Originally Posted by
HillRider
OK, now I'll try to explain.
For the big chainring you are correct, the shift gate location, etc is important but the rotational orientation is easy to determine since the chain catching pin is always placed right behind the crank arm. No ambiguity there.
For the inner chainring of a Shimano double, all of the teeth are identical in height and shape and there are no pins, ramps, gates, etc. on these rings.
Yes, you are 100% right ------ as far as you go.
HOWEVER consider that the ring has 39 teeth and 5 bolt holes. Since 39 doesn't divide by 5 it means that the teeth are in different rotational orientation at each bolt hole. That means that the phase (or timing) relationship between the teeth of the inner ring and those of the outer ring depends on which bolt hole is at 12 o'clock.
With me so far. Now as the chain leads through the shift gate and off the outer ring it's still attached, so the rotational position of the roller as it drops to the inner ring is predictable (with respect to crank angle). So the inner ring should be timed to meet the chain properly oriented (rotationally) for the chain to slip right on (the glide in hyperglide).
If you drive a car with standard transmission, it's comparable to dropping engine speed to match revs on upshifts, so the gear can slip right in.
I wish I had a good photo showing this, but you can see it on your own bike. Shift to high, then slowly downshift the front, advancing the crank until the chain (on it's secant) is just about to engage the inner ring, and STOP. Now look at whether the teeth line up ready to slide into the chain, or if they're about to bump into the rollers. Imagine if the teeth were rotated forward or back a fractional pitch (1/10") and think about whether they'd engage more or less smoothly.
As I've said a few times. This isn't critical, it just makes shifts smoother.
BTW- I took a photo showing what I'm describing. Do I need to belong to a photo sharing service to upload? Or someone might tell me how it's done.