16 ms·
Chainless electric drive system “Free Drive” for bicycles (2021)
- Lanrei 5y agoWould it still be a bicycle if the pedals aren't connected to the drive? This configuration would make it more like a scooter with a pedal generator.
- csunbird 5y agoExactly, I am not sure if they would be allowed to be in the bicycle lanes in European countries (or at least in Germany)
- mikepurvis 5y agoDo they not allow e-bikes in those? That seems surprising, given how important e-bikes and pedal-assists have been (or have been represented as being) for broadening the accessibility of cycling. A quick googling led me to this statement about Germany and e-bikes: "Insurance and license plates are required. The maximum motor output is 500 watts for e-bikes. Also, e-bike drivers must use bike lanes unless there are none, in which case they are allowed to ride on the roads." From: https://adoebike.it/en/an-ultimate-guide-to-e-bike-laws-in-europe/ https://adoebike.it/en/an-ultimate-guide-to-e-bike-laws-in-e...
- lm28469 5y agoOf course e bikes are allowed, but as you said you need an insurance + they must be capped to 25kmh and only provide power when the user is actively pedalling.
- throwaway14356 5y agoin NL we have 3 kinds rly. 1) electrical assist that is limited to matching the riders output and limited in speed. They are cosidered bicycles. 2) e-bikes that are considered mopeds and 3) electric motorcycles.
- croon 5y agoE-bikes are generally allowed as they 1) work like regular bikes. 2) only allowed if they're pedal assisted, ie not mopeds.
- Fradow 5y agoThat differs from country to country. In France, pedelecs (pedal-assist, assist is capped to 25km/h) are allowed on bike lanes, but speed bikes (not limited to 25km/h, they can reach the same speed as cars in cities, often 50km/h, and do not require pedaling) are not allowed on bike lanes. For all intent and purpose, a speed bike (or however you call it) is an electric moped.
- buovjaga 5y agoSee these: https://leva-eu.com/eu-commission-finally-confirms-series-hybrid-cycles-are-epacs-excluded-from-l-category/ https://leva-eu.com/eu-commission-finally-confirms-series-hy... https://downtown-mag.com/en/new-eu-regulation-declares-ebikes-with-serial-hybrid-system-to-bicycles/ https://downtown-mag.com/en/new-eu-regulation-declares-ebike... Quote from the second article: The EU Commission’s statement ends an almost 5-year problem that has unsettled and set back manufacturers investing in the development of series hybrid bicycles. Undoubtedly, this technology is still a niche product but manufacturers such as automotive supplier Schaeffler are increasingly investing in this technology.
- usrusr 5y agoElectric scooters are already allowed (unfortunately for cyclists...), why would adding a charge-crank change that? I'd be more concerned about how a drive system like that would subtly fail to tickle the endorphine feedback loops the way a bike does. I don't think you'd notice on a conscious level, chances are you might even think you enjoy getting button-press acceleration from the battery buffer, but riding a bike has an immediacy that is close to the walking/running evolution has wired us for. I doubt that an e-scooter fueled by an ergometer generator would come anywhere close. I'm somewhat involved with a cycling vacation business and the way a day of being exposed to those feedback loops makes everybody involved happy that business feels almost like cheating. I really doubt that "ergometer driven e-scooter" could ever come anywhere close to that. But, well, Schaeffler isn't aiming at recreational cycling at all, just at the last mile delivery industry. I could not even guess wether it would make those jobs even more miserable or not.
- masklinn 5y agoIgnoring that they're not too different from an e-scooter, these could trivially only power the cycle when the user is actively pedalling, which would be the main issue with respect to pedelec classification. Though interestingly it could also put a hard limit on upper speed, since "motor assistance" has to cutoff at 25km/h. A series hybrid bike might be considered to only work off of motor assistance (ignoring downhill).
- WelcomeShorty 5y ago"it’s an electric motorcycle with a foot-operated charging crank."
- DonHopkins 5y agoAnd you can use it to charge your OLPC XO-1! ;)
- msandford 5y agoI'd be more interested as an alternative to chain or belt drives but I'm not sure how you incorporate gearing without power electronics. Chains can be up to 98% efficient and I'd prefer not to throw away 10% just to eliminate gears.
- svnpenn 5y agoWith these, you could store energy by pedaling, even when going downhill. If you pedal downhill with a chain, that's just wasted energy.
