A novel low background apparatus  was employed to greatly suppress the signal originating from muons which miss the sample. Previous SR studies have been plagued by a large background signal which contributes significant spectral weight to the measured line shape. This has been a problem particularly for the high-Tc compounds because of the small size of good quality samples. For instance, the ratio of the cross-sectional area of the sample to the beam spot area is typically 1/3. Since the average magnetic field of the background signal is close to the average field in the sample, the background signal always appears near the middle of the measured internal field distribution of the sample.
The experimental arrangement is shown in Fig. 5.1. The crystals were mounted on a thin piece of aluminized Mylar stretched over a hollow cylindrical aluminum sample holder. A small amount of Apiezon N grease was used to attach the crystals to the Mylar. The crystals were mounted so that their -axes were parallel to the magnetic field and muon beam directions. The muons were injected with their initial spin polarization perpendicular to the magnetic field direction. A thin scintillation counter (M) was used to detect the incoming muons. The signal from the M counter started a clock provided that no muon had entered the sample in the previous time interval (). The clock stopped when a decay positron was detected. A cup-shaped veto counter (V) placed behind the sample was used to suppress the unwanted background signal from muons which missed the sample. The decay positrons from muons stopping in the sample were detected by overlapping cylindrical forward (F) and backward (B) counters. As shown in the lower illustration of Fig. 5.1, the F and B counters consisted of four individual counters F1, F2, F3, F4 and B1, B2, B3, B4, respectively. A valid muon stop was defined as and a valid positron event was defined as where or 4. The dashed lines in the top illustration of Fig. 5.1 show the solid angle subtended by the overlapping F and B counters. Note that the arrangement of the F, B and V counters is such that decay positrons originating from muons which miss the sample and stop in the back of the V counter are rejected. Since these ``bad'' muons register neither a valid muon stop nor a valid decay positron, they are logically excluded from the experiment.
The sample and veto counter were contained within a horizontal 4He gas-flow cryostat. The SR time spectra in this study were recorded under conditions of field cooling (i.e. cooled below Tc in an applied magnetic field). Typically each spectrum consisted of ``good'' muon decay events. The measurements were performed on either the M15 or the M20 beam line at TRIUMF, each of which produces a beam of spin-polarized positive muons of mean momentum 28 MeV/c.