Status of HPD Testing

1
Status of HPD Testing

• Status of Test Results (batch 7 8)
• QE measurements
• Distortion Analysis code
• Conclusions

Stephan Eisenhardt University of Edinburgh
RICH meeting, CERN, 30.05.2006
2
HPD Tests - Time Profile
Wed. 24.05.2006
Batch 8 (2911 HPD) 28.04.06 27 tested up to
24.05. (1 replacement, 1 repotted)
20-24 HPD expected (31.05.) limited by window
shortage
192
Batch 7 (26 HPD) 31.03.06 tested, accepted
(today) affected by dead column issue
162
136
109
76
results of last report (19.04.)
3
Status of HPD Test Program

• as of Wed 24.05.2006
• batch 7 provisionally accepted on 30.05.06
• H614006 arrived on 28.04.06 with batch 8 as
  replacement for H545005 (batch 5)
• H541004 (batch 4b) arrived repaired on 28.04.06
  with batch 8
• H542020 (batch 4b) expected repaired on 31.05.06
  with batch 9
• H539012 (batch 2) and H546002 (batch 4b) sent to
  DEP in May06, accepted for replacement on
  29.05.06

4
Batch 7 8 issues I

• H608001 (batch 7) chip on quartz window side ?
  complaint to DEP
• presumably from handling in vacuum chamber
  (DEP)
• window used due to shortage
• in LHCb 90 covered by Teflon tape ? fine HV/DC
  behaviour ? accept HPD
• H610004 (batch 7) Indium blob on covar ring ?
  mech. test failure
• DEP cut with scalpel, as done by DEP
• not done yet ? retest of mech. HV/DC needed ?
  conditional acceptance
• H612008 (batch 8) Indium bobs ? mech. test and
  DC OK
• feed-back to DEP request to take more care
• H612020 (batch 8) HPD leaning ? marginal mech.
  test
• just touching the jig
• Teflon tape not damaged ? accept HPD

H608001
chip
H610004
too big blob
H612008
H612020
blobs
touching!
5
Batch 7 8 issues II
Alice mode DC 0.9 kHz/cm2

• H605006 (batch 7) always-on column in LHCb mode
  only
• LHCb mode not tested at DEP
• masking Alice pixels doesnt help
• chip problem not in responsibility of DEP ?
  accept HPD
• but class F
• H607002 (batch 7) bad read-out in LHCb mode only
• looks similar to dying pixel problem
• is cured using the new tuned parameters of
  pixel chip ? class A
• 10-15 difference in npe is normal

LHCb mode DC 431 kHz/cm2
Alice mode 200k events, 4.1 npe
LHCb mode 200k events, 0.0013 npe
LHCb mode 200k events, 3.6 npe
eu_VBNG 0 eu_VBNLHCB 0
eu_VBNG 50 eu_VBNLHCB 20
6
Batch 7 8 issues III

• H607001 (batch 7) PC induced feature in DC
• total DC 1.6kHz/cm2, i.e. well within specs
• radial local structure not present without HV
• never seen before, not compatible with micro
  discharges
• originates from PC or electrodes
• by criteria class A
• ? accept HPD
• H612030 (batch 8) similar DC feature chip
  offset
• total DC 0.97kHz/cm2, i.e. very good
• no cross-check DC run without HV
• second such linear structure
• large offset of chip
• will investigate further

7
Batch 7 8 issues IV

• H610006 (batch 7) variable leakage current
  (27-2.5mA) pattern reproducible
• above 22.5mA anode does not deplete ? no signal
  response
• LC decreases with time (t 30min)
• ? anode defects anneal under bias voltage

conclusion HPD response fine after 1h ? accept
HPD, ? class E
3-day study
HV ramp-up while LC decreases
LC18.5mA
LC22.5mA
response fine for LClt10mA
no response LCgt22.5mA
15 min
8
Status of HPD Qualification

• change in qualification rules
• was class A ? E ? F for bad pixel count of 80
  400 (i.e. 1 5)
• now class A ? B ? E ? F for bad pixel counts 80,
  200 400 (i.e. 1, 2.5, 5)
• ? re-categorisation 8 HPD with 80-135 bad
  pixels E ? B

currently 5x F rejected ? -1x repaired, turned
to A ? 1x replaced, stays in list 1x F cond.
accepted (In-blob) 1x F accepted (column) 6x F
now ABE rate96.8 further expectations -1x
fixed In-blob, turned to A -1x repaired, turned
to B 2x replacements 4x F in near future (ABE
rate97.9)
batch 1-7 batch 8 (-4 HPD) 188 HPD
9
QE Measurements

• 4 existing and 3 newly acquired filters for l
  270, 340, 400, 440, 520, 640, 800nm
• 12 HPD measured 6 HPD at 7 wavelengths, 4 HPD
  measured multiple times (2-4x)
• targeted HPD with high QE_at_270nm to cross-check
  DEP data
• ? 11x good agreement, 1x -20
• observed temperature sensitivity ? logging of
  lamp case temperature

340nm between DEP data points systematically
lower
can reproduce the shape of DEP data in valley!
aims - more data - more systematic studies -
summary next RICH meeting
new filters
10
QE _at_ 270nm - Evolution

• QE delivered by DEP gets better from batch to
  batch!

batch 1-8 integrated
batch 1-8 batch by batch
batch 1-8 averages
11
Distortion Analysis Code

• aim get relative demagnification, rotation and
  shift of HPDs
• test HPD against reference HPD ? (taken with
  same pinhole setup!)
• find cluster centres of pinhole images ?
• from fit to position and size of PC image of LED
  run apply 1st order correction
• corrects for most of the chip offsets
• necessary to ensure good starting point for
  subsequent fit
• find link to nearest match ? ?
• fit to optimise demagnification, rotation and
  xy-shift ??

see 25.02.2006

• result
• 4 parameter transformation to match HPD with
  reference
• residuals ??
• stand-alone works
• (to be) integrated into analysis code
• to be applied to existing data

1st order corrected cluster centres reference
cluster centres after fit reference
12
Conclusions

• HPD testing is progressing well - according to
  schedule
• results from batch 7 ( 8) were throwing up an
  unusual number of smaller issues
• but tubes qualify fine eventually
• expected gap in HPD production will be
  (partially) filled with CERN HPDs
• good progress on analysis of distortion map and
  QE measurement
• and DEP surpasses itself with higher and higher
  QE