Upload
others
View
24
Download
2
Embed Size (px)
Citation preview
SiPM BIAS and AMPLIFIER CIRCUIT
G. FELICILNF-INFN
EDIT2015–FRASCATIOCTOBER20-29 2D.BADONI-G.FELICI
SIPM POWER SUPPLY !
SiPM BIAS (I)
EDIT2015–FRASCATIOCTOBER20-29 3D.BADONI-G.FELICI
The Problem u 8 SiPMsu Each SiPM has a different working point TWO POSSIBLE APPROACHES
16 lutego 2011 Wojciech Kucewicz 8
FLCFLC--SiPMSiPMFLCFLC--SiPMSiPMThe The SiPMSiPM gain strongly depended of the bias voltage, so it has to be gain strongly depended of the bias voltage, so it has to be
adjusted for each diode.adjusted for each diode.An 8An 8--bit DACbit DAC was added in parallel to each input of preamplifier.
The DAC provide up to 5 V bias voltage moderation.
HV=VBREAKDOWN+VREF DAC
VSIPM = HV - VDAC
1st : use a DAC to move SiPM reference• Very cheap using multi-channels DAC• Easy to integrate in ASIC design• Very low ripple required in main HV generator• SiPMworkingpointsetupisabittricky(singleSiPMworkingpointsarenotindependent)
SiPM BIAS (II)
EDIT2015–FRASCATIOCTOBER20-29 4D.BADONI-G.FELICI
2nd : use Main Voltage Generator and different regulators to adjust the single SiPM working point u The regulators behave also like a filter to reduce the Main Voltage Generator rippleu Independent SiPM working points adjustmentu More components and PCB space required (size of the board increased)
3.2. Layout of the front-end electronics 37
October 2013. In this section, the circuit layout is introduced. The full layout isprinted in the appendix, see fig. A.3. In figure 3.10, a schematic version is shown inorder to explain why this layout is better than the default circuits proposed by theSiPM manufacturers. Details are mentioned in the following subsections.
The default layout proposed by several manufacturers [23] [22] [26] is quite easy tounderstand, see figure 3.9: A low-pass at the cathode of the SiPM is used to limitthe current and to filter high frequencies from the constant bias voltage. The SiPMsignal output has already been introduced in the last chapter and is based on a shuntresistor and an (inverting) op-amp.
3.2.1 Beissel voltage controlling circuit
The Beissel voltage controlling (BVC) circuit has originally been introduced by FranzBeissel († 2012) at III. Physikalisches Institut, RWTH Aachen University. Figure3.11 shows a very simplified version of the BVC circuit. The simplified version canbe used to explain the DC performance of the voltage controller.
The di↵erential voltage at the op-amp is zero, which means that the op-amp is tryingto pull the voltage in the inverting node to V
r
. A current flows through the resistorR
7
which is equal to
I
7
= V
r
/R
7
.
The op-amp draws no current in the input, that is I� = 0. The same current I
7
flows through the resistor R6
. Therefore,
�
+
V
r
I�
R
7
I
7
R
6
I
6
L
b
I
b
V
b
C
b
R
E
V
i
C
E
I
6
+ I
b
Figure 3.11: Highly simplified circuit of the Beissel voltage controlling circuitmodified in this thesis, see fig. A.3 for detailed circuit elements. The voltage atnode V
b
is set by the op-amp according to eq. (3.2). This simplified circuit is validfor DC performance only and especially when no current is drawn at the V
b
node,i.e. I
b
= 0.
Simplified version of Beissel voltage controlling circuit
HOW DOES IT WORK ?u Assuming an ideal operational amplifier the voltage at the inverting node we’ll be Vr, then I7=Vr/R7
u The same current will flow through R6 (ZIN=∞)à Vr/R7= (Vb –Vr)/R6
u Then if Vi > Vb (the transistor must be polarized) à Vb=(1+R6/R7)*Vr
u Adjusting the R6/R7 ratio or Vr value we can adjust the SiPM working voltage
IDEAL AMP OPu gain=∞u ZIN=∞u ZOUT=∞u BW=∞
EDIT2015–FRASCATIOCTOBER20-29 5D.BADONI-G.FELICI
PREAMPLIFIERS !
