Transcript Control Module production

Status of
Control Module integration
Summary:
• LCM integration & PRR
(in particular: integration rehearsal of last week)
• Preparation for xCM production
Control Module production
1/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Control Module production
It consists of the following processes:
(transportation)
Reception
Storage
Component
level
Integration
Module
level
Test
Packing (+storage)
Control Module production
2/39
(transportation)
M. Circella, ANTARES-SC meeting, CPPM April 2004
Control Module production: scenario
All activities under definition based on the following considerations:
• Time constraints (schedule request: one working LCM produced in two
days)
• Manpower needs (one person?)
• Safety requirements: power & grounding during functional tests, ESD
protection, fibre handling
• Quality Control approach: detailed procedure for each operation,
development of tools
• Traceability & test needs
Target: Full system defined and working in time for the construction of the MILOM
modules (according to the current schedule: by end of June 2004)
LCM integration test for LCM PRR: necessary step to check the overall design of the
module from the integration point of view and in particular to spot out all possible
(minor) modifications to the module components before the mass productions start
Control Module production
3/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration tests for PRR
• First test performed in week 12 (March): report available at
http://www.ba.infn.it/~circella/LCM/LCM_int_first_report.pdf
• Complete rehearsal of the integration performed in week 16 (last week):
procedure and report available at
http://www.ba.infn.it/~circella/LCM/LCM_integration_test_prr_10.pdf
http://www.ba.infn.it/~circella/LCM/LCM_int_second_report.pdf
Conclusion after both tests:
***NOT READY***
(but not so far away from readiness…)
Control Module production
4/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM PRR integration test report
Difficulties concerned:
1. Missing or wrong specifications on screw sizes, screwing torque, thread locker
decisions made on the site
2.
Not all components available: the nylon ring and the 1 M resistor for connection to
the flange, a few screws, the plastic tubes and their brackets were missing
spare parts (most from the PSL) recycled or adapted
3. Not all boards in final configuration:
- ‘glue’ problem on the backplane
- LCM_CLOCK without fibre tray
- cooling plates of the crate out of specifications?
not possible to test the final LCM configuration
4. Integration supports not available
not possible to test the final procedure
Most serious problem: electrical contact between crate and titanium (flange + tube)
Control Module production
5/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Module integration (summary)
0. (Get all components and tools at hand)
1. Assemble a complete module before making any tests
2. Power and optical margin test: check that the voltage and
current values are acceptable when powering up the module
with a test board (SUMO) inserted between the power box
and the crate and that the optical transmission work with
proper attenuators inserted along on the optical fibres. Then
remove the SUMO and mount the power box on the module
3. Functional test
Control Module production
6/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
(0. Get all proper components and tools at hand)
Result step 0: not all components available (+ a few specification errors)
crate
1. assemble the crate (backplane included), and fix the gas
venting tube to it
Option: fix the fibre spiral tube as well!
Comments:
• no supports needed for this step
• it is not convenient (or possible?) to mount the spiral tube at this step
Control Module production
7/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Result step 1:
test
(details)
• backplane: excessLCM
of glueintegration
on coaxial cables,
coaxial
cables not labelled
• ground cables from backplane: what is the ideal position?
• insertion of the coaxial cables in the trays of the backplane holders difficult
• brackets for the plastic tubes can not be mounted without modifying the crate
possible modifications to the crate and backplane
Control Module production
8/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
2. Fix the crate to the titanium flange
crate
flange
Comments:
• preferred orientation is horizontal (supports needed)
• no (more) alignment holes in the flange and couronne 1
Control Module production
9/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Result step 2:
• difficult alignment of flange and crate (rehearsal made with crate vertical
due to missing supports)
• difficult insertion of the kapton ring
• missing 1 M resistor
• wrong screws used for the resistor (due to missing specifications)
• difficult to get crate and flange isolated (due to flange’s washers? plastic
inserts of the screws?)
• improve documentation
• develop integration supports
• fix the nylon rings to the crate?
• check again all screws with
torque specifications
• modifications to flange or crate?
