Could E-CVT Save the Motor Gearbox Unit (MGU)?

[slickrock]

​

I’m starting to worry a bit about motor gearbox unit (MGU) uptake in the EMTB market. EMTBs have been a significant growing segment of the MTB space for a decade now and nothing seems to be stopping it, especially with more powerful yet compact motors and battery tech coming out as of late. We are also starting to see gearbox traction in the MTB space growing, notably in the downhill sector and if that market takes up then there might be some gearbox trickle- down into the enduro area as well. But the unification of these two technologies, as compelling as they are conceptually and making intuitive sense for EMTB (with motor and gearing closely integrated into the same sealed package), MGUs seem to be getting a rough break out of the starting gate. While well more than a handful of of MGUs were announced at Eurobike over the last 4 years, really only the Pinion MGU has taken a small foothold and major players (as in Bosch, SRAM, Shimano, etc.) are nowhere to be found.

Now, there are always growing pains for a Gen-1 product in a new product space like the Pinion MGU, but initial good faith is to put those issues aside and celebrate what a no-derailleur, belt-drive design can do for a full suspension EMTB, where the suspension is utterly opened up, without sensations of pedal kickback and other chain-whip effects hampering performance. Earlier adopters have rung these sentiments, and while I don’t have such a bike, I fully understand this as I’ve used internal geared belt drive MTBs for year now.

But I can’t quite past some glaring design characteristics with the Pinion and a few other MGU’s in the pipelines that in my mind will keep this tech from taking off, namely: it’s weight, size, and mechanical complexity. Gearboxes are mechanically complex and e-bike motors are electro-mechanically complex and what I see in the design of the Pinon MGU is basically one mechanism grafted onto the other and put in a single box. It feels like an additive solution rather than a reductive solution: IOW, more weight, more complexity, and all inside a bigger box. This theme can be seen in other MGU solutions, like the positively huge Effigear/Valeo MGU. Intradrive, another player, seems to have realized this and is trying to keep their MGU paired down by providing far less gearing than what is available in the Pinion, which is really a kind of compromise. So what I feel the MGU segment needs is a truly reductive solution that can squarely take on the weight, size, and complexity of the whole product product space. Luckily, there seems to an MGU design approach that can achieve such a feat: the E-CVT.

For some time know I’ve been follow the MGU space and took keen interest in a few companies working on continuously variable transmissions for EMTB, but first, I have to give credit to cyclingabout, who recently produced and article and video on E-CVTs, and as far as I know, seems to have coined the term in the e-bike space, though the term can be easily found in the automotive space. I suggest you watch the video to understand what E-CVT brings to table and spares me from going over the details here. But I think the biggest thing to get over is the CVT term itself, which in many circles is stigmatic, with ideas of ramping pulleys, rubbing belts, other rotating ball mechanisms that rob power in motor cars or rob rider pedaling efficiency and feel (think Nuvinci hubs). E-CVT is nothing like this. Rather, it’s like the hybrid transmission inside of the Honda Prius.

A hybrid car has two power sources: the gas-powered engine and the battery-powered electric motor. The E-CVT elegantly combines these two power sources in a simple planetary gear layout, to produce torque to drive the wheels. Replace the gas-powered engine with a carbohydrate-powered human, and you get the exact same effect with a E-CVT MGU. Now the real magic here is in the secondary motor that drives directly controls gear ratios steplessly, but also adds torque in concert with the main motor (this vlogger video at the 2:25 to 4:30 time mark goes into greater detail). Thus, no sliding pulleys, belt ramps, rotating balls, or sprag clutches in a mechanical CVT, but also no intricate, multifarious, clanking, friction-laden, and noisy sliding gears to change transmission speeds of a mechanical gearbox; rather, a much simpler, lighter and compact design, with an order of magnitude less gears. And the E-CVT design also does away with the freewheel that’s built into existing e-bike motors.

Because the gearing is completely handled with electric motor forces, the E-CVT sports some other features not available to other gearbox (or derailleurs for that matter) solutions:

