Bases, Voltage Dividers,

and Preamplifiers
Accessories for
Scintillation Detectors

Table of Contents
Plug-on Voltage Dividers
Models P-12, P-14
Page 2
Plug-on Voltage Divider with
Integrated Preamplifiers
Models PA-12 and PA-14
Page 3
Low Noise Voltage Divider/
Preamplifier for Model 1XMP.040B
Model PX-250
Page 4
Voltage Divider Design
Considerations
Page 5,6,7
Negative High Voltage
Page 7

Voltage dividers and voltage divider/preamplifiers allow

you to supply the proper operating high voltage distribution to the photomultiplier tubes (PMTs) and to get the
signal from them to your system electronics. These devices
can be plugged onto or hard-wired to PMTs.
Plug-on models are easy to connect. You simply push them
onto the PMT bases connecting pins. Models are available
for PMTs with the standard 12-pin and 14-pin bases (the
most popular), as well as 20-pin and 21-pin bases.

Plug-on
Voltage Dividers
Standard Features of Models P-12 and
P-14 Plug-on Voltage Dividers
Aluminum housing
Positive (+) high voltage operation
BNC bulkhead connector for signal output
SHV bulkhead connector for high-voltage
supply input

P-12 Voltage Divider Schematic

800 to 1200 VDC (may be higher or lower

depending upon the specific PMT) at
<1 mA typical operating power requirements

2.2"
(55mm)
1.7"
(42mm)

Gain potentiometer
Focus potentiometer (P-14 only)
1.9"
(48mm)

P-12 Voltage Divider for 1.5" PMTs

P-14 Voltage Divider Schematic

2.9"
(73mm)
2.4"
(60mm)

2.2"
(57mm)

P-14 Voltage Divider for 2" , 3", 3.5" and 5" PMTs

Plug-on Voltage Dividers

with Integrated Preamplifier
High Voltage
Connector

Coaxial
Cable

Additional features for models PA-12 and

PA-14 plug-on voltage divider/preamplifiers
Standard hard-wired preamplifier power cable
terminating in a 9-pin connector (pin 6 for
-24VDC, pin 1 common ground)
Optional BNC (in place of standard 9-pin)
connector for preamplifier power
input; -24V at 21mA operating power
requirements

Preamp
Connector
Signal
Connector

Gain
Potentiometer

Voltage Divider Socket Compatible

Network with
with 12-pin Base
Preamplifier

Aluminum
Cover

Preamplifier specifications
Noise: Less than 50 microvolts
RMS referred to input
Input: Directly coupled via 1F capacitor
from photomultiplier tube anode to
charge integrating preamplifier input
circuit, positive pulse output.
Conversion Gain: 100mV/picocoulomb

PA-12 Voltage Divider/Preamplifier

Coaxial
Cable

Focus
Signal
Connector

Rise Time: 100 nanoseconds

Decay Time: 10 microseconds (internally
adjustable at the factory)
Gain Stability: 0.25%/C from 0 to 50C

Preamp
Connector

Linearity: 0.25% (preamplifier only)

Power Requirements: -24V +10% @21mA

High Voltage
Connector

Gain
Potentiometer

Voltage Divider Socket Compatible

Network with
with 14-pin Base
Preamplifier

Aluminum
Cover

PA-14 Voltage Divider/Preamplifier

POT2

High Voltage

1.5 M

J1
SHV

392
C4 0.05 uF
1 KV

R20 100

Pre-Amp Power

J3
BNC
9-Pin

-24VDC

D1
1N914

R13
1M

C5

R15
2.7 K

R17
1K

10P
Q1
2N3906

Q2
2N3906

Q3
2N3906
R21

J2
BNC

51
C9

D2
1N914

R14
1M

C6
10uF
35 V

R16
10 K

C7
10uF
35 V

R18
1K

R19
1K

C8

10uF
35 V

Signal Out

10uF
35 V

3
PA-12 Voltage Divider Schematic

PA-14 Voltage Divider Schematic

Preamplifier Schematic for both PA-12 and PA-14

PX-250 Low Noise Voltage Divider/

Preamplifier for Model 1XMP.040B

The PX-250 voltage divider/FET preamplifier

assembly is integrated into a Model
1XMP.040B X-ray probe. It is specially
designed for X-ray detection applications.
Plug-on versions are also available.

