Mid-Drive Drivetrains for ~1kw
- Thread starter Skirmish
- Start date
Skirmish said:
Depends on how you ride. I have a 20A bbs02 (thats roughly 1000w at full power) going through a 8 speed shimano alfine. As long as you NEVER switch under load, it works fine. Past 1200w, I would not recommend it. Also, as said, one mistimed switch and the dogs on the axle inside the hub break. You can actually feel it though the drivetrain and when it happens you know that you're screwed.
The sturdiest IGH by far is the 3 speed one - any manufacturer. Sachs, SA or shimano - they all use the same internal planetary gear mechanism that works under very high loads. 180% gear rang is very narrow though, and not nearly enough for cargo or serious hill climbing.
Sachs used to make sturdy 5 speed IGH for tandems that could work, but they are hard to come by. The new nexus inter 5 is supposed to be "specially designed for ebikes" but I never tried it - certainly on my to do list though. And finally there is still the enviolo (formerly nuvinci). They have a high power model supposed to withstand up to 200n.m-2 but its price comes near a rohloff, and they are super heavy.
Linkdrive is pure marketing bs in my opinion. The limiting factor of derailleur transmissions are not the sprockets, but the strength of the chain, and that one is limited by the flexibility required for shifting. While 6-8sp derailleurs can accomodate rather wide chains 3/32" and even some flexible 1/8" anything above that will require a too thin chain for your power level.
Skirmish, I'll cut and paste this here from another currently running thread that touches on the ebike drivetrain issue somewhat.
I work at a shop and have gotten to see industry and independent testing on all manner of things that I normally wouldn't run into. This doesn't make me an expert on any of these issues...maybe only an expert reader...LOL! However, I will mention how the advent of ebikes has yielded some testing that produced some interesting results. I apologize for not quoting the sources...I didn't take notes...but I'm sure some it can be googled for first hand reading for those interested. And some of it came from warranty warnings and advice to dealers that may not readily be "out there".
On the issue of drivetrains in emtb's and bikes in general, there seemed to be a pattern of identifying the weakest link...no pun intended...in the system as far as actual breakage and failure. I didn't see any indication that claimed a 10-speed chain was stronger than an 8 or 9-speed chain, but there was no evidence that 10 or 11 chains were weaker than 8 or 9 chains. Interestingly what they did point out was that the "weakness" cropped up in the cassette as the number increased. And apparently it wasn't the cog teeth as much as the attachment points of the cogs as the cassette grew in number...spiders, pins, etc. Now, there were caveats as to brands, models, and quality of these cassettes that can further obscure the precision of these results, and I'm only throwing this out there for thought. I'm sure the "what if's" can be significant. And remember these studies, tests, or whatever were a result of drivetrains being challenge by the new electric devil...LOL!
These findings, right or wrong, did influence me to remain at 9-speed but more importantly with an all-steel cassette...11-50 in my case...and a clutched derailleur. The upside to fewer gears from my experience so far is that my MTB doesn't need all those smaller incremental gear steps anymore with the torque of the BBSHD on tap. That's not to say one should just stick it in one gear and hit the trail. However, I'm finding that all those additional cogs just aren't as necessary. In fact if you look closely, there are quite a few newer 8-speed ebike specific drivetrains out there in that 11T,12T to 46T, 48T, 50T gear numbers like mine. I think these manufacturers have "believed" in the findings of some of these tests...and maybe just the results of observed wear-and-tear on the 10-11-12 cog sets on ebikes...and/or the lack of need for so many gears in the cassette due to the added torque available. Take all this for what it's worth, but it probably means something...LOL!
