Existing EMMA systems have provided invaluable information on speech kinematics. Two main drawbacks remain, however. First, phonetically relevant tongue movements are inherently three dimensional. Second, the necessity for midsagittal alignment severely restricts subjects' freedom of head movement, potentially compromising the naturalness of their speech. In previous work an arrangement of six transmitter coils allowing three-dimensional reconstruction of sensor coordinates (plus alignment about two axes of rotation) was designed. Now, a two-stage design strategy is reported to arrive at a full-fledged measurement apparatus, involving the development of a virtual instrument followed by its hardware implementation. The geometrical complexity of the transmitter arrangement and of the resulting magnetic fields necessitates the use of complex numeric algorithms to determine sensor location and orientation. A promising algorithm has been developed and will be briefly discussed. The crucial task is then to compare the performance of the algorithm under ideal and expected real-life conditions. For this reason a virtual instrument will simulate the entire measurement chain, to test the sensitivity of the algorithm to errors from quantization, carrier-frequency cross-talk, etc. The results of these simulation experiments will be presented and their implications for realization of the system in hardware will be discussed.