Hello,
We designed a rotary detection sensor with msp430fw427 and we need to upgrade it to sense a slightly higher distance. It works but is very unstable related to temperature mainly due to the scanif internal oscillator drift.
We implemented the algorithm for sifosc calibration when there is a change in its frequency according the Ti app note slaa288. The sifosc frequency could be recalibrated when the temperature changes, but this method is still not accurate because the "frequency step" is too high. We noticed that using a look-up table with different DAC levels for each temperature is more efficient but still not enough.
The difference between undamped and damped signals is around 100 mV, we could achieve this value only with a special inductor with a special core and wire resulting in a higher Q factor.
We also tested the provided run time calibration, but apparently this doesnt work for this signal level and this method requires a angular speed higher than 0 to work which is very bad for our application.
I am out of alternatives to solve this problem.We might decrease the frequency to minimize the sifosc drift impacts but that also decreases the signal. The optimal frequency for achieving the 100 mV signal difference was about 500 kHz.
Is there anything left that i can try?