TE-MPE-TM 27/06/2012, TE-MPE-MS section
SPS RENOVATION
Final design and status
R.Mompo, P.Dahlen, Y.Bastian, I.Romera, M.Zerlauth
1
SPS LAYOUT and Magnet Powering
• SPS (7 kms) composed of 6
sextants
TE-MPE-TM 27/06/2012, TE-MPE-MS section
• Main dipole and quad magnets
powered in series, power
converters located in respective
BA
• Auxiliary (corrector) magnets
powered individually (some
exceptions) from respective BA
• Ring-line magnets (quadrupole,
octupoles, sextupoles) powered
in series in whole SPS ring from
BB3
• Current interlock system built in
1974 (mains)  1980 (Aux)
TE-MPE-TM 02/08/2012, TE-MPE-MS section
SUMMARY
Description
Cost
Manpower
PRO
CON
Consolidation
Option #1
-
-
-
Risk of lengthy failure
Option #2
(Boolean PLC)
180kCHF
0.5 FTE (MS) +
0.1 FTE (ICE)
Simple + standard, no
cabling, diagnostic
Ring line remains, no remote test,
info only by sextants/ring
Option #3
(Boolean PLC +
cable)
300kCHF
0.75 FTE (MS)
+ 0.1 FTE (ICE)
Simple + standard,
diagnostic per demi-sextant
Additional cabling + patches, no
remote test
Option #4
(Boolean PLC + PLC
comms)
300kCHF
0.75 FTE (MS)
+ 0.1 FTE (ICE)
Simple + standard,
diagnostic per demi-sextant
Special (non-safety part in PLC, no
remote test
Option #5
(Analogue I/Os)
400kCHF
1.5 FTE (MS) +
0.4 FTE (ICE)
Diagnostic per magnet
without major
Completely new SW projects, no
remote test
Option #6
(Profibus)
>1.2MCHF
2 FTE (MS) +
0.1 FTE (ICE)
Full Monty
Considerable investment,
validation time for rad tolerant
I/Os, feasibility?
During last MPE-TM meeting, we concluded that they were 2 realistic options (#3 or #4) to
renovate the SPS interlock system !
Finally, we have chosen option #3 ! I will explain why in the following slides...
SPS Magnet families
Main Dipole and
Quadrupole Magnets
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Ring-Line Magnet
Auxiliary Magnet
4
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Principle of the SPS interlock system
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Principle of the SPS interlock system
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Principle of the SPS interlock system
Principle of the SPS interlock system
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Each magnet has an interlock box to
collect signals from thermo-switches
and for visual indication («trefle»)
SITOP
Modular power
supply
6EP1 334-3BA00
SITOP
Modular power
supply
SITOP Select
Diagnostics
module
6EP1 9612BA00
6EP1 334-3BA00
PLC
315F2DP
CP 343-1 IT
Module 8 DI / 8 DO
6ES7 323-1BH01-0AA0
Crate 2: Type 2
PLC
for SPS main circuits
Module 32 DO
6ES7 322-1BL00-0AA0
Ethernet
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
In BA1 * , BA2, BA3, BA4, BA5
and BA6:
Crate 1: Type 1
Power supplies
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
•
Configuration of WIC PLC for
the “Mains” and “Aux.
circuits.
Module 24 DI Safety
6ES7 326-1BK02-0AB0
•
Redundancy module
6EP1 961-3BA21
PLC architecture (1/2)
Profibus
From PLC power supplies
* In BA1, there is no “Aux.”
crate (circuits combined with
TT10 transfer line not
renovated during LS1).
