april 4, 2007 rolf nahnhauer1 desy-project icecube digital optical module production

April 4, 2007 Rolf Nahnhauer 1

DESY-PROJECT

IceCube Digital Optical Module

Production


IceCube StatusApril 2007


THE ICECUBE OBSERVATORY

IceTop air shower array80 pairs of ice Cherenkov tanks

IceCube deep ice array4800 optical modules on 80 stringsinstrumented volume : 1 km3first string deployed : January 2005

~ 80.000 atm. per year


2005, 2006, 2007 DeploymentsAMANDA

IceCube string deployed 12/05 – 01/06

1424 DOMs deployed to date

Next year looking for 14 to 18 strings.

Want to achieve steady state of >= 14 strings / season.

21

3029

40

50

3938

49

59

4647

48

5857

6667

74

65

73

78

56

72

IceCube string deployed 01/05

IceCube string and IceTop station deployed 12/06 – 01/07

IceCube Lab commissioned

1+ 9 + 13 = 22 strings to date


Hotwater drilling: Trend towards increasingly stable operation continued.Experienced crew

IceCube Pole Season 2006/2007


•IceCube lab commissioned, •South Pole computing system (SPS), all racks and infrastructure installed,•64 bit implemented on SPS and SPTS,•About 10 racks of computers and hubs. •Working simulation, exp control, PnF

IceCube Pole Season 2006/2007


IceCube Events

Neutrino in IC 9 Flasher Pulse in IC 22


IceTop: cosmic-ray physicsIceCube: calibration, background tagging

IceTop/In-ice coincident events


Atmospheric neutrinos

in IC 9, 2006

John Pretz, Ph.D. thesis

Ratio of data to simulation:R=1.05 +- 0.24 (syst) +- 0.09 (stat)

given at cut strength 10156 upgoing muon eventspurity of neutrino sample > 95%

First IceCube Results


Project Year 3 4 5 6 7 8 TotalDeployment Year 04/05 05/06 06/07 07/08 08/09 09/10Drill/deploy strings 4 12 16 18 18 12 80Strings in ice 4 16 32 50 68 80 80DOMs needed 400 834 1064 1242 1182 778 5500Calendar year of prod. 2004 2005 2006 2007 2008 2009Cable systems to build 5 14 24 24 15 0 82Total DOMs: build and test 400 834 1064 0 5500DOMs: DESY 60 210 280 0 1330DOMs: Sweden 50 140 180 0 880DOMs: PSL 290 484 604 0 3290

3202

1912

780510

PRODUCTION SCHEDULE

Numbers from February 2004 slightly changing with timeBaseline change: 80 strings 70 strings + 10 strings from contingency

xxx


DOM Production Site Integration Status - PY5 (April, 2006 - March, 2007)

1053

939

340372

296261

0

200

400

600

800

1000

1200

Plan integration Actual Integration Plan Integration Actual Integration Plan Integration Actual Integration

Week ending 3/11/07

PSLDESY

Sweden

Site % to Weekly Plan % PY5 completed

PSL 89 84

DESY 109 100

Sweden 88 82

Site Total Plan

PSL 1125

DESY 370

Sweden 320


DOM Production Site Integration Status - PY6 (April, 2007 - March, 2008)

1097

0

430

0

336

00

200

400

600

800

1000

1200

Plan integration Actual Integration Plan Integration Actual Integration Plan Integration Actual Integration

Week ending 1/24/07

PSLDESY

Sweden

Site % to Weekly Plan % PY5 completed

PSL 0 0

DESY 0 0

Sweden 0 0

Site Total Plan

PSL 1097

DESY 430

Sweden 336

Do not use this chart till April, 2007 - see PY5 charts for current PY5 information


DOM Production DESY IntegrationStatus - PY 07 (April 2008-March 2009)

PY Year # DOMs ready

03 4/04-3/05 60

04 4/05-3/06 160

05 4/06-3/07 370

06 4/07-3/08 430

07 4/08-3/09 120-280 ?