I will also add this. I've noticed a rather wide span of application on these ebike forums at how we use our bikes. There are mountain bikers, cargo haulers, cruiser bikers, utility bikers, road bikers, commuters, etc., etc. I'm new to ebikes but it's apparent my somewhat hardcore mountain bike needs and desires will have little in common with a cargo bike user. This causes a difference of opinion in some cases...not necessarily because someone is wrong...but because we're speaking different languages on application. For example, I would differ a little bit with qwerkus on the chain and cassette issue. There have been some results that indicate the quality chains at least up to 10 and 11 speed are as strong as the 8 speed chains and are more than adequate for most ebike applications. It's been becoming somewhat apparent that the 11-12 speed cassettes may be the weaker link due to the design of the spiders and pins necessary for cramming so many cogs into the cassette. This has been something noted and voiced by Shimano and SRAM at least from industry reports at the shop. Now...I don't claim to be any kind of insider or have seen pages of detailed scientific and engineering testing to confirm this. I'll be they're getting it from user and dealer input due to the actual failures or accelerated wear that is occurring with ebikes. And as I mentioned earlier in this post, it may also involve the lack of real need to have this many rear cogs with the additional torque of a motor. No expert here, but I'm betting it may be a mix of the two...lack of need for so many cogs and some fragility becoming evident because of combining a motor with high count rear cog sets.
I work at a shop and have gotten to see industry and independent testing on all manner of things that I normally wouldn't run into. This doesn't make me an expert on any of these issues...maybe only an expert reader...LOL! However, I will mention how the advent of ebikes has yielded some testing that produced some interesting results. I apologize for not quoting the sources...I didn't take notes...but I'm sure some it can be googled for first hand reading for those interested. And some of it came from warranty warnings and advice to dealers that may not readily be "out there".
On the issue of drivetrains in emtb's and bikes in general, there seemed to be a pattern of identifying the weakest link...no pun intended...in the system as far as actual breakage and failure. I didn't see any indication that claimed a 10-speed chain was stronger than an 8 or 9-speed chain, but there was no evidence that 10 or 11 chains were weaker than 8 or 9 chains. Interestingly what they did point out was that the "weakness" cropped up in the cassette as the number increased. And apparently it wasn't the cog teeth as much as the attachment points of the cogs as the cassette grew in number...spiders, pins, etc. Now, there were caveats as to brands, models, and quality of these cassettes that can further obscure the precision of these results, and I'm only throwing this out there for thought. I'm sure the "what if's" can be significant. And remember these studies, tests, or whatever were a result of drivetrains being challenge by the new electric devil...LOL!
These findings, right or wrong, did influence me to remain at 9-speed but more importantly with an all-steel cassette...11-50 in my case...and a clutched derailleur. The upside to fewer gears from my experience so far is that my MTB doesn't need all those smaller incremental gear steps anymore with the torque of the BBSHD on tap. That's not to say one should just stick it in one gear and hit the trail. However, I'm finding that all those additional cogs just aren't as necessary. In fact if you look closely, there are quite a few newer 8-speed ebike specific drivetrains out there in that 11T,12T to 46T, 48T, 50T gear numbers like mine. I think these manufacturers have "believed" in the findings of some of these tests...and maybe just the results of observed wear-and-tear on the 10-11-12 cog sets on ebikes...and/or the lack of need for so many gears in the cassette due to the added torque available. Take all this for what it's worth, but it probably means something...LOL!
I will also add this. I've noticed a rather wide span of application on these ebike forums at how we use our bikes. There are mountain bikers, cargo haulers, cruiser bikers, utility bikers, road bikers, commuters, etc., etc. I'm new to ebikes but it's apparent my somewhat hardcore mountain bike needs and desires will have little in common with a cargo bike user. This causes a difference of opinion in some cases...not necessarily because someone is wrong...but because we're speaking different languages on application. For example, I would differ a little bit with qwerkus on the chain and cassette issue. There have been some results that indicate the quality chains at least up to 10 and 11 speed are as strong as the 8 speed chains and are more than adequate for most ebike applications. It's been becoming somewhat apparent that the 11-12 speed cassettes may be the weaker link due to the design of the spiders and pins necessary for cramming so many cogs into the cassette. This has been something noted and voiced by Shimano and SRAM at least from industry reports at the shop. Now...I don't claim to be any kind of insider or have seen pages of detailed scientific and engineering testing to confirm this. I'll be they're getting it from user and dealer input due to the actual failures or accelerated wear that is occurring with ebikes. And as I mentioned earlier in this post, it may also involve the lack of real need to have this many rear cogs with the additional torque of a motor. No expert here, but I'm betting it may be a mix of the two...lack of need for so many cogs and some fragility becoming evident because of combining a motor with high count rear cog sets.