Control Module production
10/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
(old version support)
3. …then the (flange+crate) block is fixed on its support…
Comment: support not available during this test
Control Module production
11/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
Power box
SUMO test board
4. …the crate is connected to the power box through a
power supply test board (SUMO: Supply Monitor)
Comment: SUMO board not used during the test
5. ...then we insert the boards one at a time
Control Module production
12/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Result step 5:
• fibre spiral tube not initially fixed to crate. It needs to be fixed to the table
and elongated up to 70-80 cm
• wrong screws used for DAQ and CLOCK cooling bases (due to missing
specifications)
• improve documentation
• develop integration supports
Control Module production
13/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
Power box
SUMO test board
6. …when all boards are inserted we fix the fibre spiral
tube to the crate, then we make all connections to the
test bench (power, fibres, flange connectors) and make
the power and optical power margin tests
Option: we mount at least one of the cooling plates before the
power+optical power margin and/or the functional tests
Comment: step 6 not made during the rehearsal
Control Module production
14/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
7. …remove the SUMO board and fix the power box
to the crate
8. functional tests!!
Comment: steps 7-8 not made during the rehearsal
Control Module production
15/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM integration procedure (rehearsal of)
9. …(assuming the test is ok) prepare for
shipping
(old version LCM)
Control Module production
16/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
What to do at the line integration site
10. If necessary, remove the protection from the bottom
flange
11. Install/replace the o-rings of the bottom flange
12. Install the titanium cylinder and mount the container on
the line
Comment: only step 12 relevant
for this rehearsal
Result step 12: contact between
crate and titanium container!
Solved by putting kapton tape on
the short edges of the cooling
plates
Control Module production
17/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
PBS and bar-code
General formula for traceability:
PBSNumber/VersionNumber.SerialNumber
(rule: one individual PBS number used to identify only
components which are always completely interchangeable)
Is this THE rule? Are there exceptions?
Notes added after the meeting:
• the rule above is no longer endorsed by the Steering Committee
(therefore the discussion of page 22 needs to be updated)
• the traceability format remains valid, with the SerialNumber to be a
unique number for ONE object of a given PBS (i.e., the SerialNumber
increases independently of the VersionNumber)
Control Module production
18/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM configurations
There are 3 kinds of LCM’s (leaving aside the MILOM and IL needs):
(a) LCM connected to 3 OM’s
(b) LCM connected to 3 OM’s and a LED beacon
(c) LCM connected to 3 OM’s and an acoustic receiver
(d) LCM connected to sound velocimeter/CTD
Proposal - let us define a ‘version number’ to discriminate
among them:
1 for LCM (a)
Note: there is indeed one PBS number
2 for LCM (b)
for all LCM’s and MLCM’s
3 for LCM (c)
4 for LCM (d)
(5, 6, ... for future needs – 0 for the ‘no-version’ option)
Control Module production
19/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
MLCM configurations
There are 2 kinds of MLCM’s (leaving aside the MILOM and IL needs):
(a) MLCM connected to 3 OM’s
(b) MLCM connected to 3 OM’s and a LED beacon
Proposal - let us define a ‘version number’ to discriminate
among them:
1 for MLCM (a)
2 for MLCM (b)
(3, 4, 5, ... for future needs – 0 for the ‘no-version’ option)
Note: there is indeed one PBS number for all LCM’s
and MLCM’s (so the LCM and MLCM versions
should be enumerated together)
Control Module production
20/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
SCM configurations
There are 2 kinds of SCM’s (leaving aside the MILOM and IL needs):
(a) SCM connected to acoustic TX/RX, pressure sensor and sound
velocimeter
(b) SCM connected to the instruments in (a) + a laser beacon
Proposal - let us define a ‘version number’ to discriminate
among them:
1 for SCM (a)
2 for SCM (b)
(3, 4, 5, ... for future needs – 0 for the ‘no-version’ option)
Control Module production
21/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
PBS updates
Suggested changes:
See notes on pages 18, 19, 20
a) crates: the current PBS has only 2.1.01 LCM_CRATE => we would need two PBS numbers for
LCM_CRATE and MLCM_CRATE
b) backplanes: the current PBS has only 2.1.02 LCM_BACK and 2.2.02 SCM_BACK => we would need
three numbers for LCM_BACK-1 (for LCM-V1 and LCM-V3), LCM_BACK-2 (for LCM-V2), and
LCM_BACK-3 (for LCM-V4); 2 numbers for MLCM_BACK-1 (for MLCM-V1) and MLCM_BACK-2
(for MLCM-V2); and 2 for SCM_BACK-1 (for SCM-V1) and SCM_BACK-2 (for SCM-V2). Then, we
will add more numbers if/when needed
c) LCM flange: the current PBS has only 1.3.2 LCM_CONTAINER => we would need three different
numbers for the different flanges (for the three LCM configurations). Should we also introduce
different PBS numbers for the flanges and the other parts of the containers, as for the SCM case?