VOLTAGE PREAMPLIFIER
EDIT2015–FRASCATIOCTOBER20-29 6D.BADONI-G.FELICI
PREAMPLIFIER=INPUT AMPLIFIERu Generally located on the detector (ON-DETECTOR ELECTRONICS)u Amplify the signal optimizing the signal-to-noise ratiou Three basic configurations:
u Voltage preamplifier u Current preamplifieru Charge preamplifier
VOLTAGE PREAMPLIFIER (for sensors generating voltage signals or as second stage for sensors generating current signals)
en
AV
detC( )↓ti _det preinconntottot
inout CCCC
CQV ++==
Ctot can change as a function of detector working parameters or PT à Vout changes
CURRENT PREAMPLIFIER
EDIT2015–FRASCATIOCTOBER20-29 7D.BADONI-G.FELICI
infout iRV −≈
Rf
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
+⋅−=
−=−
−=
A
iRV
iRVVAVV
infout
infinout
inout
11
1
AV
detCVout
Vin
+-
( )↓ti
INPUT IMPEDANCE
→+
=→−=⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
+⋅−
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
+⋅−=
−=
AiR
VAV
A
iR
A
iRV
AVV
infinininf
infout
inout
111
1
11
1
AR
Z fin +=1
Cin CONTRIBUTION
RCjRZin ω+
=1
u InputsignalisconvolvedwithanexponenXal.
u IncreasingRfincreasesboththepreamplifiersensiXvityandτ.
C R( )↓ti
CHARGE PREAMPLIFIER
EDIT2015–FRASCATIOCTOBER20-29 8D.BADONI-G.FELICI
( )
( )
⎟⎟⎟⎟
⎠
⎞
⎜⎜⎜⎜
⎝
⎛
+⋅−=
−=−
−=
ACj
iV
CjiVV
AVV
f
inout
f
ininout
inout
11
1ωω
ωω
fC
Vout+
-
( )↓ti
( )f
inout Cj
iVωω
−≈
t
Vout
t
-Q/Cf( ) ( )tQti inin δ⋅=
INPUT IMPEDANCE
Vout = −AVin Vout = −iin ω( )jωCf
⋅1
1+ 1A
⎛
⎝
⎜⎜⎜
⎞
⎠
⎟⎟⎟
−iin ω( )jωCf
⋅1
1+ 1A
⎛
⎝
⎜⎜⎜
⎞
⎠
⎟⎟⎟= −AVin →Vin =
iin ω( )jωCf ⋅ 1+ A( )
( ) fin CAC ⋅+= 1
HOW MUCH SIGNAL ARE WE GOING TO LOOSE ?
( ) ( )
( ) f
tamp
tff
amp
CACQQ
CCAQ
CAQ
⋅++
=
+⋅+=
⋅+
11
11
Es. A=103; Cf=1pF
Ct=10 pF àQampl/Q=0.99
Ct=100 pF àQampl/Q=0.90
tC ( ) fCA ⋅+1( )↓ti
EDIT2015–FRASCATIOCTOBER20-29 9D.BADONI-G.FELICI
WHICH AMPLIFIER ?!
VOLTAGE, CURRENT OR CHARGE AMPLIFIER ?
EDIT2015–FRASCATIOCTOBER20-29 10D.BADONI-G.FELICI
u SiPMs output charge of the order of hundreds of fCu SiPM devices have excellent timing properties. u SiPM devices exhibit single photon resolution, but have a quite poor lineari response
Voltage or Current amplifiers configurations can be used as head-stage amplifiers
+V
+
_
R1
R2
RS
(a) Pre-amplifier
+V
+
_
RF
CF
RG
(b) Transimpedance
Figure 2: Two amplifier setups, not shown are the capacitively decoupledpower lines which all high speed circuits should have.
called a voltage amplifier. The impedance of the photodiode is largely ir-relevant as it is normally much larger than RS. Increasing RS increasesthe voltage generated by the photodiode, however this also increases thevoltage noise generated by the input current noise of the op-amp and alsodecreases the bandwidth of the photodiode. The gain in non-inverting setupis set by one plus the ratio of the feedback resistor to the resistor to groundG = 1 + R2/R1. The bandwidth (cuto↵ frequency, f�3dB) of the circuit isnormally set by the gain bandwidth product of the op-amp as the photodiodenormally has a much larger bandwidth.
