chain length argument (long)

I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be two
methods of determining correct chain length. First is the Sheldon Brown and Shimano method: wrap the
chain around the two largest cogs and add one more full link. Second is the way my boss wants it
done: wrap the chain around the two smallest cogs and adjust the chain length so that there is a
small amount of pull on the rear derailleur arm so that the chain doesn't bump into itself or the
pulleys (which will be in a horizontal line). This seems to work equally as well, and I had no
problem until yesterday, when a man brought in his kid's bike, which had a 6-speed rear der but no
front der. The problem was that the chain jumped off the front chainring too often (chain wasn't
stretched). I told him that the chain was too long and that the removal of a few links would help.
My boss (who built the bike) took me aside and emphatically told me that (according to his method)
the chain was the correct length. I went home and thought about it and realized that the small-to
small method would only work if there was a double or a triple front chainring. Couldn't a single
chainring in front be construed as a large ring instead of a small ring (thereby throwing off the
small-to-small system)? Wouldn't the one-extra-link method work better in the case of a single front
chainring? Wouldn't the other method result in a chain that was too long and slack, thereby causing
the problem? So my question is, what is the best way to determine chain length on a system that has
a rear der but a single front chainring?

Gary Smiley wrote:
> I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be
> two methods of determining correct chain length. First is the Sheldon Brown and Shimano method:
> wrap the chain around the two largest cogs and add one more full link. Second is the way my boss
> wants it done: wrap the chain around the two smallest cogs and adjust the chain length so that
> there is a small amount of pull on the rear derailleur arm so that the chain doesn't bump into
> itself or the pulleys (which will be in a horizontal line). This seems to work equally as well,
> and I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> rear der but no front der. The problem was that the chain jumped off the front chainring too often
> (chain wasn't stretched). I told him that the chain was too long and that the removal of a few
> links would help. My boss (who built the bike) took me aside and emphatically told me that
> (according to his method) the chain was the correct length. I went home and thought about it and
> realized that the small-to small method would only work if there was a double or a triple front
> chainring. Couldn't a single chainring in front be construed as a large ring instead of a small
> ring (thereby throwing off the small-to-small system)? Wouldn't the one-extra-link method work
> better in the case of a single front chainring? Wouldn't the other method result in a chain that
> was too long and slack, thereby causing the problem? So my question is, what is the best way to
> determine chain length on a system that has a rear der but a single front chainring?

there's a reason that Mr. Brown and Shimano recommend the big-big combo
- it prevents things from jamming and getting broken. Although you should never really be in those
gears anyway, if you accidentally shift to that position and your chain isn't quite long enough,
kiss your rear derailleur and possibly derailleur hanger goodbye.

If the chain is too slack in the small-small position, it means that the rear derailleur doesn't
have enough length (or tension, if its kidna crusty) for the combination of gears. Thats why they've
got a maximum tooth difference rating, and why there are different lengths of rear derailleurs.

Do it the big/big+1 way, and work on convincing your boss that you're right - its definitely the
best way to go.

Jon Bond

Gary Smiley wrote:
> I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be
> two methods of determining correct chain length. First is the Sheldon Brown and Shimano method:
> wrap the chain around the two largest cogs and add one more full link. Second is the way my boss
> wants it done: wrap the chain around the two smallest cogs and adjust the chain length so that
> there is a small amount of pull on the rear derailleur arm so that the chain doesn't bump into
> itself or the pulleys (which will be in a horizontal line). This seems to work equally as well,
> and I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> rear der but no front der. The problem was that the chain jumped off the front chainring too often
> (chain wasn't stretched). I told him that the chain was too long and that the removal of a few
> links would help. My boss (who built the bike) took me aside and emphatically told me that
> (according to his method) the chain was the correct length. I went home and thought about it and
> realized that the small-to small method would only work if there was a double or a triple front
> chainring. Couldn't a single chainring in front be construed as a large ring instead of a small
> ring (thereby throwing off the small-to-small system)? Wouldn't the one-extra-link method work
> better in the case of a single front chainring? Wouldn't the other method result in a chain that
> was too long and slack, thereby causing the problem? So my question is, what is the best way to
> determine chain length on a system that has a rear der but a single front chainring?
>
>

well - if you were to try your boss's method on my offroad bike I'd be back in a day asking you to
repair the drive train you just totaled! i use a short cage rear der on a 12/32-22/44 setup; give
that the right tension small/small and it'll tear it up big/big.