Crate 3: Type 1
Power supplies
for deported unit
SITOP
Modular power
supply
6EP1 334-3BA00
To Power
converters
SITOP
Modular power
supply
To BIC
SITOP Select
Diagnostics
module
6EP1 9612BA00
6EP1 334-3BA00
6 “new” racks
From
Magnets
PCs
status
To Power
converters
Module 32 DO
6ES7 322-1BL00-0AA0
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
5 existing racks
=
cohabitation 
Module 24 DI Safety
6ES7 326-1BK02-0AB0
Crate 4: Type 3
Deported unit
for SPS aux. circuits
Module 24 DI Safety
6ES7 326-1BK02-0AB0
Impedance
adaptation crates
ET 200 M
TE-MPE-TM 27/06/2012, TE-MPE-MS section
•
Redundancy module
6EP1 961-3BA21
Magnets and
PCs status
Remote
test
PLC architecture (2/2)
Configuration of WIC PLC for the Ring-Line located in BB3:
Crate 1: Type 1
Power supplies
Redundancy module
6EP1 961-3BA21
•
SITOP
Modular power
supply
6EP1 334-3BA00
SITOP
Modular power
supply
SITOP Select
Diagnostics
module
6EP1 9612BA00
6EP1 334-3BA00
Module 8 DI / 8 DO
6ES7 323-1BH01-0AA0
Module 32 DO
6ES7 322-1BL00-0AA0
CP 343-1 IT
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
PLC
315F2DP
Module 8 DO (relay)
6ES7 322-5HF00-0AB0
Crate 2: Type 2
PLC
for SPS Ring Line circuits
Module 24 DI Safety
6ES7 326-1BK02-0AB0
Ethernet
Profibus
From PLC power supplies
Module 24 DI Safety
6ES7 326-1BK02-0AB0
Crate 3: Type 4
Deported unit
for SPS Ring line circuits
Module 24 DI Safety
6ES7 326-1BK02-0AB0
1 existing rack
=
cohabitation 
ET 200 M
TE-MPE-TM 27/06/2012, TE-MPE-MS section
PCs status
From
Magnets
To Power
converters
To BIC
Impedance
adaptation crates
Powering scheme of the main dipoles
• If a dipole magnet overheats (or
any powering failure):
 All 14 PCs must be switched off!
TE-MPE-TM 27/06/2012, TE-MPE-MS section
• If we put a WIC in each BA (to
protect 1 sextant), how do we
propagate the info to the PCs
located in the other BAs?
• Do we need to introduce a PLCPLC communication between
the WICs ??? (=> use of nonsafety PLCs ).
• Long distances: Issues on
reliability/ dependability/
reaction time?
• How do we send a «Beam
Dump» signal to the BICs?
EDMS: 681304
Powering scheme of the main quadrupoles
• In a way simpler, as all PCs are
located in 1 BA (BA3).
TE-MPE-TM 27/06/2012, TE-MPE-MS section
• But the WIC installed in BA3
must also receive the info from
all other sextants?
EDMS: 681372
Powering scheme of the “Ring-Line”
• The “Ring-Line” is composed of:
o Quadrupole, sextupole, octupole
magnets
o Divided in 4 circuits (166
magnets, 8 PCs)
TE-MPE-TM 27/06/2012, TE-MPE-MS section
o All PCs are in BB3
• One of the main objective of this
renovation (despite replacing
the old electro-mechanical
crates) is to split the “Ring-Line”
in half sextants
o In 2012, we lost 24h to identify
the origin of a problem in the
Ring-Line (impedance of the line
changing due to ageing of the
TS).
From the WIC side, it is seen as:
4 x (6 x 2) = 48 circuits !
Powering scheme of the “Ring-Line”
• Since we had to put in place patch panels for the «Ring-Line», it became obvious
that we could also centralise all the quadrupole signals in BA3 (where the PCs are
located)
=> Advantage: Avoid using crates belonging to EPC
NER48 «Radiation-tolerant» cable
NE48 (2x24p) cable x2
Patch in
pos. 42250
Patch in
pos. 41450
«Bouchons» to cut the Ring Line:
4 «bouchons» on each side
Dedicated PLC in BB3
RA 8518
Patch in
pos. 32100
-
BA4
Patch in
pos. 51450
+
2x4 NE4 per sextant
+ 2x1 NE4 per sextant
Dedicated PLC in BA3
RA 1777
Patch in
PA3 RA2301
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Patch in
pos. 31482
-
+
BA3
BA5
-
Patch in
pos. 52250
+
Ring line as is today
(4 circuits = 4x NE4 cables
going around the machine)
SPS-RING
Patch in
pos. 22100
-
+
BA2
Patch in
pos. 61450
Dipole & Quadrupole
interlock line
(Split by ½ sextant)
BA6
-
+
Patch in
pos. 21450
-
+
BA1
Patch in
pos. 62250
Patch in
pos. 12100
Patch in
pos. 11310
This doesn’t solve our
communication pbm
for the dipoles!
