Minutes of meeting on trigger symmetrisation and normalisation
held on 29/8/1994
Present:
M.Fidecaro, P.Bloch, M.Dejardin, C.Touramanis, A.Schopper, T.Ruf, 
E.Aslanides, P.Kokkas

Concerning the KK events: All events with two SCbarS, Pt cut kaon
candidates are latched into the register, independent of their charge.
The events pass the same kinematical cut as Kpi events.
The dE/dx kaon cut of HWP2 acts only on the kaon in register 1,
(needs to be confirmed by Renaud).
If matching OK then the "KK" event passed trigger under the same conditions 
as a Kpi event and should be used for offline analysis.
If the kaon in register 1 is a false kaon, then the event is rejected
either online or afterwards offline.
During the discussion of the trigger, Philipp asked the question
what happens to tracks with "undefined" charge. This couldn't be answered
and will be found out before the next meeting. 

A study of M1-M2 asymmetries was presented by Thomas.
Nothing was found which could explain the the phasedifference of 
10 degrees between M1 and M2. 
Since we believe the difference comes from a geometric effect it
would be more appropriate to study asymmetries of K0bar and K0 
with the same curvature dependences: K0bar(M1) - K0(M2) and
K0bar(M2) - K0(M1). This asymmetries are independent of the geomtrical
effects. Whereas the overall normalisation in M1 is 1.093 and in M2
is 1.134, it becomes 1.113 and 1.114 by using events with the same 	
curvatures. The result for the phase is: 42 degrees and 45 degrees
with an error of +-4 degrees. (No LUTs were used!).
Asymmetries of the kind K0(M1)-K0(M2) and K0bar(M1)-K0bar(M2) 
should in principle show clearly the geometric effect to the 
lifetime distribution. But no effect was found.
A comparison done by Guido suggests that the trigger symmetrisation
(simulating different field) doesn't work as it should.
This needs confirmation by Renaud, because he showed sometime ago
a plot which proofed that it works.

Christos looked for a correlation of the decay products and the 
charged kaon. He showed the angular distribution of positive pions
in the K0 rest frame (transverse plane only) for 2,3,4 prim. events
and M1, M2. Within the statistics 2,3 together with M1 and M2 are
symmetric whereas the distributions for 4 prim. events not only 
allow you to tell which field you used but also define the strangeness.
The origin of this effect is unknown. It seems connected with the 
probability to loose a track with a large angle with respect to the 
radius, which we know is curvature dependent.
By removing events with a track which as an angle > 40 degress  
the fitted phase in 3,4 prim events changes completely.

The micros are now available for P24. The next step will be 
to create LUTs for the mixed field asymmetries and to look also
at the P24 results. 
Differences in M1 and M2 might be due to a different behavior of the 
detector. This will be studied by using minimum bias data and to look
at the momentum distribution of positive and negative tracks in the 
two different fields.
   
The other possibility is that the difference is the effect of our
selection criterias (online and offline) which are curvature dependent.
This should to some extent be compensated by the LUT method, but 
using LUT or not using LUT, the phase remains the same. 
Since the Pt of the kaon, p of primary pion and K0 momentum are forced
to be flat using the LUT, this would suggest that the normalisation 
is a function of the radius and that this dependence changes with the 
field. This could be studied by looking at the difference of the lifetime 
or radius distribution of K0(M1) and K0(M2), and K0bar(M1) and K0bar(M2).
Although by eye there is no difference, one should use more 
sophisticated statistical methods to measure the difference.    

Marc will complete his MiniMC program which should allow to simulate
the effect of the trigger cuts to the momentum, radius, lifetime...
distribution.   

I suggest that we meet again next monday 4pm in the CPLEAR conference room 
wherever it will be. 

Thomas