DOM integration at DESY will finish 7/08 to 11/08

decision how many DOMs for how many stringshave to be produced, will come in summer 2007


FUTURE PRODUCTION and TESTING IceCube DOM Production CY2004 - CY2008 for 80 strings installed

0

10

20

30

40

50

60

70

80

90

24.0

7.20

04

24.0

9.20

04

24.1

1.20

04

24.0

1.20

05

24.0

3.20

05

24.0

5.20

05

24.0

7.20

05

24.0

9.20

05

24.1

1.20

05

24.0

1.20

06

24.0

3.20

06

24.0

5.20

06

24.0

7.20

06

24.0

9.20

06

24.1

1.20

06

24.0

1.20

07

24.0

3.20

07

24.0

5.20

07

24.0

7.20

07

24.0

9.20

07

24.1

1.20

07

24.0

1.20

08

24.0

3.20

08

24.0

5.20

08

24.0

7.20

08

Week ending

Qu

anti

ty o

f st

rin

gs

read

y fo

r in

stal

l

Block 1

Block 2

Block 3

Final Integrated DOM PRR (early June 2006)

PMTs, spheres ordered - qty 1200

Place orders for main board, HV suite, flasher board, delay board, penetrator assembly, harness, etc - qty 1200, Block 3A

Block 3 Design freeze

qty = 1200, 3AStrings installed in ice at Pole

Known major changes between Blocks 2 and 3:1. CAEN supplied HV control board2. Toroid on HV base3. Pigtail and Harwin connector on penetrator assembly

DOM Subsystem Final Design Verification and CDR - best case: good to order 3600 units, must order 1200 min. Aug 29 - 30

qty = 2400, 3B

qty = 700, 3C to 80 strings

Block 3A readiness review, March 2006

Must order last 700 units to continue to 80 strings Feb, 2008

Must order Block 3B 2400 units to continue uninterrupted in Nov, 2006

Note: this plan reduces quantity of DOMs to ship to the Pole in CY2006 from 16+ to 14+ to accommodate later production start


DOM Production and Testing at DESY


DOM PRODUCTION at DESY

Production of up to 1300 Optical Modules until mid 2008

Production comprises:

Gel mixing, filling and potting PMTs

Collar mounting and assembly of electronics

Sealing of DOMs at low pressure

Harness DOM with suspension

Finally pack DOMs and ship them to the pole


PROJEKT-ABLAUF


Dr. H. Waldmann 100%

Dr. J. Bolmont ( postdoc) 65%

xx

S. Henze xxxJ. PieperS. Ladegast


OMP-Projekt Investitionen

#OM

605

620

545

2107


OMP-Projekt PersonalbedarfWeeks kEuro

Weeks kEuro

Produktion 07 314.8

Physiker 48 62.8Postdoc 33 33.6Ingenieure 24 26.4Techniker/Werkstaetten 96 192.0

Produktion 06 222.8

Physiker 35 38.5 Postdoc 23 24.5 Ingenieure 18 19.8 Techniker/Werkstaetten 70 140.0


DOM TESTINGElectronic and optical requirements

Reboot- and communication over a wide temperature range from +20°C to -45°C

Single photo electron detection

Wide dynamic signal range – capable to handle large light pulses with up to several 1000 photo electrons per microsecond

Time resolution better than 5ns for single photo electron pulses

High voltage calibration of the PMT better than 5%

Optical sensitivity within low variations for different DOMs

Dark noise rates less than 1kHz in ice

Mechanical requirements

Vibration and pressure fluctuation during transport

Rapid temperature variations from +20°C to -45°C

Very high environment pressure up to 650 bar


DOM FINAL ACCEPTANCE TEST

Reversed Hypothetical FAT

-60

-50

-40

-30

-20

-10

0

10

20

30

1-Jun 6-Jun 11-Jun 16-Jun 21-Jun

Time

Te

mp

erat

ure

X Y

X + Y > 180 h

A full set of different tests is performed for defined temperatures

Test of the electronics (mainly runningdiagnostic programs, checking the hardware components)

