Log entry time 16:43:02 on February 13,1998

Entry number 40

The following plots show the Beam Energy in the Missing Mass plots only for LH2.

  
Comments for fpiog 40 by Dave Mack.
     
        The purpose of making these plots at this point is to see if we are making any
gross errrors in kinematics (beam energies) or energy loss corrections (I don't even
know if they are turned on). Only the 1H(e,e'pi+)n runs were examined. These fits
aren't yet done "right". That is, we haven't subtracted the randoms and aren't fitting
with a function with the correct shape. For now we want something convenient,
reproducible, and reasonably accurate. The presence of a radiative tail biases the
centroid from a single Gaussian fit to higher missing mass values. Ilkyoung did a
two-Gaussian fit; the wider Gaussian mocks up the randoms and radiative tail somewhat.
I would say his fits give a centroid which is no more than 1 MeV higher than the
correct value. That is just one bin on the zoomed MM .hbook plot in replay_1. 

  Improvements compared to online(see hclog 2663):
1. Used only the arc energy measurements for the beam energy. 
2. More careful to avoid biases created by the random background and radiative tail.
3. Include nearly all 1H(e,e'pi+)n runs in the experiment. 


Results:
0. No "freak" runs were found with ridiculous centroids or sigmas which couldn't be
explained by poor statistics. That suggests that the magnets were set close to the
correct values for all the settings with positive HMS polarity. 

1. If we subtract 0-1 MeV from the centroids in Figure 1 it is clear that the data 
are all roughly at or below the neutron mass, 939.6 MeV/c2. If you blame it all on 
beam energy (no reason to do that at this early stage!) it would mean that our
value of the beam energy from the arc measurements was 0-6 MeV/c2 TOO HIGH (for
relative errors of 0 - .17%). That +- 3 MeV/c2 scatter is about HALF that seen from
semi-offline results in hclog 2663. This is VERY good. 

2. The MM centroids furthest from the nominal value occur when the SOS momentum was 
above 1.6 GeV/c. However, there are also runs at Psos=1.59 GeV/c which look fine. 

3. In Figure 2 the MM peak resolutions increases with beam energy. It also appears
that the energy resolution gets significantly worse for Psos .gt. 1.6 GeV/c.


Still to do:
 
1. Check the energy loss corrections for electrons, pions, and protons and enable them. 
2. Calculate, histogram, and correct for beam energy drifts.
3. Make plots of MM versus x'target, etc, to look for SOS momentum dependent aberrations.
4. Do the same for D runs, both pi+ and pi-.
5. Check whether e+p elastics tell a similar story.