Novel multi-layer magnetic Vernier group method for multi-turn encoding

Multi-turn encoders receives great attentions recently, especially the gear-type ones with mechanical structure and permanent magnets, which has the function of multi-turn mechanical memory. However, the current gear-type multi-turn encoders still have many challenges, such as complex structure, large volume, complex calculation method, etc. To provide a more flexible multi-turn magnetic encoder that can encode position information in an analytical way, this paper studies the principle, system model, encoding algorithm, and design policy of a novel multi-layer magnetic Vernier group (MVG) method. The weakness of the conventional multi-turn magnetic Vernier encoder is firstly analyzed. Then, the new principle, system model and general encoding algorithm of the investigated multi-layer MVG method are introduced. The design policy of the novel multi-layer MVG method are carefully discussed: (1) from the tooth number of the minimum common period (TNMCP) perspective, all possible combinations of two lower-layer MVGs to upgrade to a higher-layer MVG are analyzed. (2) The gear and tooth number decision policy for lower layers are investigated when the top-layer TNMCP is decided. A gear-saving scheme is established, and the corresponding parameter selections are analyzed. (3) The general process of upgrading the current MVG into a higher-layer alternative is shown out. Finally, a two-layer MVG encoding system with three gears is designed as an example for experimental verification. The results demonstrate that angle information and multi-turn encoding can be realized according to the proposed multi-layer MVG method. The correctness of the gear-saving scheme is also verified. The turn values decoded from the multi-layer MVG are consistent with those from the reference motor encoder.