- samstave 5y agoUnless you downshift and peddle hard to gain a lot of speed :-)
- a4isms 5y agoYou're joking, but to answer the question seriously, the problem with slamming it into your biggest gear and pedalling downhill is that wind resistance increases with the square of your velocity. When going downhill, you are going faster thanks to gravity. A small increase in speed requires a disproportionally large increase in effort. Hand-waving over the inefficiencies involved in an electric power train, you are far better off pedalling and storing the energy on the downhill, then "spending" that stored energy on the flats or even saving it for the next uphill. With respect to going downhill faster with this system, I doubt it can do anything useful in a straight line descent, but in a long, switchback descent of the type seen in the big Grand Tour races on mountain stages, cyclists need to brake into the switchbacks and accelerate out of them. Regenerative braking followed by assisted pedalling out of the corners would be a huge win.
- a4isms 5y ago> Regenerative braking followed by assisted pedalling out of the corners would be a huge win. ...Until the next ascent, where you have to haul the weight of the battery and what amounts to two electric motors uphill... Although it seems sacrilegious to imagine energy-recovery units in UCI bicycle racing, innovation is strongly influenced by the big manufacturers who are trying to sell bikes. It's unlikely there will be a UCI-sanctioned ERU any time soon, but if there was, the key to adoption would probably be the UCI minimum weight regulations. We are now at the point where high-end bikes often need weights added to meet the minimum, and if the rest of the bike ever gets light enough, it could be possible to add an ERU without compromising the total weight for climbing. But that would require breakthroughs in the culture of bicycle racing and multiple technologies.
- sandgiant 5y ago> The Free Drive system works by converting the rider’s pedaling power into electrical energy via a small generator housed between the pedals. It then delivers this energy to the rear wheel (or wheels) via cables strung inside or outside the frame of the bike, rather than sending it mechanically through a chain or belt. Excess energy created by pedaling is fed back into the battery. The end result is a power system with fewer moving parts to complicate construction. https://www.theverge.com/2021/9/2/22653697/schaeffler-free-drive-e-bike-system-chainless https://www.theverge.com/2021/9/2/22653697/schaeffler-free-d...
- ashika 5y agopedal by wire
- kazinator 5y agoNo different from how most diesel trains work.
- uoaei 5y agoBikes handle much differently from diesel trains. This may work for some but I can see it occupying some kind of uncanny valley in the riders' psyche for a while.
- Fatnino 5y agoYou ever ride in one of those bar on wheels things you find in touristy places? It's got "barstools" along both sides and the passengers/revelers are supposed to turn the pedals under their seats while the bartender/driver steers. In reality, the pedaling charges the battery somewhat but the vast majority of the battery power comes from being plugged in at a charger before the ride. You can't actually rely on a bunch of drunk sods to keep it moving.
- kazinator 5y ago> Bikes handle much differently from diesel trains. So, like, no nudging the right handlebar forward to turn right? Bummer!
- lm28469 5y ago> Networked, flexible, sustainable, and environmentally friendly Adding batteries, more waste and breaking points is now "environmentally friendly" because "electricity"
- emerged 5y ago
- infecto 5y agoElectric is generally a lot better than driving a scooter with an ice around.
- lm28469 5y agoBut a "dumb" bike is miles better than an electric one in term of ecology. Replacing a ICE scooter is an electric bike is positive, replacing a regular bike with an electric one is a net negative. The argument only works if it replaces a more polluting option. Just like a Tesla can be a net negative depending on what it replaces
- samatman 5y agoIt depends on how you do your accounting. The electric bike is more efficient than the pushbike, just as the pushbike + human is more efficient than the pushbike alone. The carbon cost of a joule of food is higher than a joule of electricity, and joules are what get you where you're going. You can make a case that it's nice exercise you needed anyway; I like bikes as well. But the claim you're making isn't obviously correct and I would say it's more false than true.
- Tostino 5y agoIt also doesn't account for reduced car use due to having a more convenient mode of transportation than a standard pushbike.
- maxerickson 5y ago
- nradov 5y agoI would like to have a bike with this drive system for doing structured training workouts outdoors. Cyclists and triathletes often do workouts involving multiple steps with a prescribed power output, for example 10 one minute intervals at 300 W with two minutes of recovery at 140 W plus 15 minutes of warm up and cool down at 170 W. It's simple to execute that workout on an indoor smart trainer in erg mode since it will dynamically adjust the resistance to hit the exact target power, but even with a game like Zwift indoor training is super boring. You can do the same workout outdoors on a real bike, but then you have to keep watching the power meter to stay in the target range which is distracting and somewhat of a safety hazard. So this system could offer the best of both worlds if the target power is externally controllable through ANT+. Of course that would probably be only a small niche market.
