Author Topic: Di2 Failure mode  (Read 38214 times)

Re: Di2 Failure mode
« Reply #350 on: 02 May, 2019, 11:12:52 am »
The old style was Dura-Ace 7900. Ultegra has only ever been canbus*, although the 10 speed was orphaned fairly quickly.

(* Or what people call canbus. I've not seen been able to find any reverse engineering of what protocol or signalling is actually used. Sending power and data over two cores doesn't really match normal canbus)

jiberjaber

  • ... Fancy Pants \o/ ...
  • ACME S&M^2
Re: Di2 Failure mode
« Reply #351 on: 02 May, 2019, 11:24:32 am »
The old style was Dura-Ace 7900. Ultegra has only ever been canbus*, although the 10 speed was orphaned fairly quickly.

(* Or what people call canbus. I've not seen been able to find any reverse engineering of what protocol or signalling is actually used. Sending power and data over two cores doesn't really match normal canbus)

Yep - I found this reference to the old DA 4 core approach https://road.cc/content/blog/67385-xi2-anyone-hacking-shimanos-electronic-gears-three-peaks

Canbus does have a single wire mode (CAN-SW) There is some info on the signaling on the link I put in the previous post. (ETA: http://carltonbale.com/shimano-di2-everything-you-need-to-know/comment-page-11/#comment-134283 )
Regards,

Joergen

Re: Di2 Failure mode
« Reply #352 on: 02 May, 2019, 11:44:07 am »
Canbus does have a single wire mode (CAN-SW) There is some info on the signaling on the link I put in the previous post. (ETA: http://carltonbale.com/shimano-di2-everything-you-need-to-know/comment-page-11/#comment-134283 )

Which system is that? One of your comments references a fourth wire and separate power and shift wires. Modern Di2 only has two for everything.

Re: Di2 Failure mode
« Reply #353 on: 02 May, 2019, 12:18:14 pm »
I am about to give up on Di2 and convert the bike to mechanical gears. I built this bike in 2012 and fitted it with the then state-of-the-art Ultegra 10 speed Di2. All was well for 5 years apart from the occasional accidental cable disconnect which was easily spotted and sorted. However, I have had 3 failures since then. All occurred suddenly when out on a ride and the cause was not obvious at the roadside, without access to mains power and/or the diagnostic computer.

Fault 1 was damage to the connection box near the bars when a cable was nipped by the clamp of a Garmin mount (my fault really) but it took a while to wear through to break the wire.

Fault 2 was sudden failure of the internal junction box in the down tube. I needed the LBS to diagnose that one as a visual examination of system didn’t provide any clues.

Fault 3 was sudden battery failure the day after showing steady green and only a short 15 mile ride the day before. Shifting both front and rear stopped dead after about 7 miles the following day. The described FD followed by RD failure didn’t occur. Got home, charged battery and it all works again, but for how long?
It’s a 7 year old battery so I would expect some loss of charge.

None of these issues cost much to fix but the hassle required to diagnose the causes has made me lose confidence in the system so I’m swapping to mechanical gears now.

Di2 is an expensive luxury, and when it works it is so good that you dont even notice it. However, by its very nature any faults are likely to be more hassle and expense to solve than mechanical, and anybody buying the system has to bear this in mind.

jiberjaber

  • ... Fancy Pants \o/ ...
  • ACME S&M^2
Re: Di2 Failure mode
« Reply #354 on: 02 May, 2019, 12:55:55 pm »
Canbus does have a single wire mode (CAN-SW) There is some info on the signaling on the link I put in the previous post. (ETA: http://carltonbale.com/shimano-di2-everything-you-need-to-know/comment-page-11/#comment-134283 )

Which system is that? One of your comments references a fourth wire and separate power and shift wires. Modern Di2 only has two for everything.

Old style (as identified by you as DA 7900 I think and the link  https://road.cc/content/blog/67385-xi2-anyone-hacking-shimanos-electronic-gears-three-peaks) had 4 wires - no CANBus
New(er) style (Ultegra and later DA) has a 2 core cable implementing something similar to CAN-SW (data and power over 1 pair)

We probably got confused by the side conversation regarding can CANBus exist in a single wire form....
Regards,

Joergen

Re: Di2 Failure mode
« Reply #355 on: 02 May, 2019, 01:32:27 pm »
YOAV reports all was well for five years; no idea what the usage was, but maybe that is about the intended design life for this kind of system?

