The new Shimano Dura Ace groups are official now, and you’ve likely had a chance to see and read a little bit about them.  Having already had the chance to play with the pieces and ride them, this seems like a good chance to to talk about them.   It’s a massive amount of information and I can’t really decide what to leave out.  So I think the best approach is to just sit down and start typing and see where it goes.  The only thing I know for sure is that there will be a few occasions in this post where I will have to eat my words, and that this will be a long post, or 3.

I also think it’s important to point out that while I am a fan of Di2,  I have not liked 7900 and have been a harsh critic of it.  Honestly, I was not expecting to like the new 9000 group, but I have to admit after riding it I am thoroughly impressed.

Dura Ace 9000

This is the new mechanical Dura Ace.  The first thing I noticed was that Shimano skipped ‘8000’ entirely, and there are some guesses as to why but in the interest of keeping things short we’ll skip it for now.  The next thing of course is the price: MSRP is $2699, a 5% increase from 7900.  It’s completely disproportionate from the performance increase, which is honestly quite enormous.  The group shaves 77 grams overall from 7900, but the focus for this group wasn’t on shaving weight, it was on increasing performance.  There has been talk on the net about the new group, much of which focused on the 11 speed aspect and the look of the new crank.  In reality these two things are pretty much just incidentals.  The shifters are the primary focus, although every piece in the group has been re-worked and changed with some definite influence from the Dyna-sys MTB series.

Shifters

Dura Ace 9000 Shifter

Since the shifters were the primary focus it seems as good a place as any to start.  The new shifter gets ergonomic changes such as decreasing the overall volume of the hood almost back to 7800 dimensions.  Overall the shape is very close to the Di2 hood.  The rubber hood itself gets a dual density compound.  The weight drops 14 grams from 7900.  The setup gets easier with the mounting bolt and reach adjustment  moved to more accessible positions. The cable guide system is redesigned so the cables just slip into place without any struggle.   The lever blade gets more outward lean for its starting position.  The release is a 30% shorter stroke.  But the big change is something Shimano is calling Vivid Index.  This is basically a two part system and the most noticeable improvement to the lever.  The first aspect of Vivid Index is a decreasing resistance as you shift up the cogs.  Through some mechanical magic and by using a dynamic pull ratio Shimano has reduced the effort it takes to shift by a claimed 47%.  It takes roughly half the pressure to shift as previous 7900 versions, but it goes further than that as well.  Normally as you shift to larger rear cogs the resistance in the mechanism increases with each shift, so shifting from say a 12 to 13 tooth is noticeably easier than shifting from 23 to 25t.   However with the decreasing resistance the shifting on 9000 gets easier the further you move up the cassette.  So it starts with less effort than previous versions and with each shift the effort involved to move it to the next cog is less still.  Ok so maybe people are thinking, “big deal, it’s not like it was all that difficult to shift 7900“. And maybe it wasn’t, but the change is so drastic you can’t help but notice it.  The shifting effort of 9000 is so light it really can be compared to Di2.   The next thing I think some will say: “if it’s too soft it lacks that tactile racing feel”.  This leads us to the second aspect of Vivid Index, the detents.  The detents (clicks as some people like to call them.) are more defined and stronger than previous versions.  So while it takes less effort to shift, it gives more feedback as to when it hits and locks into a gear.  The combined change is dramatic.  The feeling is very ‘hydraulic’ — the best I can equate it to is a hydraulic brake on a MTB where it takes practically no effort to move the lever but there’s no question as to when it engages.

Rear Derailleur

Dura Ace 9000 Derailleur

Of the entire group the rear derailleur saw perhaps the fewest changes.  Obviously it’s quite changed aesthetically but in terms of performance not much has changed.  The pull ratio has been matched to the new shifter.  The 4mm pinch bolt location is closer to the lower parallelogram linkage to help cover the wider 11 speed range. The pivot bolt is changed and the pinch bolt mounting is cleaned up and moved a bit more out of the way.  The rear derailleur drops 8 grams of weight.