- user32489318 5y agoI see your point, and I can see some application in group rides, social events with a less fit SO, but 1) in TT, or even in any regular bike climbing, holding constant power is an important skill to master. Try doing a 45min varying gradient climb at 5-15% below FTP. It is not only your legs, but also your brain anticipating the gear shifts. Give the Alpe du Zwift a try, turn ERG mode off, and you'll understand what I mean. Variance in the power output wears you down. (that's why we use Normalised Power numbers for fatigue). You've to train that, not by doing3min on-1min off intervals, but by riding on real tarmac. Also, after few thousands hours of training you don't need to look at your head unit to know if your power is constant or not. 2) Power is not the only bottleneck for riding 300W, 400W, 450W for 10min, that where the handling skills of your bike come in play. That one of the reasons why we do 'speed training', cycling behind a buddy on a scooter/car. To get used to the cadence, braking, steering, ... Imagine participating in a race, riding 50+km/h without any fast cornering/breaking experience, that's a hazard! 3) Imagine doing an all-out interval session on TT bike, in a full aero position, and getting passed by an elderly on e-bike. Everyone will think "that guy is a bit weird", won't they? 4) Max. power output of this motor is 250W, I believe this to be an EU regulation. That is 20% less power on the rear wheel than the FTP of an average amateur racer. There are also speed regulations in EU, 25-28 km/h. That's certainly the Strava numbers to show off! You can make it a motorcycle/scooter and remove these limits, but then you'll need to wear motorcycle helmet in many EU countries. Also, you are no longer allowed on the bicycle lanes. 5) People who are serious enough about their cycling are generally quite serious on the bike position. For example, the Q-factor is the distance between your shoe cleats. Favero released power meter pedals for Shimano SDP-SL cleats, but added 1cm to the q-factor, just look at the flame war it caused. 6) Most people train on their 'race-day' bike. Some have a spare 'bad weather' bike. Now, you want to train on an additional training bike, with a different geometry to the race-day bike? Or are you expecting to have a TT bike with this drive train + a race day TT-bike? TT-bike will run into 5-15k USD, I might like my hobby, but won't put extra 5-15k for this nice to have feature. I think 'niche market' is an underestimation, since there are already assisted racing/MTB bikes (limited to 28km/h), just slap power meter pedals on and you're ready to go.
- yohannparis 5y agoI like innovation like this, but I will never buy one. A chain or belt will always work when the battery is dead. Or something went wrong. I need a fail-safe mechanism, and this idea does not offer one.
- a4isms 5y agoAn escalator can never break: it can only become stairs. You should never see an Escalator Temporarily Out Of Order sign, just "Escalator Temporarily Stairs. Sorry for your convenience." --Mitch Hedberg
- ryukafalz 5y agoThis is a funny quote, but for those who don’t know it’s also wrong. Escalators can indeed break, and violently: https://gizmodo.com/catastrophic-escalator-failure-in-rome-sends-crowd-spee-1829954495 https://gizmodo.com/catastrophic-escalator-failure-in-rome-s...
- Toutouxc 5y agoThere's also the fact that an escalator makes for some terrible stairs — the steps are much taller then what would be comfortable (and allowed, AFAIK) for stairs.
- nradov 5y agoConventional mechanical drive trains are very efficient, but those with multiple gears can be fragile with exposed moving parts subject to damage and misalignment. Most of us who ride a lot have experienced broken shifter cables, bent derailleur hangers, etc. Hardly fail safe. A well engineered electric drive system could potentially be more robust and reliable, but obviously more expensive and less efficient.
- yohannparis 5y agoI agree, but even the most sophisticated or rusty/bent/damage drivetrain is only unridable if a chain-link is broken and you don't have a chain-link tool to shorten it. Even with broken cables, you can select a gear by moving the chain and go forward.
- germinalphrase 5y agoDoes peddling really generate enough watts to fully power the rear wheel at comparable speed/torque level?
- messe 5y agoI don’t see how it couldn’t. I mean, the wheels spin when you pedal on a normal bike, don’t they?
