Muons-On-REquest (MORE) -- piM3 Area
(A full description
of the MORE setup is given in Ref. )
In conventional time-differential µSR
experiments, an ideal lifetime spectrum would contain only start-stop events
belonging to the same muon. In real experiments using continuous muon beams,
however, there is always a certain number of muons escaping detection,
e.g. those stopping near but not hitting the muon counter. This
results in a random rate of events in the start and stop channels and a
random background in the muon-decay histograms, thus limiting the useful
time interval to about 10 µs and excluding investigations of low
muon-spin precession frequencies or slow relaxation processes.
Pulsed muon beams, on the other hand, deliver
many muons per pulse at low repetition rates (e.g. 50Hz at ISIS).
In this technique, the time resolution is limited by the finite length
of the muon pulses (80ns at ISIS compared to 1ns on the GPS instrument
at PSI). However, the background is typically three orders of magnitude
A method to solve the background problem
at continuous muon beams has been proposed earlier (see 
or ) for surface-muon beams at TRIUMF or KAON
but has never been realised. The basic idea is to extract only one muon
at a time out of a continuous beam by means of a fast-switching electrostatic
deflector ("kicker") on request from a µSR instrument. This makes
sure that no other muon reaches the spectrometer until the extracted one
has been processed. Moreover, the concept of "Muons-On-REquest" (MORE)
does not significantly reduce the intensity of the original beam which
is therefore available for the simultaneous use by a second spectrometer,
i.e. MORE produces twice as many results of even higher quality
(lower background in one spectrometer) than the conventional technique.
Experimental Setup in
the piM3 Area
The main components installed for MORE
in the surface-muon beam line piM3 at PSI are shown in the Figure
1 together with the two instruments GPS (General Purpose Surface-muon
spectrometer) and LTF (Low Temperature Facility).
Layout of the piM3 surface-muon beam line at PSI for the muons-on-request
The kicker contains two 1m long, 20cm wide
electrodes 20cm apart (see Figure 2). Two power
supplies for dc voltages up to +5kV and -5kV are connected to the
electrodes via fast switches, giving a difference of 20kV between the two
field directions and a separation of the muon trajectories of about 5cm
at the intermediate beam focus in front of the septum magnet located about
5m from the kicker exit. Each switch consists of a series of 15 high-voltage
MOSFET transistors type IXYS 6N100.
(a) Schematic circuit of the fast-switching
deflector ("kicker") for muons on request. Power per switch <600VA,
repetition rate <40kHz, voltage |U|<10kV between electrodes, switching
time: <25ns (10-90%).
(b) Driving circuit for one switch.
Delay between trigger signal input and high-voltage output ca. 40ns.
The muon detector (M-counter) in the spectrometer
(GPS or LTF) is used to trigger the kicker. The kicker is switched to the
spectrometer running in "MORE mode" (say, GPS) for a maximum of 5µs
at a fixed repetition rate (max. 40kHz). The signal of the first muon hitting
the trigger detector (M-counter) after a minimum delay of 200ns is used
to switch the kicker back to the spectrometer running in "parasitic mode"
(LTF in this case). The delay is necessary to avoid damage to the power
Either instrument, GPS or LTF, can be used
in MORE mode while the other one is running in "parasitic" mode. Two variable
slits (FS301 and FS302 for LTF and GPS, respectively) and located in the
intermediate focus plane of the beam in front of the septum magnet are
used to adjust the event rate in both legs individually. The two instruments
can also be used simultaneously in conventional "shared" mode by means
of a horizontally defocused beam spot in the septum magnet slit plane.
Figure 3 shows
an example of µSR in silver in an external magnetic field of 10mT,
taken with the GPS in MORE mode. For comparison a conventional spectrum
is shown taken at the same event rate. The background in MORE mode is at
least a factor of 100 lower than in conventional mode, thus allowing the
study of muon-spin precession and relaxation easily up to 20µs.
: µSR in silver in a magnetic field of 10mT measured at the GPS facility
instrument in conventional and in MORE mode.
Insert: Reduced asymmetry plot for
the first 2µs in MORE mode (function fitted for t > 0.4µs).
|Time resol. [ns]
|Event. rate [106/h]
Table 1 : Comparison of results
obtained with GPS in conventional and in MORE mode (using the GPS M-counter
as trigger). Values for pulsed µSR (ISIS) are also shown.
In MORE mode with M-counter trigger we
observe a small distortion in the spectra at times t < 450ns (insert
in Figure 3) which is due to the delay between
the passage of the muon and the arrival of the trigger signal at the kicker.During
this time, additional muons can enter the spectrometer and "kill" events
through pile-up rejection. Fitting the distorted spectra is possible
using different values of the fit parameters N0 (normalizing
constant) and B0 (background) in the two regions. However,
the MORE technique is preferentially used for slow signals where cutting
off the first channels does not affect the data analysis.
The unique new feature of the system is
that - due to the high time resolution - it opens the possibility of studying
slow relaxation phenomena at high magnetic fields and to resolve
close spin-precession frequencies (see the striking example obtained
in UPt3 ).
R. Abela, A. Amato,
C. Baines, X. Donath, R. Erne, D.C. George, D. Herlach, G. Irminger, I.D.
Reid, D. Renker, G. Solt, D. Suhi, M. Werner and U. Zimmermann,
Hyperfine Interact. 120/121 (1999) 575.
Hyperfine Interact. 66 (1990) 1137.
Z. Phys. C 56 (1992) S258.
A. Yaouanc, P. Dalmas
de Réotier, F. N. Gygax, A. Schenck, A. Amato, C. Baines, P. C.
M. Gubbens, C. T. Kaiser, A. de Visser, R. J. Keizer, A. Huxley, and A.
Phys. Rev. Lett. 84 (2000) 2702-2705.