Di2 - why wired and expensive?

[lobstermash]

I don't have Di2. I've recently 'got into' cycling after being much more utilitarian with the pedals for a few years. One of my other hobbies is remote control planes.

I understand what people love about Di2 - smooth, reliable, precise shifting every time. But am I the only person that thinks the system, wired and clunky that it is, is stupidly expensive? I know, I know, monopolised market and all that...

But some perspective, RC vehicles use radiowaves transmitted between a transmitter that sends the proportional inputs of the operator to a receiver that uses that signal to adjust servos and phase timing accordingly. The most commonly used radio frequency for RC nowadays is 2.4Ghz, which has a lot of 'channels' within it. Most RC systems hop between channels and use unique, randomly generated coding to ensure that only the signal from the actual operator is being translated into a controlling signal. These low power systems (generally between 70-200mW) are robust to interference and good for 500-2000m of range. The servos used for RC are extremely efficient at generating torque - a 40g servo can generate >4kg of torque in <0.15s, using just 5V at ~0.2A. Servos are also stupidly cheap - you get what you pay for to an extent, but one with the example stats would cost around $5-$30. As for transmitting and receiving devices, one of the most robust systems retails at ~$60 for the pair (FrSky for those who know RC). Good, reliable and safe Li-ion (and LiFePO4) batteries are also cheaply available for about 1/5 the price of a Di2 pack.

I can't see why a lighter (being wireless and lower power) and cheaper (should be achievable for <$200) system hasn't been produced... Does Shimano have a patent on the concept of electronic shifting, or is it just that nobody else has thought to come up with an alternative system?

[jules21]

i have a Di2 Ultegra groupset. the F&R mechs. and shifters are probably where most of the cost is. the electronics are extremely basic - it's just switches and servo motors. but derailleurs are precision instruments (at least mine was, until i crashed on it - still shifts well though!). the Di2 derailleurs are quite different to the mechanical versions.

having said that, there's no doubt shimano are putting a big markup on them.

[human909]

For starters prices are based on what the market will pay, NOT what the equipment costs to manufacture. It should come as no surprise to anybody but the margins on the premium products are extremely high. I'd expect at least 1000%. DuraAce and Ultegra are both high end products.

That said Shimano is top dog for a reason. If manufacturing high quality groupsets was easy we would have plenty of low cost competition. However we don't. We have SRAM which probably prices itself slightly lower. Campy prices itself itself higher. There are other big manufacturers of really cheap gear but they don't get a look in in the top tier.

I would hazard a guess that you are underestimating the R&D required and the manufacturing tooling necessary to produce high quality reliable parts. Otherwise there would be other players entering the market.

[lobstermash]

Again from an RC perspective, there are heaps of people who do open source stuff for the fun of it. Some of the most unique and powerful RC systems, like (autonomous) flight controllers and signal diversity systems are and were developed through open source. Retrofitting electronic shifting to current mechanical systems (that are certainly not easy to produce without specialised tooling) should be quite simple. I certainly see how I could do it from the perspective of fitting servos to operate the derailleurs. Simplifying a transmitter/receiver system is a little beyond my ken, but would be a no brainer for some of the open sourcers out there compared to many of the devices they've produced.

[rkelsen]

I can't see why a lighter (being wireless and lower power) and cheaper (should be achievable for <$200) system hasn't been produced.

You should do it.

[george-bob]

There have been several attempts at electronic shifting (mostly by mavic I believe) they were rubbish, I believe the most recent one was banned by the UCI as the levers were so large. So, only shimano and campy have been able to make a marketable groupset so far.

Several cost push the price up for Di2 over mechanical:

-Newness. Di2 is only on its second generation (arguably third I suppose) the manufacturing and development costs are still going to be higher than mechanical groups.

-Different manufacturing techniques: shimano mainly builds mechanical equipment, retooling (or contracting out) electronics work will be more expensive than building in house mechanical gear.

-Reliability: mechanical systems are inherently simple, electronics require connections, waterproofing, batteries etc. these are all on top of the mechanical elements of the system, making the system more complex and expensive.

-Number of components: Di2 essentially replaces ONLY the mechanical components, the running hear portion of the groupset still has the same degree of complexity as a mechanical set, i.e. most of the mechanical elements remain in place, on top of these are placed the electronic elements.

[george-bob]

I can't see why a lighter (being wireless and lower power) and cheaper (should be achievable for <$200) system hasn't been produced...

Wireless would create a whole heap of issues. You would need four batteries instead of one, each unit would require additional transceivers to be included without impacting latency. I really doubt you could include four batteries (and charging ports/hardware) for a lower weight than the wiring harness. Also, in terms of robustness, when you are a bike mechanic suddenly you need to debug a wireless system!

