There is a misunderstanding held by some that Di2—Shimano's revolutionary electronic shifting system—is wireless. It's not. It's only cableless. Were it wireless, it would require a lot more on-board power and your bike would need a bigger battery than the remarkably small and light one that the Di2 system requires.
So, one thing I hear sometimes from folks is, "If it's not wireless, then what's the point?" What's the big deal if you're just substituting a cable with an electronic wire?
Fair question. The answer, when it comes to triathlon in particular, is, often: Because you can shift from either the pursuit position or the bar ends.
Yes, that's true, and yes, that's nice, but, that's not in my opinion the primary benefit of Shimano's Di2 electronic shifting. The prime benefits are these: Shifting is perfect, every time and—yet more important—the front derailleur never misses a shift and trims itself automatically to normalize for the chain's placement on the rear cluster.
Now, a lot of naysayers might comment that index shifting—introduced by Shimano as a Dura Ace 6-speed product in 1984—by itself makes each shift by definition perfect. As opposed to "friction" shifting, index shifting is like (if you will) frets on a guitar. There is no necessity to know where a note is on a stringed instrument, the fret establishes exactly where the note is. You just have to place your fingers in the general vicinity of the note, just upstream from the fret.
But, if any of you are guitar, banjo, or mandolin players, you know that it's not quite that easy and, in fact, it is possible to mis-shift with an index system.
However, I'll grant you that practice (shifting a bike or playing a guitar, but in this case, let's keep it to shifting) makes perfect or, at least, nearer to perfect.
In the rear of the bike.
It's in the front of the drive train—the front derailleur—that mistakes happen, and if Shimano's electronic shifting simply functioned on the rear of the bike, that shift system would be all hat and no cattle. It's how Di2 functions around that crank that makes it special.
There are two ways to install Di2 onto a tri/TT bike, as on a road race bike: with the wires internal or external (just as with cabled bikes). But—just as with cabled bikes—there are frames built for internal Di2 wiring. This requires a different wiring harness, and a different battery mount, so, make sure you specify the bike this is going to go on before you make a purchase of a Di2 groupkit.
I installed mine on a Trek Speed Concept 9.9. Trek did an expert job of building a bike that accepts with exceptional ease Di2, and you can follow that link to read about the process of installation.
There are certain bikes in addition to the Speed Concept—Felt's DA as one example—that are built with the idea of an internally-wired Di2 kit. If the bike is not designed, from the ground up, with Di2 in mind, you might have a hard time with an internal route. Specifically, there are wiring junctions and harnesses that would require a pretty large aperture in the frame, and if the frame's not built with this in mind, one ponders whether drilling and cutting ad hoc apertures is indicated.
Otherwise, the wiring runs externally, and Shimano provides tape-like stuff to cover up the wires adjacent to the frame, to smooth the look, enhance the aerodynamics, and keep wires hidden. The advantage to having an external route is that it's easier, I would think, to troubleshoot and fix any wiring problem that might happen in the future.
That said, this is a closed system, completely waterproof, and if you use the "shrink tubes" that encompass and "hug" each wiring connection, connectivity problems will be very rare.
If you install a Di2 group, well, this is no longer bike building. Not in the traditional sense. It's helpful to have associated man skills. Like, if you were an electrician, or an appliance repair person, or a general handyman, I think you'd benefit from the interdisciplinary nature of your work. This is by no means a requirement, just, I'm used to floating through bike builds half-paying attention, on auto pilot. No can do here.
It's just like the battery in your cordless drill. You have a charging station for recharging, plugged into a 110v outlet, and when the battery is low you take out off the bike, stick it into the charging station. Ninety minutes later you're charged and ready. There is a junction on the bike with LED lights (pictured just below, which I affixed with 2-sided tape to the Speed Concept's top tube) that tell you if the battery's low.
The battery life is admirably long—many, many rides—and the shift system doesn't just stop cold if you run your battery entirely down. The front derailleur ceases to shift, but you still have the rear. It should never get to that, because it's easy to check for a battery that has less than 25 percent power. As long as you have more than 25 percent remaining, you have nothing to worry about as long as you're not immediately commencing RAAM.
The basic functions of the two derailleurs are unchanged. There are stlll hi/low limit screws, and the rear derailleur still needs to be adjusted to match the placement of the cluster on the wheel. On a typical cabled bike, you do this by adjusting the cable tension on a barrel adjuster, located at spot on the derailleur where the cable enters.
On a Di2 bike, you have one wiring junction that serves as a master control. You push a button (it's this same junction, pic adjacent, that you use to check battery life), a red light turns on, and this signals you're ready to adjust the rear derailleur. You then press the shift buttons just as you would to shift gears up or down, but the shifters are now programmed to move the derailleur in tiny increments. Move the derailleur up or down until the gnashing noise between the chain and cog disappears. Hit that junction button again, the red light goes off, presto, derailleur is now adjusted.
This is so easy, you can—and I did—perform a derailleur adjustment on the bike, while riding.
If there ever was a Shimano group that deserved to enjoy "gruppo integrity" by its customers, this is the one. And, the Di2 group I hung was entirely Di2/Dura Ace. Retaining integrity means the group is by no means inexpensive. An entire group will cost you between $4500 and $5000 complete. Keep in mind many bikes today—like the Speed Concept—come with their own brakes, so you don't need the brake calipers that come with this Di2 kit. That'll save you some coin.
Di2's absolute necessities are the front and rear derailleurs and shifters (both bar-end and pursuit), wiring kits, battery and charger. The chain, cassette and chainrings, while engineered with Di2 in mind, are not absolute Di2 requirements. None of these appear on Felt's B10, and the Di2 on that bike by all accounts functions fine.