- nradov 5y agoNo it's obviously less efficient than a mechanical drive. The point is that it gives more flexibility in design and energy management, and may be more mechanically robust. So the benefits can be worth the loss of efficiency for certain applications.
- germinalphrase 5y agoFair enough. It’s certainly interesting.
- Tostino 5y agoThe applications where additional batteries are not preferable to this complicated system are vanishingly slim IMO. I've seen this proposed over the years over and over, with people designing similar systems, and i've always ended up disappointed in the objective performance characteristics of them.
- user32489318 5y agoI would say that will depend on the rider and the level of pain they want to go through. Some ballpark figures: A pro-level male athlete can comfortable put out 200W without breaking a sweat. All-out effort for an elite rider (male) is upwards of 400W for 20+ min effort. (According to Strava records Van der Poel has put out 450W for 20min) A pro-level female rider can put out 300-350W continuously for an hour (and longer) in all-out effort. A semi-untrained (male) individual can put out ~150-200W on average for an hour of all-out effort. 180W is ~30km/h for an amateur rider in a comfort position on a endurance geometry racing bike with system weight of 80kg. A semi-untrained rider would consider 90-130W acceptable for a leisure ride. These figures are quite less useless, you've to put them in perspective with the system weight (bike + rider), just can't escape physics.
- bikemike026 5y agoI live on a bike. It's my primary mode of transportation. This could be a very interesting choice. Getting around is not about speed. It's about ease. If this system can get people around more easily then it's worth pursuing. There are bicycles with chains, belt drives, pedal assist electric, and throttle electric. I would be interested to know if this system could work without plugging in, get me up a steep climb, and do it cheaply. I use chain drives but belt seems to be the winner at the moment.
- Tostino 5y agoI cannot see why anyone would prefer this for a bicycle that is at all shaped like a normal bike. An integrated mid-drive ebike is just all around much more efficient. Direct hub motors for large (bike size) diameter wheels frankly suck at this point in time. You need to have much more copper to efficiently drive the motor at the power levels required, because RPM is so much lower with hub motors. Not being able to use RPM to your advantage is such a huge efficiency hog at low speed acceleration, which is a lot of how people use ebikes. If this is using an internally geared hub, my point is moot; however that comes with it's own drawbacks. It's incredibly hard to get heat out of a geared hub motor efficiently. Your motor is encased inside the hub, with no direct connection for heat to escape, and you generally have way less copper available to heat soak. Mid drives get the advantage of mechanical gearing, and can be built in such a way to allow very little heat generation to begin with, but you also can easily cool something in the frame compared to a spinning hub.
- kazinator 5y ago> I cannot see why anyone would prefer this for a bicycle that is at all shaped like a normal bike. There are already e-bikes out there with a design that puts a motor into the bottom bracket; so there are already frames for this. That's probably what is being targeted. In the absence of such frame designs existing already, this idea would be hard to pitch. The path already seems paved though. We have e-bikes with bottom bracket motors, which assist the pedals, and drive a chain. We also have bike wheel designs with a hub motor that can retrofit into ordinary bikes. This looks like it just combines the two: take a bike which has a bottom bracket motor, and replace the motor with a pure generator. Scrap the chain and sprockets, and just deliver electricity to a hub motor. Chains and sprockets get dirty and require cleaning, except in fully-closed systems that require a complicated transmission. Chains can slip and break. Chains wear out and require replacement, usually together with the rear sprockets. Front rings wear also; about once every three times you change a rear sprocket, you have to change the front rings which, for entry level bikes is usually most cheaply done by getting the entire crank set. Typically, a multi-speed rear cassette must be removed in order to replace a broken spoke, which is a PITA. This is because it is larger than the hub, and is right next to it, blocking access to the spoke entry holes on the drive side. You need chain whip to prevent the hub from turning, while you apply a wrench to a special lock ring tool, using a great deal of force. From the looks of most wheel builds with a hub motor, it looks like the spokes are easily accessible without removing any difficult part from the wheel: just get the wheel out of the bike. Internal gears alleviate some of the issues with chains. Without a derailleur system, chains can be encased to protect them from the elements. An electric transmission is going to be more efficient and quieter than internal gears, though, and require no maintenance.
- tux1968 5y agoWould be nice to have one of these pedal units installed on a footstool, so that I could charge my phone while watching a video on the couch.
