Avalanche Photo-Diodes (APDs)

1

• Avalanche Photo-Diodes (APDs)
• K.Deiters, Q.Ingram, D.Renker, T.Sakhelashvili
• Paul Scherrer Institute, Villigen, Switzerland
• J.Grahl, I.Kronqvist, R.Rusack, A.Singovski,
  P.Vikas
• University of Minnesota, Minneapolis, USA
• I.Britvitch, A.Kuznetsov, Y.Musienko, S.Reucroft,
  J.Swain
• Northeastern University, Boston, USA
• Z.Antunovic, N.Godinovic, I.Soric
• University of Split, Croatia
• Honorary assistent J-L.Faure

2
APD History
Late 1992 1st Hamamatsu prototype Early
90th Push for a homogeneous calorimeter 1995 Tes
t of an APD on a PbWO4 crystal in H4
1996-97 APDs chosen for ECAL
A comparison of the response to 80 GeV electrons
of a lead tungstate crystal with a PIN diode
(top) and an APD (bottom) read-out. The tail to
the right of the peak in the PIN diode spectrum
is due to particles leaking out of the back of
the 18 cm long crystal and passing through the
diode
3
APD properties
4
APD properties cont. (1)
APDs are insensitive to magnetic fields, are
compact, cheap and have a small nuclear counter
effect. They are potentially radiation hard. But
they have a small area (5x5 mm2) and they were a
widely unknown device
5
Distance of operating to breakdown voltage
Stable since 2000
Side remark Thousands of APDs have been tested
and sometimes accidents happened. APDs were
biased with the wrong polarity for a long period
or the bias voltage was far too high (3000 V
instead of 300 V). No APD ever died due to such
an event.
.
6
Reliability tests

• Weak or bad APDs are identified by a number of
  tests after the 60Co irradiation
• is there a change of the breakdown voltage (dVb)?
• is the dark current at gain 50 within the limits
  (Id)?
• Is the dark current divided by the gain falling
  continuously (Id/M vs M)?
• Is there an abnormal high noise?
• Is the APD coming from a bad position on the
  wafer (mask problem)?
• Is the APD coming from a bad wafer (problem in
  the neutron irradiation)?
• The same tests are performed after the annealing
  with the exception of the noise.
• The noise is later on measured in Lyon when the
  APDs have been mounted on
• the capsule.
• How do we measure these parameters? Yuri
• How efficient is the screening? Quentin
• What are the rejection criteria? Sasha

7
PROVISIONAL FINAL TECHNICAL SPECIFICATION

• General conditions VR (operating voltage for
  gain M 50) and VB (breakdown voltage, at which
  the dark current exceeds 100 ?A), as
• all other values measured at T 25? C. For
  radiation hardness neutron (1 MeV) fluence
  2x1013 n/cm2
• Sensitive area 5x5 mm2
• Survival rate after irradiation 99.9 ? 0.1
• Survival rate after 2 months at 90º C _at_ M50
  99.9 ? 0.1
• Survival rate after 1 day at 20 ?A (V VB 20V)
  (1) 99.9 ? 0.1
• Passivation layer Si3N4
• Charge collection within 20 ns (2) 99 ? 1
• Capacity at VR (mean value) 65 85 pF
• Spread of capacity at VR ? 5 pF
• Capacity at VR Capacity at VB ? 3 pF
• DM/dV1/M (gain sensitivity to voltage) at VR ?
  3.5 /V
• Quantum efficiency (mean value) (3) 75 ? 5 _at_
  430 nm
• Spread of measured quantum efficiency (3) ? 7
• Operating voltage VR 340 440 V
• Difference VB - VR ? 37 V
• Nuclear counter effect leff (4) lt 7 ?m
• Excess noise factor at VR 2 ? 0.3 _at_ 430 nm
• Dark current Id at VR ? 50 nA