Caesium Magnetometer test results

The use of Caesium Vapour (CV) sensors utilising optically pumping technology has been heavily criticised by other manufacturers, specifically claims are made that heading errors on the basic sensor can be as high as 20nT. They claim these can be reduced to 1 to 2 nT by the use of split beam technology, or a polarising filter, but at the expense of system sensitivity I.e. loss of basic resolution.

Geometrics totally refutes this, and believes in letting their technology answer these criticisms for themselves.

Furthermore, CV sensors can be used with great success in airborne land and marine applications, subject to the same limitations that would also be applicable to any other sort of magnetometer sensor.

Below are some test results from typical CV magnetometer sensors, which show the actual measured heading errors. A second test of actual noise measurements, and showing the superb bandwidth of this technology, can be accessed via the links on this page. In magnetometers, bandwidth is defined as how great a change in the field strength can be tracked over a set period of time.

It is also stated by competitive manufacturers that Caesium Magnetometers need to be returned to the factory periodically for expensive re-alignment of the optical components. This is not the case. All components in the Geometrics design of CV sensors are fixed in place at the time of final testing, and do not ever need to be returned to the factory for re-alignment unless serious physical damage has occurred to the sensor.

If CV sensors are fitted to an aircraft for airborne survey purposes, the use of an active or compensation system is required to reduce the influence of the aircraft itself on the measured magnetic field, such a compensation system is required irrespective of the type of sensor being utilised to measure the magnetic field.

Before presenting the test result, some description of the testing procedure is in order which Geometrics conducts at the AMES test facility built and maintained by NASA.

The Heading Error Test Site:

The Magnetic Test Laboratory was constructed in the early 1960s by the United States National Aeronautics and Space Administration after extensive magnetic surveys were conducted at the site to locate a magnetically suitable location. Originally built to test and measure magnetic properties of missiles and spacecraft, the facility has been utilized in support of many programs beginning with Voyager and, most recently, the Magnetic Probe Spacecraft deployment. Stanford University has been a routine user of the capabilities of this location and several military organizations specialising in magnetic detection work and missile research.

The location chosen for construction has a natural magnetic gradient of less than 1nT per meter. Construction materials were carefully specified and screened to be non-magnetic so as not to interfere with sensitive measurements necessary for the space program. All electrical and mechanical hardware used in construction were aluminum or brass. Interior and exterior finishes, roofing materials and fasteners were all carefully screened for magnetic content. All of this allows precise measurement and evaluation of magnetometer systems performance as they are rotated through all positions under test.

The facility was equipped with a twelve foot three axis Helmholtz coil and power supplies necessary to manipulate the apparent magnetic field of the Earth at the “sweet spot” for equipment under test. The earth’s magnetic field vector can be rotated in any direction, nulled, or even reversed. This capability is not available anywhere else in the Western US and been used by Geometrics Engineering to evaluate and improve performance of magnetometer systems.

As requirements of the US space program and defense applications wound down during the 1990s this facility became available for US Government sponsored applications, which required precision magnetic measurement and instrumentation. Geometrics initiated a Cooperative Research and Development Agreement (CRADA) with the Naval Research Laboratory (NRL) to develop and deploy a multi-sensor towed array detection and classification system for unexploded ordnance in 1996. Because of the precision magnetic measurements required for development of this system Geometrics was granted permission to utilize this facility for the duration of the CRADA. Upon expiration of the CRADA in 1999 Geometrics applied for and was granted exclusive operational control of the facility under a Space Act Agreement (SAA). All significant development and testing of magnetometer systems have been accomplished there since that time.

Magnetometer sensors are tested for heading error at the test laboratory. The sensors under test are mounted on a precision turntable which can be rotated about all three axis’ relative to the earth’s field vector. The turntable rotation angle is monitored and recorded using a special nonmagnetic optical encoder mounted on the turntable axle. Two stationary magnetometers are mounted on each side of the sensor under test to measure and record the diurnal variation in the magnetic field. The data from all of the above is recorded to disk by computer. Data recording is started / stopped with a foot switch.

By subtracting the sensor under test data from one of the stationary sensors and using the optical encoder data we then plot variations in the measured field as a function of rotation angle (heading error). In addition as a quality assurance measure, we also subtract the two stationary sensors from one another and verify that their difference remains at zero. While collecting data an audible alarm sounds if they do not precisely track each other – usually indicating the operator has forgotten to remove keys, belt buckle, or other magnetic objects before beginning the test.

The process of collecting and plotting the heading error data is automated. The collected data is examined by software before plotting to paper for several things such as acceptable noise levels, the presence of magnetometer data at all angles (no angle gaps), that any recorded dead zones occur where they should, and that the two reference magnetometers track each other. For troubleshooting purposes any data that fails the above tests can still be plotted but the plots are clearly marked as having failed.

Typical test results:

Figure 1: This plot shows the output of a sensor spun around its pole, and shows a heading error of approximately +/- 0.25nT

Figure 2: This plot is the output from a sensor which has been tumbled about its equator, again the total heading error is in the region of +/- 0.25 nT. The gaps in the plots are caused by the well known dead zone phenomena.

Follow the link above to see results of a test to measure noise and bandwidth of Caesium Sensors, or click here.