Study of Ion Feed Back and QE

1
Study of Ion Feed Back and QE

• Outline Batch 0 History
• IFB rates from Strobe Scans
• IV curves
• QE Measurements
• Plans
• Conclusions

Stephan Eisenhardt University of Edinburgh
HPD phone meeting, 30.05.2007
2
Batch 0 History

• Batch 0 HPD have worst case history for vacuum
  quality
• produced in a rush relatively bad vacuum quality
  from the start
• 1 HPD produced in 03/2004
• 8 HPD produced around 1011/2004
• relatively rough treatment in tests and testbeam
• from production to 0709/2005 stored under normal
  atmosphere where Helium could diffuse into the
  tube and worsen the vacuum
• storage under N2 atmosphere from 2005 onwards
• 5 tubes at CERN yes
• 2 tubes at Edi yes
• 2 tubes at Gla no
• Candidates to study the effect and development of
  IFB
• He as main residual gas ? ionisation ? secondary
  emission at PC
• gain 1 e- ? 1 e- 1 e- 1 ion ? 2 e- n e-
  with n2060(?)

3
IFB rates from Strobe Scan

• tube 2004(CERN) 2005(PDTF) 2007(PDTF) storage
• batch 0
• H441001 (Edi) ? 0.17 0.49 N2
• H439004 (Edi) ? 0.70 0.96 N2
• H428001 (Gla) ? 0.043 1.11 air
• H446005 (Gla) ? 0.013 0.17 air
• batch 0 (5 HPD) ? n.a. to be measured N2
• batch 1 (9 HPD) n.a. 0.10.6 n.a. N2
• batch 1 (12 HPD) n.a. 0.010.08 n.a. N2
• batch 2 (27 HPD) n.a. 0.010.08 n.a. N2
• batch 3 onwards n.a. 0.0010.08 n.a. N2
• batch 5 H527009 n.a. 0.002 to be measured N2

2.9
1.4
26
13
improving!
4
IV-curves for QE-Photocurrent

• Measurement of the photocurrent wrt. bias voltage
  used for charge collection in the HPD
• Expect charge collection efficiency to plateau
  from some V

V
I
5
Reference Measurements

• 12/2006 syst. check for QE (l440nm)
• absolute current amount of light QE
• voltage limit 120V, due to power supply
• low IFB tubes
• H634003 (batch 14) IFB 0.017
• H516002 (batch 2) IFB 0.024
• plateau from 20V onwards

120V
std. V for QE measurement

• 05/2007 more detailed study (l270nm)
• power supplies matching results
• Agilent 0120V
• Kepco 0500V
• low IFB tube
• H527009 (batch 5) IFB 0.019
• plateau from 15V onwards
• 20230V 4.8 pA/100V
• 230500V 0.7 pA/100V

0.7pA/100V
4.8pA/100V
0120V match
500V
6
Reference Measurements II

• 12/2006 syst. check for QE (l440nm)
• absolute current amount of light QE
• voltage limit 120V, due to power supply
• low IFB tubes
• H634003 (batch 14) IFB 0.017
• H516002 (batch 2) IFB 0.024
• plateau from 20V onwards

120V
std. V for QE measurement

• 05/2007 more detailed study (l270nm)
• power supplies matching results
• Agilent 0120V
• Kepco 0500V
• low IFB tube
• H527009 (batch 5) IFB 0.019
• plateau from 15V onwards
• 20230V 4.8 pA/100V
• 230500V 0.7 pA/100V

0.7pA/100V
4.8pA/100V
0120V match
500V
7
Interpretation of Slope in Plateau .