Certainly you haven't been able to buy several key parts for Ultegra 10s Di2 for several years now; for some time shimano's 'repair method' for a broken system of this type has been to replace most of it. In truth if you can't find good used parts that are compatible you would be as well off to replace the entire system, simply because a complete groupset can often be bought new for about the same amount of money as half the parts would cost if bought piecemeal.

FWIW all modern cars are equipped with fairly sophisticated electrical systems.  One of the more common failure modes is that an older vehicle develops an electrical fault which either can't be fixed or can't economically be fixed; the result is that the car goes off to the scrappies. This (presumably) reduces the average age of the vehicle fleet on the roads (which may or may not result in lower emissions....) but constitutes an appalling waste of resources.   Bicycles aren't that bad because you can usually bolt on different parts and the bike will still work; long may this continue.

Are there many 'Di2 -only' framesets?  Would you buy one?

cheers

Kim

  • Timelord
    • Fediverse
Re: Di2 Failure mode
« Reply #356 on: 02 May, 2019, 02:28:02 pm »
FWIW all modern cars are equipped with fairly sophisticated electrical systems.  One of the more common failure modes is that an older vehicle develops an electrical fault which either can't be fixed or can't economically be fixed; the result is that the car goes off to the scrappies. This (presumably) reduces the average age of the vehicle fleet on the roads (which may or may not result in lower emissions....) but constitutes an appalling waste of resources.

I'm not convinced it's affected the average age that much, simply because the rise of electronics has coincided with other technology that's reduced the effect of corrosion and so on.

But just like with Di2, the problem with car electronics isn't that they're electronics, it's that they're proprietary and not designed to be serviceable.  Back in the days when people were routinely setting points and adjusting carburettors, consumer electronics came with service manuals, and easily-replaceable discrete components were readily available.  I remember dismantling a broken television at a formative age and discovering a slot on the inside of the case containing a full schematic with in-spec voltages and oscilloscope traces.

Technology has moved on, generally trading serviceability for low manufacturing cost and increased reliability, and today's electronics is harder to service without specialist tools and skills.  But that's not the barrier that people tend to assume it is.  The real problem is usually that the electronics is encased in glued-together plastic, specialist parts may not be commercially available (or only available in prohibitively large quantities) and you can forget about obtaining protocol specs (everything's a serial bus now!) or whatever to diagnose faults without having to resort to reverse-engineering techniques.

You see this in cars too, with hard-to-repair and largely cosmetic bodywork damage writing off otherwise functional cars.  This does at least go some of the way to solving the proprietary parts problem.


Quote
Are there many 'Di2 -only' framesets?  Would you buy one?

Closest thing is probably the mid-drive motor specific framesets.  And if I wanted an e-bike I might, on the basis that with such a system the frame represents a relatively small part of the bike's cost.


The solution for bicycle electronics is for someone to come along and make an open system a selling point.  Falco did this by not requiring a proprietary battery (you give the motor DC volts on a standard commercially available connector and it gets on with it).  At some point someone will no doubt decide that they might sell a few more widgets (to the tinkerers, motor-de-restrictors, free software enthusiasts and the molsihers of unconventional cycles) if they make their groupset Arduino-compatible and hold the case together with screws.

But ultimately, proprietary electronics is no different to derailleurs/shifters with non-standard pull ratios.

Cudzoziemiec

  • Ride adventurously and stop for a brew.
Re: Di2 Failure mode
« Reply #357 on: 02 May, 2019, 02:39:31 pm »
https://www.statista.com/statistics/299951/average-age-of-cars-on-the-road-in-the-united-kingdom/
Graph shows flat trend around 7 years on the road this century.

https://www.smmt.co.uk/industry-topics/sustainability/average-vehicle-age/
Quote
The average age of a car at scrappage in 2015 reached 13.9 years, which is on a par with the 2014 performance. The lowest scrappage age, 13 years, was recorded in 2009, a result of government’s scrappage scheme.

