Log entry time 23:59:42 on November 9,2005

Entry number 159648

Followups:

• 159654 11/10/05 03:09 lkaufman Cavity monitors have clear position and charge crosstalk

To calibrate the position cavity monitors, I plot the predicted position 
difference using the stripline data along the x axis (units are microns); and 
along the y axis, I plot the measured position difference in the cavity (units 
are channels).  The slope gives me the calibration constant in units of 
chan/um, so I take the inverse of the slope as the calibration constant so that 
I can convert the position differences measured by the cavities in channels to 
microns.

Both figures are only for slug 1: Figure 1 shows calibrations for cav2 and 
figure 2 shows calibrations for cav3.  The top two plots in each figure show 
the linear fit to the data from which I calculate the calibration constant for 
the cavities.  The bottom two plots show the residual of the the fit in one 
dimension.  The x axis of the residual plots is in units of channels.  It is 
obvious from these plots that the fit for the y cavities (right column) is much 
worse than for the x cavities (left column).  The residuals are not only wider 
in y, but they are also double-peaked.  This double-peaking can also be seen in 
the raw or regressed position differences as measured by the y cavities (shown 
in halog 158096) and seems to be a problem with the cavities (and not the 
striplines) based on the results I showed in halog 159607.
 
I know that there is a lot going on here, and we are just beginning to try to 
understand what's going on, so stay tuned.