HAPPEX III Electron Analysis is already good enough. Photon integration as test. Photon counting analysis should be good. –Fix present analysis –Taka data at different beam rates to study rate effects See analysis talk
Compton must integrate at 850 MeV Serious accidental issues now seen at an energy of 3 GeV Lineshape is an issue Both problems much worse at 850 MeV See analysis talk
Spectra at low energy Low energy spectra have a very complex shape. Large light output doesnt help.
Noise with Integration Need pulses near saturation. Flip helicity at 120 Hz or more? Pedestal noise complex.
Need Linearity with Known Small Corrections Electromagnetic MC shows linearity Photo-Neutron production at ~8 Mev is a potential of few % nonlinearity. Need Monte Carlo of neutron production How big are systematics Can we design a feasible test (straightforward test to 1% subject to systematics.) Linearity
Light output of detector Need just enough for statistics;>10PEs/Compton edge event. Spectrum has messy shape at low energy.
Speed of Detector Response Must be fast for 3 GeV photon counting analysis to keep accidentals low
Cavity and Stripline Studies 4ax4bx C2xC3x Positions measured and computed from Dithering 4ax 4bx C2x C3x Positions computed by regression (One coefficient has the wrong sign.) There is a waist at 4bx with a width near the electronic noise level
Explanation Typical Regression coeficient = = + Uncorrelated noise reduces coefficients Need correlated noise to explain data = = +
Electronic noise If σ B ~ σ E, electronic and beam noise can be unfolded (without large subtraction in quadrature.) Regression and dithering give different results: Regression is unreliable Dithering implies that both have ~1 micron noise Correlations probably invalidate regression analysis
Interpretation Correlated noise is significant noise limit. May be due to bandpass of instrumentation! Increases helicity reversal might help. Need more analysis (LUMI with both cavities and BPMs?)