1. AMS timing

From SVTSPYMON stored histograms from TAKA's special run (4/5) 159183 and 159184
/cdf/code-common/cdfonline/svttest/spymon/root_files/SVTSPYMON_159183.root
/cdf/code-common/cdfonline/svttest/spymon/root_files/SVTSPYMON_159184.root
Look at GB timing in 0-50 microsec scale, 1microsec per bin.
Using default AMS microcrode

Using new (v2.0) AMS microcode

  1. A clear shift of about 1 bin toward smaller times is visible in the second histograms, consistent with 700nsec improvement expected.
  2. Timing average is dominated by fluctuactions in the tail. A different problem.
  3. Note that 20 microsec to pass an empty event is unbelievable.
  4. All this effort to make it 19... looks so irrelevant.
  5. Compare with 15 microsec to go through one wedge. How do we manage to spend 5 microsec on empyt events on final mergers and GB ?
  6. I do not see urgency to check new microcode timing on LSA
  7. Will still try to do it, esp. to have a test stand for possible future work, still need a special cable and time: on the backburner.
  8. Propose to make AMS microcode v2.0 the default.
  9. But... will make SVT output road list different then simulation
  10. Do we want to wait until this is fixed ? Bill ? Stefano ?
For comparison, here's the same timig histogram as before for 4/4 (i.e. the previous physics run) 159180

2. SVT multiplicity (timing?) vs. XFT cut

Studies on Run 156456

Thanks to Bill, Alex, Rolf and Liz S-K

This run's L1 triggers:
This run's L2 triggers:
  1. I forgot to exclude short tracks. Thanks Rolf
    Here's new plot:
  2. This work is just starting
  3. Plan to understand a bit better high multiplicity sources and effect. Then repeat with cut on Pt in XTF. Hope to be done by friday.