PMT high voltage calibration

Rate monitoring while DOMs are illuminated with light of different wavelength

Dark rate monitoring

Data taking with a DAQ system similar to the final low level south pole DAQ(Linearity and time resolution tests)

Timing scenario:


DARK FREEZER LABLarge cooling chamber (4 x 6 x 2 m)

Temperature control with coolingaggregate and heaters

Minimal temperature for test cycle is -45°C(in the US -55°C for IceTop DOMs)

Optical fibers and mirror system installed on each test station

DOMs sit on top of cylindric cans

Cans are taped with aluminum foil todistribute the light

DOMs are covered with black plastic bags to keep them as dark as possible for the measurements


TEST ENVIRONMENT SETUPDark Freezer Lab (DFL) with 64 test stations

Same DAQ and wiring as for the South Pole system

Simulated cable length up to 3km

Light is distributed equally to the DOM stations via optical fibers

Time synchronization of multiple domhubs with a global GPS clock

Light system allows event simulation

Different light sources:

Laser for time calibration, pulsed LED for linearity test, DC lamp with monochromator for optical sensitivity test


DFL STATION CALIBRATION

Light yield differs for all stations

There is a wavelength dependence of the optical system

For studies of the optical sensitivity the stations have to be calibrated

During test runs a well characterized monitor PMT takes direct signals from the light sources

Together with the calibration information the relative optical sensitivity of the DOMs can be evaluated


ELECTRONIC TESTSA self test of the electronic components runs on the DOM

performed on bare mainboards before delivering

performed on integrated DOMs as functional test

Reboot tests are performed during the complete test cycleDOMs are turned on and off more than 100 times over a test runin addition tests of software uploads are performed

For the tests of the local coincidence chain the DOMs in the DFL are connected pairwise like on a real string


DOM CALIBRATION I

ATWD Calibration :

Reconstruction of PMT voltage waveform from ATWD data requires:

Voltage calibration of ATWD

Measurement of amplifier gain for ch0, ch1, and ch2

Frequency calibration of ATWD

Baseline measurement

Performed using theinternal pulser and oscillator


DOM CALIBRATION IIDark noise events are used to record the charge spectrum of single photo electron events

Calculation of the mean charge for different PMT high voltage settings reveals the gain

Finally this results in a linear fit of the gain versus high voltage relation

Working voltage yields gain = 107

Requirements:

Peak to valley ratio > 2.5

Gain > 5107 at 2000 V

overall relative error < 5%

Charge Spectrum of single photo electron events at 1400 V

Gain versus voltage[V]


TIME SYNCHRONIZATIONEach DOM runs with an individual clock

A global time synchronization has to be performed regularly to match the DOM-time to a global time including signal run-time

The RapCal method :

Time stamped bipolar pulses are sent from the DOR card to the DOM and vice versa

This gives a coarse time synchronization of the order of 50ns (local clock speed)

In order to improve the time synchronization accuracy the bipolar pulses are fitted at arrival, giving an exact time e.g. for zero crossing of the waveform

With that information the round trip time can be calculated and the accuracy of the time calibration is improved to better than 5ns


TIME RESOLUTION Pulse laser sends accurate signals (~75ps wide) to the DOMs

A bare mainboard records the absolute pulse time (sync pulse)

Analyze the distribution of time differences of DOM’s and sync pulse

The width of the distribution gives an upper limit for the DOM time resolution

Fraction of late pulses identify noise and time calibration problems

An overall comparison of the signal time offsets of all DOMs give information about the time resolution stability

Light and signal travel time


LINEARITY

DOMs are illuminated with different light intensities using a pulsed LED with different power settings

Different filters are brought into the light path to attenuate the amount of light

Plots show the charge distribution for different filter settings (low to high attenuation)

Plotting the mean charge versus light intensity gives the linearity characteristic for a DOM