Preamplifier Specifications
Noise: Equivalent noise charge of 1.25 x
103 RMS electrons referred to the input
(typical)
Input: Capacitively coupled from photomultiplier tube anode to charge integrating preamplifier input circuit, positive
pulse output.
Charge Sensitivity: 667 mV/picocoulomb
(typical)
Rise Time: 100 nanoseconds
Decay Time: 33 microseconds

High Voltag
e Divider A low-leakage curoltage
rent, 12-pin photomultiplier socket is wired
with resistors to provide the proper interdynode voltages. Capacitors C1, C2 provide
peak currents when large signals are encountered at the last two dynodes. D1 and
D2 provide protection against negative high
voltage: Resistors R-13, R-14 along with C-4
and C-5 filter out noise from the high-voltage power supply. R12 is the load resistor
for the photomultiplier tube.

Pr
eamplifier The preamplifier is a
Preamplifier
charge sensitive circuit. Pulses from the
photomultiplier tube are coupled to Q1, a
field-effect transistor which provides for
low-noise operation. The drain of Q1 is
coupled to the emitter of Q2, a common
base amplifier. The collector output of Q2 is
connected to the Darlington Pair emitter
follower comprised of Q3 and Q4. Feedback
is provided through C6. DC stabilization is
provided through R-15 and R-17 by connecting the emitter of Q4 to the gate of Q1. R30 and C-13 filter the -24 Volt power supply.
C-14 couples the signal to the output connector.
Test input should be made through a 1 pF
capacitor to pin 6 of the photomultiplier
socket when the PMT and high voltage are
not connected.

Gain Stability: 0.25%/C from 0 to 50C

Power Requirements: High voltage
Typical 900 to 1400 VDC.; Preamp power
24 VDC at 12MA
Cables supplied: 12' of RG-59B/U with
SHV plug; Preamp power 12' of RG58/U with 9-pin plug; Signal 12' of
RG-58/U with a BNC plug

510 K

390

PX-250 Voltage Divider/Preamplifier Schematic

Voltage Divider Design Considerations

Voltage dividers used by SGCD are resistive devices specially
designed to accelerate electrons from the cathode to the anode of a
phototube. Or, to say this in another way, a resistive voltage divider
is a chain of resistors linked end to end (in series). One end of the
chain or string of resistors is attached to a voltage source and the
other end attached to the voltage source return (ground) so that a
complete circuit or loop is formed. The connection points between
resistors in the string are attached to the elements of a phototube
to cause electrons to be accelerated from the cathode to the anode.

Let :

A.

In order for electrons to be accelerated through a phototube, each

element must be more positive in voltage than the element that
immediately precedes it. See Figure 1 for a typical resistive voltage
divider and phototube circuit.

H.V.
RG
R
RL
C
CC

=
=
=
=
=
=

+1200 volts
1.2 M
0.1 M
0.1 M
0.01 F
0.001 F

The vvoltag
oltag
e bec
omes mor
e positiv
e as w
e mo
ve fr
om K to F
oltage
becomes
more
positive
we
mov
from
to D1 to A.
If the wiper of RG is set to point Y, then 1200 volts is applied between points Y and X. This 1200 volts is applied across 12R;
therefore, each R drops 100 volts. Starting with the cathode atground or 0, F is at 100 volts. D1 is 200 volts, D2 at 300 volts,
etc. Each element is 100 volts more positive than the preceding
element.
To help in looking at some of the other characteristics of this
positive divider string, an equivalent circuit combining Figures
1 and 2 is shown in Figure 3.

Figure 1.

The circuit in Figure 1 is typical for either a positive or negative high

voltage system. Note that in this circuit there are points at the ends
of the string labeled X and Y and S. Connections to these three labeled points make the phototube operate at positive or negative
high voltage with direct or capacitive coupling to the preamp. We
will review operation at positive high voltage first and then go on to
negative high voltage. To use positive high voltage with a gain adjustment potentiometer, make connections to points X, Y and S as
shown in Figure 2.