This isn't to say you can slap anything on a Di2-shifted bike and it'll work. As with any bike, bad chainrings are bad chainrings, and the bike won't shift well. Felt made sure, when it spec'd Vision's TriMax crank on its B10, that it also spec'd Vision's best chainrings, so that the shifting was crisp and clean.
Conversely, on the rear it doesn't appear that the cassette even needs to be Shimano. Reports from the field say that SRAM cassettes shift admirably well on a Di2 bike.
The stuff you absolutely, positively need to make a Di2 bike—shifters, both derailleurs, wiring, battery and charger—will cost you between $2500 and $3000. How much of the rest of the bike you want Dura Ace determines how much of the $2000 difference between the bare essentials and complete Dura Ace you want to spend.
Please allow me a backward and sideways digression. What I do not know yet is whether you could leave off the Di2 pursuit position shifters/levers and have the system still function. You'd have an open circuit there—a plug that expects a shifter, but, no shifter for it to plug into.
Not that I think this is a wise idea, but, this shifter is $500 at retail, so, if you really wanted to do Di2 on the cheap, this is one possible omission. Of course, shifting at the pursuit position is supposed to be the big deal for triathletes—this is what makes Di2 so cool for us. Honestly, I rarely use this function, because I'm always in the aero position. I can see where time trialists might use it more often: accelerating out of the saddle from the start, and again after rounding a corner.
But, as a triathlete I don't do that, and I climb almost always while remaining in the aero position. So, I guess this feature is technique-specific. If you climb with hands on the pursuits, then the ability to shift while in the pursuits becomes a more valuable feature.
I don't know what its downside is yet. I don't know what sort of failure rate has been experienced and, by "failure rate", I mean you're just riding along and you push the button and the bike won't shift.
Anecdotally, I haven't heard of a single failure while riding. Not one. And I've asked. I can imagine, thought, that failures can be somewhat problematic on internally routed bikes. You should expect to pay more for a bum wire—both in labor and parts—than a shift cable that went bum on your traditionally shifted bike. There are three wiring kits that makes up the Di2 gruppo and none cost less than $150. I don't know if there's a lifespan and, if so, what it is. (It's certainly longer than the lifespan of a shift cable.)
My mind ran to everything that might happen to a Di2 bike. I'm recharging my battery, I throw my bike in my car and head to my race, forgetting to take the battery out of the charger and replace it on my bike. That would be unfortunate. Just know that you should never leave your battery off the bike except for the brief 90 minutes of charging, because, as noted, this is a closed and waterproof system—but only when everything (battery included) is on the bike. Once you remove the battery, now you've opened this system to the elements.
Also of concern to me is the specter of what might happen during transport to an event, putting the bike in a box for an airplane ride. A wiring junction coming undone. Is this a logical concern? Why am I afraid of this? I don't know. Maybe I'll get over it.
Look, I'm sounding a little panicky, even just listening to myself, and I know it's because Di2 is a paradigm shift (pardon the pun) in how a bike functions. Like anyone who's got two feet firmly planted in tradition, I'm suspicious of new; I'm afraid of new.
I remember vividly, in 1984, dismissing the very idea of index shifting. "If you don't know how to shift your bike," I said, "You've got bigger problems than Shimano can ever solve for you."
I said something similar when STI shifting debuted. Likewise the idea of 6 speeds on a cassette—I was perfectly happy with 5 speeds and did not see the need for a 6th. Same when Shimano moved to offer cassettes with 7 speeds. 8 speeds. And so on.
I'm slowly learning my lesson. While Shimano doesn't always recognize other companies' improvements (threadless headsets, and we can argue about chain quicklinks), the fact remains that Shimano's improvements are, always, improvements.
Ultegra Di2 will at some point debut, though it will not be compatible with Dura Ace Di2—no mixing and matching. But, you can see the trend. Pretty soon we'll have 105 electronic shifting, and some day maybe mechanical shifting will be a thing of the past.
I'm not going to get caught on the wrong side of history this time. Count me in. No more front derailleur trimming, no more chain suck, no more mis-shifts. If the art of shifting will now devolve into simply the art of pushing a button, I'm okay with that.
Di2 is one of Shimano's current strongholds in triathlon; this, and a current hegemony over the mid-price point in cabled systems (Ultegra). SRAM, however, seems to be gaining a strong position in high-end cabled (RED) and they're poised to take some of Shimano's Ultegra biz if and when they get traction on their narrative about Force (a very nice, but still rather anonymous, gruppo positioned to complete with Ultegra).
SRAM does well in tri because it did what Shimano should have, but failed, to do for one decade and maybe two: make a tri-dedicated bar-end shifter and plug-in brake lever. Ironically, Shimano now makes both, and they're both expertly engineered and executed, but, they're both a part of its electronic group.
SRAM will probably continue to eat into Shimano's market share in cabled tri bikes unless Shimano either addresses its lack of an up-to-date cabled bar-end shifter, or unless it brings electronic shifting downstream—and down-priced—so fast that can sidestep a need to make a new cabled bar-end shifter.
Electronic shifting has not ruined cabled shifting for me. I can very happily move from one to the other, just as in my garage are both a stick-shift diesel pick-up and an automatic transmission touring sedan—I drive each, I like each, each has its place.
Still, the way the front derailleur shifts and self-trims is the unexpected surprise and delight of Shimano's Di2. Front derailleur performance is what places this system ahead of any shift system now on the market.