2 Transimpedance amplifier
In the transimpedance amplifier setup the amplifier itself is acting as thecurrent-to-voltage convertor (this is often called a current amplifier), withthe output voltage gain set by the feedback resistor. A feedback capacitor isalso often needed control the frequency response of the circuit as at higherfrequencies the signal increases with a gain of
3
In voltage amplifiers increasing RS increases input amplifier voltage, but amplifier input voltage noise increase and bandwidth decreases.Current (or transimpedance) amplifier allows to overcome voltage amplifier limitations in terms of gain and bandwidth still maintaining a good signal-to-noise ratio (but stability problems could arise because diode parasitic capacitance).
EDIT2015–FRASCATIOCTOBER20-29 11D.BADONI-G.FELICI
AMPLIFIER STABILITY !
SiPM AMPLIFIER (I)
EDIT2015–FRASCATIOCTOBER20-29 12D.BADONI-G.FELICI
SiPM PS
16
© 2004 National Semiconductor Corporation
The Photodiode Amplifier:
• Oh yeah, addsome feedback:
-10 Volts
Light +
-
+-
This pin stays at ground so output goes more positive with more light.
This current makes positive voltage here
-
+
16
© 2004 N
ational Sem
iconductor Corporation
Th
e P
ho
tod
iod
e A
mp
lifier:
•Oh yeah, add
some feedback:
-10 Volts
Light+ -
+-
This pin stays at ground so output goes m
ore positive w
ith more light.
This current makes
positive voltage here
- +
The SiPM current makes positive voltage at the output of operational amplifier
This pin at GND force this pin to GND (virtual GND)REMIND à GAIN=∞
BUT most likely if you use high bandwidth amplifier looking at the output with a scope
17
© 2004 National Semiconductor Corporation
The Photodiode Amplifier:
• So it oscillatesand/or clips,what is wrong?
-10 Volts
Light +
-
THE AMPLIFIER OSCILLATES !! WHY ???
SiPM AMPLIFIER (II)
EDIT2015–FRASCATIOCTOBER20-29 13D.BADONI-G.FELICI
-
+
THE PROBLEM: oscillation are caused by SiPM parasitic capacitance on input
17
© 2004 National Semiconductor Corporation
The Photodiode Amplifier:
• So it oscillatesand/or clips,what is wrong?
-10 Volts
Light +
-
-
+
16
© 2004 National Semiconductor Corporation
The Photodiode Amplifier:
• Oh yeah, addsome feedback:
-10 Volts
Light +
-
+-
This pin stays at ground so output goes more positive with more light.
This current makes positive voltage hereSiPM current source
forces output to change
SiPM parasitic capacitance lag the feedback signal
FEEDBACK LAG IS BAD
SiPM AMPLIFIER (III)
EDIT2015–FRASCATIOCTOBER20-29 14D.BADONI-G.FELICI
-
+
THE SOLUTION: insert a capacitance to compensate for the phase lag
NEXT PROBLEM: noise (current noise & voltage noise)
CURRENT noise important on this pin
VOLTAGE noise important on this pin
AMPLIFIER NOISEu Hard to match low current noise and low voltage noise requirement on the same deviceu JFET amplifiers have low current noiseu Bipolar amplifiers have low voltage noise
u More on this in Noise Introduction slides ….
BIBLIOGRAPHY
EDIT2015–FRASCATIOCTOBER20-29 15D.BADONI-G.FELICI
[1]Front-endElectronicsforSiliconPhotomulXpliers,JohannesSchumacher[2]Photodiodeamplifier–G.Lochead