Analysis of time spectra run 3 and 4

Overview Analysis Results Interpretation

Overview

Time spectra from run 3/4 were analyzed. Muon stops were defined by beam scintillators
and muon tracking, electron were defined by pc and scintillator coincidences with tracking.
The analysis procedure and preliminary results are presented.

Analysis procedure

Preliminary results

Tentative explanation

The changing effect suggests 2 possible explanations Given the overall information from the data and the beam properties the second hypothesis seems
excluded.
In order to obtain more information about the first explanation I analyzed the detailed structure
ot the time spectra. An instructive example is given here (left colums mu-, right column mu+,
first row: Pu time spectrum, second row: *e(+l0 t), third row: as second, but mgood required).
While the mu- spectra behave like expected, the mu+ spectra show a clear oscillation. These spectra
were systematically displayed for all conditions and preliminary fits with the form (mut#msrall)
            msr1=1.-par(4)*cos(par(5)*x+par(6))*exp(par(7)*x)
            msr=par(3) +par(1)*exp(par(2)*x)*msr1
were performed.
Oszillation parameters are about:
 
omega (1/ch) T (us) B(muonium) B(mu+)
run 3 0.60 5.5 0.34 G 34 G
run 4  0.23 2.1 0.13 G 13 G

using 1 us == 0.72 G (muonium) and 1 us == 74 G (mu+), respectively. Since the required mu+ fields
are high, probably we see muonium.
Moreover it is encouraging that the lambda deviations get reduced to about < 0.5 % level with this
MSR fit (a more careful analysis has to be done).
If one assumes that the inititial phases of the muonium and mu rotation are close to zero, one can
estimate the effect of free mu rotation based on the observed muonium signal.  In general the rate
variation in the z direction for mu+ is
Rz = l0 exp (-l0 t) ( 1 + P alpha sin (w t)) ~ l0 exp ( l0 t ) (1 + P alpha w t )
where P alpha ~ 0.1. The overall correction to lambda then becomes
P alpha omega / r0 which gives 0.1 0.23/0.2 e-2 /0.45 = 0.25 e-2.
(Compare Fleming et al PR A26 (1982) 2527 for muonium/mu+ formation in H2 and other gases).
This effect is similar to the deviations we presently observe. We have to perform more careful fits
of our MSR spectra to limit the rough <0.5 % level quoted above. For instance, I have not fitted
up/down sum spectra etc.

The changing magnetic field is a bit surprising. Nevertheless a rather consistent picture is
emerging. In the end we need the field defining coils for fast rotation, but it does not seem
to be an unsurmountable MSR problem.


 
Peter Kammel / UC Berkeley and LBNL / kammel@sseos.lbl.gov

Lawrence Berkeley National Laboratory, MS 44B
1 Cyclotron Road, Berkeley, CA 94720
Tel (510) 486-5021 / Fax (510) 495-2111