Zero crossing ?
There is no evidence that .25v -> 70mA will make 7v -> 2A.
There is no requirement for 0v -> 0mA
If a circuit was devised such that .25v -> 70mA and 7v -> 170mA then it would follow that an increase of .25v will cause an change of 100mA/27 (about 3.7mA)
0V
in would therefore result in a current of 66.3mA
This gives .25v -> 70mA
and is linear (assuming my hasty maths is correct).
For small, well defined ranges like this a simple, single common emitter transistor will be near enough to linear, good enough for a couple of octaves. I'm sure I have synth circuits on file that use this technique.
You need to define the maximum current you require from a 7v input, try to imply this from how many octaves this .4 to 7v range represents and the kind of rise-times (cap charge times) this requires.
the lowest input voltage ... ... ... is around .4V and the highest is 7V. so I need the capacitor to charge up to 7V for the whole input voltage range
doesn't make sense to me.
Surely the cap will charge to some (constant) target voltage regardless of input voltage - only its
time will alter, thus resulting in some voltage to frequency law.
Also you refer to
the V/I converter from the MIDI circuit on the synth
Surely this will be a MIDI-to-voltage converter?