[DoubleSpeeded]

good point

shimano seem to be the leaders and have bought out electronic shifting.

like you said, it would probably be more affordable to some and more competition if other manufacturers produce them also.

problem is ive heard theres no manual over-ride. so if its out of battery, theres manual shifting.

a thing of the future, but it appears to add some weight

[lobstermash]

My 8 channel RC radio system, with LED backlight and 200mW transmitter, consumes about 0.1A/hour (from 12V converted to 5V, so some power wasted there). Servos (only two required) run on even less power (also 5V). This could be achieved with quite impressive endurance with quite small batteries. Thinking even bigger (concept wise), three small solar panels would have no trouble supplying power (much like the old solar powered calculators). With small backup batteries, you'd never run out of power...

RC systems don't ever need calibrating after the first time (and they use as pathetic as 8-bit processors with about 128 bytes (yes, bytes, not kb) of flash memory).

I'm no engineer, but to me, the Di2 system design seems pretty lazy (wiring purely for the sake of using a single battery) and over-complicated (pretty customised parts that over-complicate the mechanism IMO). The cynic in me says it's to patent proprietary components and price gouge for something that should be much simpler, and creates a few red herrings for competitors design-wise.

[Dimis]

Me thinks you may missed the point...

If it was that far progressed to begin with, they wouldn't be able to repeatedly fleece the consumer who has to have the latest and will pay for the privilege.

Now that comment may upset those who have gone out and purchased a Di2 system, but you just stand by and watch how upgrades to these systems will be rolled in.

Mechanical shifting for me thanks.
My bike doesn't need a green footprint. Haha

[RonK]

My 8 channel RC radio system, with LED backlight and 200mW transmitter, consumes about 0.1A/hour (from 12V converted to 5V, so some power wasted there). Servos (only two required) run on even less power (also 5V). This could be achieved with quite impressive endurance with quite small batteries. Thinking even bigger (concept wise), three small solar panels would have no trouble supplying power (much like the old solar powered calculators). With small backup batteries, you'd never run out of power...

RC systems don't ever need calibrating after the first time (and they use as pathetic as 8-bit processors with about 128 bytes (yes, bytes, not kb) of flash memory).

I'm no engineer, but to me, the Di2 system design seems pretty lazy (wiring purely for the sake of using a single battery) and over-complicated (pretty customised parts that over-complicate the mechanism IMO). The cynic in me says it's to patent proprietary components and price gouge for something that should be much simpler, and creates a few red herrings for competitors design-wise.

Why, it would be a piece of cake then, and very profitable too.

I wonder why only three manufacturers have attempted electronic systems and only two have succeeded.

Maybe they are just dumb, or plain lazy.

But here's your chance - build your own and make a fortune.

[joshnh]

What is the benefit of having a wireless system? No cables? Bikes have brake cables anyway, so I'm fine with another set of cables if that saves on additional batteries.

[rebilda]

Wireless cycle computers are sort of the normal thing now - no one uses a wired one now, eh?

The trouble is, if you go out at night or early in the morning and switch on your High Intensity LED front light, the wireless computer suddenly stops working because the frequency from the light interrupts the wireless signal.
I'd hate for that to happen to my Di2 - wireless

I have been using a Di2 bike for 2 years. The cost of the whole package (Di2 Bike, complete) was only marginally more than buying the groupset on it's own. I am still really enjoying the bike, and would buy another one in a heartbeat.

[lobstermash]

I wonder why only three manufacturers have attempted electronic systems and only two have succeeded.

Maybe they are just dumb, or plain lazy.

You look at when they attempted it... Mavic's last attempt was 1999 and Campy had a play in the early 2000s. Technology has moved on and become a lot cheaper in the last decade...

2.4Ghz, or 5.8Ghz (both freely available and easy bandwiths to access), are pretty much impervious to interference over the range of a bicycle frame. I don't think you'd have to worry turning your light on.

[lobstermash]

Just did a quick YT search. Nice to know I'm not completely mad in thinking this could be done with some RC servos and a couple of buttons. I don't rate his wiring and battery pack though... http://www.youtube.com/watch?v=xXwDmX9Ly_U

[eeksll]

I'd love to see what electronic/rc hobbyist could come up with.

I have a mate who is a electronics/rc hobbyist. I reckon he is really only making the RC cars because of his interest in electronics. We just need those guys to like cycling, but like the OP said the electronics complexity is probably so low it wouldn't really spark any interest in him.

[andione1983]

My other hobby is RC helicopters electronic gadgetry in them is amazing and $$$ just a couple of mine

Sent from my Galaxy far far away....

[RonK]

I wonder why only three manufacturers have attempted electronic systems and only two have succeeded.

Maybe they are just dumb, or plain lazy.

You look at when they attempted it... Mavic's last attempt was 1999 and Campy had a play in the early 2000s. Technology has moved on and become a lot cheaper in the last decade...

2.4Ghz, or 5.8Ghz (both freely available and easy bandwiths to access), are pretty much impervious to interference over the range of a bicycle frame. I don't think you'd have to worry turning your light on.

As H909 has already suggested, you seem to underestimate the R&D that has gone into electronic transmission systems.

