Cocoa Plans

1
Cocoa Plans
James N. Bellinger University of
Wisconsin-Madison 13 February 2008
2
Cocoa Plans

• Working with Cocoa and ME1
• Cocoa with root input

3
General Findings w/ ME2/SLM2

• The laser error was 0.200 microns instead of 200
  microns ? Fit tried to keep the laser at the
  default position come hell or high water.
• Including the offsets and pixel calibrations
  reduced the measurement contribution to the chi
  squared by about 5

4
First studies with ME1

• Fitting all 6 HSLM at once
• Internally treated as 3 SLM with only partial
  info for each laser
• Default MAB and DISTANCEMETER values for now
• From original Maeshima file
• Data from 10-Nov-2006, pixel-size corrected

5
ME1 dataset isnt ideal

• It was handy
• 3 DCOPS had no data
• Of remaining 84 profiles
• 6 were shadowed
• 4 were saturated
• 3 had low signals

6
First Pass Results Cocoa

• Tried to move 11 CCDs more than 400 microns
• Tried to change the upCCDXtoDowel2 etc on several
  transfer plates
• (not correlated with the above shift!)
• 2nd item suggestive Put in calibrated offsets

7
With Calibrated Offsets

• Cocoa tried to move chamber 1/3/9 by over 14
  meters and twist it around
• Serious stability problem here
• No obvious errors in SDF file
• Looking for pilot error

8
Cocoa ME1 Plans

• Understand/fix stability problem
• Day?
• Use better data
• ½ day
• Mock up data for and vary included devices to
  study effects
• ½ day
• Locate data for included devices and test
• ½ day
• Mock up non-included devices, esp. tilt meters
• Week? 2 Weeks?
• Locate data for these and test
• 2 days?

9
Connect with Link System

• Goal Run side Cocoa fit using ME1 DCOPS and
  analog systems and at least some subset of the
  Link system
• DCOPS
• DCOPS Z-sensors
• DCOPS Z-sensors minimal set of Link
  components
• Add tilt meters
• Add R-sensors
• Add as much of Link as reasonable
• No simultaneous and side DCOPS yet

10
Root input to Cocoa

• Discovered that what I earlier thought was good
  output wasnt. Not all DCOPS root data was
  getting into Cocoa data structures
• Must revisit this required for final operation
• Cocoas default operation is to read input data
  event by event and try to converge on a final fit
  of simulated positions.
• Not ideal for our debugging work
• ?Root input is not a high priority for our
  testing
• Scripts to create text file data exist for final
  config
• Scripts to create files from MTCC data need
  testing