Front Derailleur

Visually the longer lever arm is the most notable

The front derailleur sheds merely one gram of weight, but like so much of the group gains a lot in performance.  The most noticeable difference is the much longer actuation lever arm.  This drops the required shift effort by a claimed 43%.  After riding it, I have no doubts about the honesty of that number.  The lever has a shorter stroke.  The front change is equally as noticeable as the rear and perhaps even more Di2-esque.  The mechanical front now adds in a secondary support screw like the one introduced on Di2 a few years ago.  That makes for a more rigid mount and in turn quicker shifting.  The inner cage also gets a replaceable plastic bushing which seems to serve no purpose other than to deaden noise, particularly when cross chaining.  One thing that will probably escape most product announcements is the inclusion of a cable mode converter.  The short throw of the shifter pulling on the long leverage arm of the derailleur needs to have some adjustability for differences in frames and their braze-on locations.  The mode converter works as a small eccentric which changes the pull ratio of the front derailleur based on its position.  The cable wraps around the converter and goes to either of two small tabs where the cable pinch bolt is located.  Put it in one location for on and rotate it to the other tab for off.  The front shifting is perhaps even far more striking of a change than the rear.  There is no arguing that Shimano has for years dominated in the front shifting arena, but with this new setup they have gone leaps and bounds beyond the previous versions.  The action is so amazingly light and precise that their is no arguing that they’ve taken back some serious ground on gains made by Di2.  The difference in front shifting between the 9000 and Di2 are actually quite minimal.

Brakes

New Dual Pivot Front Brake

The 9000 brake, which is being called Ev-control, also got a massive reworking from the ground up.  Pre-production pictures that have been floating around the net have led some to speculate that the brake is a 3 pivot, which is not the case.  The brake remains a dual pivot design, but the main bolt now keeps a fixed position and is no longer one of the pivots.  The change aligns the pivots to a mirror-image position with each other, which reduces the distance from pivot to the brake pad from previously 39mm down to 22mm.  The shorter span means stiffer arms which translates to a stiffer brake.  Shimano claims it is a 10% stronger brake with more modulation.  All told Shimano is claiming a 20% increase in braking power, but only half of that comes from the actual caliper.  The other half comes from a change in the brake lever as well as a reduction in friction in the entire system through the use of a new cable and routing method.  The brake is backwards-compatible and is confirmed to work with a 7900 lever.  Any friction where the arms meet is reduced through the use of a rolling bushing which decreases resistance more.  We tested the brake with 23mm wide rims and found it works well with the new range of wide rims.  The brake is also available in a direct mount version for aero/TT use which includes a remote quick release for the system.  The only complaint we had with the new brake was a minor one.  The main mounting bolt is a slip fit T-bolt which does not back directly to anything.  So while trying to tighten the mounting nut, the T-bolt has a tendency to rise up out of its housing.  This was solved by filling the empty space with something while installing the brake, in our case we used a plastic tire lever.  Like with it’s front shifting Dura Ace has also been the gold standard for brakes.  It’s the brake that others benchmark for comparison.  My personal opinion is that it worked so well it really didn’t need improving, but I’m certainly not going to complain about it being improved.  The overall feel of the brake is strong and well-modulating as one would expect; it’s hard to say if I noticed an increase in power or modulation but the previous one shined so well in those areas that it would be a hard thing to notice.  For me the thing I noticed the most was the increase in how smoothly it operated.

Cranks

Four Arm Crank For Both Compact and Standard Ring Sizes

The new crank is probably the most visibly noticeable change in the group.  The arm is massive, and while we don’t yet have one that we can do deflection testing with, we expect to see an increase in both stiffness and a jump in stiffness to weight ratio coming from both the increase in stiffness and the drop in weight.  In fact most of the weight savings of the group comes out of the crank and bb, which together shave 52 grams.  Thanks to their continued refinement of super stiff hollow forged chainrings, Shimano could eliminate one of the spider arms and go to a 4 arm arrangement.  The placement of the 4 arms aligns with the power delivery of  the rider’s pedal stroke, with the gap placed at the ‘dead spot’. And the stiffer rings are able to handle the longer spans of unsupported ring.  The 4 arm design does away with different spiders for compact and standard and replaces them with a proprietary bolt circle for all configurations.  Chainring options will be  50/34, 52/36, 52/38, 53/39, 54/42, and 55/42.  I still don’t know whether they’ll have a differing A-B series or if all of each size ring will be the same.  We can hope that there will be only one series, which would give the ability to mix and match combinations.  We suspect that won’t be the case but also suspect that mixing, even though not ideal, will still work just fine.   The spindle remains the same 24mm diameter while the bottom bracket itself gets smaller.  The cup shrinks but will come with a plastic sleeve that increases its outer diameter during installation so that existing Shimano BB tools will work with it.

