Dura Ace triple build question/chainline/shim cassette?

I am having difficulties getting derailleurs adjusted on a Dura Ace triple. This exact kit has
worked fine on at least one other frame where I could get all gears with very little rub. On this
new frame, the same kit does ghost shifting when on middle ring in front and largest two cogs in
rear. In other words, without me shifting it goes down to small ring and then comes back up to
middle ring. Up down, up down. Two different shops have not been able to adjust properly. Looks like
rear end of chain line would benefit from about 1 mm of outward ajdustment. Can/should this be done
by putting a shim behind cassette? Or, is there any way to move drive side of bb inward? Does anyone
have to settle for not being able to get these two ratios?

 

eflayer-<< On this new frame, the same kit does ghost shifting when on middle ring in front and
largest two cogs in rear. In other words, without me shifting it goes down to small ring and then
comes << Looks like rear end of chain line would benefit from about 1 mm of outward ajdustment.
Can/should this be done by putting a shim behind cassette? >

Doubt it, have you had the frameset checked for alignment?

Or have the BB shell checked to ensure it's centered in the frameset?

It should work...it sounds like the crank is too far out, either because of the BB shell of the rear
triangle is skewed left.

Peter Chisholm Vecchio's Bicicletteria 1833 Pearl St. Boulder, CO, 80302
(303)440-3535 http://www.vecchios.com "Ruote convenzionali costruite eccezionalmente bene"

 

"They all do that sir."

Actually, they don't anymore, but early production D-A triples had a strong tendency to do exactly
what you describe until the chain had worn in a bit (500 miles), and even then, the biggest
sprocket/middle ring combo was always "iffy". Shimano later introduced a modified middle ring that
seems to eliminate the problem (I tested it pretty carefully with a brand-new chain when I replaced
my original middle ring).

So, your first port of call is back to your dealer, to ask him to get you the new type middle ring
(I am not sure there is any identifying code on the rings that tells you which version is which,
but there are some clear differences to some of the tooth profiles if you have one of each type in
your hand).

1mm difference in chainline won't make any difference - the sprocket spacing is way more than that.
However, there is something you can do to improve the chainline. I fitted an XTR 116.5mm bottom
bracket instead of the standard 118.5 D-A triple bottom bracket. You might think that would only
make 1mm difference each side, but you would be wrong! Shimano's road BB's are offset to the right,
the MTB ones are symmetrical. Fitting the XTR BB got me about 3mm difference in chainline, which
fixed the ghost shifting problem on the 2nd sprocket (the 1st sprocket was still a problem) before I
got the new middle ring.

The "value" of switching to the XTR BB is that you no longer have your feet offset to the right by
about 3.3mm (6.5mm relative difference - I felt I was slightly twisted in the saddle). A lot of
modern cranksets (especially triples) have this sort of offset - I think it's due to the silly
obsession with Q factor at the expense of all else. I'd rather have a few mm more Q and symmetry,
personally, but YMMV.

Note that the XTR BB comes with various spacers (so it fits both 73mm and 68mm BB shells and can be
offset to handle fat seat tubes) - you *must* use the two wide ones (one each side) on a road frame
(which has a 68mm BB shell). I actually used one wide and one narrow spacer on the drive side,
because my frame needed the extra 1mm clearance. Don't worry about this - the XTR cups are meant to
be used this way in a 68mm BB shell and are extra-deep to handle this. You even get a slight benefit
in the form of wider bearing spacing.

I found front shifting slightly better with the new chainline (but you do have to adjust the
vertical position of the FD slightly, as well as adjusting the cable and limit screws).

 

> Actually, they don't anymore, but early production D-A triples had a strong tendency to do exactly
> what you describe until the chain had worn in a bit (500 miles), and even then, the biggest
> sprocket/middle ring combo was always "iffy". Shimano later introduced a modified middle ring that
> seems to eliminate the problem (I tested it pretty carefully with a brand-new chain when I
> replaced my original middle ring).

Could also be that he's trying to use it on a bike with extremely-short chainstays. Your 5500 has
415mm chainstays and works OK; we've tried them on bikes with 405mm chainstays and found they simply
didn't work. What's so magical about chainstay length that a difference of so few percent would make
it go or not is curious, but that seems to be the case.

