Technical question about XTR M952 bb's

Hi all,

I have two XTR M952 BB's with different spindle lengths, one is 112.5 and another is 116:



The left one (112.5) is definitely OK and I used it for some time. As for the right one, I grabbed it yesterday and there are two things which seem quite strange to me. First of all, it pobably has a wrong left cup: it is marked "112.5" and manual says there are different cups for 112.5 and 116. Am I right?

OK, suppose I simply got bb with a wrong cup. But another thing is much more strange. When you compare right sides you can see that spindle segments on the right are absolutely equal:



But how that's possible? 112.5 bb is for 47.5mm chainline, and 116 bb should give 50mm chainline by adding only one 1mm spacer. 50-47.5=2.5 is not equal to 1! I'm really confused.

Any ideas?

Because spacers are used on these bottom brackets, they'll set the slight difference in chainline.
116 - 112.5 = 3.5
3.5/2 = 1.75 either side
Assuming symmetric axle
They may just be aiming for more overall clearance, who knows, so more on the non-driveside and a touch on the drive.
What are theory like fitted.

I know about spacers, and there is a link the manual in the first post. For 112.5 one should use 2.5mm spacers on each side with 68mm bb tube and no spacers with 73mm bb tube. For 116mm there should be 3.5mm on each side for 68mm tube and 1mm on each side for 73mm. In both cases, 68mm and 73mm, the difference is just one 1mm spacer on each side. It won't move the crank by 2.5mm and even by 1.75mm, it will only move it by 1mm!

Sorry my calculation was to show it wouldn't happen anyway and that their aim was probably something else.
1mm drive side 2.5mm non-drive side, it may well be 1.5 and 2.
Keeping to the shifting for the top group might have been more important, tolerances on the splines I would guess is much easier than the wide tolerances on a square taper that are aimed at.

It seems a bit like the jump from 107 to 110 in the square taper world when all but .5mm is on the non-drive side iirc.

If they where doing it fully it would be 112.5 and 117.5 like the 113 to 118 in low profile cranks.

Just found (non-official) Q-factor data for Shimano cranks here. It seems that you are right, 1mm goes to drive side and 2.5mm to non-drive side. So what Shimano says about chainline is probably not exact, the actual chainline will be 48.5mm for the long spindle. I wonder if modern front mechs optimized for 50mm chainline will work well with 48.5... We'll see.