Acoustic ringing

probe (Lakeshore [92]) with care being taken not to rotate the probe between measurements. The pseudo-homogeneous region between around 5mm and 15mm above the probe surface revealed the profile shown in Fig. 2.17.

Figure 2.16: Measuring theB0 field profile using a Hall effect probe. The probe was

positioned vertically with the aid of a Perspex jig. A Vernier height gauge used for the height measurements had an aluminium arm clamped onto it to keep the metal gauge at a distance from the magnet.

The damaged Mole probe was taken to Wellington and re-mapped along the

z-axis using an automated 3D mapper. The central magnet was then removed and replaced. The damaged magnet was 4.47mm thick and measured (115±2)mT at the center of the face. The replacement magnet measured 4.49mm and (129±2)mT.

The z-axis B0 profile was then re-measured at different central magnet heights

until the z-axisB0 field remained constant. Interestingly, by raising or lowering the

central magnet the B0 profile exhibited an almost linear gradient along the z-axis

irrespective of the height of the central magnet. Thirteen profiles were measured at quarter-turn rotation intervals of the central magnet holder using the automated mapper. The results are shown in Fig. 2.18. The B0 field strength between 5mm

and 15mm varied from 88mT down to 82.7mT at one extreme and from 73.3mT up to 75mT at the other extreme. The profile ofB0 off thez-axis was not investigated.

2.5.5

Acoustic ringing

After replacing the central magnet the Mole was returned to Palmerston North and re-tested. A single spin-echo with 500µs echo time had previously returned a time

62 Chapter 2. Experimental apparatus 0 5 1 0 1 5 2 0 2 5 3 0 3 5 4 0 6 6 6 7 6 8 6 9 7 0 7 1 7 2 7 3 7 4 7 5 B 0 f ie ld a lo n g z -a x is ( m T ) H e i g h t a b o v e M o l e s u r f a c e ( m m ) B ₀ f i e l d s t r e n g t h a l o n g z - a x i s v e r s u s h e i g h t a b o v e M o l e s u r f a c e

Figure 2.17: Damaged Mole B0 profile measured along the z-axis.

domain echo, but now returned an echo heavily modulated with∼40kHz oscillation. Extending the echo time to 1,500µs largely eliminated the oscillation. The cause of the oscillation was unknown at this time and further damage to the Mole was suspected, however this did not explain the absence of the 40kHz oscillation at the longer echo time. After researching ringing in NMR magnets it was discovered that coil disease/acoustic ringing was possibly the cause of the oscillation (Buess and Petersen (1978) [25], Fukushima and Roeder (1978) [58]).

Acoustic ringing in the context of NMR probes describes an event in which a voltage is induced in the B1 coil by an ultrasonic wave in a nearby metal object. In

this instance the object was the Mole’s central magnet. The ultrasonic wave was believed to originate from theB1 coil’s linearly oscillating magnetic field generating

eddy currents within the skin depth of the central magnet. According to Buess and Petersen, circulating electrons experience a Lorentz force due to the B0 field

2.5.5. Acoustic ringing 63

Figure 2.18: Mole B0 profile along thez-axis for differing central magnet heights. The

central magnet height was adjusted by a quarter-turn rotation at each step. The change

inB0 was approximately linear between 5mm and 15mm above the probe surface.

which is perpendicular to the B1 field. This produces a transverse wave in the

magnet that propagates along the direction of the B0 axis before being reflected

back toward the B1 coil. Upon returning to the surface from which it began, the

wave generates an electric field which is detected in theB1 coil.

It should be noted that Buess and Petersen considered acoustic ringing in aluminium (and brass, but minimal acoustic ringing occurred at their <10kG fields), and Fukushima and Roeder investigated a range of metals. Neither of their papers referred to acoustic ringing in rare earth magnets.

To confirm the presence of acoustic ringing in the Mole probe central magnet, the magnet was removed and placed directly on the B1 coil on the lab bench and

a spin-echo experiment was performed again. Measurements were difficult due to EMI from the surroundings so the B1 coil was mounted in the base of a small

64 Chapter 2. Experimental apparatus

cable shield significantly reduced the EMI pickup and sitting the box lid on top of the box reduced it further. Oscillations were present with the central magnet sitting directly above and below theB1 coil and absent with the magnet removed

suggesting that oscillations were not coming from the aluminium box. Increasing the distance between the magnet andB1 coil by 3.3mm reduced the amplitude of

the oscillations by around 80%. Phase cycling also mostly cancelled the ringing.

Buess and Petersen observed ringing in 1.0mm thick aluminium at six regularly spaced frequencies below 10MHz. They also observed acoustic resonances 90kHz apart in the same 1mm thick sample. The Mole probe central magnet was scanned over a range of excitation frequencies between 3.570MHz and 3.650MHz. The wide variation in tuning required soldering additional capacitors to the B1 circuit to

obtain the required tuning and matching conditions at each excitation frequency. It was found that at the lowest and highest frequencies the ringing was around 10µVp-p and increased to around 35µVp-p at the center frequency. The oscillation

frequency was also markedly higher at the outer frequencies in similar fashion to Buess and Petersen’s observations.

Inserting a piece of 50mm×50mm copper foil between the B1 coil and magnet

shifted the tuning of the B1 circuit thus again requiring additional capacitors to

obtain the correct tuning and matching. Upon re-testing, the copper foil was found to have significantly reduced the oscillations, irrespective of whether or not it was grounded. Wrapping the central magnet in copper foil eliminated the oscillations.

The central magnet holder originally had silver foil on the inside base and around the walls as shown in Fig.2.19(a), however this was no longer functioning to stop acoustic ringing. This may have been due to damage to the electrical continuity of the foil during magnet removal, or perhaps to a slightly different sized central magnet leading to different electrical connections. To shield the entire magnet using 0.05mm copper foil required removing a thin layer of material from the inside of the holder in order to fit the foil and magnet. This was achieved using a lathe, then copper foil was cut and placed in the base and around the perimeter of the holder and carefully soldered as shown in Fig.2.19(b) without melting the magnet holder. Acetone was used to remove the adhesive backing from the foil. Another copper disc was cut slightly larger than the magnet diameter to make a top cap with a 1.5mm lip and was fitted over the new magnet. This eliminated the acoustic ringing problem.

Buess and Petersen note other methods used to solve acoustic ringing. These include coating the metal surface with an acoustically lossy substance such as RTV (silicon) or using brass rather than aluminium as it is a much poorer transmitter of