Skirmish said:
Skirmish, I'll add some specific suggestions as to your question aside from long winded post...LOL!
If you're mountain biking, I'd look at the Box 3 drivetrain kit. It is a 9-speed setup with a clutched derailleur, shifter, and chain. It's very affordable. The cassette is all steel, even the spiders, pins, etc. You can get 11-46 or 11-50. Don't be put off by the warning that the kit isn't ebike compatible. It's the shifter which allows multi-shift with one push that they claim disallows it as ebike capable. The solution...just don't push the shifter all the way in. They also make the exact same kit with a one-push shifter that makes it ebike capable. I've had no problem in avoiding the extra couple of gear shifts in the multi-shifter. They are pretty far into the lever push to get to them. There is even an 8-speed package for ebike from Box.
Microshift is offering a similar package that seems a good value. Both these companies seems to be focusing on the lower cog count/extra strength gear sets to accommodate ebikes specifically and all bikes generally. Again, probably mainly MTB driven.
Again, I'm coming from strictly a more MTB venue.
TNC said:
I'd like to see the source for that one. Here the standart bike chain width taken from wikipedia bicycle chain article:
6 speed – 7.3 mm (9⁄32 in) (Shimano HG), 7.1 mm (9⁄32 in) (SRAM, Shimano IG)
7 speed – 7.3 mm (9⁄32 in) (Shimano HG), 7.1 mm (9⁄32 in) (SRAM, Shimano IG)
8 speed – 7.3 mm (9⁄32 in) (Shimano HG), 7.1 mm (9⁄32 in) (SRAM, Shimano IG)
9 speed – 6.5 to 7.0 mm (1⁄4 to 9⁄32 in) (all brands)
10 speed – 6.0 to 7.0 mm (1⁄4 to 9⁄32 in) (Shimano, Campagnolo)
10 speed (Narrow) – 5.88 mm (7⁄32 in) (Campagnolo, KMC)
10 speed (Narrow, Direction) – 5.88 mm (7⁄32 in) (Shimano CN-5700, CN-6700, CN-7900)
11 speed – 5.5 to 5.62 mm (7⁄32 to 7⁄32 in) (Campagnolo, KMC, Shimano CN-9000)
12 speed - 5.3 mm (13⁄64 in) (SRAM)
Single speed: 8mm
Using the same steel from the same manufacturer, there is no way a 6 mm wide chain will EVER be stronger than a 7.3mm chain. 6-8sp chains are best for ebikes; funny thing is that a lot of people - at least in europe- noticed this, and retrofit their super expensive bosch mid drive e mtb with 8 sp derailleurs. That's why cheapo shimano acera 8sp or deore 8sp cogs are basically sold out at least until the end of the year.
I think the future for e mtbs is integrating the gearbox into the mid drive, as kervelo ist trying to do it. If manufacturer want to keep lightweight derailleurs, they need wider hubs with space for wider chains.
qwerkus, you probably already know that the physical size is only one element of chain design, and in MTB's it's surprising to many that it isn't what makes a chain strong within the same confines of pitch at .5" for the chains I think we're mainly talking about here. For one, you mention "using the same steel" as a benchmark. That doesn't happen. The raw materials and manufacturing process for Shimano making an XTR chain and an Acera level chain are different, so that already muddies the water.
Also the wear point on these chains is mainly the bushing...well, really the interface of the bushing and the pin...and the interface of the bushing and pin are almost identical between 8-speed and 12-speed. It's the width of the roller and pin that mainly changes with cassette gear numbers. These are not the weak link...pun intended. And let's clarify that the term "bushing" is somewhat nebulous in the we usually think of a bushing as kind of an inner roller component. In most of the bike chains we're talking about here, the bushing is the inside of the side plate that the pin is riding in.
So, while I'm not trying to provide a class in chain construction and wear, it has to be pointed out that the strength of the bike chains we're talking about here...mainly derailleur designs...are not really dependent on how wide the chain is...within reason of course. And as I mentioned earlier, the level of quality in the material and manufacturing also changes this issue as many/most of the higher cog set chains are manufactured to better standards than a lot of the lower speed cog sets for the usual reasons of economy. Even this, however, is changing to some degree as some manufacturers are filling a need for ebikes with lower rear cog set numbers like 8-speed while providing stronger chains and cassettes manufactured to a higher standard rather than just relying on the fact that the chain is wider because it's 8-speed. My Box-3 setup has a very strong 9-speed chain and a steel cassette with steel spiders and pins instead of lighter alloys.