d) LCM_DAQ/SC: the current PBS has only 2.1.17 LCM_DAQ/SC => we would need three different
numbers for MLCM_DAQ/SC, LCM_DAQ/SC and SCM_DAQ/SC
e) MLCM_DWDM: the current PBS has only 2.1.16 MLCM_DWDM => we would need EITHER 5
different PBS numbers OR the version number used to specify the different wavelengths (or are there
other options?)
f) SCM_REP: the current PBS has only 2.2.04 SC_REP => we would need two different numbers for
SCM_REP1 (serving 3 sectors) and SCM_REP2 (serving 2 sectors)
g) LPB: the current PBS has only 2.2.11 POWER_BOX. It is indeed the same object used inside the
LCM/MLCM/SCM, but in case of the xLCM it will be powered at +380 V, in the SCM it needs to be
powered at +48V: the difference is made by the connector => shall we give a PBS number to the
connector and its cable?
[h) should we also worry about the different SCM flanges?]
Further suggestion: double-check all labels and avoid using numbers in labels when not strictly needed
(so: LCM_BIDICON => MLCM_BIDICON ; SCM_WDM1 => SCM_WDM ;
SCM_BACK1 => SCM_BACK-1/SCM_BACK-2 ;)
Control Module production
22/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
• LCM-V1 (configuration with just 3 OM’s) is composed of:
- 1 LCM_CRATE PBS-2.1.1 (*)
- 1 LCM_BACK-1 PBS-2.1.2 (*)
- 1 POWER_BOX PBS-2.1.11 (*)
- 1 titanium flange PBS-1.3.2 (*) [the rest of the container will be installed at the
level of line integration]
- 1 COMPASS_MB PBS-2.1.3
- 3 ARS_MB PBS-2.1.4
- 1 LCM_DAQ/SC PBS-2.1.17 (*)
- 1 LCM_CLOCK PBS-2.1.7
Then,
• LCM-V3 (configuration with 3 OM’s and an acoustic receiver) comprises in addition:
- 1 ACOUST_RX_PREAMP PBS-2.1.8
- 1 ACOUST_RX_DSP PBS-2.1.9
- 1 ACOUST_RX_CPU PBS-2.1.10
• LCM-V2 (configuration with 3 OM’s and a LED beacon) comprises in addition:
- a fourth ARS_MB PBS-2.1.4 (but we have to replace LCM_BACK-1 with LCM_BACK-2)
• for LCM-V4 LCM_BACK-1 should be replaced with LCM_BACK-3)
[Beware: the flange is different for each configuration]
Control Module production
23/39
Note: PBS-related issues exist for items marked (*)
LCM main components
M. Circella, ANTARES-SC meeting, CPPM April 2004
MLCM main components
• MLCM-V1 (configuration with just 3 OM’s) is composed of:
- 1 MLCM_CRATE PBS-2.1.1 (*)
- 1 MLCM_BACK-1 PBS-2.1.2 (*)
- 1 POWER_BOX PBS-2.1.11 (*)
- 1 titanium flange PBS-1.3.2 (*) [the rest of the container will be installed at the
level of line integration]
- 1 COMPASS_MB PBS-2.1.3
- 3 ARS_MB PBS-2.1.4
- 1 MLCM_DAQ/SC PBS-2.1.17 (*)
- 1 LCM_CLOCK PBS-2.1.7
- 1 MLCM_DWDM PBS-2.1.16 (*)
- 1 MLCM_SWITCH PBS-2.1.18
- 1 (M)LCM_BIDICON PBS-2.1.05
Then,
• MLCM-V2 (configuration with 3 OM’s and a LED beacon) comprises in addition:
- a fourth ARS_MB PBS-2.1.4 (but we have to replace MLCM_BACK-1 with MLCM_BACK-2)
[Beware: the flange is different for the two configurations]
Note: PBS-related issues exist for items marked (*)
Control Module production
24/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
• SCM-V1 (configuration without a laser beacon) is composed of:
- 1 SCM_CRATE PBS-2.2.01
- 1 SCM_BACK-1 PBS-2.2.02 (*)
- 1 POWER_BOX PBS-2.1.11 (*)
- 1 titanium flange PBS-1.2.6.2 (*)
- 1 COMPASS_MB PBS-2.1.3
- 1 SCM_DAQ/SC PBS-2.1.17 (*)
- 1 SCM_CLOCK PBS-2.2.05
- 1 SCM_WDM PBS-2.2.03 (*)
- 1 SCM_REP1 + 1 SCM_REP2 PBS-2.2.04 (*)
- 1 SCM_DWDM PBS-2.2.16
- 1 ACOUST_RXTX_EM PBS-2.2.07
- 1 ACOUST_RXTX_PREAMP PBS-2.2.08
- 1 ACOUST_RXTX_DSP2 PBS-2.2.09
- 1 ACOUST_RXTX_DSP1 PBS-2.2.10
- 1 ACOUST_RXTX_CPU PBS-2.2.11
- 1 ACOUST_RXTX_POW PBS-2.2.12
Note: PBS-related issues exist for items marked (*)
SCM main components
Then,
• SCM-V2 (configuration with a laser beacon) comprises in addition:
- 1 ARS_MB PBS-2.1.4 (but we have to replace SCM_BACK-1 with SCM_BACK-2)
[Beware: the flange is different for the two configurations]