time was when having the chain as loose as possible was a good thing. the old ders 'liked' slack -
the new stuff prefers a thighter chain. don't you think that just maybe the manufacturer of the
shifters/ders/cogs etal have a 'slight' clue as to how to setup their stuff?

Gary Smiley wrote:

> I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be
> two methods of determining correct chain length. First is the Sheldon Brown and Shimano method:
> wrap the chain around the two largest cogs and add one more full link. Second is the way my boss
> wants it done: wrap the chain around the two smallest cogs and adjust the chain length so that
> there is a small amount of pull on the rear derailleur arm so that the chain doesn't bump into
> itself or the pulleys (which will be in a horizontal line). This seems to work equally as well,
> and I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> rear der but no front der. The problem was that the chain jumped off the front chainring too often
> (chain wasn't stretched). I told him that the chain was too long and that the removal of a few
> links would help. My boss (who built the bike) took me aside and emphatically told me that
> (according to his method) the chain was the correct length. I went home and thought about it and
> realized that the small-to small method would only work if there was a double or a triple front
> chainring. Couldn't a single chainring in front be construed as a large ring instead of a small
> ring (thereby throwing off the small-to-small system)? Wouldn't the one-extra-link method work
> better in the case of a single front chainring? Wouldn't the other method result in a chain that
> was too long and slack, thereby causing the problem? So my question is, what is the best way to
> determine chain length on a system that has a rear der but a single front chainring?
>
>

i forgot to ask - you work for that f@^%ing idiot bill parker @ beckwiths? there isn't a *Worse*
mechanic (who thinks he's the greatest) in the world.

No I don't, but I'm not mentioning any names.

James Connell wrote:

> Gary Smiley wrote:
>
> > I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be
> > two methods of determining correct chain length. First is the Sheldon Brown and Shimano method:
> > wrap the chain around the two largest cogs and add one more full link. Second is the way my boss
> > wants it done: wrap the chain around the two smallest cogs and adjust the chain length so that
> > there is a small amount of pull on the rear derailleur arm so that the chain doesn't bump into
> > itself or the pulleys (which will be in a horizontal line). This seems to work equally as well,
> > and I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> > rear der but no front der. The problem was that the chain jumped off the front chainring too
> > often (chain wasn't stretched). I told him that the chain was too long and that the removal of a
> > few links would help. My boss (who built the bike) took me aside and emphatically told me that
> > (according to his method) the chain was the correct length. I went home and thought about it and
> > realized that the small-to small method would only work if there was a double or a triple front
> > chainring. Couldn't a single chainring in front be construed as a large ring instead of a small
> > ring (thereby throwing off the small-to-small system)? Wouldn't the one-extra-link method work
> > better in the case of a single front chainring? Wouldn't the other method result in a chain that
> > was too long and slack, thereby causing the problem? So my question is, what is the best way to
> > determine chain length on a system that has a rear der but a single front chainring?
> >
> >
>
> i forgot to ask - you work for that f@^%ing idiot bill parker @ beckwiths? there isn't a *Worse*
> mechanic (who thinks he's the greatest) in the world.

On Mon, 05 May 2003 03:04:08 +0000, Gary Smiley wrote:

> ... I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> rear der but no front der. The problem was that the chain jumped off the front chainring too often
> (chain wasn't stretched). I

Neither of these systems is perfect in all applications. I certainly make sure that the system
will shift into the big-big combination, even though I never plan to use it, since if that does
not fit, you will break something. On the other hand, I also check that the small-small
combination is not slack.