I also mentioned that one other element of requiring less incremental gear choices between gears with an ebike that also indicates some motivation to go to a lower number of rear cogs. I don't shift nearly as much with my BBSHD on my mountain bike on an actual trail as I did with pedal-only power. You still use gears of course, but the jump from one to the next just isn't as critical with this much torque. And yes...one shouldn't abuse the motor by pushing this benefit too far.
As to the source you asked for that supports chain strength or weakness based on width, cassette gear numbers, materials, and manufacturing quality, there are plenty of examples over time provided by reliable internet sources when this question is googled. I won't poo-poo Wikepedia on content, but it is often a compilation of fact, supposition, and anecdotal experience.
Also the wear point on these chains is mainly the bushing...well, really the interface of the bushing and the pin...and the interface of the bushing and pin are almost identical between 8-speed and 12-speed. It's the width of the roller and pin that mainly changes with cassette gear numbers. These are not the weak link...pun intended. And let's clarify that the term "bushing" is somewhat nebulous in the we usually think of a bushing as kind of an inner roller component. In most of the bike chains we're talking about here, the bushing is the inside of the side plate that the pin is riding in.
So, while I'm not trying to provide a class in chain construction and wear, it has to be pointed out that the strength of the bike chains we're talking about here...mainly derailleur designs...are not really dependent on how wide the chain is...within reason of course. And as I mentioned earlier, the level of quality in the material and manufacturing also changes this issue as many/most of the higher cog set chains are manufactured to better standards than a lot of the lower speed cog sets for the usual reasons of economy. Even this, however, is changing to some degree as some manufacturers are filling a need for ebikes with lower rear cog set numbers like 8-speed while providing stronger chains and cassettes manufactured to a higher standard rather than just relying on the fact that the chain is wider because it's 8-speed. My Box-3 setup has a very strong 9-speed chain and a steel cassette with steel spiders and pins instead of lighter alloys.
I also mentioned that one other element of requiring less incremental gear choices between gears with an ebike that also indicates some motivation to go to a lower number of rear cogs. I don't shift nearly as much with my BBSHD on my mountain bike on an actual trail as I did with pedal-only power. You still use gears of course, but the jump from one to the next just isn't as critical with this much torque. And yes...one shouldn't abuse the motor by pushing this benefit too far.
As to the source you asked for that supports chain strength or weakness based on width, cassette gear numbers, materials, and manufacturing quality, there are plenty of examples over time provided by reliable internet sources when this question is googled. I won't poo-poo Wikepedia on content, but it is often a compilation of fact, supposition, and anecdotal experience.
iirc plate width on a single speed chain is going to be a bit thicker, perhaps giving the interface a bit more "meat" to wear down before it affects performance?
The questions I'd have in that case would be:
Assuming no shifting under load & using whatever parts give the most km to $
Nexus inter 5-e sounds like an attractive option for being able to at least handle the lower power range: ~750w.
The questions I'd have in that case would be:
- Over the same no. of km would replacement cogs & chains cost more or less with a single-speed than a 7-9 speed?
- How does the IGH hold up compared to a 7-9 speed derailleur setup?
Assuming no shifting under load & using whatever parts give the most km to $
Nexus inter 5-e sounds like an attractive option for being able to at least handle the lower power range: ~750w.
Skirmish said:
Oh, OK...I think most of my recommendations are more directed at trail riding. You have many more options at your disposal for your application. I'm not very "up" on hub motors, but from a lot of exposure to threads about them they seem like a good option to consider for your situation.
Grantmac
10 kW
Skirmish said:
That's geared hub motor territory in my opinion.
