Title: A discrete event system: the CMS Tracker interlock
1The CMS Tracker interlock system
- The CMS Tracker and its environment
- The interlock system and its role in our work
frame - Input sensors and conditioning cards
- Outputs, system modularity and hardware
integration - Information flow, communication and software
integration - Status
- Conclusions
2The CMS Tracker
3The CMS Tracker
3 sub detectors 6 partitions 107 readout
channels 8,000 voltage channels gt105
monitoring channels Power dissipation 50 kW
4The CMS Tracker
5The Trackers environment
- The Tracker will have to be maintained at lt-100
C, while its surrounding environment will be at
180C. - The Tracker will be exposed to irradiation doses
foreseen to be between 10 and 65 kGy and neutron
and charged particle flows ranging from 1012 to
1014 particles/cm2. - The Tracker will share the CMS experimental
cavern environment and resources. Power losses
will be unavoidable as will be other harmful
incidents (fire, flooding, etc).
6The Trackers environment
- The Tracker has to be protected from temperature
excesses
Temperature interlock
- The Tracker has to be protected from the presence
- of humidity.
Humidity monitoring
- The Tracker has to be protected from repeated
power - on/offs the power should be switched on and off
in - a predefined sequence.
Power control
- The Tracker has to be protected from all cavern
- environmental hazards.
DSS connection
- The Tracker has only control over its own power.
7The Interlock system must
- Protect the Tracker by switching power off in the
- correct method and in time
- Evolve with the Tracker accompany it during the
- testing/integration period and during all its
life at - LHC.
- Have optimized sensor/interlock resolution.
- Be able to be integrated to the CMS experiment on
time.
8The inputs
- conditioning only at
100m - The sensor requirements radiation tolerant, tiny
(mm) - no magnetic field
effects (4T) - The temperature sensors pt1000s,thermistors
(3types) - The humidity sensors radiation tolerant, able
to sustain magnetic field.
Digital inputs from external sensors
9The inputs
- The sensor signals travel inside the Tracker on
specially designed control cables (1,600) for
100m in order to enter in the control system.
- The thermistor and humidity sensor signals have
to be - conditioned before entering the readout of the
control - system the conditioning systems have to be
under the - surveillance of the control system.
10The Outputs
- The output of the control system logic is
distributed to the 200 power supply crates of the
Tracker in the form of interlock signals - The power supplies provide bias both for the
detectors and for the readout electronics chain. - Each crate accepts up to four different interlock
signals - The granularity of action is that of a crate.
- There is a hierarchical way of distributing the
interlock.
11The inputs numbers
The control system (based on PLCs of the SIMATIC
S7-300 family)
Temperatures 800
Read inputs
The Power Supply
Other 100
Do logic
RH 200
Set outputs
12System modularity-hardware integration
Tracker
Thermal screen
13DES
We try to treat the Tracker control system as a
Discrete Event System. A Discrete Event System is
a discrete-state, event-driven system, that is,
its state evolution depends entirely on the
occurrence of asynchronous events over time. We
treat the Interlock system as such. EVENT should
be thought of as occurring instantaneously and
causing transitions from one discrete state value
to another. E.g. a cable is disconnected a
cooling line gets clogged or a dry air pipe blows
up.
14GRAFSET
15Information flow, communication and software
integration
DCS layer (Diagnostic Reconfiguration)
Measurements
Measurements
cond. DB
cond. DB
Drivers
PLC Plotter (java)
Hardware
DCU Plotter (java)
T, V, I
via DCU readout
T, H, other systems
hardwired
16Status Conclusions
- The Six large systems have to start being
delivered in January 2006. Three shall remain for
integration tests at CERN and three shall be
shipped to laboratories outside CERN. They shall
return home in order to form the control racks
of the CMS Tracker. - Six miniatures of the large systems are being
deployed at CERN and in existing laboratories for
more restricted roles but with exactly the same
I/Os and functionalities. - We are very tight on time schedule because we
would like to have everything completely
operational by spring 2006.
17Status Conclusions
- The CMS Tracker infrastructure and environmental
controls are fully based on standard industrial
hardware. Although we needed to proceed to
certain developments in order to integrate those
standards to the general software and other
requirements of a large HEP sub detector, the
overall modularity, scalability and the up to now
performance are satisfactory.