You probably wouldn't understand if you're not one of those folks into close ratios, but here's the deal - the physical limitations of a derailer drivetrain (few rings, many cogs) impose a suboptimal situation regarding the gaps between gears.
Everyone is familiar with the way cassette tooth counts are spaced - with gaps of say, 4t, 3t, 2t, 2t, 1t, 1t, 1t, etc. If you plot the resulting ratios, you find that once you're down to 1t gaps, the gaps between ratios widen as you go faster, which is the opposite of desirable, given that air resistance is proportional to speed cubed.
My half-step concept halfway fixes this, making the gaps pretty close to consistent. Of course, nobody would ever bother to shift this manually, but it might be worthwhile if an electronic system could take care of it. I had an Arduino programmed to control a couple of servos to suit, but there was no way to ascertain the viability of it without CAD-optimised front shifting.
The two big rings could be spaced about 6mm apart, and perhaps timed shifts with the aid of a crank position sensor could be smooth and reliable enough for it to be a goer, although timed shifts could only happen once per rev given the 1t ring difference.
You probably wouldn't understand if you're not one of those folks into close ratios, but here's the deal - the physical limitations of a derailer drivetrain (few rings, many cogs) impose a suboptimal situation regarding the gaps between gears.
Everyone is familiar with the way cassette tooth counts are spaced - with gaps of say, 4t, 3t, 2t, 2t, 1t, 1t, 1t, etc. If you plot the resulting ratios, you find that once you're down to 1t gaps, the gaps between ratios widen as you go faster, which is the opposite of desirable, given that air resistance is proportional to speed cubed.
My half-step concept halfway fixes this, making the gaps pretty close to consistent. Of course, nobody would ever bother to shift this manually, but it might be worthwhile if an electronic system could take care of it. I had an Arduino programmed to control a couple of servos to suit, but there was no way to ascertain the viability of it without CAD-optimised front shifting.
The two big rings could be spaced about 6mm apart, and perhaps timed shifts with the aid of a crank position sensor could be smooth and reliable enough for it to be a goer, although timed shifts could only happen once per rev given the 1t ring difference.