Design And Performance Tests Of Magnetic Circuits For The Magnetic Feedback Structure Of A Micro-machined Accelerometer

This project originates to meet the electro-magnetic feedback control requirements ofa micro-machined accelerometer. Generally, force feedback systems of accelerometers arecategorized as either the electrostatic feedback or the magnetic feedback type.Electrostatic feedback features high precision and easy control, but the electrostatic forceis relatively small. Magnetic feedback can provide greater feedback, but the uniformityand stability of the magnetic field is not as good as electrostatic feedback. According tothe feedback requirements of the spring-proof-mass accelerometer feedback system, weaim to implement a magnetic field with uniformity between10-2and10-3,magnetic fieldstrength is0.7T in a6.5mm*18mm rectangular area.This thesis describes the design of the magnetic feedback system, the implementationof the magnetic circuits and its performance. The first magnetic structure was designed byreferring to PhD thesis of Sunil Kumar of the British Imperial University. The first set ofmagnetic measurements showed the magnetic field was uniform in the middle partbetween the magnetic poles, but the magnetic field strength exhibited a tendencydownward to the edge of the magnetic pole both in ANSYS simulations and experimentalmeasurements. The first set of magnets used a cross bracket to fix the relative positionbetween two poles in the structure center by the attractive magnetic force, but the two polesurfaces were not facing each other in perfect parallel. To improve the parallelism betweenthe upper magnet and down magnet and the uniformity of the magnetic field, weredesigned fixation brackets of the second version for better poles parallelism, andincreased the length of the magnetic circuit for higher uniformity. In order to meet thepackaging requirements of the modified accelerometer, we designed brackets of the thirdversion, which could adjust the distance between the upper magnet and down magnet.In a6.5mm*18mm rectangular area, the magnetic field strength of the second versionmagnetic circuit was measured to be (0.672±0.005)T, with the uniformity of10-2and thepower spectrum of10-6T/Hz at0.1Hz; the magnetic field strength of the third versionmagnetic circuit was (0.688±0.005)T, with the uniformity of10-2and the power spectrumof10-6T/Hz at0.1Hz. Compared with the second version, magnetic circuit of the thirdversion has larger magnetic field strength, better stability and similar uniformity. Generally, given the same area of interest for various magnet designs, the larger themagnetic field strength, the worse the uniformity. Therefore, the third version magneticcircuit demonstrated better properties than the second version. Temperature modulationexperiments showed that the corrected temperature coefficient of the second versionmagnetic circuit was (1438±213±5)ppm/oC and the temperature coefficient of the thirdversion magnetic circuit was (1561±213±173)ppm/oC. Knowing the temperaturecoefficient of the gauss meter probe is67(213)ppm/oC, the temperature coefficient of thesecond and third version magnetic circuit is consistent within the accuracy errors.