HF PMT ISSUES

1
HF PMT ISSUES

• By
• Ugur Akgun
• The University of Iowa

2
Outline

• Introduction to HF PMTs.
• PMT Selection Process
• Complete Tests on 2000 PMTs
• Relative Gain, Gain vs HV
• Afterpulse Rate Issue
• What is PMT Afterpulse and its rate?
• The University of Iowa Afterpulse Tests
• Hamamatsu Afterpulse Tests
• TB2004 Analysis for PMT Afterpulses
• Conclusion

3
HF PMT Specs

• Window Material Borosilicate glass
• Eff. Pho.cath. dia. 22-28mm, head-on
• Quantum efficiency gt15 400-500 nm
• Photocathode lifetime gt200 mC
• Anode current vs position lt/-20 with 3 mm spot
  scan
• Gain 104 to 105, at lt0.75 x V(max)
• Single pe resolution rms/mean if single pe peak
  50 or better
• Pulse linearity /- 2 for 1-3000
  photoelectrons (g4X104)
• Anode pulse rise-time lt5ns
• Transit time lt25 ns preferred
• Transit time spread lt2 ns preferred
• Anode pulse width lt15 ns FWHM
• Gain (1/2)-lifetime gt1500 C
• Gain recov. (2000pe pulse) within 10 of nominal
  (g104) in 25 ns
• Average current Ik lt1 nA (g104)
• Average current Ia lt10 microA (g104)
• Anode dark current lt2 nA (g104)
• Stability lt/- 3 within any 48 hr. period
• Envelope opaque and -HV conductive coating

4
Candidate PMTs and Tests

• Photonis
• XP3182/D1
• XP2960
• Electron Tubes
• D843WSB
• D844WSB
• Hamamatsu
• R7525
• Testing, Evaluation and Results Reported at
• CMS-IN 2002-026,
• CMS-IN 2002-030,
• CMS-IN 2002-032,
• CMS-NOTE 2003/029, 
• IEEE Transactions on Nuclear Science. Vol. 51,
  No. 4, August, 2004

5
Testing 2000 Hamamatsu R7525

• Iowa Measurements
• Timing (Pulse Width, Rise Time, Transit Time,
  TTS)
• Linearity, Relative Gain, Dark Current, Single
  Photoelectron Resolution
• Hamamatsu Measurements
• Cathode Luminous Sensitivity, Anode Luminous
  Sensitivity, Anode Dark Current _at_1500V, Cathode
  Blue Sensitivity
• Anode Dark Current and Anode Voltage (_at_ Gain
  5xe4)
• Database is on the web.
• All info sent along with the PMTs and sorting
  tables.
• Complete tests of 2000 Hamamatsu R7525HA
  Phototubes for the CMS-HF Forward Calorimeter
  CMS-NOTE 2004/019, Submitted to IEEE Transactions
  on Nuclear Science

6
Relative Gain
Reference PMT, HV vs Gain measurements 900V-1650V
(30 measurements) Max 4 variation.
Relative Gain vs HV (E. Gulmez, CMS Note Draft)
7
For 2000 PMTs
Hamamatsu Measurements
Iowa Measurements
8
TB2004 Data
EM PMTs, Variation is 9-10 after RG Corrections
Corrected and Normalized Output
Normalized Output RG values
Andrei Gribushin's SPE data for W-6
HAD PMTs, Variation is 11-13 after RG
Corrections
Corrected and Normalized Output
Normalized Output RG values
9
TB2004 Data
W-13, 100 GeV electron data, with 1150V and 1350V

• The variation is 13 on both
• HV values. (4 more than spe)
• Due to optical differences??
• -Qie channels
• Light guide coupling
• Fiber lengths
• Leakage
• HV variations
• Connectors
• Table position uncertainity
• Beam variation

10
PMT Afterpulse 101

• Definition Afterpulses are spurious pulses that
  appear in the wake of true pulses.
• FACT Every true pulse may be followed by one or
  more afterpulses.
• FACT The afterpulse size(charge) does not vary
  with respect to the incoming pulse height.
• FACT For bigger incoming light more afterpulses
  appear, So total charge ratio stays the same.
• i.e. Finding the amplitude ratio of afterpulse
  to any main pulse, at any given light intensity,
  is NOT RATE.. because IT IS DIFFERENT for every
  incoming light intensity.
• i.e. Finding the number of afterpulses for any
  main pulse, at any given light intensity, is NOT
  RATE... Because IT IS DIFFERENT for every light
  intensity, as well.
• Only meaningful AFTERPULSE RATE definition is
• RATE (SQafterpulse /SQmain pulse ) x 100

