bit DDC-1.

• Eight-channel, 10-bit DDC-8.

• Firmware and software.

• University of Rochester: 25% of the grant work.

• Data link between DDC-8 and universal logic XLM.

• University of Rochester: education and research projects.

• Performed in addition to this grant, extremely valuable.


Single-channel, 12-bit DDC-1

Signal OUT

JTAG connector

USBprocessor connector

FPGA

Signal IN

ADC 65 MHz * 12 bits

Fast reconstruction DAC 65 MHz * 12 bits

Variablegain amp

Designed and built by WS.Used in several student projects during last 2 years.


Diagnostic OUT40 MHz * 10 bits

JTAG connector

microprocessor

FPGA

ADC 40 MHz * 10 bits(8 channels)

16 bidirectional TTL lines + 1 in(fast parallel interface to XLM)

Analogsignal IN8 channelswithdigital offsetand gain control

RS-232

Logic connectors NIM 16 lines IN, 8 lines OUT

USB

RAM500 kB

ECL clock IN(optional)


Gain=1, noise below 1 LSB

Intrinsic noise of the DDC-8 board

2 mV/ADC count

Time step (25 ns/step)220.0 240.0 260.0 280.0 300.0

516.0

517.0

518.0

519.0

520.0

pulse ADC gain = 1

ADC waveform 0

0.24 mV/ADC count

Time step (25 ns/step)340.0 360.0 380.0 400.0 420.0

530.0

535.0

540.0

545.0

550.0

555.0

560.0

pulse ADC gain = 8

ADC waveform 0

Gain=8, noise ~3 LSB (peak-peak)

Intrinsic noise is excellent


Short filter, pulser resolution 0.37 keV

DDC-8 + research pulser Ortec 448

Dynamic range of 18 bits obtained with 5µs running sums

120

100

80

60

40

20

0

Cou

nts

17.98x103 17.9617.9417.9217.9017.8817.86

Measured pulser amplitude (nat. units)

ADC channel 2, maximum gain, 90% amplitude step

0.5µs trapezoidal filter

Solid line: gap 0.5 µsSdev = 5.02943 Mean = 17901.9Resolution = 0.028 %

Pulser leading edge = 20ns

40

30

20

10

0

Cou

nts

179.05x103 179.00178.95178.90178.85178.80178.75

Measured pulser amplitude (nat. units)

ADC channel 2, maximum gain, 90% amplitude step

5µs trapezoidal filter

Pulser leading edge = 20ns

Solid line: gap 0.5 µsSdev = 20.8234 Mean = 178899Resolution = 0.012 %

Long filter, pulser resolution 0.16 keV

Pulser peak = 179,000 ==> 18 bits

Maximum gain Maximum gain


DDC-8 + 1-inch NaI(Tl)

Threshold = 5keV

Trapezoidal filter with 5µs running sumsCounts

Energy (keV)0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0

0.0

200.0

400.0

600.0

800.0

Compton back-scatter

33 keV, Ba X-ray662 keV

137 Cs Small NaI(Tl)Energy

W.Skulski GRETINA Electronics Workshop, Argonne, 24-25 July 2004Diagnostic channel

JTAG

Link connector to XLM-80Signal IN

8 channels

etc...

NIMconnector

NIM-level signals16*IN, 8*OUT

DDC-8 + XLM link board

W.Skulski GRETINA Electronics Workshop, Argonne, 24-25 July 2004 VME connectors

Linkdaughtercard

XLM-80motherboard

XLM-80 + DDC-8 link board (J.Toke, UofR)


Digital link DDC-8 <=> XLM

Sustained transfer rate = 75 Mbytes/sec

Digital link: data transfer and handshake signals DDC-8 <=> XLM


Documentwindow

Both software development and DAQ in the same environment

DDC-x software development and DAQ system

DAQwindows


Education and R&D projects at Physics and Astronomy

• S.Zuberi, Digital Signal Processing of Scintillator Pulses in Nuclear PhysicsTechniques, Senior Thesis, Department of Physics and Astronomy, University ofRochester. Presented at Spring APS meeting, April 2003, Philadelphia, PA.

•Awarded the Stoddard prize for the best Senior Thesis in the Department.