OPTICAL SENSITIVITYSensitivity of integrated DOMs differs due to transmission characteristic of glass (sphere,PMT), gel and the PMT

quantum and collection efficiency

Do rate measurements while the DOM is illuminated with light of varying wavelength

Data has been already recorded, analysis is in preparation

Measure behaviour of subsystems: e.g. transmission of gel samples with respect to air

Allows relative comparison of different production batches


DARK NOISEReadout on board scaler – discriminator crossings of PMT signal

Requirements:

Mean noise rate < 3kHz No outliers within 5σ

(Noise rate in ice is much lower (~700Hz))

Search for spikes in noise ratebackground – maybe an indication for problematic PMTs or HV generators

Coincidence spikes are likely caused by some outside influences

[Hz]

Histogram of the dark noise ratesfor a single DOM


DOM - Passport

Decision June 2005:

use formalized DOM-passport for characterization

and qualification of DOM’s (ready early 2006)


DOM-Production 2004-2008

YearYear DOMs

produced

DOMs OK

finalDOMs

shipped

2004 60 45 0.75 28

2005 160 159 0.99 160

2006 257 255 0.99 224

2007 480

2008 183-343

only ~1% of DOMs are cannibalized – goal was 5% or bettergood components are used in next years production


2004 DOM TESTING + USAGE

time load good bad comment

FAT1 20.9.-5.10.04 25 20 5 3 gel, 1 MB, 1 HV

FAT2 4.11.-17.11.04 21 8 13 13 gel, 1 MB

FAT3 10.1.-20.1.05 20 17 3 1 MB,3 FB,1 HV

FAT4 24.2.-7.3.05 8 8 0 use 5 CAEN HV

location number

quality purpose

Pole 42 good deploy

DFL 3 good standard

Madison 8 bad Check

Dortmund

1 bad HV-tests

- 6 bad cannibalize

Production efficiency:

P2004 = 75 %

Wanted:

P = 95 %


2005 DOM TESTING Time load good bad FAT

FAT5 10.06.-27.06. 51 37 14 0.73

FAT6 12.07.-8.08 54 49 5 0.91

FAT7 29.08.-12.09 57 47 10 0.82

FAT8 14.9.-30.09 26 20 6 0.77

FAT9 14.11.-01.12 6 6 0 1.00

*) in all cases +3 permanent DOM’s from FAT3 for comparison

FAT 7

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 8

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 9

Passed

STF

FB

Dark Rates

Comm/other

GelFAT5

FAT6

FAT7

FAT8

FAT9


2006 DOM TESTINGtime load good bad FAT

FAT10 19.06- 30.06 57 46 11 0.81

FAT11 19.07-04.08 57 50 7 0.88

FAT12 09.08-25.08 57 53 4 0.93

FAT13 05.09-20.09 51 48 3 0.94

FAT14 12.10-27.10 55 49 6 0.89

FAT15 27.10-15.11 14 12 2 0.86

FAT 10

Passed

STF

FB

Dark Rates

Comm/other

Gel FAT 11

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 12

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 13

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 14

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 15

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT10

FAT11

FAT12

FAT13

FAT14

FAT15


2007 DOM TESTINGtime load good bad FAT

FAT16 29.01-14.02 57 50 7 0.88

FAT17 28.02-17.03 57 49 8 0.86

FAT18 22.03-xx.yy 34

FAT 16

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT 17

Passed

STF

FB

Dark Rates

Comm/other

Gel

FAT16 FAT17

No dominant error source could be identified in FATs

Different small problems appeared for all FATs


SummaryDOM first year mass production is running without problems

Production team well trained

Delivery of components mostly smooth

Storage space sufficient but at the limit, many transports betweenZeuthen and Wildau necessary

Many small problems contribute to test failure rate

Total failure rate after rework sufficiently small (better than expected)

DESY well prepared for continuous mass production until autumn 2008

april 4, 2007 rolf nahnhauer1 desy-project icecube digital optical module production

Documents

strings season

icecube eventsneutrino

total number of doms

icecube lab commissioned1

strings instrumented

production planning

final production run

minimal production activities