" Campy had a play in the early 2000s" - well no, in fact Campy produced their first version in 1992, and since have spent 20 years developing EPS to the point where they were satisfied it was ready to release to the market. You can see the timeline here.

Yeah, maybe you can throw together a crude system with a few batteries, switches and servo motors, but that is a long way from producing a system that can gain marketing acceptance. For example, Campy put much effort into making the tactile "feel" of the buttons identical to the mechanical Ergopower levers. You may think this isn't important, but when you read reviews like this one it's not hard to understand why Campy put so much effort into getting it right.

[Duck!]

Shimano recruited a lot of technical boffins from the watchmaking industry in order to get the expertise in working with such small precision parts. Rolexes aren't cheap either.....

[warthog1]

I reckon the general premise of your assertion is correct; the level of complexity is low and the price high.

This is typical of cycling, as has been pointed out.

For starters prices are based on what the market will pay, NOT what the equipment costs to manufacture.

As far as wireless goes, why replace one battery with 4?

AFAIC cable works very well with no complexity. Why make something more complex than is necessary anyway?
That said those who use di2 all seem to like it.

[warthog1]

Rolexes aren't cheap either.....

Yes they are

[mtbkym01]

You'd need pretty strong servos to shift the rear I'd reckon, and they're not that small, and quite expensive to get the precise movements required for smooth shifting. Yes there are many cheap servo's on the market from somewhere like hobbyking for example, but to get perfect precision movements required here, youd be looking at KO Propo or similar, meaning big $$$.

I'm also an RC hobbyist, and have seen issues with 2.4 getting glitches which its not supposed to do, i've also seen systems unbind themselves from time to time, so it's not as reliable as it should (and would need to) be

JMO

[Alien27]

My 8 channel RC radio system, with LED backlight and 200mW transmitter, consumes about 0.1A/hour (from 12V converted to 5V, so some power wasted there). Servos (only two required) run on even less power (also 5V). This could be achieved with quite impressive endurance with quite small batteries. Thinking even bigger (concept wise), three small solar panels would have no trouble supplying power (much like the old solar powered calculators). With small backup batteries, you'd never run out of power...

RC systems don't ever need calibrating after the first time (and they use as pathetic as 8-bit processors with about 128 bytes (yes, bytes, not kb) of flash memory).

I'm no engineer, but to me, the Di2 system design seems pretty lazy (wiring purely for the sake of using a single battery) and over-complicated (pretty customised parts that over-complicate the mechanism IMO). The cynic in me says it's to patent proprietary components and price gouge for something that should be much simpler, and creates a few red herrings for competitors design-wise.

I cant imagine the servers and mechanical components on an RC plane are under even a hundredth of the stress the components of a group set on a bike are, nor would RC plane servers need anything like the power the ones on a group set need. The devil is always in the detail. Producing something for a hobbyist that can tinker and accept the odd failure, is an entirely different prospect to producing something that simply can not fail under just about any circumstance. The group set needs to be very robust, very strong, light, have a long battery life and be completely water and vibration proof.

I would imaging getting the last 1% waterproofing and reliability costs as much if not more than the first 99%.

I'm fairly sure shimano make their best money on their top group sets however as has been said before the electronic group sets have to have nearly all the same mechanicals as mechanical group sets, with the only difference being cables replace wire, plus there is the extra server, battery and computer components.

[lobstermash]

I cant imagine the servers and mechanical components on an RC plane are under even a hundredth of the stress the components of a group set on a bike are, nor would RC plane servers need anything like the power the ones on a group set need.

On contrare, RC helis are subject to much greater forces than a derailleur.

There are people that documented using servos quite successfully online, including Hobbyking ones. The resolution required for shifting on a derailleur is much less than most control surfaces on RC vehicles. Waterproofing servos and RC systems is also pretty easy. I actually think some of the points in one of the other Di2 threads are realistically why nobody's really bothered developing a smart, cheap system.

[Alien27]

I cant imagine the servers and mechanical components on an RC plane are under even a hundredth of the stress the components of a group set on a bike are, nor would RC plane servers need anything like the power the ones on a group set need.

On contrare, RC helis are subject to much greater forces than a derailleur.

There are people that documented using servos quite successfully online, including Hobbyking ones. The resolution required for shifting on a derailleur is much less than most control surfaces on RC vehicles. Waterproofing servos and RC systems is also pretty easy. I actually think some of the points in one of the other Di2 threads are realistically why nobody's really bothered developing a smart, cheap system.

Happy to be corrected, I have absolutely no experience with RC planes and the like. But if I change up the front derailer with the bike on the stand and no chain, using my left hand on the shifter while trying to stop the derailer moving with my other hand I'm pretty sure the shifter wins the battle. I can't imagine the server in a remote control plane being that strong, wouldn't you just break a component or stop its movement?

And I also agree with a lot of the other posts as well. Just thought the big factor is that last bit of system waterproofing and reliability and I also doubt it's any laziness on shimano's part using wires and only one battery. Just doesn't fit with their company's image or reputation.