I have been scratch building a road bike using a Trek compact frame,and mostly Shimano 105 components. Mostly I build older vintege bikes, so its not often I touch 10 speed gears.

Normally I measure new chain length using big to big +2 links, but I had read that for 10 speeds you should use small to small and thread the chain through the derailleur, so I used that method.

However, when I cam to set up the SI, I could not get it to work. The derailleur did not seem to be indexing and would not reach the two largest cogs. I adjusted the angle in case the derailleur was hitting the cogs, I double checked parts numbers in case I had a non 10 speed, changed the outer cables, as I had been a bit tight - no change.

The only change I had made was to measure the chain. So I split the chain, re-measured my normal way and there was a 4 link difference (too long). I rechecked the new way, and I appeared tom have that correct. I had images to check against as well as instructions.

As soon as I shortened the chain, everything worked, but I cannot get my head around how this impacted on the SI. Nothing else was changed.

Can anyone provide an explanation.

What's an SI?

By "10 speeds" do you mean 2x5 gearing, a.k.a. 10-speed bicycle, or that newfangled system where they name them based on the number of sprockets on the rear? Either way, I would think the danger of breaking the RD when you shift to larger sprockets or chainring would be the same as on any other gear combination.

Cannot reason why chain length would impact shifting unless the limits of the RD are reached. Even then, it doesn't make sense to me. I challenge the "Nothing else was changed." Maybe not knowingly.

+1

If you make the chain longer, you're going to have less A tension opposing B tension, so you may need to let some B tension off by backing off the B screw. But that's just on some derailleurs, not SRAM, for instance.

Chain length can affect shifting and the big/BIG +2 method is not always optimal. Sometimes the shimano small/BIG vertical jockey wheels method works better.

Sorry to state the obvious but did you check your limit screws? I usually set these up first, as in the old friction shift days. Pull on the cable manually to shift. Will it reach those cogs then?

I don't know what SI is either. Do you mean STI? Indexing?

Assuming a typical Shimano derailleur and small/small vs big/big chain lengths, what is going to cause the shifting quality to be affected by the angle and tension of the jockey pulley cage if the indexing, stops and B settings were made correctly?

Chain length, even a link or two, will affect the wrap and the position of the jockey cage. It's kind of hard to describe, but a chain will often shift better with a link added or removed. There's always an optimum, despite the fact that often times a range of chain lengths will actually work. It's mostly in the smaller cogs in back that the differences are obvious.

AFA the why, I don't know and don't frankly care. There are a lot of aspects of bike mechanics where experience trumps theory. This is one.

I see no reason why you can't move the pulley to where you need it with the B screw.

You've got some adjustment with the B screw of course, but it isn't everything. Chain length matters more IME. I suppose we'll have to agree to disagree.

The SI is the most important parameter with modern drivetrains. It boggles my mind that people who would step into this discussion do not know what it means.

Clearly. It is nearly as critical as RL.

​​​​​​shift index?

If you know what SI means, why not tell us? Not a common or C&V acronym.

I've adjusted many thousands of derailleurs and drivetrains without knowing, so apparently it's not as important as you think.

My knowledge of BSA is clearly lacking. I'll need to take a TO from this BDT. CUL.

Did you reconnect the chain with the special pin? You can't just push a pin back in, the "rivet" head is broken off. I always use quick links instead of the supplied pin.

That's interesting that the shifting was so bad and improved so much. I've never had that happen to me, and my previous bike came with a too-long chain.

(From the context: "SI" shift indexing? I've never heard that term...)

I commented on the SI term as a hoax to elicit some explanation from the OP, but thinking about it I may have been right.

I seem to recall from a seminar at Interbike many years ago an in-house quantitative term used to describe the performance of shifting performance/robustness that related to how much of an adjustment range would still give acceptable performance.

That term just may have been the "selectivity index".

But it seems that the OP is going to let us toss this one around for a while.

Yeah, I realized the hoax after my first post. Hence the second less serious comment. I can be slow before coffee...

It seems the OP is referring to indexing, but we don't really know.

There have been a few silly in-house terms introduced at Interbike over the years. I still think Q factor is kind of dumb. Chosen cuz it sounds serious I guess. Q factor means something very specific in engineering/science, and it is not pedal to pedal width.

And closely correlated with BS, one suspects.

It's a metaphor. How do the chains (of capitalism) affect the Solomon Islands (SI)? OP is trying to start a discussion regarding how the economic imperialism of western nations contributed to the widespread economic, political and social instability experienced by many former colonial possession upon gaining Independence, which is illustrated by the ethnic violence between the Guales and Malaitans in the late 1990s, and how modern day expansion of Chinese business interests continues to contribute to the instability in the region.

When OP refers to splitting the chain, and then shortening it, he is referring to the frequent need for deployment of Australian and New Zealand troops to the region to quell simmering ethnic tensions. Yes, it solves the immediate problem, but fails to address the underlying conditions leading to the problems in the first place.

My state of bogglement remains unabated. But now I have an enlightened insight - it is just one of many forms of BS.

I'm starting wonder about the Tr0LL factor w.r.t. SI. And it isn't even April.

In the old days before chain pins were supposedly peened we just pushed the pins back into place with ye ol' chain tool, and they worked just fine. I figure that if a pin has as much or more friction in the side plate now as it did then then it ought to work. My experience with new 8-spd chains like the SRAM PC870 and KMC (Is that what it is? I fergit 'xactly...) is that they do take a lot more force to push the pin out and a fair amount to push it back in too. So it should work as well or better than in the old days. So I've treated them the way I always did (which cussing at the extra effort it takes). The only time I've had a chain break was when a quick-link came apart.

Considering that the side plates have gotten thinner to make the chains narrow, I suspect narrow chains have much less friction between the side plate and pin than they did on 7 speed chains.