CORRELATION BETWEEN LEAKAGE
CURRENT AND NOISE
Talk by Edmund WIDL
(HEPHY Vienna)
presentation can be downloaded from: wwwhephy.oeaw.ac.at/u3w/e/ewidl/www/talks/NOISE.ppt
INTRODUCTION
For detectors measuring absorbed energies it is
useful to quantify the noise in terms of
Equivalent Noise Charge (ENC)
This ENC corresponds to the amount of charge
that would have to traverse the detector to yield
a signal as big as the actual noise:
noise = ε . ENC
ε ... calibration constant
AMPLIFIER NOISE
• main noise source: input transistor at the amplifier
• contributions of further stages usually neglected
• load capacitance plays an important role due to
the integrating nature
In a simple approach, the amplifier noise can be
described by the sum of:
• a constant value (parallel noise) and
• a part which scales with the load capacitance C
(series noise)
ENC APV = ENC APV, parallel + C ⋅ ENC APV, series
250 + 36 pF −1
(peak )
ENC APV [e] = 
−1
+
400
60
pF
(deconvolut ion)

ADDITIONAL NOISE SOURCES
equivalent network for a single strip & readout:
Ileak
Rpoly
C
Rstrip
TP
... fraction of the detector leakage current
seen by one strip
... polysilicon resistor
... strip capacitance
... line resistance of one strip
... peaking time
numerical noise equations:
ENCleak [e] = 106 ⋅ Ileak [nA] ⋅ TP [µs]
TP [µs]
ENCpoly [e] = 758 ⋅
Rpoly [MΩ]
R strip [Ω]
ENCRS [e] = 0.395 ⋅ C ⋅
TP [µs]
The total noise figure is the square sum of the
individual contributions, since the individual sources
are uncorrelated:
ENC2 = ∑ ENCi2
CORRELATION FOR OB2-MODULES
OB2-sensors:
• Rpoly = 1.5 MΩ
• C = Cint + Cback
w strip
= 0 .8 + 1 .6
pF cm−1
pstrip
⇒ C = 11.5 pF
• Rstrip:
wstrip=1.15 ⋅ wimplant= 52.6 µm
dstrip≈1.5µm
Lstrip=9.44 cm ⇒ R strip = ρ Al
L strip
= 31.7 Ω
w strip ⋅ dstrip
OB2-modules:
• 2 electrically daisy-chained sensors
• Rpoly of each sensor in series
• C of each sensor parallel
• APV: TP = 0.05 µs
This yields:
1066
(peak )

ENC APV [e] = 
1759 (deconv.)
ENCpoly [e] = 138
ENCRS [e] = 319
⇒
 (561.8 ⋅ Ileak [nA ] + 1 254 949 )1 2
(peak )
ENC(Ileak ) [e] = 
12
(
)
⋅
+
561
.
8
I
[
nA
]
3
214
805
(deconv.)

leak
RESULTS
In order to correlate the noise from different
sensors the relative noise of each sensor is used:
noise (Ileak )
ENC (Ileak )
=
noise (Ileak = 0 ) ENC (Ileak = 0 )
left side of the equation: data from 57 OB2 modules
• Ileak for each strip from the tracker database
• noise(Ileak= 0) is taken as the noise from daisychained strips where the sum of Ileak is less than 2
nA (minimal error – less than 0.5‰)
right side of the equation: theoretical relative noise
following strips were not included:
• strips #1 & #128 of each APV (are known to
have higher noise)
• strips with an Ileak-entry in the database higher
than the total detector current at 400V
• unbonded strips
28623 strips remaining
leakage current distribution:
(in bin 0-10 nA)
leakage current vs. strip noise (peak mode):
theoretical correlation
leakage current vs. strip noise (peak mode):
theoretical correlation
leakage current vs. strip noise (deconv.):
theoretical correlation
leakage current vs. strip noise (deconv.):
theoretical correlation
MODULES WITH CM-PROBLEMS
Several
modules
showed drastically
increased noise on
single APVs, obviously caused by
single,
extreme
leaky strips.
None of these strips showed an abnormal
behavior prior to assembly: