From:	FNALD::MARRINER      9-SEP-1993 17:01:16.05
To:	RAGAN
CC:	JZH,LISS,SCHLABACH,MARRINER
Subj:	BFIELD Changes

Ken,

   I have a new version of BFIELD that you should be able to just 
plug in.  The calculation is unchanged if either abs(z)<259.08
or r<25.0.  I think that this means that the change only affects
muons.

   The CMP field is changed to be computed from the geometry.  The
computation yields B(in return yoke) = -.7758 times the central field.
The old program hard coded a constant that was equal to -.6663.
The justification for the new constant comes from the statement that
the net flux crossing the boundary at z=0 should be zero.  We assume
that the field is constant for r<149.1 cm.  We assume that the field
in the hadron calorimeter (207.26<r<345.09) is constant and equal to -6.6%
of the central field.  The return yoke dimensions are 4 pieces
60.96 cm high x 285.75 cm wide.  Assuming that the field is constant
in the return yoke and zero everywhere else, one obtains that the
return yoke field must be -77.5789% times the central field.  I
believe that the old constant was calculated with the same reasoning,
but it appears to me that it is not correct.

   The change in the return yoke field should affect the matching in
the CMP.  Since the lever arm is nil, only the angle matching should
be affected.  I would think that the affect should be appreciable -
one should see a significant mu+ mu- asymmetry with the wrong value.
I believe that Tony Liss is the CMP czar and he should say whether
my theoretical argument is borne out by the data.  If the CMP data
seems to prefer another value, one can hard code (for example)
BYOKE=775.7894 (example to show proper placement of decimal point and
sign). 

   The CMX data is spoken for by Phil Schlabach.  He (with Jorge Troconiz)
has looked at the data and determined that the "improved" end field model
results in better matching and removes the differences in mu+ and mu-
matching.  I have asked him to double check with the new routine, but that
is basically just checking that I didn't screw anything up in the coding.

    I have spent most of the last day testing, and I am confident that
the above description accurately describes the new version.  I think that
it would be an improvement to replace the current version of BFIELD with
the one now located in USR$ROOT:[MARRINER.MUON.SIM]BFIELD.CDF.

    I think I am done with this project.  Let me know if I can be of any
further assistance.

Regards, John Marriner