Cassette

Carbon and Aluminium Spiders Keep Weight Down

The obvious change here is the addition of an 11th cog. Even so, the cassette only gains 3 grams in a given size.  The 5 largest cogs are titanium and the 6 smaller cogs are steel.  The two largest cogs are attached with a traditional aluminum spider.  And the next three cogs are attached to a carbon fiber spider which helps shave some of the weight.  The teeth of the cassette have also visibly changed in their orientation. Many of the teeth are rotated, leaning sideways and rounded.  From above the cassette looks more like a chipper/shredder than a bicycle cassette.  The thickness of each cog remains unchanged, while the distance between them is only very slightly decreased.  Enough so that it appears you can use an 11 speed group with a 10 speed cassette, though I can’t image why you’d want to.

Chain

New Symmetric and pfte Coated Chain

With cog thickness not changing and cog spacing not changing much, the chain didn’t really have to change much, yet it did nevertheless.  The asymmetrical chain of 7900 is no more and we’re back to a symmetrical chain.  I’m told this was possible through a combination of improvements in the cassette, rings, shifters and derailleurs.  The result being that an asymmetrical design just isn’t needed anymore.   The chain has also been pfte coated.  This is said to reduce chain noise, reduce friction and increase chain life.  The chain is expected to last longer than a 7800 chain.  The new chain also includes a new bullet-shaped connecting pin.   There is also a new chain tool which is spaced slightly narrower to hold the thinner links in place, however we suspect that while maybe not ideal, one could use an existing chain tool on the chain.

Cables

Improved Coating Resists Wear and is Sliiiiiiiiiick

Even the cables get a change.  The traditional pfte coating had a tendency to scratch off and ball up inside the housing, causing an increase in system friction.  These cables have a new coating that they say won’t scratch off.  It definitely is slick though, I kept trying to wipe the grease off my fingers after touching the cable, but couldn’t because there wasn’t any grease to wipe off, it just felt like a greased cable.

Overall the entire group is really impressive.  It seems every piece has been reworked with serious increases in performance and some touches to make the group even run with less noise.  Everything about this group just feels better.

So to come back to what I said about having to eat my words.  A couple years ago when electronic shifting started to pick up steam, I made the comment that mechanical groups were basically zombies (living dead).  The 3 manufacturers had all pushed them just about as far as they could go.  The only thing left was to shave a few grams here and there, make them slightly more aero, but that no major improvements could be expected from any of them.  As to that, I have to say I was wrong.   The new group is truly amazing and is leaps and bounds better than 7800. (That’s right, I’m not even comparing it to 7900 because 7800 was to me the pinnacle of the Dura Ace line.) And now I’ve seen that the  9000 group takes it way beyond all that have come before it.

The other words I have to eat are words that I’ve said over and over since the introduction of Di2: “I’ll never have mechanical on my own personal bikes again.”   Again I was wrong.  I won’t give up my Di2, especially not with 9070 coming, but I will certainly want to add a 9000 group to the arsenal.

In the past when building a project and helping someone to choose the right group,  I’ve traditionally said that the differences are small and that all the top shelf groups work really well.  I tell people to choose a group based on how the lever feels in your hands, the weight and the looks of the group.  Now I have to reconsider that as 9000 has a definite performance advantage over the others.

So that’s it, it’s not everything but it’s a pretty good introduction to Dura Ace 9000 and should give you a good idea of how much engineering went into this group and what I think of it.  We did have to wonder though, “why now?”  Why spend the time and resources to so massively overhaul an unfortunately dying breed?  Perhaps that’s just it though, perhaps it’s because it is dying.  While Shimano has no official plans to kill off mechanical, I think the feeling with a lot of people including the engineers at Shimano is that this could perhaps be the last high end mechanical group that Shimano produces.  And if it’s going to be the last why not end on the highest note possible.

Read More: Dura Ace Di2 and Shimano Wheels