CAL CALIBRATION Overview and Stability

Thomas Schörner-SadeniusHamburg University

ESCALE MeetingDESY, 7 June 2005

DESY 7 June 2005 TSS: CAL Calibration 2

INTRODUCTION TO THE ZEUS UCALDepleted Uranium for calibration and compensation

PrincipleDepleted Uranium – scintillator calorimeter; analog pipelinedPMT readout. Analog electronics on front-end cards on detector;digital electronics in 3rd floor of rucksack.

Division in F/B/RCAL with 2172/2592/1154 cells, 2 PMTs per cell EM cell size: 520 (1020) cm2 in F/BCAL (RCAL) HA cell size: 2020 cm2.

Resolution: (e)/E=17%/E, (h)/E=35%/E

Claim: absolute energy scale known to 1%.

98.1% 238U, decay to 234Th, decays to 234PA and then 234U, cascades in between.

Emax()=2.3 MeV, E()=10-1000keV

(U)=4.5 Gy (Giga-years) rather stable signal of ZEUS lifetime.

Use UNO (Uranium noise) signal to monitor CAL behaviour with time

Calibration

Uranium

Idea

DESY 7 June 2005 TSS: CAL Calibration 3

READOUT OVERVIEW Necessary for understanding of calibration constants.

shaper pipeline buffer

shaper pipeline buffer

To trigger

55pC

IUNO

20ms integration time over UNO current

Qhigh

Qlow

VDAC

DAC

Vref

Vref=1.67V

UNO

Energy Q Current IPMT Charge Q Voltage V ADC counts

Digitalelectronics

Necessary: Calibration of particle/jet energy to ADC counts

PMT

DESY 7 June 2005 TSS: CAL Calibration 4

CALIBRATION IDEA Use stable Uranium noise as calibration signal

Uranium activity stable in time.

UNO signal stable in time to about 1% (CERN testbeam)

e/UNO or h/UNO for given Ee, Eh stable in time (studying the ratio cancels some uncertainties more precise result

e,h response linear in energy (CERN testbeam)

Assumption 1

Assumption 2

Assumption 3

Assumption 4

STEP A

STEP B

Keep UNO signal stable trimming of HV settings UNO scale factors (offline GAFs)

From known and linear e/UNO (h/UNO) then estimate energy of e,h.

One of the many complications: UNO signal and (fast) physics signal go through differentsignal paths on front-end card!

• UNO[ADC] fixed

• e(E)[ADC]/UNO[ADC]: CERN!

• then e[ADC]e[GeV]

DESY 7 June 2005 TSS: CAL Calibration 5

“1%” CALIBRATION Refers to rather different things

Intermodule/interregion calibration

After UNO calibration: compare various modules in their response to well-defined input energies (test beam) spread of various modules ~ 1%

1% - the first

1% - the second Determination of absolute scale

E(particle) ADC: after UNO calibration (UNO gives precise ADC count): absolute scale delivered by (using the test beam results):

signalUNO

signalparticle

_

_

Two important questions/tasks here: -- Keep the UNO signal to the nominal as closely as possible (UNO scale factors, but also other smaller corrections for front-end, signal path etc.) -- Derive offline correction factors from physics data (kin. Peak, E-pz, etc.)

‘one-to-one relation betweenADC and energy. Within onemodule response from towersis gaussian with width ~1%! in testbeam fix scale to 1%!

DESY 7 June 2005 TSS: CAL Calibration 6

UNO SCALE FACTORSCAL offline GAFs – the one that always stop the reconstruction

For all CAL regionsmeans within fewpermille around 1.Widths below 1%.

Channel-by-channel comp-arison of two UNO GAFs from 010305 and 080505.Means ~1permille.Widths ~0.5%.

Several HV adjustmentsbetween the two dates.

No systematic trends, distributions gaussian absolute calibration preserved at 1%-level!

DESY 7 June 2005 TSS: CAL Calibration 7

UNO SCALE FACTORSModule by Module comparisons of relative UNO differences

No significant changes between modules intermodule calibration still at 1%-level!

DESY 7 June 2005 TSS: CAL Calibration 8

FURTHER OFFLINE CORRECTIONS Motivated by physics; take into account dead material etc.

PHANTOM routine escale03.fpp (default in ORANGE) applies corrections for F/B/RCAL, separately for EMC and HAC; in addition cell corrections for some RCAL cells

FEMC: 1.024FHAC: 0.941

BEMC: 1.003*1.05BHAC: 1.044*1.05

REMC: 1.022RHAC: 1.022

Corrections derived from kinematic peakevents and DA measurements. -- repeat in newer data?-- dependence on physics case?-- any manpower currently involved?

DESY 7 June 2005 TSS: CAL Calibration 9

SUMMARY and possible outlook

Started to look into long-term stability of CAL calibration (UNO scale factors). absolute calibration seems stable to within 1% over time. intermodule calibration within about 1%

if initial absolute energy calibration good to 1%, then this quality is probably preserved until today.

Needed: Better understand of calibration in detail; only then can judge on quality of calibration.

Important: control of offline CAL regional (caltru) correction factors use physics events for that (kin. peak, E-pz, DA method etc.)