Control Module production
25/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Control Module minor components (1)
For integration of each xCM the following components are also needed:
1. small parts for the crate assembly, namely:
1.1 the screws to fix couronne 1 to the titanium flange;
1.2 the plastic inserts for these screws;
1.3 the nylon rings to be put between couronne 1 and the flange for
LCM/MLCM;
1.4 the screws to fix the LPB to couronne 2;
1.5 the mini-screws to mount the backplane into its support bars;
1.6 the bolts to fix the ground connections from the backplane to the crate;
1.7 the screws to fix the backplane to couronne 2
1.8 the brackets to install the two plastic tubes (with screws+washers);
1.9 the screws for the cooling bases of LCM_DAQ/SC and LCM_CLOCK
(LCM/MLCM/SCM), MLCM_DWDM and MLCM_SWITCH
(MLCM), SCM_DWDM and SCM_WDM (SCM)
Control Module production
26/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Control Module minor components (2)
Then we need also:
2. the 1 MΩ resistor to be installed between the crate and the flange for
LCM/MLCM (with its screws)
3. the LPB connector terminated by the wires to be soldered to the
backplane (with its screws)
4. the HV cable for the LPB (with the screws for the connector)
5. the gas venting tube and the fibre spiral tube
6. the thread locker(s)
7. the thermal gel for the cooling plates
Control Module production
27/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Component reception: assumptions
Very strong recommendation:
Every single board should arrive to the xCM integration site
ready to be integrated in a Control Module
It implies that at the xCM integration sites there will be:
- no modifications to any board
- no wire or component soldering
- no check of board configuration
implications for component
delivery to the integration sites
traceability implications (PBS)
Control Module production
28/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Implications for component delivery to
the xCM integration sites
We expect that:
- the titanium flange arrives with all cables terminated with the connectors to plug to the backplane; it
should have no o-ring (or, at least, no ‘final’ o-ring) and possibly it should be provided with proper
protections for the connectors and the inner surface
- the backplanes arrive equipped with all components properly installed (as needed!), including the power
wires and the connector to plug to the LPB, the coaxials and any other cable
- the LPB arrives equipped with its HV connector and wires
- each board arrives exactly in the configuration it should be (i.e., as identified by its traceability label:
PBSNumber/VersionNumber.SerialNumber)
- for all components we expect to receive an accompanying test report and to be able to trace the product
history through the bar code label (we will use it as well to update the history of the integrated modules)
- ... (this list will be updated according to the integration test results)
Control Module production
29/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Component reception at the
xCM integration site
Therefore we need that:
- each component is accompanied by a proper test
report (= a PBS compliance certificate)
- traceability is possible through the bar code label
Actions upon reception:
- visually inspect the components
- check the accompanying documentation
- interact with the database to update the product history
Control Module production
30/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Component storage at the
xCM integration site
- ‘Standard’ laboratory conditions needed for storage and
integration
- ESD protection required for all sensitive components (i.e.,
electronics boards and LPB) at all times