RG
R
RL
Cc

BNC
Signal
Connector

Positive
High
Voltage

Figure 2. Connections to points X, Y and S of Figure 1 for

operation with positive HV
As an example, we will now go through this positive high voltage
circuit (Figure 1 and Figure 2) to show the following:
A.
B.
C.
D.
E.

The voltage becomes more positive as we move from K to F to

D1 ... to A.
The approximate range of gain adjustment made possible by
the use of RG .
The change in voltage at the BNC signal connector when 137 Cs
gamma ray is detected.
The reason for the capacitors on D9 and D10.
An average phototube current and the voltage divider current.

First, we assign some reasonable values to the various circuit

elements.

1.2M
0.1M
0
0.1M
0
0.001F
0

Figure 3.
B.

Ground

=
=
=
=

The appr
oxima
te rrang
ang
e of gain adjustmen
appro
ximate
ange
adjustmentt made possible
by the use of RG.
If we assume that the resistance of the phototube is much,
much greater than 12R, then the resistance of the parallel
current paths shown by IPMT* and IDIV is approximately equal
to 12R. The applied HV is then dropped across the series combination of RG plus 12R and IDIV is equal to IT. A simple ratio
can be derived to show how much of the applied HV can be
dropped across RG. Since for every change of 100 volts across
the phototube (volts dropped across RG), the gain changes by
approximately a factor of 2. We can calculate the range of
gain adjustment made possible by the use of RG. The voltage
dropped across the gain pot is equal to the applied high voltage times the resistance of the gain pot divided by the voltage divider resistance plus the gain pot resistance.
RGain Pot

V(gain pot) = V(applied) x

RDivider + RGain Pot

In the example:
V(gain pot) = 1200 x

1.2

= 600 volts

1.2 + 1.2

*Current here is defined to be the flow of electrons.

This is contrary to the normal definition of current.

Voltage Divider Design Considerations (continued)

The gain pot in this circuit can change the phototube gain by a factor of approximately 64. This large gain adjustment range may be
desirable when trying to match the gain for a group of phototubes,
but it is not a good design practice. As a rule, the gain pot resistance should be less than or equal to one half the voltage divider
resistance. Look at the values used in the SGCD standard P-14:
Gain Pot Resistance = 1.5 Mohms

If the anode is connected through the coupling capacitor to a transmission line terminated in 50 ohms, then the voltage
developed across this 50 ohm load is given by:
Vout = I x R = (1.7 x 10- 3) x 50 = 85mV
Case 2.

Voltage Divider Resistance = 6.6 Mohms

V(gain pot) = 1200 x

Case 1.

If the output signal is connected to the input of a charge sensitive

preamplifier through a coupling capacitor, the equivalent circuit is
shown in Figure 4. (Note that R1 and C1 correspond to RL and Cc in
Figures 1, 2 and 3.)

1.5 = 200 volts

8.1

Using our standard P-14, the phototube gain can be changed by a

factor of approximately 4. Notice the range of gain adjustment
changes as the applied voltage changes.
The initial assumption that the resistance of the phototube is
much, much greater than 12R, can be checked as follows:
RPMT =

V(Applied)

I(Dark Current)

1200
1 x 10-9

BASE

= 1.2 x 1012 ohms

PREAMP

R1 = 100K
R2 = 1M
R3 = 680K

C1 = .001F
C2 = .01F
C3 = 100pF

Cs = 45pF

and 1.2 x 1012 ohms is much, much greater than 12R(C1.2 Mohms)..

Figure 4.

C. The chang
e in vvoltag
oltag
ea
ay
change
oltage
att poin
pointt S when 137C s gamma rra
is detec
ted.
detected.
When no signal is present, the voltage at point S (the anode) is
equal to the applied high voltage minus the voltage across the gain
pot. The coupling capacitor keeps this voltage from being present
on the BNC signal connector. The voltage change at point S caused
by the total absorption of a 137 Cs gamma photon can be calculated
as follows.

