Currently I have a limited access to an electronics lab, but from this point my options are:
1) assume that there is the other half of the voltages divider inside the helix, and calculated its value
2) go the lazy way, but make your expression pedal generic and work with any modeler:
sample the read value to some applied analog voltage at, say 100 points, and whenever you need some value in between these points do a simple linear extrapolation. You assume this should give a resolution good enough for most applications.
Still didn’t have the time to test the sampling rate tho..