We were asked whether moving gas cylinders on the floor above a large TEM would cause a problem with its magnetic field environment. We made measurements of the DC field fluctuations at different positions as a large gas cylinder was wheeled past on a trolley. The results show that a non-uniform field fluctuation was created by the cylinder, which would exceed the specifications of the microscope. It would require a dual magnetic field cancelling system to bring the field back into specification. It should be noted that the cylinder was a standard type used in many laboratories, but we only made measurements for one cylinder. Other cylinders may give different levels of field.
A large high resolution TEM with monochromator and energy analyser can image with sub-Angstrom resolution, but is very sensitive to magnetic fields in the environment. Typically the field is required to be stable to within 30 nT Pk-Pk, but some models have tighter requirements.
We were asked whether moving gas cylinders on the floor above a large TEM would cause a problem with its magnetic field environment, so we decided to make some measurements. It was not convenient to measure the field below the cylinder, so we measured the field above it.
The steel gas cylinder was made to BS 5045-1. It was 1.5m long and 0.22m diameter. It was found to be slightly magnetised, with fields up to about 4 times the Earth’s field on the surface of the cylinder. No special effort was made to magnetise or de-magnetise the cylinder, so it may be regarded as typical. It was placed on a standard steel trolley designed to carry gas cylinders.
A Spicer Consulting 3-axis DC sensor was placed at various different heights. The cylinder trolley was pushed under the sensor for a distance of 5m in the Y direction relative to the sensor. The Earth’s field is X=120, Y= 140, Z=-440 mG, relative to the sensor. The sensor was reset each time to remove the Earth’s field from the measurement.
The SC11 Chart Recorder was used to chart the magnetic field waveform in 3 axes. The resulting waveforms were extracted from the chart recorder results and plotted using SCplot. The peak to peak change reported by SCplot was adjusted to remove the AC (50 Hz) that broadens the trace on the chart. The height of the cylinder was subtracted from the height above ground to obtain the distance from the cylinder. The results are summarised in Table 1 and Chart 1 below.