Analysis of Machine Studies Measurements of 22 August 2005
==========================================================

Following the brief initial study by JPS and me on
<a href="http://cesrelog.lns.cornell.edu/MSLOG/cesrc/2005/cesrc_050812.html"> 12 August</a>
(see <a
href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/16aug05/report_17aug05.html">analysis</a>),
MGB and I made a number of detailed systematic studies
on <a href="http://cesrelog.lns.cornell.edu/MSLOG/longrbb/cesrc_pc_050822.html"> 22 August</a>
to make sure the observed orbit distortions really were caused by the BBI and to
see if the BMAD calculations remained in agreement under differing
sets of conditions.

After repeating the original measurements using e+ t2-6.b1-5 2 mA/bunch with 1.6 mA in e- t1.b1
(albeit now in the 4030 lattice rather than the 3970 lattice) and obtaining the distortion as
before, we made five systematic tests. A web page of analysis results can be found
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/index.html">here</a>).

I presented the analysis in last Friday's machine studies meeting. There I incorrectly
reported that the measurement had been done in HEP conditions. In fact, we found we had to
do them in elcsr conditions in order to maintain beam lifetimes, similar to what JPS and I had
found. In the meeting, I also presented a plot comparing a BMAD calculation with 1.6 mA/bunch of
positrons to orbits measured at 1.9 mA/bunch. Having corrected this, the measurement/calculation
agreement improved.

The analysis web page contains plots comparing BMAD calculations (red lines) to the measured orbit
differences (green dots) for the five systematic studies:

1) we turned off PRZ13, as in the BMAD calculation. It made little difference to the
     orbit and for purposes of comparison with the calculation we made the subsequent measurements
     with it off (though we needed to turn it on for injection)

2) we lowered the electron current to 1 mA (negligible effect),

3) we changed the positron bunch current to 1.6 and 0.9 mA (linear effect on distortion, as expected
     and calculated):
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/P002.html">1.9 mA</a>
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/P001.html">1.6 mA</a>
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/P003.html">0.9mA</a>,

4) we lowered the pretzel amplitude by 20%, again obtaining a result in agreement with
     the BMAD calculation:
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/P004.html">PRZ1
3277-->2659, 2 mA/bunch</a>,

5) we changed the positron bunch configuration to 3.6 mA/bunch in t9.b1-5, which the
     BMAD calculation predicted would yield a qualitatively different orbit distortion.
     Indeed, the orbit changed in the way BMAD said it would:
<a href="http://www.lns.cornell.edu/~critten/cesr/injection/notes/30aug05/all/P005.html">0.9mA</a>.

These studies cover a range of orbit distortion magnitudes from 0.4 to 1.4 mm. The agreement between
the orbit measurements and the BMAD calculations is better than 0.05 mm per point.
This yields a high degree of confidence that we are indeed measuring the magnitudes of the dipole
kicks from the BBI and that they are correctly simulated in BMAD.


========================================================
James Crittenden                   Tel. (607) 255-9424
Wilson Synchrotron Laboratory      Fax  (607) 255-8062
Cornell University
Ithaca, New York 14853-8001
========================================================