--Mike-- Chain Reaction Bicycles http://www.ChainReactionBicycles.com

 

Subject: one more thing-dura ace triple question

I just did a semi accurate measurement with eyeball and clumsily used caliper. The distance of each
crank from its respective chain stay is quite different. In this case, the drive side crank at the
pedal end is quite a bit further from the stay than the one on the non-drive side. So you think xtr
bb not only evens this out but also will allow me to move drive side in to improve chain line and
shifting??? This frame is a TST ti frame with a
1.250 seat stay.

Thanks again. It's challenging to figure out who to believe and how to approach trouble shooting.

I really appreciate your response to my question. I have now spoken to a tech at Shimano twice
asking about updated middle chain ring. He said the only frames that these do anything for are Treks
and those are the only ones covered under warranty. He said the new ring fixed a derailment problem
and would not address the ghost shifting problem. What brand of frame was your DA triple on?

I will definitely check out your idea re: xtr bb. Would any shimano bb do the same thing -- or
something special about xtr?

It seems like moving the middle and small ring in a bit would do the trick. What do you think
about a spacer under the middle ring bolts to bring both the middle and small cog inward? Seems
the front d has the necessary range to cover a couple of extra mm of inward movement onto middle
and small ring.

[email protected] (Peter Headland) wrote in message
news:<[email protected]>...
> "They all do that sir."
>
> Actually, they don't anymore, but early production D-A triples had a strong tendency to do exactly
> what you describe until the chain had worn in a bit (500 miles), and even then, the biggest
> sprocket/middle ring combo was always "iffy". Shimano later introduced a modified middle ring that
> seems to eliminate the problem (I tested it pretty carefully with a brand-new chain when I
> replaced my original middle ring).
>
> So, your first port of call is back to your dealer, to ask him to get you the new type middle ring
> (I am not sure there is any identifying code on the rings that tells you which version is which,
> but there are some clear differences to some of the tooth profiles if you have one of each type in
> your hand).
>
> 1mm difference in chainline won't make any difference - the sprocket spacing is way more than
> that. However, there is something you can do to improve the chainline. I fitted an XTR 116.5mm
> bottom bracket instead of the standard 118.5 D-A triple bottom bracket. You might think that would
> only make 1mm difference each side, but you would be wrong! Shimano's road BB's are offset to the
> right, the MTB ones are symmetrical. Fitting the XTR BB got me about 3mm difference in chainline,
> which fixed the ghost shifting problem on the 2nd sprocket (the 1st sprocket was still a problem)
> before I got the new middle ring.
>
> The "value" of switching to the XTR BB is that you no longer have your feet offset to the right by
> about 3.3mm (6.5mm relative difference - I felt I was slightly twisted in the saddle). A lot of
> modern cranksets (especially triples) have this sort of offset - I think it's due to the silly
> obsession with Q factor at the expense of all else. I'd rather have a few mm more Q and symmetry,
> personally, but YMMV.
>
> Note that the XTR BB comes with various spacers (so it fits both 73mm and 68mm BB shells and can
> be offset to handle fat seat tubes) - you *must* use the two wide ones (one each side) on a road
> frame (which has a 68mm BB shell). I actually used one wide and one narrow spacer on the drive
> side, because my frame needed the extra 1mm clearance. Don't worry about this - the XTR cups are
> meant to be used this way in a 68mm BB shell and are extra-deep to handle this. You even get a
> slight benefit in the form of wider bearing spacing.
>
> I found front shifting slightly better with the new chainline (but you do have to adjust the
> vertical position of the FD slightly, as well as adjusting the cable and limit screws).

 

> the drive side crank at the pedal end is quite a bit further from the stay than the one on the
> non-drive side.

That's normal for the D-A triple (and various other triples on the market).

> So you think xtr bb not only evens this out

I don't think, I know. I have this setup fitted on my bike. I measured before and after.

> but also will allow me to move drive side in to improve chain line and shifting???

See above.

