optocrowbar

1
Gas detector Crowbar circuit Present status
and planning
Contents Principle of crowbar Some major sub
systems and calibration Crowbar scheme circuit
and design Testing performance and Observations
Justification of location and scheme Justificatio
n of Appropriate Protection Scheme Implementation
plane
M. R. Dutta Majumdar 6th June, 2008
2
Principle of crowbar
Principle whenever spark is sensed, quickly
stored charge of the ALICE module may
dumped outside using a crowbar system. This may
help to reduce over energy dumping to FEE boards
and save them due to slow discharge or Electrical
Over Stress (EOS).

• Causes of damage and remedy
• ESD and fast discharge case partially series
  resistance improves it and ESD protection upto
  HBM (Human body model) and partly MM (Machine
  model) also.
• EOS (electrical over stress) where slow discharge
  degrades FEE
• in the long run and may result occasional damage.
• EMI (electromagnetic interference) results minor
  damage possibility during sparks lots of EMI
  produced and which may go through (as spikes) LV
  supply rail. This can damage any of surrounding
  channel or chip. Probability of this type of
  damage is less compare to other two above
  mentioned damages.
• So remedy should take care of Both ESD and EOS.

3
Some major sub systems and components
1- HV current sensor using opto-coupler LED and
Photo transistor encapsulated inside light tight
encolsure using otical glue. Bypass diode is used
to protect LED from reverse voltage surge. 2-
opto-electrical transformer current source High
efficient photo diodes in series combination
coupled with LED (parallel) in light tight
enclosure. 3- HV transistor switching in
series Each stage HV transistor collector emitter
is connected with high resistance (200 meg) to
equalize voltage distribution during off
condition.
HV current sensor using opto-coupler
-ve HV
LV
Calibration 1volt/150mA
LED
Bypass diode
PT
100K
Limitation- low current (below 5ma) not
available in market
current
Single stage of HV transistor switching using
opto-electrical transformer current source to
achieve isolation
Calibration at input 1 mA,output current nearly
20 mA
-ve HV
1K
1M
1K
200M
1 ms
perspex
HV Tr
Assembly of photo diode and LED
Prev ref http//arxive.org/abs/physics/0512227
4
Calibration of opto-couplers
OPC
OPCF
5
Schematic circuit of crowbar for Gas Detector FEE
protection
HV cable length Very small or 3.5m
HV filter
HVPS
detector module
Shorting connectors
D.U.T
comp
10M
ref
1ms
Mono shot
Modification Previous logic circuit modified to
avoid multiple trigger during short interval,
which was causing over discharge and tripping HV
occasionally.
Mono shot
500ms
Previos ref http//www.veccal.ernet.in/pmd/ALICE
/extprotection.ppt
6
Present Fabricate Circuit size
Device Under Test configurations
B
A
Crowbar action with AB used For data collection
RS1K
10M
10M
5 High voltage Transistors Length 175mm, W
125mm and H 25mm
C
Feasibility of C also checked with limitations
2 High voltage Transistors Length 125mm, W
80mm and H 25mm After getting few 800 volt HV
transistors
RS
3V
5 High Voltage Transistor configuration working
nicely without damage of HV transistor,
availability also good and voltage sustainability
also good (2000V with 400X5). So this is
preferred.
10M
-3V
7
Configuration of Crowbar System
A
B
ALICE PMD Module
ALICE PMD Module
HV Filter
HV Filter
OPC
OPC
HVPS
HVPS
CB
CB
AB were initially thought of but later on
found delay introduced due to HV filter. So less
effective.
C
C was tried out where signal of spark sensing
time reduced using optical fibre based
optocoupler. But needs additional Optical fibre
cable. So other right options also explored.
ALICE PMD Module
HV cable3.5m
HV Filter
OPCF
HVPS
CB
D
E
ALICE PMD Module
ALICE PMD Module
HV cable3.5m
HV Filter
OPC
HVPS
HV Filter
OPC
HVPS
CB
D away from module by 3.5m HV Cable and works.
Data shown.
DE options are for implementation
E close to module and works much better. Data
shown.
CB
8
Location of Crowbar in PMD stand
Close to detector module Present HV filter
Box with extended length, if possible supply with
- 2.5 V LV or 3 V floating, needs SMD
components and optimized space with HV transistor
integration, HV filter opto-coupler. This
configuration is much faster as delay of filter
capacitor combination avoided. Opto-coupler is
between HV filter and detector HV wire. This
circuit will sit on top of each module. Away
from detector module 1 to 5 meter from module,
block of 24 circuit in a fox with separate LV
supply. Delay may be introduce due introduction
of inductance of 1 to 5 meter length of HV
cable. Present way HV filter is near to detector
module, but in principle HV filter can be placed
inside crowbar PCB.
9
Crowbar circuit PCB layout when circuit is on top
of module (provisional)
Proposed Box L190 mm H25 mm D12 mm Only
length will Increase of HV Box.
HV connector
HV filter optocoupler
Optoelectrical transformers HV transistors
LV logic circuit
10
Testing performance and Observations
Crowbar circuit performance testing Earlier
reported circuit logic single retrigger mono-shot
was used. So possibility of multiple triggering
of crowbar was there. Now using dual single shot
and AND gate possibility is completely removed.
Once crowbar (duration 1ms) triggers that will
not retrigger with in 500 ms time. This avoids
false triggering and over discharge of detector
capacitor charge which cause unnecessary tripping
of HVPS. Present circuit working without any
failour on LV circuit or HV part and major tests
of guinea pig are done after new circuit.