TNC said:
I don't want to hijack this thread, so I'll keep it short. While I agree with your comment about derailleur contruction (steel sprockets are paramount!) I think your take on chains is plain BS and exactly the reason why many are so confused. Unless I'm utterly mistaken wider chains have more wearable surface, it's as simple as that. Here some reading about chain construction:
https://www.sheldonbrown.com/chains.html
https://www.sheldonbrown.com/chain-wear.html
That's the reason why a single speed chain of the same build quality will last longer than a derailleur chain, and why ebike builder should look for less gears and stronger chains.
Grantmac
10 kW
Look at the worn links on that page, it's the shoulder (bushing) which wears and that is the same for all sizes. The modern chains are also better built in terms of material and design.
Chalo
100 TW
Grantmac said:
Lower gear count, wider chains (both bushing type and bushingless) have more bearing surface between the pin and the bushing to work with. That's lower specific load and less wear.
Bushing chains are acknowledged to last much longer than bushingless (when well lubricated, which is admittedly harder to do); they just don't flex as much and thus don't shift as well as bushingless. They also don't tolerate adverse chainline (a new-hotness bike signature feature) as well as bushingless.
If you spent as much time in the shop as I do measuring and replacing chains, you wouldn't propagate nonsense like "narrower chains last longer". They really don't. They just cost more. They're made of harder material because that's the only way they can manage to work.
Chalo, I don't think there's a claim that it's the narrower chain design that is making the chain stronger. In the cases where they are proven stronger by actual testing, it's much more likely to be as a result of material and manufacturing quality. The takeaway is that it's more likely that none of these chains are necessarily weaker, and in some cases stronger, because of the chain width...at least within the chain groups we're discussing here. These are the kinds of patterns that are cropping up in many tests and wear histories in the bicycle industry, bike shops, and individual anecdotal experience.
These chains' weak link, so to speak, is the plate bushing and pin interface. Until you get to the single speed chain comparison mentioned by qwerkus, you can see that the plate bushing/pin interfaces are pretty much identical in size and relative design. Look at the many side-by-side pic comparisons of all of these chains ranging from 5-speed to 1/8" single speed chains. The rollers among all the mulit-speed chains are surprisingly not that much different in size. This is where the width as it applies to bearing surface, contact patch, or whatever is not dramatically different enough to make this wear issue the critical element. The critical element is the plate bushing and pin interface.
You can see why a single speed chain will or should have longer wear life at this location...materials and construction being of some notable level of quality...because is has more material at the pin to interact with the plate bushing...and...it indeed does have a wider roller to actually contribute to the bearing surface. It will take longer and more force to eventually cause the pin and plate to separate. Yes, chains can and do break in the middle or elsewhere in the side plates that don't involve separating at the pin and plate bushing, but it's not the common death of a chain. When those catastrophic plate breaks occur there are a few other causes usually at play that don't really apply to the discussion here for now.
qwerkus, those Sheldon Brown articles actually seem to support the wear factor having much more to do with wear at the plate bushing and pin rather than being dependent on a chain's width and bearing surface...at least in multi-speed chains. Also, some of that info may be a little old.
These chains' weak link, so to speak, is the plate bushing and pin interface. Until you get to the single speed chain comparison mentioned by qwerkus, you can see that the plate bushing/pin interfaces are pretty much identical in size and relative design. Look at the many side-by-side pic comparisons of all of these chains ranging from 5-speed to 1/8" single speed chains. The rollers among all the mulit-speed chains are surprisingly not that much different in size. This is where the width as it applies to bearing surface, contact patch, or whatever is not dramatically different enough to make this wear issue the critical element. The critical element is the plate bushing and pin interface.
You can see why a single speed chain will or should have longer wear life at this location...materials and construction being of some notable level of quality...because is has more material at the pin to interact with the plate bushing...and...it indeed does have a wider roller to actually contribute to the bearing surface. It will take longer and more force to eventually cause the pin and plate to separate. Yes, chains can and do break in the middle or elsewhere in the side plates that don't involve separating at the pin and plate bushing, but it's not the common death of a chain. When those catastrophic plate breaks occur there are a few other causes usually at play that don't really apply to the discussion here for now.
qwerkus, those Sheldon Brown articles actually seem to support the wear factor having much more to do with wear at the plate bushing and pin rather than being dependent on a chain's width and bearing surface...at least in multi-speed chains. Also, some of that info may be a little old.
Similar threads