> This frame is a TST ti frame with a
> 1.250 seat stay.

But how long are the chainstays (measure centre of the crank to centre of the axle)?

> I have now spoken to a tech at Shimano twice asking about updated middle chain ring. He said the
> only frames that these do anything for are Treks

Gosh, I wonder how the chainring knows what brand of frame it is fitted to? If he had said that this
fix only works for a certain range of frame geometries, and maybe only for fairly stiff frames, I
would think that reasonable, but this "Trek only" story smells like BS.

> and those are the only ones covered under warranty.

Yes, well Trek has enough "clout" to demand a solution. If your frame is within Shimano's technical
specifications for this drivetrain (chainstay length, chainstay to seat tube angle, seat tube
diameter), then Shimano should stand by their product and make it work (or refund your money). If
your frame is outside their parameters, all bets are off.

> He said the new ring fixed a derailment problem and would not address the ghost shifting problem.

I must be thick - I don't understand the difference between "derailment" and "ghost shifting". What
you described is exactly what I experienced - the chain shifted down to the small ring (call that
"derailment", if you want), and then either shifted back up or rubbed horribly (because the FD was
still positioned for the middle ring).

> What brand of frame was your DA triple on?

Trek 5500. But see above - any frame with similar geometry should benefit from the fix, IMO.

> I will definitely check out your idea re: xtr bb. Would any shimano bb do the same thing -- or
> something special about xtr?

Only XTR 116.5 will do. The other MTB BBs have a different spline. All the road triple BBs have the
same offset as the one you already have.

> It seems like moving the middle and small ring in a bit would do the trick.

I estimate that, to eliminate the problem without chainging the middle ring, you would have to move
the chainline in about 8mm. Even if your frame had that much clearance (which it doesn't), that
would mess up the chainline from the small and middle rings to the small sprockets.

> What do you think about a spacer under the middle ring bolts to bring both the middle and small
> cog inward?

A small spacer (0.5mm - 1mm) won't help. Even a small spacer may spoil shifting and/or cause
chainsuck. Any significant extra space where you describe would give the risk of dropping the chain
in between the big and middle rings (and you *really* don't want that to happen). Plus you'll run
out of thread on the chainring bolts and end up with a severely weakened crankset. Don't go there!

> Seems the front d has the necessary range to cover a couple of extra mm of inward movement onto
> middle and small ring.

It does. These FDs are designed to work on a variety of seat tube diameters, so they are quite
tolerant of lateral displacement relative to the rings.

 

Thanks Peter for your detailed response. I would like to be prudent in how I spend my money to
determine next steps. It seem like first replacing used chain with new and if that does not work to
try bb replacement.

Should a good shop or frame builder, I have spoken to Mikkelsen here in the SF Bay area, be able to
do the measurements to determine if the frame is within Shimano specs?

Your response about "does the derailer know which frame it is on" makes sense and does suggest bs
from Shimano.

And as it relates to the difference between ghost shifting and derailment, the tech at Shimano said
the Trek problem was that somehow the chain would be derailed onto the bb rather than ghost
shifting--going up and down on its own from middle to small chain ring.

Still curious what frame your da triple was on???

Thanks for engaging in this dialogue.

