Measurements of new Geiger mode APD samples from Hamamatsu

1
Measurements of new Geiger mode APD samples
from Hamamatsu
2007/1/19ACC/RICH meeting Y.Mazuka,
M.Yamaoka Nagoya-Univ.
2
New samples from Hamamatsu
New 4 samples have come to Nagoya from Hamamatsu
CPTA sample
New HPK sample
Old HPK sample
20mm
Check the performances Noise Gain Photon
Detection Efficiency
3
Dark noise Gain
Dark noise and gain as a function of
Vover(Vbias-Vbreakdown)
Noise rates are counted with 0.5pe thresholdGain
is calculated from ADC 1pe peak
4
Dark noise vs Gain
T192?
new
old
100pixel
400pixel
gain
With the same gain, Noisenew ? 1/3 less than
Noiseold
5
Measurement of PDE
PDEPhoton Detection Efficiency

• QE(l) probability of electron-hole pair
  creation
• egeom a fraction of the total photodiode area
  occupied by sensitive pixels (new sample 0.6)
• eGeiger(V) probability for a photoelectron to
  initiate the Geiger discharge

PDE is measured by comparison of current of
reference PD and Geiger mode APD
setup
movable stage
mono-chromater
filter
A
Geiger mode APD new sample HPK400-08 Vbias
on measured current is divided by gain
Ref. PD
A
halogenlump
Geiger mode APD
pin holef200mm
Light shield box
First, we measured PDE as a function of
Vbias at l500nm,800nm
6
PDE(l500nm or 800nm) result
Compare with HPK measurement (see 400 pixel
type)
PDE as a function of Vover
100 pixel type
400 pixel type
1600 pixel type
PDE increases with Vover, but it may include
growing cross talk When we adopt Vover2.1V,
absolute value of PDE looks similar to HPKs
results
Noise rate 350kHz
7
PDE(l) result preliminary
Next, we measured PDE as a function of wavelength
at Vover2.1V
HPK measurement (see 400 pixel type)
Our result are similar to HPKs at
lgt460nm region
Actually between 440nm and 460nm point, setup is
a little changed ? Well check it and take more
precise data
8
Summary

• We got new Geiger mode APD samples from Hamamatsu
• Noise rate against each gain became 1/3 less than
  old samples
• PDE(l) at long wavelength(gt460nm) is consistent
  to Hamamatsu data
• PDE 40 _at_500nm, Vover2.1V, Noise rate 350kHz
• It may include cross talk
• They seems to have good performance for RICH
• Next
• Check the PDE at short wavelength region again
• Measure another samples
• Remove cross talk effects