Dario Menasce, Stefano Magni, Lorenzo Uplegger

1
Uffizi upgrades and new plug-in features
Dario Menasce, Stefano Magni, Lorenzo Uplegger
2
General phylosophy
We are reaching the end of a long and difficult
upgrade of the functionalities of the Pomone DAQ
system aimed at upgrading the efficiency by which
we take data and at the same time giving more
flexibility in controlling the process.

• The three most important new items are
• a new configuration file editor (Louvre)
• an on-line pattern recognition
• a plug-in technology to add-remove additional
  functionalities to the
• read-out without changes to the code and
  consequent compilation-debugging
• cycle required

In the meantime some red items in the DAQ todo
list have been fixed
3
The new Louvre configuration file editor
Why do we need a new configuration file editor?

• The DAQ now has an expanded range of items to
  control
• new FSSR chips
• new module assemblies (several chips per module)
• new PTA design (future)

The existing Louvre was really difficult to
upgrade I designed a new one which makes it
really easy to upgrade in the future with more
and more functionalities (basically I disjoined
the GUI aspect from the algorithmic implementation
, the code is now made largely independent from
graphics a new skin can be designed with very
little change to the code)
4
The Louvre cover
Like the old version, both a stand-alone program
exist and a component bundled with the Uffizi GUI.
5
The Louvre main page

• The widget is divided in two
• parts
• a global one (quantities
• global to the whole DAQ)

• a set of tabs, one per
• mezzanine board, with
• detector-related things
• (two detectors per board)

Each board, in turn, has global quantites and
detector-specific ones (thresholds, pulser, etc)
6
The Pulser widget for mezzanine slot A
Each input is properly validated (e.g. numerical
fields no longer accept arbitrary characters)
Numerical quantities can be selected by sliders
and appropriate fine-scale spin-boxes.
7
The Pulser widget for mezzanine slot B
Adding new properties to a detector now means
just adding a new tab (plus very limited code)
8
The detector widget for mezzanine slot A
An FPIX1 detector is lodged in slot A of
mezzanine in branch 0 an FPIX1 has few
controls, so a single widget can lodge them all.
Each detector has an associated kill-inject mask
editor
9
The kill mask for an FPIX1 detector

• Two tabs
• kill mask
• inject mask

10
The inject mask for an FPIX1 detector

• This widget has several
• functionalities
• restore a previously defined
• configuration
• kill everything at once
• kill nothing
• set kill mask to complement
• of the corresponding inject
• mask

11
The complement feature
12
An FPIX2 chip example

• An FPIX2 has a richer set of
• configurations to control.
• These are further grouped in a
• set of sub-tabs
• Bias voltages
• Threshold voltages
• Spatial information

13
An FPIX2 chip example

• An FPIX2 has a richer set of
• configurations to control.
• These are further grouped in a
• set of sub-tabs
• Bias voltages
• Threshold voltages
• Spatial information

14
An FPIX2 chip example

• An FPIX2 has a richer set of
• configurations to control.
• These are further grouped in a
• set of sub-tabs
• Bias voltages
• Threshold voltages
• Spatial information

15
A new FSSR chip example
An FSSR has an even richer set of configurations
to control.
16
A new FSSR kill-inject mask
An FSSR sees the strip detector as a pixel
detector with 8 rows x 16 columns
17
The dynamic load of external plug-ins

• Uffizi has a new editor
• to control the dynamic
• loading of external plug-ins.
• Each plug-in is a pre-compiled
• library containing a component
• that can be used by the
• consumer at event-build time.
• These can be
• histogram filling
• track reconstruction
• calibration

18
The dynamic load of external plug-ins
19
The Control Room Logbook interface

• Uffizi has a new way to
• communicate with the CRL
• a message is sent to the CRL
• each time the following happens
• a run is started ended
• an error occurred or a
• message of a user-selected
• priority is generated

This page shows the content of the last sent
message
20
Current status of the project

• Everything shown is currently in cvs and
  available at MTEST
• Lorenzo and Mauro are debugging it (Im doing it
  in Milano with a more limited hardware)
• The pattern recognition was made a plug-in
  yesterday and still suffers from bugs Im
• currently fixing them

To do

• Use information generated by the Circus to
  influence filling of histograms
• Improve type of information sent to CRL to
  include Circus data
• Finalize remaining small items in the DAQ list of
  todos