Log entry time 03:35:18 on August22,2005

Entry number 152841

Fig 1 shows the how the cavity BPM differences correlate to 
stripline differences.  These are helicity correlated differences
for six slugs (25,26,29,36,37,38) where things worked.  A few
subtleties :
1. The cavities are more correlated to some striplines than others,
so I picked the most correlated pair to plot.  E.g. cavity 2X is
well correlated to stripline 4AX but not as tightly to 4BX.
I haven't tried doing a detailed calibration or linear interpolation,
Lisa is presumably doing that.
2. The signs of the correlations flipped sometimes when cavities
were power cycled.  For example, in slug 29 cavity 2X is positively
correlated to 4AX, but in slug 38 they are anti-correlated.  I have
adjusted all signs so that the correlations come out positive.
3. The cavities are NOT yet calibrated.  Units are ADC channels.

The "bottom line" is that the BPM diffs are correlated and therefore
probably real physical beam differences (not just electronics).

Another handle on this is to compare the detector asymmetries for
regression (I didn't do it for dithering).  So, here is a table
showing the det1 and det3 asymmetries for some typical "good" runs.
The "bottom line" is that cavities and striplines regress to the
same result which again suggests they are measuring the same h.c.
beam parameters.  All units are ppm.  Results read from ROOT files
*run*regress* and *run*regress*cavity* with "ok_cut":

Run ..... Det1 Asymmetry .......... Det3 Asymmetry ......... Error
......... Stripline .. Cavity .... Stripline .. Cavity
4547 .... 13.5 ...... 13.2 .......... 14.4 .... 14.6 .... +/-14
4554 ..... 3.62 ..... 3.63 ......... -7.02 ... -7.00 ... 9
4562 ..... -8.74 .... -9.06 ........ -2.80 ... -2.62 .... 7
4566 ..... -9.43 .... -9.70 ........ 1.31 .... 1.73 ..... 7
4568 .....  9.12 .... 9.40 ......... 5.33 .... 4.97 .... 7.5
4572 ..... 3.33 ..... 3.30 ......... 5.72 .....5.86 .... 7