[email protected] (Peter Headland) wrote in message
news:<[email protected]>...
> > the drive side crank at the pedal end is quite a bit further from the stay than the one on the
> > non-drive side.
>
> That's normal for the D-A triple (and various other triples on the market).
>
> > So you think xtr bb not only evens this out
>
> I don't think, I know. I have this setup fitted on my bike. I measured before and after.
>
> > but also will allow me to move drive side in to improve chain line and shifting???
>
> See above.
>
> > This frame is a TST ti frame with a
> > 1.250 seat stay.
>
> But how long are the chainstays (measure centre of the crank to centre of the axle)?
>
> > I have now spoken to a tech at Shimano twice asking about updated middle chain ring. He said the
> > only frames that these do anything for are Treks
>
> Gosh, I wonder how the chainring knows what brand of frame it is fitted to? If he had said that
> this fix only works for a certain range of frame geometries, and maybe only for fairly stiff
> frames, I would think that reasonable, but this "Trek only" story smells like BS.
>
> > and those are the only ones covered under warranty.
>
> Yes, well Trek has enough "clout" to demand a solution. If your frame is within Shimano's
> technical specifications for this drivetrain (chainstay length, chainstay to seat tube angle, seat
> tube diameter), then Shimano should stand by their product and make it work (or refund your
> money). If your frame is outside their parameters, all bets are off.
>
> > He said the new ring fixed a derailment problem and would not address the ghost shifting
> > problem.
>
> I must be thick - I don't understand the difference between "derailment" and "ghost shifting".
> What you described is exactly what I experienced - the chain shifted down to the small ring (call
> that "derailment", if you want), and then either shifted back up or rubbed horribly (because the
> FD was still positioned for the middle ring).
>
> > What brand of frame was your DA triple on?
>
> Trek 5500. But see above - any frame with similar geometry should benefit from the fix, IMO.
>
> > I will definitely check out your idea re: xtr bb. Would any shimano bb do the same thing -- or
> > something special about xtr?
>
> Only XTR 116.5 will do. The other MTB BBs have a different spline. All the road triple BBs have
> the same offset as the one you already have.
>
> > It seems like moving the middle and small ring in a bit would do the trick.
>
> I estimate that, to eliminate the problem without chainging the middle ring, you would have to
> move the chainline in about 8mm. Even if your frame had that much clearance (which it doesn't),
> that would mess up the chainline from the small and middle rings to the small sprockets.
>
> > What do you think about a spacer under the middle ring bolts to bring both the middle and small
> > cog inward?
>
> A small spacer (0.5mm - 1mm) won't help. Even a small spacer may spoil shifting and/or cause
> chainsuck. Any significant extra space where you describe would give the risk of dropping the
> chain in between the big and middle rings (and you *really* don't want that to happen). Plus
> you'll run out of thread on the chainring bolts and end up with a severely weakened crankset.
> Don't go there!
>
> > Seems the front d has the necessary range to cover a couple of extra mm of inward movement onto
> > middle and small ring.
>
> It does. These FDs are designed to work on a variety of seat tube diameters, so they are quite
> tolerant of lateral displacement relative to the rings.

 

> first replacing used chain with new

My experience was that the problem reduced as the chain wore. Then I put a new chain on (I rotate
chains for cleaning) and the problem was right back to square one. So, I'd expect a new chain to
make the problem worse for you.

> Should a good shop or frame builder, I have spoken to Mikkelsen here in the SF Bay area, be able
> to do the measurements to determine if the frame is within Shimano specs?

I would hope so.

> And as it relates to the difference between ghost shifting and derailment, the tech at Shimano
> said the Trek problem was that somehow the chain would be derailed onto the bb

That was not my experience. I had exactly what you have.

> Still curious what frame your da triple was on???

We already covered that:

> > > What brand of frame was your DA triple on?
> >
> > Trek 5500.

 

I have a much better set of trouble shooting options. Peter, just to let you know there is someone
on the Serotta board who works in a shop back east who totally agees with you. He sells Treks has
had this exact problem and says the new ring made the difference. He did not say anything about
changing out the bb, but agreed it was a ghost shifting problem and not a complete derailment issue.

Thanks again, Eddie

[email protected] (Peter Headland) wrote in message
news:<[email protected]>...
> > first replacing used chain with new
>
> My experience was that the problem reduced as the chain wore. Then I put a new chain on (I rotate
> chains for cleaning) and the problem was right back to square one. So, I'd expect a new chain to
> make the problem worse for you.
>
> > Should a good shop or frame builder, I have spoken to Mikkelsen here in the SF Bay area, be able
> > to do the measurements to determine if the frame is within Shimano specs?
>
> I would hope so.
>
> > And as it relates to the difference between ghost shifting and derailment, the tech at Shimano
> > said the Trek problem was that somehow the chain would be derailed onto the bb
>
> That was not my experience. I had exactly what you have.
>
> > Still curious what frame your da triple was on???
>
> We already covered that:
>
> > > > What brand of frame was your DA triple on?
> > >
> > > Trek 5500.