- mrfusion 5y agoCould this let us put a kill switch in bikes? Maybe enforce exclusion zones?
- VectorLock 5y agoIs anybody selling these in a complete bike yet?
- botswana99 5y agoNorthern climate bike commuter here. I am getting tired of commuting in the snow and rain. I want a 'winter bike' that is fully enclosed, allows me to use my bike lane to work, and get my exercise without being covered in sleet, rain, or snow when I get home. When I was younger, the machismo of winter/rain commuting was a fun brag at parties. Now, when I look out the window after a long day at work. I want my exercise miles minus the macho. This could help me get that winter bike.
- Epa095 5y agoSounds like podbike https://www.podbike.com/ https://www.podbike.com/
- aidenn0 5y agoYou can look up velomobiles. They are quite expensive though.
- asdff 5y agoThe enclosure is a lot of mass. Easier to get the same sheltering effect with better rain/snow clothing that you can just strip off and shove in a pannier or something.
- masklinn 5y ago> The enclosure is a lot of mass. It's also a lot of sail surface.
- wolverine876 5y agoIsn't it more aerodynamic than a human on a bicycle?
- geocrasher 5y agoNo, not really. Not at the speeds most cyclists, especially commuters, travel at. You gain quite a lot more by going recumbent when it comes to drag. In fact, a recumbent bike, for the same effort, is a lot faster. Putting a fairing on one of those is another story.
- Dave_Rosenthal 5y agoI wonder how good they can make the feel without a direct connection to the road? For example, as you stand on a bike stopped at an intersection you have your foot resting on the petal, ready to take off. As you launch, you put a lot of force on that petal and rely on the feeling of connection to the ground to get going. This motor is going to have to hold that force statically and have a control system with sufficient power and bandwidth to emulate the familiar feeling of the ground. Not impossible I guess, but I wonder how well it works. Regarding efficiency, I think this is a smaller issue than it seems. Let's say a chain is 95% efficient and their system is 85% efficient (they claim 5% less efficient than a chain, apparently, but I'm sure that's a stretch). Most energy cycling, if not grinding up a steep hill, is dumped into air resistance. But speed only goes up with the cube root of power with regard to air resistance. This means (given the hypothetical numbers above) that you'd go 3.8% faster with a physical chain (or 1.7% if you believe their loss numbers). That difference is not perceptible on a bike and not an issue for 'getting around' use cases.
- bryanlarsen 5y agoA chain is 95% efficient in theory, but in practice they are usually a lot less.
- sudosysgen 5y agoIn theory a chain can be up to 99% efficient. In practice, a clean modern chain can easily hit 95%. Chains are by far the most efficient drivetrain.
- asdff 5y agoHow do belt drives stack up? I see more ebikes going belt drive which makes me think theres some efficiency gain there or at least its able to handle more torque perhaps.
- bryanlarsen 5y agoBelt drives are less efficient than a well maintained chain but more efficient than a poorly maintained chain. E-bikes are less concerned about maximal efficiency and more concerned about ease of use. Belt drives don't need cleaning and greasing like chains do.
- loonster 5y agoAn AWD bicycle would be pretty neat.
- throwawayboise 5y agoAWD motorcycles have been done. They are useful in extreme off-road scenarios. http://www.christini.com/awd-technology/about-the-tech http://www.christini.com/awd-technology/about-the-tech
- sandworm101 5y agoTake away the chain and you are on the slippery slope to electric motorcycle. This thing is clearly meant for an electric moped. Up the batter/motor power a little and the human input becomes irrelevant. You are then driving an electric motorcycle with some pedals for topping up the battery. Absent the chain-to-pedal drive, the only difference become e-bike and e-motorcycle is some arbitrary line about battery/motor capacity.
- analog31 5y agoSome of the technology choices are driven by: 1) Regulatory distinctions between bikes and motorcycles (and scooters, etc.), how they are licensed, where they can be used. 2) Providing the experience of low intensity exercise. Don't remember where I read it, but e-bikes are becoming prevalent enough in some parts of Europe, that there is a bit of a regulatory backlash going on, for instance requiring a bike to get at least 1/2 of its power from the rider.
- Steltek 5y agoe-bikes around here (Boston) seem to be mostly motorcycles with vestigal pedals. If the rider stopped pedaling, I'm not sure there would be any visible change in vehicle behavior. I may not like it but I prefer it to having more cars on the road.