11
The References

• As long as the gain is not too high, the ratio
  does vary much with the number of true pulses, or
  the amount of charge they contain. When the
  charge transferred by each true pulse is very
  small (i.e. SPE level), that transferred by each
  afterpulse may be as large or even larger.
  However, as proportionally fewer true pulses are
  then followed by afterpulses, the charge ratio
  remains the same. (Phillips, Photomultiplier
  Tubes Principles and Applications, page 4-42).
• Definitions and effects
• Hamamatsu and Philips PMT books.
• B.H. Candy Rev. Sci. Instr. 56, 183 (1985)
• G.A. Morton et al. IEEE Tran. Nucl. Sci. NS-14
  No.1, 443, (1967)
• R. Staubert et al., NIM 84, 297 (1970)
• S.S. Stevens et al., IEEE Tran. Nucl. Sci. NS-19
  No.1, 356, (1972)
• Suppression ideas
• S.J. Hall et al., NIM 112, 545 (1973)
• G.P. Lamaze et al., NIM 123, 403 (1975)

12
The University of Iowa Afterpulse Tests

• The tests are performed on 3 different PMT Types
  (R7525, R6427, R1398).
• We Tested 83 R7525 (HF PMTs).
• Blue LED (with 420 nm peak) is used as light
  source. The light intensity set to be 2TeV.
• The LED driver is running with 100 kHz, providing
  the gate signal as well.
• The gate signal is generated by LeCroy 222 Dual
  Gate Generator.
• The measurements are taken by LeCroy 2249 ADC.
• Every measurement contains 10,000 ADC counts.
• We DID NOT use any amplifier to eliminate the
  noise factor.

13
Iowa Afterpulse Test Setup
14
Iowa Measurements
R7525

• Every PMT we tested shows afterpulses.
• There is no magic PMT that gives no afterpulses.

R6427
R1398
15
Iowa Afterpulse Timings
R7525
Different PMTs, similar afterpulse delay times.
2000 events each. With high voltage change peak
positions shift.
300ns 400ns
R6427
300ns
300ns
R1398
16
R7525 Afterpulse Delay Time
R7525 Afterpulses appear to have 3 distinct delay
time regions, possibly due to H2, He and CH4
17
Iowa Afterpulse Rate Results
All afterpulses in 150ns-2µsec range
Different HV values
Different light intensity

• Result R7525 total afterpulse rate (around 2µsec
  range) is always 3.
• It is constant with different light
• Intensities, but tends to increase to 5 at
  1750V.

18
Rates of Afterpulse Peak Regions
Hamamatsu R7525 Only
Rate from peak regions
Single Photoelectron level No charge
accumulation observed over pedestal level, from
afterpulse regions 10000 ADC counts
19
Iowa Afterpulse Conclusion

• There are afterpulses for R7525 as every other
  PMT.
• The afterpulses appear to localize at 3 distinct
  time delay regions.
• Afterpulse rate is 3, for 150ns-2µsec region.
• The charge accumulation rate due to the
  afterpulse regions mostly less than 1.
• It is less than Dark Current charge ratio.
• It does not vary with respect to the batch
  number.
• It does not vary with respect to the incoming
  light intensity or frequency.
• It does not vary much with PMT high voltages.

20
Hamamatsu Tests
Charge Mode

• Hamamatsu performed afterpulse tests on R7525
  PMTs in Charge Mode and Counting Mode.
• They used 400 GeV light for charge mode, 80 GeV
  for counting mode.

Counting Mode
21
Hamamatsu Charge Mode Setup
Charge coming from the afterpulses are integrated
on a 9500 ns range!!
22
Hamamatsu Counting Mode Setup
Number of afterpulses coming 150ns after the main
pulse are counted
23
Hamamatsu Results
Counting Mode and Charge Mode give Around 3-5
over 9500ns range.
Different light intensities.
Different PMT gains.
24
Hamamatsu Conclusion

• R7525 Afterpulses are not different than any
  other PMT Hamamatsu produces.
• The afterpulse tests done in 2001 give the same
  result with 2004 tests. So there is no change in
  time.
• Afterpulse rate does not change with PMT gain
  (HV) or incoming light intensity.
• In the charge mode the integrated afterpulses on
  9500ns range produce 2-3 rate.

25
TB2004 PMT Afterpulse Analysis
60 Time Slice Data for Wedge 13, at 1150V and
1350V
60 Time Slice data The 3 peaks can easily be seen.
20 Time Slice Data, afterpulses.
26
Integrated PMT Afterpulse Rate

• When we take 1.1µsec afterpulse region
• - 1150V gives 2-4 afterpulse rate
• - 1350V gives 2-6 afterpulse rate
• On both cases, Tower 20 is
• significantly higher than the others

Fit to the summation of these two graphs gives
3.2 afterpulse rate.
27
W-13 Afterpulse Rate for 25ns
We calculated the charge accumulation rate of
every 25ns bin of every tower with respect to the
biggest 25ns signal bin.
In any given 25ns time, afterpulse charge
accumulation rate is rarely More than that of
Dark Current.
28
Summary on HF PMTs

• PMT selection and testing is well documented. And
  Iowa-Hamamatsu measurements are in database, on
  the web.
• The relative gain values (taken at 1100V) tend to
  have bigger variation as we increase the HV.
• PMT afterpulse charge accumulation rate is around
  that of PMT dark current.