Transcript ppt

The VPT HV System
 Overview of the architecture
 The HV filter card
 Radiation tolerance
 Power supply specification
 Fault analysis
 Summary
EE EDR Workshop CERN March 2002
R M Brown - RAL
1
VPT Gain vs bias voltage
DG<5% for DV=10%  Single operating voltage for all VPTs
(apart from handling fault conditions)
12
10
Gain
8
6
V(A)=1000V
V(A)=800V
4
2
0
0
200
400
600
800
1000
Dynode Voltage
EE EDR Workshop CERN March 2002
R M Brown - RAL
2
Overview of EE HV System
EE EDR Workshop CERN March 2002
R M Brown - RAL
3
HV Distribution inside a Dee
One quadrant is shown
EE EDR Workshop CERN March 2002
R M Brown - RAL
4
HV Distribution inside a SC
Three of five filter cards are shown
EE EDR Workshop CERN March 2002
R M Brown - RAL
5
Filter network in SC
Signal to FPPA
Anode HV
VPT
Dynode HV
EE EDR Workshop CERN March 2002
R M Brown - RAL
6
5-Channel filter card
Anode filter side
Dynode filter side
EE EDR Workshop CERN March 2002
R M Brown - RAL
7
HV filter cards in SC
EE EDR Workshop CERN March 2002
R M Brown - RAL
8
Radiation tolerance
HV filter card components have been irradiated under bias to:
> 200kGy (60Co)
(~ 4x 10 year dose at  = 3)
~ 1015 n/cm2
(~ 10 year neutron fluence)
No failures occurred
(A capacitor showed DC ~ 17% - to be followed-up)
Irradiation of HV cables is just starting
Further studies are planned of filter operation under irradiation
EE EDR Workshop CERN March 2002
R M Brown - RAL
9
Fault analysis - typical examples
Fault
Consequence
Remedy
Power supply
Lose 1 Quadrant
Change power supply
(In control room)
Cable
Lose 1 Supercrystal
Use spare cable
(Requires access Inside or outside of Dee?)
VPT
(A or D short to Earth)
Lose 1 Crystal
(Voltage drops by 50V
on other 24 Xtal in S/C)
Change calibration for
24 channels.
(Failed channel is lost)
Filter component
(open/short circuit)
Affects 1 Crystal
(Noisy or lost)
None
HV discharge
(capacitor/cable/VPT)
Noise induced triggers
Reduce HV on one S/C
EE EDR Workshop CERN March 2002
R M Brown - RAL
10
Fault analysis (2)
‘Extreme case’:
1% of all VPTs develop a short circuit
from Anode/Dynode to earth.
 single S/Cs: 19.5% with 1 bad VPT  I= 50mA, DV= -50V
2.4% with 2 bad VPT  I=100mA, DV= -100V
0.2% with 3 bad VPT  I=150mA, DV= -150V
If Anode and Dynode both short to earth this is not a problem
(the loss in gain on the other VPTs is < 5% for DV= -100V)
However, if the Dynode shorts, but not the Anode, then VA-VD
increases on the other VPTs  potentially damaging
Therefore, need possibility to change VA or VD on one S/C.
(Should be incorporated in HV fan-out in Control Room)
EE EDR Workshop CERN March 2002
R M Brown - RAL
11
Power supply specifications
Output voltage: +1200V maximum
remotely set with DV = 10V (or less)
Output current: > 5mA (source) (to handle fault conditions)
> 5mA (sink)
Voltage ripple: 50mV maximum
Ramp rate:
~10V/s (2 mins total) (Rise and Fall)
Set in hardware (S/W control in addition?)
Format:
Rack-mounted (CAMAC or NIM/RS232)
Integrity:
Must be powered via UPS system
EE EDR Workshop CERN March 2002
R M Brown - RAL
12
Ramp-up/down
Ramp-down
OPAL experience: Graceful ramp-up/down essential
(Several VPTs lost before going to UPS and 2 min ramp-down)
RIE experience: Many VPTs draw large currents at 0.2 < B < 0.8T
 Need to ramp-down HV following a fast quench (t ~ 280s)
Ramp-up
at 10V/s, current to charge cables is ~ 250mA (not a problem)
a VPT may draw ~100nA for a short time (<< 200mA/quadrant)
VA-VD
Must not exceed 300V  Anode and Dynode Power Supplies
must be interlocked
EE EDR Workshop CERN March 2002
R M Brown - RAL
13
Summary
 The HV system for the VPTs presents no special problems
 Power supplies meeting the specification are available offthe-shelf (eg Iseg NHQ202)
 System costs are dominated by distribution and on-detector
filters
 A systematic fault analysis has been made
 Initial irradiation studies of components with g and n
are encouraging
EE EDR Workshop CERN March 2002
R M Brown - RAL
14