GIF tests status report Work of several people: Test setup design - - PowerPoint PPT Presentation

▶
gif tests status report
SMART_READER_LITE
LIVE PREVIEW

GIF tests status report Work of several people: Test setup design - - PowerPoint PPT Presentation

Jun 21, 2023 470 likes •800 views

GIF tests status report Work of several people: Test setup design & construction Oliver, Hubert, engineers, technicians, ... Data taking & installation at Cern Jurgen, Albert, Felix, Carl-Friedrich, Julia, Federica

slide-1
SLIDE 1

02/08/08

  • F. Legger

1

GIF tests status report

  • Work of several people:

– Test setup design & construction

  • Oliver, Hubert, engineers, technicians, ...

– Data taking & installation at Cern

  • Jurgen, Albert, Felix, Carl-Friedrich, Julia, Federica

– Analysis framework

  • Felix and Oliver

– Data analysis

  • Albert and Federica

– Garfield simulations

  • Albert, Federica, (Carl-Friedrich, Julia)

– MTGeant4 simulations for next GIF tests

  • Markus
slide-2
SLIDE 2

02/08/08

  • F. Legger

2

GIF analysis

slide-3
SLIDE 3

02/08/08

  • F. Legger

3

The test setup

COMPASS LHCb cosmic muon

  • ref. chambers

50 cm long standard MDT tubes 6 small tubes (1 m long) Trigger: 2 layers of 6 scintillators

slide-4
SLIDE 4

02/08/08

  • F. Legger

4

The setup at the GIF

Gas DAQ Source Cs 137 590 GBq LHCb MWPC chamber

  • 20 M events
  • Threshold scan:

34, 36, 38 mV

  • HV scan: 2700,

2745, 2760 V

  • Counting rate

scan: 50, 800, 1100, 1400 Hz/cm2

slide-5
SLIDE 5

02/08/08

  • F. Legger

5

GIF measurements

  • Run length ~12 hours (1M evts ~ 40000 hits in small tubes/run)
  • Drift time spectra
  • r-t relationship
  • Efficiency as a function of background rate
  • Resolution as a function of background rate

– Needs 1000 tracks in 1 mm slices 75000 tracks →

slide-6
SLIDE 6

02/08/08

  • F. Legger

6

Analysis status: drift time spectrum

  • Tubes 67-68-69-70 only
  • track cuts
  • CL1, CL2 > 0.02
  • |d1-d2| < 0.8 mm
  • |d1+d2| < 14.2 mm

Slightly different conditions in data and simulations (T) Drift time spectrum for small tubes is not well reproduced by Garfield refine t0? First hit or all hits?

.

d1 d2 Track in ref. chamber 1 (bottom) Track in ref. chamber 2 (top)

slide-7
SLIDE 7

02/08/08

  • F. Legger

7

Analysis status: tube swapped

slide-8
SLIDE 8

02/08/08

  • F. Legger

8

Analysis status: hedgehock cards

Upper multilayer: Hedgehock card type II Lower multilayer: Hedgehock card type I

In the analysis code both multilayers were assigned the same HH card type II

  • > big tubes were swapped
slide-9
SLIDE 9

02/08/08

  • F. Legger

9

Fix in Calib file (from Felix)

slide-10
SLIDE 10

02/08/08

  • F. Legger

10

Distance from wire vs. slope (track ref. ch. 2)

Before fix After fix Last tube

slide-11
SLIDE 11

02/08/08

  • F. Legger

11

Old plots: efficiency & r-t relationship

All tubes: still with HH card problem...will be redone

slide-12
SLIDE 12

02/08/08

  • F. Legger

12

Old plots: resolution

All tubes: still with HH card problem...will be redone

1100 Hz/cm2 σ = 240 +/- 15 µm σ (big) = 171 +/- 11 µm σ (small) = 170 +/- 18 µm

  • CL1, CL2 > 0.02
  • |d1-d2| < 0.8 mm
  • |d1+d2| < 14.2 mm

800 Hz/cm2 σ = 212 +/- 6 µm σ (big) = 147 +/- 4 µm σ(small) = 153 +/- 7 µm 50 Hz/cm2 σ = 190 +/- 2 µm σ (big) = 120 +/- 2 µm σ (small) = 147 +/- 3 µm

slide-13
SLIDE 13

02/08/08

  • F. Legger

13

Garfield simulations

slide-14
SLIDE 14

02/08/08

  • F. Legger

14

Gas gain

  • electrons on the wire/primary electrons:

– described by Townsend coefficient, however small uncertainties on Townsend coefficient lead to large uncertainties on gas gain -> hardcoded in Garfield – for each electron, gain distributed with Polya function (xαe-(1+α)x, 0<α<1)

  • Diethorn formula:

– ρ0/ρgas = 1 bar/ 3 bar – Emin(ρ0) = 23523 V/cm – ∆V = 34 V

  • Big tubes:

– E(a) = 193500 V/cm – b = 14.6 mm – a = 0.025 mm – V = 3080 V – G = 20700

  • Small tubes:

– E(a) = 195400 V/cm – b = 7.1 mm – a = 0.025 mm – V = 2760 V – G = 25500 NB: to have SAME electrical field in small and big tubes, HV = 2730 in small tubes Historically, HV = 2760 V in small tubes, which corresponds to HV = 3101 V in big tubes New Garfield simulations needed!

slide-15
SLIDE 15

02/08/08

  • F. Legger

15

Current signal

  • Big tubes:

– b = 14.6 mm – a = 0.025 mm – V = 3080 V – Max t = 4.3 ms

  • Small tubes:

– b = 7.1 mm – a = 0.025 mm – V = 2760 V – Max t = 1.1 ms

  • Only ion mobility taken into account
  • Ramo's theorem gives
  • ion mobility is constant (r > 100 µm): µ = 0.51 cm2/Vs @ 3 bar
slide-16
SLIDE 16

02/08/08

  • F. Legger

16

High rate effects

  • Electronics effects:

– Baseline shift and fluctuations – Shift effects reduced by using bipolar shaping (need the introduction of large dead time to prevent multiple hits)

  • Space-charge effects:

– Ion clouds drifting towards the cathod change the electric field and the gas gain: important for non linear gases such as Ar:CO2 – electrons drifting toward the wire see charge only within 1 cm – Ion clouds (Poisson distributed):

  • n = (1 cm) * Nc * tmax
  • Nc background rate per unit wire

length (i.e. Hz/cm2 => Hz/cm2 * tube diameter [cm])

  • Big tubes:

– b = 14.6 mm – Max t = 4.3 ms – N (50 Hz/cm2) = 0.6 – N (800 Hz/cm2) = 9.6 – N (1100 Hz/cm2) = 13.2 – N (1400 Hz/cm2) = 16.8

  • Small tubes:

– b = 7.1 mm – Max t = 1.1 ms – N (50 Hz/cm2) = 0.075 – N (800 Hz/cm2) = 1.2 – N (1100 Hz/cm2) = 1.65 – N (1400 Hz/cm2) = 2.1

x8 less ion clouds in small tubes!!!

slide-17
SLIDE 17

02/08/08

  • F. Legger

17

Electric field for high background rates

Boundary condition: Trascendental equation for k (numerical solution): If k<<b, second term is dropped and we obtain the usual electric field (1/r) If k>>r: Solution: However, one must take into account the gain drop due to the change of electric field, by calculating the line charge and insert it into the Diethorn formula -> Iterative process (convergence after a few steps)

slide-18
SLIDE 18

02/08/08

  • F. Legger

18

Electric field parameters

k >> r not valid at very high rates Average energy deposit: 36 keV

Albert's plots

Small tubes: 1000 Hz/1.5cm2 = 666Hz/cm2 big tubes: 666Hz/cm2*3cm= 2000Hz/cm

slide-19
SLIDE 19

02/08/08

  • F. Legger

19

Electric field at high background rates

no background ~1/r with background

[cm]

  • Small (big) tubes:

– E(high rate) < E(low rate) for r<2 (4) mm

  • E field variation: from less than 1% (positive, close to the wire) to a few

percent (negative, close to the wall) Albert's plots

slide-20
SLIDE 20

02/08/08

  • F. Legger

20

Garfield simulation: status

  • Implementation:

– HEED (particle interaction with the gas) – MagBoltz (electron trasport properties) – Ionisation along the track – Drift of electrons – Current signal (ion part) – Response of ATLAS electronics – Low rate: gas gain only input – High rate effects

  • Gas gain from iteration of Diethorn formula
  • Adds background electric field (parameter k from iteration)
  • Scales with number of ion clouds Poisson-distributed
  • Response of ATLAS electronics
  • Simulation with no background: 60000 evts/day/CPU
  • Simulation with background: 10000 evts/day/CPU