The charge developed at the phototube anode will be integrated on

C3 and CS provided that the coupling capacitors C1 and C2 are
much larger than C3 and CS. In this case, the peak voltage developed at the preamp input will be:

For a photomultiplier tube, a typical bialkali photocathode generates about 8 photoelectrons per keV of nuclear photon energy deposited when coupled to a good quality NaI(Tl) scintillator.

This voltage will decay with a time constant of:

Therefore, the total charge at the PMT anode corresponding to a

137
Cs gamma photon is calculated by the following formula:

where R is the parallel combination of R1, R2 and R3.

Q = N x E x G x Ce
where
N = Number of photoelectrons/keV: 8
E = Energy of Gamma Photon: 662 keV for 137Cs
G = Photomultiplier Gain: 5 X 105 typical
Ce = Charge per electron: 1.6 x 10-19 coulomb
therefore
Q = 8 x 662 x (5 x 105 ) x (1.6 x 10-19) = 4.24 x 10-10 C
This is the total charge transferred from the PMT anode to the load.
To calculate the anode current per pulse, the value of charge needs
to be divided by the time interval. Since the natural decay time
constant of NaI(Tl) is 250nsec, the anode current per pulse (Ia) is:
I a = Q/T = 4.24 x 10-10 / 250 x 10-9 = 1.7mA

V =

Q
C

4.24 x 10-10
1.45 x 10-10

= 2.92V

where C is the parallel combination of C3 and CS.

T = CR = (1.45 x 10-10)(8 x 104) = 12 sec

Note 1:
The general purpose main amplifiers designed for nuclear instruments include pulse shaping and pole-zero correction functions.
Typically, these functions require a minimum of 50sec input pulse
decay time constant.
In Case 2 of the above example, the pulse decay time constant is
12sec which is too short for the input requirements of a main amplifier. As a result, the pole-zero function will not be effective. To
correct this situation, the decay time constant of the output pulse
in Case 2 should be increased to a minimum of 50sec. One simple
way to achieve this is to increase the values of R1 and R2. If we
choose to use 2.2 Mohm for both R1 and R2, the net effective resistance would be the parallel combination of R1, R2 and R3 which is
420 kohm and the value of the time constant would be:
T = (145 x 10-12) (420 x 10-3) = 61sec
This value is well within the range of the pole-zero correction function of the main amplifier.

Voltage Divider Design Considerations (continued)

Note 2.
The count rate capability of the divider/preamp combination is affected by the selection of the pulse decay time constant which is a
function of the anode load resistor values. Additionally, the gain pot
setting will affect the decay time constant in systems which are designed with a gain pot.
Note 3.
The coupling capacitor (C1) in this voltage divider is not very large
compared to the preamp input capacitance. This coupling capacitor
will charge to a significant fraction of the input charge and will reduce the signal amplitude available at the preamp input . A suitable
value for the coupling capacitor would be 0.01F.
Note 4.
The amplitude of the output pulse in the above example is dependent on the PMT anode load. A phototube can be made to deliver
output pulses of amplitude equal to the voltage difference between
the last dynode and the anode.
D.

The rreason
eason ffor
or the capacitors to gr
ound on Dynodes 9 and 10.
ground
In Figure 1, bypass capacitors are used between D9 and D10 to
ground. (For high rate applications, other dynodes may be bypassed
also, i.e., D7 and D8 in Figure 1.) In our application, a pulse of light
generated in the NaI(Tl) crystal strikes the photocathode and causes
a pulse of photoelectrons to be emitted. This electron pulse is
amplified as it travels through the dynode structure and will reach a
very high peak value at the last few dynodes. The peak pulse current
will be much greater than the voltage divider current and must,
therefore, be provided by a secondary current source. The bypass
capacitors are the secondary current source. If the capacitors were
eliminated from the circuit, the last few dynodes would have to
draw from the voltage divider current causing a change in the
interdynode voltages which cause the tube gain to shift. In the
majority of divider circuits used, the divider current is much lower
than the peak pulse current so that, without the capacitors, the
tube linearity would also be affected.
In our standard voltage divider, we use .01 microfarad bypass
capacitors. During the peak current pulse, these capacitors discharge through the dynode structure and anode load resistor with a
time constant of one millisecond. This time constant is very large
compared to the peak pulse duration, therefore, the voltage on the
bypassed dynodes does not change. Good wiring practice requires
that these capacitors be grounded using the shortest possible wire
to the ground side of the signal connector.