• D.Miner, W.Skulski, F.Wolfs, Detection and Analysis of Stopping Muons Using aCompact Digital Pulse Processor, Summer Research Experience for Undergraduates,Department of Physics and Astronomy, University of Rochester 2003 (unpublished).

• P.Bharadwaj, Digital and analog signal processing techniques for low-backgroundmeasurements, graduate research started this Summer.

• P.Bharadwaj, W.Skulski, F.Wolfs, Developing an efficient triggering system forPHOBOS at RHIC, ongoing.


Τraditional slow-tail representation

1 cm3 CsI(Tl) + phototube Single-channel digitizer DDC-1 at 48 Msamples/s * 12 bits natTh radioactive source

PID = TAIL / TOTAL

Note energy-independent PID

Particle ID from CsI(Tl)Senior Thesis by Saba Zuberi

Best Senior Thesis 2003Dept. of Physics and Astronomy

University of Rochester


Energy spectrumCounts

Filtered energy (arb. units)0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0 900.0 1.0E+3

0.0

50.0

100.0

150.0

200.0

250.0

Energy 1

662 keV, 137Cs

NaI 2" by 2"

0.884 keV/bin

77 keV, Pb X-ray

33 keV, Ba X-ray

Compton back-scatter

Signals from a Bicron 2”x2” NaI(Tl) detectordigitized with DDC-1 at 48 Msamples/s * 12 bits

137Cs

Response to scintillator pulses: NaI(Tl)Senior Thesis by S.Zuberi

Best Senior Thesis 2003, Dept. of Physics and Astronomy, University of Rochester


Detection and Analysis of Stopping Muons#

BC-400 5” x 6”& phototube

Experiment controland data display

Digitizer board

#Daniel MinerUniversity of RochesterSummer 2003 REU

•Example of pulse processing & analysis•Table-top experiment•Several observables from one signal


ADC waveform ADC value

Time steps, 20.8 ns/step0.0 100.0 200.0 300.0 400.0 500.0 600.0 700.0 800.0

1750.0

1800.0

1850.0

1900.0

1950.0

2.0E+3

2050.0

2100.0

2150.0Transient 3183

Signals from a BC-400 5”x6” scintillator

Stopping muon

Detection and Analysis of Stopping MuonsDaniel Miner, University of Rochester, Summer 2003 REU

Muon decay


Measured muon lifetime in very good agreement with published valueAfter 4% capture correction agree to within 0.35%

Time between the leading and trailing pulsesDaniel Miner, University of Rochester, Summer 2003 REU

0

0.2

0.4

0.6

0.8

1

1.2

0 2 4 6 8 10

Delta T (microseconds)

Nor

mal

ized

Cou

nt

Normalized Fit Delta T

Normalized Experimental Delta T

Measured <τ>: 2.12 + 0.04 µsAccepted <τ>: 2.19703 + 0.00004 µs

In matter

In vacuum


Vertex and centrality definition in real time

• Analog signals: Paddles, T0, ZDC.

• Logic signals from conventional NIM. • Signal processing: on-board FPGA.

• Accept/reject event within about 1 µsec.

Online trigger for PHOBOS with DDC-8Under development

Vertex definition from TACs.T0 OR ∆t, Paddle ∆t,ZDC ∆t.

PHOBOS @ RHIC

Centrality from paddle and ZDC.


Plans

• New board is under development.• RIA, RHIC, low-background underground measurements.

• Design objectives.• High reliability.• Remote operation and diagnostics.• Low cost per channel.• High level of integration (many channels per board).• On-board DSP.• Integration with existing infrastructure (VME).

• Status: schematic almost finished.• Prototype will be assembled this Fall.


Acknowledgements

• SkuTek Instrumentation.• Joanna Klima, WS (Principal Investigator).

• University of Rochester.• Jan Toke: digital link.• Frank Wolfs: support for DDC-8 development.

• Graduate students.• Palash Bharadwaj.

• Undergraduate students.• Suzanne Levine, Daniel Miner, Len Zheleznyak, Saba Zuberi.

Documents

for Nuclear Physics Applications Digital Signal Processing ...nucalf.physics.fsu.edu/~mriley/GRETINA_ElecWG_July04/Skulski.pdf · Digital Signal Processing Electronics for Nuclear