- storage room needs under evaluation (most serious
requirement from mechanical pieces)
Control Module production
31/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Packing (and delivery)
• Transportation logistics provided by IN2P3
“Cellule Logistique”
• Flange transportation box will be used for module
delivery
Control Module production
32/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Work in progress...
• In progress (due for xCM PRR):
– PBS list
– component configuration (namely, define possible changes)
– integration procedure
– integration supports and tools (all or just what is essential?)
– documentation
Target: complete rehearsal has to be made in time for LCM PRR
Control Module production
33/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Work to be done...
• To be done (due for end June for LCM):
– definition of intermediate (power, optical power margin) tests
– transportation boxes for components and modules
– storage containers/conditions
– bar code traceability (requires reader and software)
– complete integration table
– define and implement test benches!
– define test scenario (including: failure modes, module repair line)
– documentation
Control Module production
34/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Critical items...
• What’s critical:
– Quality Control directives needed (in time for PRR?)
– Test bench status unsatisfactory (needed for production)
– the Pisa group can contribute more
– no ‘MLCM training’ currently possible
– possible conflict in Bari for finalization of the SCM and MLCM
test benches
Control Module production
35/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
Implications
• Should the LCM PRR wait to have the full Quality Control
system implemented?
• Is the integration of the MILOM LCM’s really needed in
July? (SCM and MLCM due in September)
• Can we enable the Pisa group to work on the module
integration and on the development of the test bench at Pisa?
(needed: onshore clock system, ethernet fibre connection, a
complete LCM)
Control Module production
36/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
SCM test bench development
(ongoing work at Bari)
• General definition essentially done
• Implementation delay due to unexpected problem with the ethernet link
• Implementation essentially stopped because:
- acoustics system control is not possible with the DAQ/SC software
- lack of detailed documentation on other tests
• Current status of the test bench:
- it manages SC and clock communications with the SCM
- it can drive the SCHarness through the state machine
- it can query and decode SC data
all in an integrated Labview software
Control Module production
37/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
LCM test bench development
(ongoing work at Catania and Pisa)
• General definition started
• Getting the LCM to acquire data was a tough job, due to poor
documentation
• Current status of the test bench:
- it manages DAQ/SC and clock communications with the LCM
- it can drive the DAQ/SCHarnesses through their state machines
- it can query and decode SC data
- it can drive data acquisitions with emulated signals from a pulser
all in an integrated Labview software
Control Module production
38/39
M. Circella, ANTARES-SC meeting, CPPM April 2004
MLCM test bench development
• Definition essentially linked to the LCM case (except for the MLCM-LCM
communications and ethernet switching)
• Implementation should inherit both from the SCM and LCM test benches
(not yet defined what to inherit from which test bench...)
• Hardware for the MLCM-LCM ethernet connections purchased but not yet
integrated
Control Module production
39/39
M. Circella, ANTARES-SC meeting, CPPM April 2004