E.

An a
ver
ag
e phototube curr
en
oltag
e divider curr
en
av
erag
age
curren
entt and the vvoltag
oltage
curren
entt .
Using the component values shown in Figure 3, a count rate of
10,000 per second for 137Cs (assuming that 10,000 gamma photons
are totally absorbed in NaI(Tl) per second), and an assumed phototube gain of 5 x 105, we can calculate the divider current and an average phototube anode current.
HV
12R + RG
1200
=
= 5 x 10-4 A
2400000
= 500 mA

Calculation of the average phototube anode current:

The average phototube anode current is the product of the anode charge per pulse as calculated in section C and the pulse
rate (which is the same as count rate).
Average Phototube current is:

As shown above, Q = N x E x G x Ce
where
N = Number of photoelectrons/keV: 8
E = Energy of Gamma Photon: 662 keV for 137Cs
G = Photomultiplier Gain: 5 x 105 typical
Ce = Charge per electron: 1.6 x 10-19 coulomb; and
R = Count rate: 10000 c/s
therefore,
IPMT = 4.24 x 10-10 x 10,000
= 4.24 x 10-6A = 4.24 A
For good phototube gain stability and linearity, the divider current should be 50 to 100 times the average phototube anode
current. In this example,
IDiv
500A
=
= 118
IPMT
4.24A
which is a good ratio for PMT gain and linearity operation.

Important!! Negative High Voltage

The basic voltage divider circuit of Figure 1 can be wired for negative
high voltage operation by making the connections shown in Figure 5.
The coupling capacitor has been removed and an electrostatic shield
in close contact with the glass bulb of the phototube has been
added. Connections to points X and Y are the reverse of the positive
high voltage system. The 10M resistor is simply a current limiter that
is operative in the event of accidental grounding of the shield which
is at cathode potential.
All of the calculations and information given previously apply
equally well to a negative high voltage divider.
For fast timing applications, the RL may be changed to
50 ohms. Some users make RL=0. This can present a hazard to a
preamplifier if it is connected when the PMT is operating and a large
potential has built up on the anode. Instead of RL=0, a large value
bleeder resistor is recommended. Gauging applications do not use
the 50 ohm resistor, but should use a large value bleeder resistor also.
X

Shield

10M

Divider Current = IDIV =

The divider current increases to 1.0 mA as the wiper of RG is moved

to point Y.

IPMT = Q*R

Neg H.V.

BNC

Figure 5. Connections to points X, Y and S of

Figure 1 for operation with negative HV.

Scintillation Products
Bases, Voltage Dividers, Preamplifiers

USA
Saint-Gobain Crystals
17900 Great Lakes Parkway
Hiram, OH 44234
Tel: (440) 834-5600
Fax: (440) 834-8767
Europe
Saint-Gobain Crystals
104 Route de Larchant
BP 521
77794 Nemours Cedex, France
Tel: 33 (1) 64 45 10 10
Fax: 33 (1) 64 45 10 01
P.O. Box 3093
3760 DB Soest
The Netherlands
Tel: 31 35 60 29 700
Fax: 31 35 60 29 214
Japan
Saint-Gobain KK, Crystals Division
3-7, Kojimachi, Chiyoda-ku,
Tokyo 102-0083 Japan
Tel: 81 (0) 3 3263 0559
Fax: 81 (0) 3 5212 2196
China
Saint-Gobain (China) Investment Co.,
Ltd.
15-01 CITIC Building
19 Jianguomenwai Ave.
Beijing 100004 China
Tel: 86 (0) 10 6513 0311
Fax: 86 (0) 10 6512 9843
www.detectors.saint-gobain.com

Manufacturer reserves the right to alter specifications.

2003-8 Saint-Gobain Ceramics & Plastics, Inc. All rights reserved.

(12-08)