I don't have my Feickert with me (on long-term loan), but, if I recall correctly, it offers both a one-point (no need to iteratively move the protractor) or the two-point (like the Mint mentioned above) alignment (protractor must be moved).
For the one-point alignment, you will be select the curved line on the protractor that matches the type of alignment you want to use. You MUST then use the tool that also determines the pivot to spindle distance to correctly align the whole protractor. You should use something to wedge and hold the platter firmly in this position because this alignment of the protractor must not be moved. Next, for the alignment you want to use, look along the curve to find the marking for 212mm (your pivot to spindle measurement). At that point along the curve, you will also see a thin crossing line which is what you use to align the cantilever. You must not move the protractor at all using the one-point method.
As a check on this alignment, the Feickert also offers two-point alignment using cross hatched squares around the two null points (after all, getting tangency at these two points is the purpose of the alignment). Using the two point method does require moving the protractor to see if, when the need is placed at both points, the cantilever achieves tangency at both points. I think that the two-point method is more accurate because it does not require perfectly accurate measurement of the pivot to spindle distance, it does not require that the protractor be in the perfect position (the measurement spike perfectly aligned over the center of the pivot), and it will achieve the desired alignment even if the pivot-to-spindle distance falls somewhere off of the perfect 210, 211, etc. that is provided for using the one-point curves.
For the one-point alignment, you will be select the curved line on the protractor that matches the type of alignment you want to use. You MUST then use the tool that also determines the pivot to spindle distance to correctly align the whole protractor. You should use something to wedge and hold the platter firmly in this position because this alignment of the protractor must not be moved. Next, for the alignment you want to use, look along the curve to find the marking for 212mm (your pivot to spindle measurement). At that point along the curve, you will also see a thin crossing line which is what you use to align the cantilever. You must not move the protractor at all using the one-point method.
As a check on this alignment, the Feickert also offers two-point alignment using cross hatched squares around the two null points (after all, getting tangency at these two points is the purpose of the alignment). Using the two point method does require moving the protractor to see if, when the need is placed at both points, the cantilever achieves tangency at both points. I think that the two-point method is more accurate because it does not require perfectly accurate measurement of the pivot to spindle distance, it does not require that the protractor be in the perfect position (the measurement spike perfectly aligned over the center of the pivot), and it will achieve the desired alignment even if the pivot-to-spindle distance falls somewhere off of the perfect 210, 211, etc. that is provided for using the one-point curves.