But most of the time you have some leeway between these two extremes. I generally prefer a shorter
chain over a longer one, all else being equal. A shorter chain will shift better. But it also shifts
better if it is not over-tightened.

But are you sure this chainring wasn't worn out? Maybe this will never happen with a kid's bike, but
it could be an unusual kid.

--

David L. Johnson

__o | And what if you track down these men and kill them, what if you _`\(,_ | killed all of us?
From every corner of Europe, hundreds, (_)/ (_) | thousands would rise up to take our places.
Even Nazis can't kill that fast. -- Paul Henreid (Casablanca).

As I mentioned, the chain wasn't stretched (i checked it). I believe it was too slack, but my boss
likes to stick with his system.

"David L. Johnson" wrote:

> On Mon, 05 May 2003 03:04:08 +0000, Gary Smiley wrote:
>
> > ... I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> > rear der but no front der. The problem was that the chain jumped off the front chainring too
> > often (chain wasn't stretched). I
>
> Neither of these systems is perfect in all applications. I certainly make sure that the system
> will shift into the big-big combination, even though I never plan to use it, since if that does
> not fit, you will break something. On the other hand, I also check that the small-small
> combination is not slack.
>
> But most of the time you have some leeway between these two extremes. I generally prefer a shorter
> chain over a longer one, all else being equal. A shorter chain will shift better. But it also
> shifts better if it is not over-tightened.
>
> But are you sure this chainring wasn't worn out? Maybe this will never happen with a kid's bike,
> but it could be an unusual kid.
>
> --
>
> David L. Johnson
>
> __o | And what if you track down these men and kill them, what if you _`\(,_ | killed all of
> us? From every corner of Europe, hundreds, (_)/ (_) | thousands would rise up to take our
> places. Even Nazis can't kill that fast. -- Paul Henreid (Casablanca).

With your single front chain ring and multiple rear cog example, I would put the chain around the
largest rear cog. With the chain threaded through the derailleur adjust the chain so that the
derailleur is at its maximum extention (that should be about at a 45 degree angle). That would yield
the shortest best shifting chain that won't damage the drive-train. Depending on the derailleur
capacity it may or may not take up all the slack in the small cog.

Having a dogmatic approach as in either of the two methods you outlined only works on modern proper
spec drive-trains. As soon as drive-train componants become mixed and matched (or just poorly
spec-ed) you need to be prepared to compromise shifting performance at either the high end or the
low end. In order to prevent damage I prefer to make the large/large combo as short as safe and then
the compromise if any will be in the small/small.

"Gary Smiley" <gasmiley@ihatespam.attbi.com> wrote in message
news:3EB5D4AF.733714A0@ihatespam.attbi.com...
> I recently took a part-time job at an LBS, and I replace a lot of chains, etc. There seem to be
> two methods of determining correct chain length. First is the Sheldon Brown and Shimano method:
> wrap the chain around the two largest cogs and add one more full link. Second is the way my boss
> wants it done: wrap the chain around the two smallest cogs and adjust the chain length so that
> there is a small amount of pull on the rear derailleur arm so that the chain doesn't bump into
> itself or the pulleys (which will be in a horizontal line). This seems to work equally as well,
> and I had no problem until yesterday, when a man brought in his kid's bike, which had a 6-speed
> rear der but no front der. The problem was that the chain jumped off the front chainring too often
> (chain wasn't stretched). I told him that the chain was too long and that the removal of a few
> links would help. My boss (who built the bike) took me aside and emphatically told me that
> (according to his method) the chain was the correct length. I went home and thought about it and
> realized that the small-to small method would only work if there was a double or a triple front
> chainring. Couldn't a single chainring in front be construed as a large ring instead of a small
> ring (thereby throwing off the small-to-small system)? Wouldn't the one-extra-link method work
> better in the case of a single front chainring? Wouldn't the other method result in a chain that
> was too long and slack, thereby causing the problem? So my question is, what is the best way to
> determine chain length on a system that has a rear der but a single front chainring?