Rate (Hz/cm2) 800 1100 1400 Evts 300000 60000 160000 OK To be changed To be changed OK OK To be checked

slide-21
SLIDE 21

02/08/08

  • F. Legger

21

Conclusions

STILL A LOT TO DO

slide-22
SLIDE 22

02/08/08

  • F. Legger

22

Spare slides

slide-23
SLIDE 23

02/08/08

  • F. Legger

23

Electric field change due to space charge

Gauss theorem: Differentiating both sides: Q = 36 keV/ 22 eV e

slide-24
SLIDE 24

02/08/08

  • F. Legger

24

Tracking resolution

  • CL1, CL2 > 0.02
  • |d1-d2| < 4 mm
  • |d1+d2| < 14.2 mm
slide-25
SLIDE 25

02/08/08

  • F. Legger

25

GIF runs

  • 14 M + 4 M + 2 M + 120k = 20 M evts collected
  • Threshold scan (34-36-38 mV) ~ 1M evts each;
  • Counting rate scan (0, 800, 1100, 1500 Hz/cm2) ~ 1 M evts each;
  • HV scan (2700V, 2745V) with source on/off ~ 1 M evts each;
  • Atlas settings for electronics with source on/off ~ 1 M evts each;
  • Time over Threshold scan with source on/off and threshold scan

(34-36-38-40-42-44 mV) ~ 10k evts each.

slide-26
SLIDE 26

02/08/08

  • F. Legger

26

GIF counting rates (I)

SOURCE Threshold Nb. evts Hit rate <- (small tubes) (mV) (Hz/cm2)

  • att. inf. 38 3772021 54.6174
  • att. Inf. 36 1099982 68.9975
  • att. inf. 34 1256475 71.7997
  • att. 1 (shielding) 38 3772021 1032.94
  • att. 1 (shielding) 36 1199981 1129.75
  • att. 1 (shielding) 34 1256475 1169.96
  • att. 2 (shielding) 38 1387269 806.269
  • att. 2 (shielding) 36 1060907 807.611
  • att. 2 (shielding) 34 1199979 863.51
  • att. 1 (NO shielding) 38 2481841 1427.56
  • att. 1 (NO shielding) 36 1182973 1462.7
  • att. 1 (NO shielding) 34 1399971 1559.61

To get the counting rate/tube multiply by 150 cm2

slide-27
SLIDE 27

02/08/08

  • F. Legger

27

GIF counting rates (II)

SOURCE Threshold Nb. evts Hit rate <- (small tubes) (mV) (Hz/cm2)

  • att. 1 HV 2745 V 38 1545879 1277
  • att. 1 HV 2700 V 38 1499971 1224.25
  • att. inf. HV 2745 V 38 1099981 68.02
  • att. inf. HV 2700 V 38 1007010 68.0713
  • att. inf. Atlas sett. 38 1199987 68.3454
  • att. 1 Atlas sett. 38 1199981 1457.01
  • att. inf./1 ToT scan 44 10000
  • att. inf./1 ToT scan 42 10000
  • att. inf./1 ToT scan 40 10000
  • att. inf./1 ToT scan 38 10000
  • att. inf./1 ToT scan 36 10000
  • att. inf./1 ToT scan 34 10000

To get the counting rate/tube multiply by 150 cm2

slide-28
SLIDE 28

02/08/08

  • F. Legger

28

GIF tests: shielding

LHCb MPWC

drawing not to scale! reference chamber lower multilayer shielded by Pb wall small tubes and upper multilayer shielded by LHCb chamber

slide-29
SLIDE 29

02/08/08

  • F. Legger

29

GIF tests: counting rates per tube

  • from closest to farthest

tube (from source) we get ~ 30% attenuation

  • due to increasing distance

from source (~140 cm -> 164 cm)

  • to be taken into account

during analysis

  • collect more statistics with

highest irradiation hits/small tube (source on, att 1, no shielding, 36 mV) closest to source farthest from source close close far

slide-30
SLIDE 30

02/08/08

  • F. Legger

30

Time over Threshold scan

Drift time [ns] ToT [ns] time time time time time time time hits with shorter drift times have longer pulses and viceversa

slide-31
SLIDE 31

02/08/08

  • F. Legger

31

Time over Threshold scan

length of pulse ~ 200 ns, does not depend on threshold