• Assumption IFB dominated by Helium gas in tube
• diffused through Quartz glass and sintered Al
  body
• He ionisation cross section
• rises with e- energy 25130V
• falls with e- energy gt130V
• Secondary yield
• rises with ion energy
• gas gain is exponential
• Electrostatic field in HPD
• 3D-field with local potential distribution 0V
  and geom. e- probability in HPD
• Slopes in IV plateau become plausible
• integrate over processes and probabilities
• 20230V rise due to increased He X-section and
  secondary yield
• 230500V slow-down He X-section and secondary
  yield compensate

130V
24.6V
500V
20V
Montague, Harrison, Smith J.Phys.B 17(1984) 3295
8
High IFB Measurements

• Batch 0 H441001
• IFB 0.49 (H527001 26)
• plateau 1020V
• Photocurrent rise
• 20230V 57 pA/100V
• 230500V 359 pA/100V
• Interpretation
• increased He pressure ? more secondary e-
• behaviour like H527001 but gain due to more
  secondary e-12

offset due to different GND
339 pA/100V
57pA/100V
500V

• Batch 0 H439004
• IFB 0.96 (H527001 51)
• plateau 1020V
• Photocurrent rise
• 20230V 200 pA/100V
• 230500V exponential !
• Interpretation secondary e- ionise gas gain!

instable measurements
link of two sessions
stable
200pA/100V
500V
0120V match
9
IFB and DC
DC 0.02kHz
DC 1.25kHz
DC 1.87kHz
H527009
H441001
H439004
IFB (DC) 0.03
IFB (Strobe) 0.002
IFB (DC) 5.5
IFB (Strobe) 0.49
IFB (DC) 6.1
IFB (Strobe) 0.96

• IFB(DC) is relative to DC
• IFB(Strobe) is delayed and relative to light
• high IFB(Strobe) HPDs can still be low in DC!

10
QE Measurements low IFB

• H527009 (batch 5) IFB0.002
• QE(100V)/(22V) to be meas.
• photo curr. 100V/20V 1.05
• QE (100V) 2007/2006 1.04
• IFB 2007/2006 to be meas.

• H516002 (batch 2) IFB0.024
• QE(100V)/(22V) 1.10
• photo curr. 100V/20V 1.09
• QE (100V) 2007/2006 1.04
• IFB 2007/2006 to be meas.

match
11
QE Measurements high IFB

• H441001 (batch 0) IFB0.49
• QE(100V)/(22V) 1.34
• photo curr. 100V/20V 1.35
• QE (100V) 2007/2006 n.a.
• IFB 2007/2006 2.9

match

• H439004 (batch 0) IFB0.96
• QE(100V)/(22V) 1.76
• photo curr. 100V/20V 1.75
• QE (100V) 2007/2006 n.a.
• IFB 2007/2006 1.4

match
12
Work Plan

• tubes IFB/DC IV QE
• CERN 2005/6 2007 0120V 0500V CERN
  2006 100V 22V
• batch 0 H441001_at_ Edi ? ? ? ? ? ?
  n.a. ? ?
• batch 0 H439004_at_ Edi ? ? ? ? ? ?
  n.a. ? ?
• batch 0 H428001_at_ Gla ? ? ? ? ?
  n.a. ? ?
• batch 0 H446005 _at_ Gla ? ? ? ? ?
  n.a. ? ?
• batch 0 (5 HPD) _at_ CERN ? n.a. ? ? ?
  n.a. ? ?
• batch 2 H516002 ? ? ? ?
  ? ?
• batch 4 H542006 ? ? ?
• batch 4 H545002 ? ? ? ?
  ? ?
• batch 5 H527009 ? ? ? ? ?
  ? ?
• batch 22 H614025 (rec.) ? ? ?
  n.a. ?(?) ?
• batch 23 _at_ mid-air (rec.) n.a. ? ?
  n.a. ? ?
• ? measured, data compiled ? measured,
  results to be compiled in
• ? results planned for Tue 05.06.
• ? results planned after Tue 05.06. to do BP
  measurement ? n of secondary e-

13
Conclusions

• When corrected for effect of IFB, the two batch 0
  HPD still show same QE as measured by DEP.
• Bulk of HPD starts out with very low IFB rates
• HPD properly stored under N2 atmosphere show only
  small increase of IFB
• IFB 1.53 / year if stored under N2
• IFB 1030 / year if stored in air
• IV behaviour is consistent with He being the
  dominant gas
• High IFB HPD still can have low absolute DC _at_
  HV20kV
• QE measurement very sensitive to IFB rate due to
  time integration
• can be disentangled by measuring at different
  bias voltages
• we will look very carefully into recovered HPD