2009.08.11 - Gain Mapping for BPMs 9W, 10W

Gain mapping data was collected for BPMs 9W, 10W on 2009.06.01 (see eLog entry). Analysis was performed using the "bpm_gain_anal" program, using both the Poisson and "wire_calc" interpolation grids. (See nonlin_bpm studies summary for descriptions of the two interpolation methods). Variables for the bpm_gain_anal fitting are:

• Four button gains (renormalized after fitting s.th. gain(1) = 1.)
• (x,y) for each measurement
• Button pedistals (offsets) for all four buttons

The quality-of-fit is judged by a "gain_sigma" parameter, defined to be: \sigma = \sqrt(\frac{\sum((b_i^{meas}-b_i^{fit})/b_i^{meas})^2}{N}), N = number of (x,y) grid points.

Fitting with the Poisson interpolation grid

BPM	Fitted Gains	Button Sigmas	Button Offsets
9W	{1.000, 0.881, 0.981, 0.861}	{0.26%, 0.30%, 0.38%, 0.44%}	{92, 336, 142, 236}
10W	{1.000, 0.915, 0.993, 0.967}	{0.29%, 0.30%, 0.40%, 0.42%}	{115, 163, 133, 160}

• Bpm_gain_anal output file for 9W, 10W
• Plots of data grids:

Fitting with the "wire_calc" interpolation grid

BPM	Fitted Gains	Button Sigmas	Button Offsets
9W	{1.000, 0.837, 0.933, 0.790}	{0.29%, 0.31%, 0.42%, 0.51%}	{-107, 350, 72, 247}
10W	{1.000, 0.869, 0.944, 0.888}	{0.29%, 0.29%, 0.43%, 0.39%}	{-103, 122, 36, 131}

• Bpm_gain_anal output files for 9W, 10W
• Plots of comparing fitted data grids for 9W, 10W:

We find there is a shift between the two interpolation grids' results, present both before and after fitting.