Furthermore, the average age of a vehicle on the road has increased, from 6.8 years in 2003 to 7.8 recorded in 2015. This reflects both slower fleet renewal and the increased longevity of vehicles. This trend works against the uptake of new vehicles, which would bring greater environmental benefits. Newer vehicles also incorporate  more advanced occupant and pedestrian safety features.
Obvious self-interest in the last two sentences, but stats seem broadly in line with the above.

https://wolfstreet.com/2018/08/21/average-age-of-cars-trucks-vehicles-by-household-income-vehicle-type/

Steady increase in years on the road from the early 1970s, also longer life than UK.
Riding a concrete path through the nebulous and chaotic future.

Re: Di2 Failure mode
« Reply #358 on: 02 May, 2019, 02:45:46 pm »
The problem with cars is that we make too many of them - there are about 30 million on the road, we sell about 2.5 million a year, so if they last longer than 12 years there are more of them than drivers. I bet most cars that get sent to the scrapyard are perfectly functional, or just need a relatively minor repair that exceeds the near zero value of the car.

While we're talking wastage, what's the life expectancy of a mechanical gear cable?

Kim

  • Timelord
    • Fediverse
Re: Di2 Failure mode
« Reply #359 on: 02 May, 2019, 02:51:18 pm »
I wouldn't be surprised if the proliferation of active safety technologies (automatic braking, speed limiting, etc) lead to an increase in average car age when fewer of them get written off in crashes.

And of course electric vehicles are the elephant in the room, being mechanically and electronically simpler than combustion engine cars.  I expect there will be enough demand that keeping early models on the road (eg. by refurbishing/upgrading batteries) will make economic sense for quite some time, unless the manufacturers radically change their approach.

Re: Di2 Failure mode
« Reply #360 on: 06 May, 2019, 05:54:39 pm »
My Di2 was the first generation of the current 2 core Di2 cabling. Unusually, for Shimano, the Ultegra version came out before the Dura-Ace version. The battery is external, mounted on the down tube. Apart from being 10 speed only, it’s fine, if it works. Replacement parts such a cables and connectors are easy to get and I can even buy a new external battery on eBay for about £60. 10 speed mechs are harder to get and expensive. As I said above, I’ve simply lost confidence in the system not failing and leaving me stranded somewhere miles from home.
I am often asked, what does YOAV stand for? It stands for Yoav On A Velo

zigzag

  • unfuckwithable
Re: Di2 Failure mode
« Reply #361 on: 21 June, 2020, 11:41:33 pm »
i can only feel sorry for Emma (on her multiday adventure), but she's a fighter that doesn't give up easily: https://www.instagram.com/p/CBtodhwFKNh/?utm_source=ig_web_copy_link

"The 13 stages of a mechanical defect:
1. Shock
2. Denial
3. Hack
4. Failure
5. Innovation
6. Failure
7. Actual mechanical intervention based on google / friends’ advice
9. Failure
10. Anger
11. Grief
12. Acceptance
13. Ride singlespeed until reaching a bike shop where a qualified mechanic can run the diagnostics and new Di2 battery can be purchased

Another wonderful day for the views, the smiles, the kindness of strangers, the support of friends, 🥰 and cycling!

But no pizza 😢 and the 3 hours at the roadside checking every connection turned out to be a waste of time. Still, I finally learned how to thread through some of those elusive sh1tty little cables.

Thanks to everyone who stopped and offered to help!"

Re: Di2 Failure mode
« Reply #362 on: 22 June, 2020, 10:43:07 am »
i can only feel sorry for Emma (on her multiday adventure), but she's a fighter that doesn't give up easily: https://www.instagram.com/p/CBtodhwFKNh/?utm_source=ig_web_copy_link

"The 13 stages of a mechanical defect:
1. Shock
2. Denial
3. Hack
4. Failure
5. Innovation
6. Failure
7. Actual mechanical intervention based on google / friends’ advice
9. Failure
10. Anger
11. Grief
12. Acceptance
13. Ride singlespeed until reaching a bike shop where a qualified mechanic can run the diagnostics and new Di2 battery can be purchased

Another wonderful day for the views, the smiles, the kindness of strangers, the support of friends, 🥰 and cycling!

But no pizza 😢 and the 3 hours at the roadside checking every connection turned out to be a waste of time. Still, I finally learned how to thread through some of those elusive sh1tty little cables.

Thanks to everyone who stopped and offered to help!"

A quick bit of self-documented cheating at step 7!

Re: Di2 Failure mode
« Reply #363 on: 22 June, 2020, 12:46:03 pm »
If you adopt stage 13 from the outset you’ll have no need for DI2 ever again