Transcript Nuovi Partitori per Fototubi NA62 - Indico

New dividers for LAV
photomultipliers
G.Corradi,D.Tagnani,C.Paglia, M. Raggi
LAV working group
CERN 3/2/2010
1
Motivations
o The modification of the divider improves the performance of ToT
electronics removing some spurious effect on PMT signal:
• Ringing of the signal (due to last 3 dynodes inductance)
• Shoulders of the signal (due to anode disperse inductance)
o Clamp of the first dynode rise the potential to 240-250 V
• Improves the first dynode multiplication factor and therefore total
gain
• Reduce the probability of multiplication starting from 2nd dynode
(lower spread)
o Include the 10 KW HV decoupling resistor
o Better precision in divider components:
• Resistors precision 1% (wrt 10% in old divider)
• Resistors stability 50 ppm (wrt 100 ppm in old divider)
• Improves gain stability wrt to the one measured during construction
2
Layout
First dynode clamp
Quenching of d10 d11 d12 and anode
Quenching capacitors
HV decoupling resistor (10 KW)
3
First dynode clamp
o The voltage drop on the first dynode is generated by 2
diodes 1N4150 in series selected to provide a reverse voltage
in the range 120-130 V each
o This solution allows us to produce a low noise Zener voltage
using low doping in the junction
o The differential resistance is on the other hand increased
4
New PCB production status
5 foils 20 dividers each already produced
Component mounting is ongoing
5
Printed Circuit Board
New “Star” like shaping allows faster positioning
of the divider on the dynodes pins
6
PCB components mounting
o Improved thermal stability 50ppm (wrt 100 ppm in old divider)
• Improved gain stability wrt the one measured during
construction
o Resistors precision 1% (wrt 10% in old divider)
7
Air thermal test
V = -1500 V
Tamb= 27 °C
ΔT = 17 °C
Imon = 420μA
was (Imon=460 μA)
15% less power
consumption wrt
old dividers
8
Vacuum test setup
The operating P:
Operating voltage:
Test Duration:
Vacuum window:
IR Termocamera:
4·10-4 - 8·10-5 mbar
1500 V
30 h
ZnSe
Flir E320
The ZnSe window is not fully transparent in the IR spectrum
The calibration of the window has been made looking at the
same
object
with
and
without
the
window
The difference in temperature has been estimated to be 6 °C
The calibration has been shown to be stable at different
temperatures
9
Vacuum thermal test
After 1.5 h @ 1500 V
Tcorr =46.6+6 = 52 C
After 2.5 h @ 1500 V
Tcorr = 46.9+6 = 52.9 C
After 30 h @ 1500 V
Tcorr=47.5+6 = 53.5 C
The value obtained for old dividers was 60 °C.
The difference observed is in agreement with the
15% reduction in the divider power consumption
10
Divider discharge test
V = -1000V @ T = 27°C
V= -1500V @ T = 27°C
Persistency set to 5 minutes
Persistency set to 5 minutes
11
Cosmic ray signals
V = -1400 V
10 mV vertical scale 
V = -1400 V

50 mV vertical scale
12
HV divider quality test
Vpartitore
500
V(V)
475
450
425
400
375
350
325
300
275
250
Elet t ro do
225
0
2
4
6
8
10
Tested on 5 different dividers;
Maximun non linearity = 1%
Vg = -500V
12
14
13
Conclusions
Mounting of the divider is very fast (5 to 10 minutes/PMT)
o First HV test 2500V for 1 min did’nt shown any discharge
o Power consumption reduced by 15%, operating
temperature in vacuum reduced (from 60°C to 54°C)
o Resistors stability 50ppm (was 100ppm)
o Resistors precision 1% (was 10% )
o Divider non linearity below 1%
o Spread of dividers characteristics is around 2%,
negligible wrt intrinsic gain spread of PMTs
o
14