Timing and Cross-Talk Properties of BURLE/Photonis Multi-Channel MCP PMTs

S.Korpar

a,b

, I. Adachi

c

, R. Dolenec

b

, P.

Križan

d,a

, R. Pestotnik

b

, A. Stanovnik

d,a

a

University of Maribor, Slovenia,

b

Jožef Stefan Institute, Ljubljana, Slovenia,

c

KEK, Tsukuba, Japan,

d

University of Ljubljana, Slovenia

Abstract

We report on on-the-bench studies of the two types of BURLE multi anode micro channel plate (MCP) PMTs, one with 64 channels and other with 4 channels, both with 10 m m pores. A possible applications of this tubes are RICH and time-of-flight counters. We have investigated the timing properties of the tubes and studied various cross-talks and their influence on the timing and spatial resolution. We have also performed tests in magnetic field up to 1.5 T.

Motivation

For the Belle particle identification system upgrade, a proximity focusing RICH detector with aerogel as radiator is being considered. One of candidates for the detector of Cherenkov photons is a microchannel plate PMT. With its excellent timing properties, such a counter could serve in addition as a time-of-flight counter. A prototype of this novel device using BURLE 85011 64-anode, microchannel plate PMT with 25 m m pores, was tested in the test beam at KEK. Excellent performance of this counter could be demonstrated (left). In particular, a good separation of pions and protons was observed in the test beam data with a time-of-flight resolution of 35ps (right).

Modeling of processes in the MCP PMT

Photo-electron:    d 0,max ~0.8 mm t 0 Δt 0 ~ 1.4 ns ~ 100 ps g

e − d 0

a

l

Parameters used: - cathode to MCP potential difference U = 200 V - photocathode to MCP distance L = 6 mm - photoelectron initial energy E 0 = 1 eV g

d 1

Photo-electron range, projected Backscattering:   d 1,max =2L=12 mm t 1,max ~ 2.8 ns b

e − l

Charge sharing Distributions assuming scattering by angle b that back is uniform over the solid angle.

Photo-electron travel time ~ 12mm ~ 2.8ns

Timing

back-scattering

t

1  2

t

0 sin b

d

1  2

l

sin 2 b TDC distribution has three contributions: - prompt signal ~ 70% - short delay ~ 20% - uniform distribution ~ 10%

Present study:

-measure detailed timing properties and cross-talk, -determine their influence on the position resolution and time resolution.

Experimental set-up for on the bench tests

BURLE 85011 MCP-PMT: - multi-anode PMT with two MCP steps - 2x2(8x8) anode pads - 10 m m pores - bialkali photocathode - gain ~ 0.6 x 106 - collection efficiency ~ 60% Outside dark box: - PiLas diode laser system EIG1000D (ALS) - 404 nm and 635 nm laser heads (ALS) - neutral density filters (0.3%, 12.5%, 25%) - optical fiber coupler (focusing) - optical fiber (single mode,~4 m m core) Signal processing: - laser rate 2kHz (~DAQ rate) - amplifier: 350MHz (<1ns rise time) - discriminator: leading edge, 300MHz - TDC: 25ps LSB( s ~11ps) - QDC: dual range 800pC, 200pC - HV 2400V Inside dark box mounted on 3D stage: - optical fiber coupler (expanding) - semitransparent plate - reference PMT (Hamamatsu H5783P) - focusing lens (spot size s ~ 10 m m) NIM amplifier ORTEC FTA820A signal splitter passive 3-way discriminator Philips model 806 TDC Kaizu works KC3781A CAMAC QDC CAEN V965 VME ALS PiLas controller PC LabWindows CVI

Tests in magnetic field

Measurements of gain, uniformity of response snd cross talk Magnet at KEK with B field up to 1.5 T Light source - laser: •wavelength 439 nm •spot size < 0.5 mm •pulse timing 90 ps (FWHM)

Gain in magnetic field

.

.

Gain as a function of magnetic field for different operation voltages and as a function of aplied voltage for different magnetic fields.

70% 1 1 20% 10% 2 2

s = 40ps s = 37ps s = 39ps s = 38ps Time resolution of the main peak is dominated by the photo-electron time spread ( 26ps rms, estimated from the model); other contributions are laser timing (15ps rms) and electronics (12ps rms).

Time-walk corrected TDC distributions of all four channels of 2x2 MCP PMT.

Timing with a signal from the second MCP stage

If a charged particle passes the PMT window, ~10 Cherenkov photons are detected in the MCP PMT; they are distributed over several anode channels.

Idea

: read timing for the whole device from a single channel (second MCP stage), while 64 anode channels are used for position measurement MCP second stage output

s < 40

ps Single anode Timing resolution as a function of light intensity

Uniformity of response at B=0T and B=1.5T

Surface response of PMTs is fairly uniform. Multiple counting is observed at pad boundaries due to charge sharing.

Scans across a 64 channel PMT without (top) and with magnetic field (bottom). B = 0 T, HV = 2400 V B = 1.5 T, HV = 2500 V

2 x 12mm

= range of back-scattered photo-electrons Slice of 2D distribution shows a uniform response within the pads, and the long range cross talk due to photoelectron back scattering and photon reflection.

Number of detected hits on individual channels as a function of light spot position.

In the presence of magnetic field, charge sharing and cross talk due to long range photoelectron back-scattering are considerably reduced.

#

ch

(rad)