• Protection testing Guinea pig tests
• Guinea pig test seems good for under standing
  basic damage problem and help FEE board testing
  with crowbar later on.
• ALICE module with shorting connector, DIP switch
  connected with D.U.T, this is a small signal low
  voltage diode (germanium) used in radio receivers
  or silicon diode.
• High Voltage applied upto 1500V under gas
  flushing in PMD lab VECC using CAEN N 470
• The circuit does not trip with proper setting of
  HV current setting, trip time, crowbar trigger
  duration and drive current. Typical values are
  15mA, 5 sec, 1 ms and 4 mA

11
Guinea pig (germanium diode) test 01 Crowbar
location in configuration A
A
Crowbar Passive
Ratio of reverse resistance of D.U.T before and
after use
10M
More damage
Crowbar in Action
26-4-08 27-4-08
Sample number of D.U.T tested in configuration A
12
Guinea pig (germanium diode) test 02 Crowbar
location in configuration A
Crowbar Resistance off (4)
B
RS1K
Only Crowbar on (3)
Ratio of reverse resistance of D.U.T before and
after use
Only Resistance on (2)
More damage
Crowbar Resistance on (1)
10M
Sample number of D.U.T tested in configuration B
13
Guinea pig test 03
Crowbar with cable length of 3.5 m using silicon
signal diode as guinea pig
Crowbar is close to detector module using silicon
signal diode as guinea pig
01s
04s
Procedure of testing a) Sequence are HV on i)
Crowbar on Res on HV off measured ratio of
resistance with previous value HV on ii) Crowbar
off Res on HV off measured ratio of
resistance with just previous value. b) Repeated
the sequence of operation 5 times for each
sample. c) Resistance measured at Reverse mode
of diode at 90 Volt, 10 below rating in
test Jig.
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
02s
05
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
03s
06
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
Number of operation done on each sample diodes
Number of operation done on each sample diodes
14
Guinea pig test 04
Crowbar is close to detector module using
germanium signal diode as guinea pig
Crowbar with cable length of 3.5 m using
germanium signal diode as guinea pig
04g
01g
Procedure of testing a) Sequence are HV on i)
Crowbar on Res on HV off measured ratio of
resistance with previous value HV on ii) Crowbar
off Res on HV off measured ratio of
resistance with just previous value. b) Repeated
the sequence of operation 5 times for each
sample. c) Resistance measured at Reverse mode
of diode with DVM.
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
05g
02g
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
03g
06g
Ratio of reverse resistance of D.U.T before and
after use
Ratio of reverse resistance of D.U.T before and
after use
Number of operation done on each sample diodes
Number of operation done on each sample diodes
15
Summery of results observed

• Initial tests done 01 and 02 are now carrying
  less importance due to improper location. Which
  is known after some studies performed. But have
  historical importance for further progress of the
  work.
• Test plots of 03 and 04 shows comparison of
  performance of crowbar away
• from module (3.5 m) and close to module using
  silicon diode, germanium diode.
• Plots shows damage is less when crowbar and
  series resistance are both implemented.
• Further close to module configuration seems more
  effective.
• Some inconsistency are due to quick rise and fall
  of HV, which normally may
• be absent with FEE board testing with seasoned
  and slow raising of HV.

16
Justification of Appropriate Protection Scheme
j

• Resistance and crowbar, avoiding external diode.
• Series resistance decreases substantially ESD and
  fast discharge
• (ns to micro sec) as seen in FEE testing (ref
  http//www.veccal.ernet.in/pmd/ALICE/HV-problems-
  in-PMD.ppt )
• which supported by guinea pig test (crowbar
  operation) also.
• Slow discharge remedy taken care by crowbar.
• Additional external diode protection need not
  require as ESD and fast discharge
• damage protection is taken care by series
  resistance and internal diode of Manas chip.
• Putting external diode may not improve the
  situation very significantly and may increase
  noise level also.
• Implementation in kapton cable is difficult to
  accommodate and handle, powering of - 2.5 V to
  diodes makes wiring cumbersome and may be another
  source of introducing EMI to LV line.
• Lastly this is not cost effective as 48 modules
  for diode scheme needs nearly 80,000 CHF where as
  for crowbar scheme total cost may be 3,500 CHF
  for 48 modules.

17
Implementation plane
Design fabrication- complete, needs more
miniaturize and low power version, procurement of
components can start.
Over all testing- guinea pig test to real test
with FEE board test and finally testing at CERN
with HVPS SY1527
Real Location and final configuration i) on
module with HV filter or ii) few meters away from
stand
Whether needs integration with DCS signal of
sparks of each module can be Used for data
cleaning and status of module during run time.
Necessary logic out put provided for
implementation.
Review documentation for ALICE need to convince
collaboration before Final integration
Time scale for implementation Online discussion