Communication between BAs
Overview of the layout of all BAs
(for the mains):
Interlock crate
(present in each BA)
Dipole & Quads PC’s
EDMS: 682971
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Around year 2000, EPC introduced
some major changes:
1. Local PLCs to pilot each PC.
2. A Master_PLC.
3. A Hardware Interlock Loop (HIL)!
Crates named «chassis CO» and
“CISBOX” are used to distribute
signals between «BAs».
Centralisation crate
for the mains Quads
Water interlock
crate (EN/CV)
Communication between BAs
Overview of the layout of all BAs
(for the mains):
EDMS: 682971
TE-MPE-TM 27/06/2012, TE-MPE-MS section
• Modifications after LS1:
Hardware Interlock Loop (HIL)
• 3 current loops
• Red: main dipole
• Green: main quadrupole
• Blue: Sextupole PCs of the
Ring-Line
TE-MPE-TM 27/06/2012, TE-MPE-MS section
• If a loop is open, all PCs of the same
loop will trip and send a beam
dump signal to the BIC in BA3.
• HIL loop and functionality remains
under the responsibility of TE/EPC
EDMS: 684948
Thanks to this HIL loop, our WIC PLCs in each BA do not need to communicate between them!
=> In case of a fault in 1 sextant, the WIC sends a «FPA» to the 2 PCs of the
corresponding BA. This will cause the opening of the HIL loop and the trip of the
other PCs!
Beam Dump (Current Situation)
•
The Beam Dump Request:
o
o
o
o
TE-MPE-TM 27/06/2012, TE-MPE-MS section
o
The beam dump requests are centralized
in BA3.
3 input signals (Dipole, Quads,
Sextupoles).
+ 1 input from the ROCS system (that
checks the consistency between the
output of the PCs and the loaded
functions)
+ Other PCs are interlocked via the
Software Interlock System (SIS).
... + inputs for all other systems.
CIB.BA3.S3.Beam
Dump Control.
Our proposal:
1. Keep those inputs as is.
2. Add a signal sent by the WIC to the BIC in
each respective BA.
In the future, we could think of replacing
this crates (EPC) by a FM352 PLC (High
speed Boolean Processor).
Status of the renovation
•
Functional Specification (EDMS 1295506):
o
•
A first version is being circulated to gather comments from all groups involved with detailed
explanation of all changes (What will stay, what will be removed, cabling convention etc...).
Underground work:
o
o
•
EN/EL has installed the cables for us in BA3 and BA5 (2/6 sextants).
13 patch panels are in place in the tunnel.
Surface work:
TE-MPE-TM 27/06/2012, TE-MPE-MS section
o
o
o
•
5/6 “new” racks installed in each BA. UPS and UTP installation scheduled.
15/31 PLC crates are cabled, remaining ones should be ready by end of August 2013 (Spie).
Patch panels (x17) are ordered, need to be cabled (+ control cables) .
PLC/PVSS programming:
David Willeman’s (EN/ICE) contract ends in Feb. 2014 
We depend on David to get the PLC programs  (… true as well for Booster, Linac4 etc...)
o
o
We agreed on deadlines to get the programs before!
•
HWC:
o
Discussion started to optimize the strategy, given the little time dedicated to the powering tests
before the SPS restart (planned for July-Aug. 2014).
•
For the final checks, we need the PC to be «déconsigné».
Conclusions
•
In the case of a renovation, to design an efficient magnet interlock system, one has
to know in details how a machine is operated... (Machine layout, naming convention,
interaction with other systems, standard operation mode as well as non standard operation
modes i.e. MD, different people).
TE-MPE-TM 27/06/2012, TE-MPE-MS section
o
“Each machine at CERN is a different Kingdom”, dixit Karel Cornelis 
•
In «old» machines (same issues for the «Booster» and the «PS») the optic of the
machine is often modified, therefore some flexibility must be anticipated at the
level of our WIC configuration (to avoid hardware & software modifications at
every changes).
•
Production of hardware and cable installation is going on as scheduled.
•
Potential issue with David Willeman’s contract ???
•
During LS1, our colleagues from EN/EL will replace all cables in sextant 1.
o
Part of the standard procedure to replace periodically cables exposed to radiation
=> We are impacted.
TE-MPE-TM 27/06/2012, TE-MPE-MS section
Thank you for your attention!