- pmyteh 5y agoIt's interesting, because this exact transition (bike+motor assist -> motorbike) has been seen before. In Britain we had the category of 'moped', complete with vestigial pedals, before they were subsumed into the small motorbikes (limit: 50cc engine). The difference is that now the motor is electric, not a small petrol engine. Will be interesting to see if we end up treating them (again) as small motorbikes, or as bikes with assist (so no compulsory helmet, insurance, or registration). Edit: around here (Liverpool) I see them styled both ways. Some look like motor scooters with pedals, others pedal bikes with a motor. Seems to be fashion rather than functionality that determines which.
- seltzered_ 5y agoSee also bike2.dk as a 'chainless electric bicycle' series hybrid bicycle drivetrain - https://www.youtube.com/watch?v=4JKOKpg21dQ&t=313s https://www.youtube.com/watch?v=4JKOKpg21dQ&t=313s (2016). Like others, it didn't at first make sense to me when thinking about efficiency losses and material complexity, but later realized it might have a place in certain battery-electric delivery bike applications where the majority of the power will be coming from a pre-charged battery. Think about how one could, in a worst off case, charge the bike pedaling in standsill position in a shaded area (while reading a book or using your phone) before trying to continue using the bike.
- masklinn 5y agoIt also can make a lot of sense if you're assuming ebike, and thus that the material complexity is already 90% there (if not more, a generator is not really a complex beast after all). And between the limited maintenance and the ebike-engine, direct-drive efficiency also becomes less of a concern.
- graycat 5y agoClassic fact from my old college E&M (electricity and magnetism) course: "A series wound electric motor has infinite torque at stall" Okay, the electric connection between the pedals and the drive wheel should act as essentially as a perfect transmission, that is, with infinitely many gears (except for the 5% or whatever is lost in efficiency). And if the wheel motor is "series wound", then should have the "infinite torque" when starting from a dead stop. That infinite torque could be nice to have when going up a steep hill -- e.g., for the last 10 speed bike I had, the lowest gear was still not low enough to let me pedal up my steep driveway and, instead, I had to walk my bicycle up that hill. Also, for that bike, the highest gear was not high enough -- on the course I was using, there was a long hill, and in the highest gear before I got to the top I was pedaling as fast as I could and wanted a still higher gear. Sooooo, for something better, if the constant torque generator had a resistance adjustment, then just increase the resistance a little, let me pedal at the same RPM as before, and get up the hill faster; that is, I would be pedaling with my maximum power and the infinite gearing would move the bike at the maximum speed for that power, e.g., the power needed for the friction and air resistance for that maximum speed. Constant torque at the generator side and infinitely many gears connecting to the drive wheel -- NICE!
- Animats 5y agoIt's really an e-bike with a pedal powered generator. But it has a control system to create the illusion that the pedals are stiffly connected to the wheels. The e-bike people try too hard to pretend they're not building light-duty motorcycles. This seems to be changing. The wheels are getting smaller and stronger, and the center of gravity is going down.
- Jemm 5y agoGonna make a guess that this is not meant to be used as a source of propulsion outside the e-bike ecosystem. To lllustrate, here in Toronto, e-bike must have pedals that can be used to actually move the bike. Electric assist must be "pedal assist" and not throttle based. My guess is that the Free Drive system is intended for that purpose.
- marcodiego 5y ago> The regenerative solution is a serial hybrid drive that converts the mechanical energy generated during pedaling into electric energy, which in turn is converted back into mechanical energy in the wheel hub motor. How much efficiency is lost compared to a direct mechanical system?
- m463 5y agoThat's what I wonder too. Honestly, the efficiency of a chain drive system is really high. (however, it should match your body's efficiency - 90 is supposed to be an excellent cadence for cardio, and is good for the knees) wikipedia says up to 99%: Mechanical efficiency From a mechanical viewpoint, up to 99% of the energy delivered by the rider into the pedals is transmitted to the wheels (clean, lubricated new chain at 400 W), although the use of gearing mechanisms reduces this by 1–7% (clean, well-lubricated derailleurs), 4–12% (chain with 3-speed hubs), or 10–20% (shaft drive with 3-speed hubs). The higher efficiencies in each range are achieved at higher power levels and in direct drive (hub gears) or with large driven cogs (derailleurs). https://en.wikipedia.org/wiki/Bicycle_performance https://en.wikipedia.org/wiki/Bicycle_performance