From Glen Marshall, Elog, 16 July 2002:

The fundamental settings for surface muons are as follows, according
to Jaap and based on some assumptions about the existence of hardware
which does not in fact yet exist (F2V). These would give a beam
emittance of something like pi*5mm*16mr in each of x and y, suited to
optimum passage across the fringe field.

F0V = 18 mm
F0H = 82 mm
F1H = 10 mm  (determines momentum acceptance)
F2H = e.g., 20 mm, or at least something a little larger that F1H.

Recall that the momentum acceptance is about 1% FWHM for each 12.5 mm
of F1H slit edge separation (width), with an expected minimum of about
1% due to beam imperfections (non-point-source, aberrations, higher
order effects, etc.).

Also it is useful to remember that the fractional range spread of
surface muons is given _approximately_ by

(\Delta R)/R = [0.1**2 + (3.5*(\Delta p)/p)**2]**0.5

This might be taken as FWHM, and is just the quadratic sum of
contributions from range straggling (10% or 0.1) and from the momentum
spread of the beam.  The latter term is from differentiation of the
relationship of range on momentum.

The setting which was used has a much larger F1H width, and a
correspondingly larger range spread, except that the F2H width should
limit it. Presumably F1H was increased in order to get higher rate for
cloud muons; if so, opening up F0V and F0H proportionally might make
more sense than opening F1H. On the other hand, a wider F1H setting
(keeping F2H larger than F1H by 10 mm or so) would be useful to spread
out the stopping distribution by increasing the momentum acceptance.

The surface muon rate for the recommended setting of (18,82,10,20) is
about 3500/s for 130 muA, which should not be too tough on the
trigger. If it is too high, it would be better to reduce F0V and F0H
proportionally (e.g. to cut the rate by about 4, use F0V=9 and
F0H=41).