1. Instantaneous and tunable torque response: the variable gearing, and hence torque response, can be electronically tuned to maximize power response, motor efficiency, or riding mode. Currently, e-bike motors do this within each gear, providing power delivery with user switchable modes (e.g. Eco, EMTB, Race), but image if these modes could also integrate gearing continuously, beyond the limits of any particular gear you happen to be in. Now this might sound a lot like automatic shifting, which currently leaves a lot to be desired (e.g. SRAM, I2, Pinion MGU) due to choosing the wrong gear or not being able to anticipate gear shifting in MTB scenarios. But having motor-driven continuous gearing sidesteps the need to force the choice into a limited number of gears and having to potentially run through the whole range before getting to the gear that makes sense for any given particular condition. The effect can drastically increase the responsiveness of the motor to rider input. For example take a look at this Eurobike video with a vlogger demoing a bike with a Revonte One. At around the 10:45 time mark, notice how quickly the bike takes off when pedal power is applied, with no pauses in shifting or marked changes in cadence.
2. Virtual Manual Shifting: Because the continuous gearing is continuous and electronic, the motor can of course provide manual-like button shifting based on rider selection like with today’s EMTBs, with the added capability to choose gear count, inter-gear range spacing, and gear skipping to suite rider styles and riding conditions. And of course the gear shifting would be instantaneous.
3. The ability to offer regenerative braking. The effect would be similar to way a Tesla operates, where pressing the accelerator speeds the car but letting off the gas slows its down, like downshifting and hence charging the battery. (Many years ago this was a strange concept compared way we drive ICE cars but now it seems second hand to BEV owners.) For EMTB, this would require a fixed rear hub to work and a spinning belt when coasting. For EMTB, it may be too strange to always pedal, except when wanting to slow down, but it may make sense for city bikes.
4. Smaller batteries: With the combination of E-CVT efficiency and possibly regenerative braking, you could live with a smaller battery without sacrificing range for given power mode. And with obvious weight savings.
5. Less maintenance. Internal electric motors themselves, tend to last, but the mechanical components of an e-bike often wear/fail first. The simpler the transmission the less likely the maintenance (grease vs. oil bath).
6. Improved miniaturization of mid-drive MGUs: there are limits to shrinking gearboxes because of the mechanical complexity, but shrinking shrinking E-CVT transmission and making them more powerful (Think TQ for size and DGI for power) is more feasible for mid-power bikes.
7. Ability to operate in reverse. An interesting parlor trick for sure, or perhaps a kind of reverse walk mode, but at least a useful feature for cargo bikes.

Now one fundamental downside of E-CVTs is the seemingly lack of a degraded mode. Currently if your EMTB motor dies, your mechanical gears are still there to get you home, but if power-fails to an E-CVT, it’s possible the gearing could just “disappear” [edit: or drop to the lowest gear] - a virtual version of losing your chain. Hopefully, as the technology evolves, there could be ways to address this issue like some kind of mechanical override, or magnetic transmission or some kind baseline reversion gear ratio mechanism, etc.

The E-CVT video cited above discusses a number of players in this space, but IMO, I feel only two of them seem to have the most promising designs:

1. Revonte One: this motor is perhaps the closest design as what I described above, but this Finnish company went into receivership 2 years ago and its IP went up for sale and was purchased, I believe, though what company bought it was not made public.
2. Villeger Dynamic: This motor sounds (suspiciously?) similar on paper to the Revonte One, but is markedly lighter. It also sports a built-in belt tensioner, which is super ingenious because gearbox and MGU transmissions are all over the place when it comes to tensioner design and placement, with some being quite crude and heavy. Unfortunately, the company has been quiet for a year now. Maybe we will see something at the upcoming Eurobike in a weeks time.

So while all of these aspects of E-CVT are promising, the current landscape looks fledgling, if not dormant. Unless a huge player like Bosch or ambitious player like DGI steps in with significant backing or product of their own, E-CVT’s may not see the light of day. The same can almost be said of MGU’s in general. If Pinion continues to be the only viable MGU player, there may not be enough market forces or competition to drive innovation needed in the space to go mainstream. E-CVTs would be a clear example of such innovation that could advance this product sector and bolster MGU adoption, which is really what us EMTB folk want in the long run.

 

[DirkWisely]

If you don't have any gearing, how can you climb in a situation where motor + human are insufficient torque? Maybe I misunderstood something.

 

[Waynemarlow]

A Planetry gear system is not really a CVT gearbox. All your murmurings of complexity of say the Pinion gearbox in truth is probably no more complex than the Revonte system where gears are still used but just in a different manner. Where the Planetry sytem wins is compactness but that same planetry system is quite poor in its ability to choose gear ratios compared to a conventional geared gearbox.

Customer resistance to new ideas in bikes ( who wants to outlay say £ 10K as a guinea pig ) is probably holding us up at the moment more than design, particularly where R&D funding is dependant on buyers wanting to buy your product.

At the end of the day a true belt driven CVT would be optimal in our use. A lot of small cars are now using it to great effect, a lot of small scooters have a very simplified version. But we as riders would have to allow the feel of constant cadence peddling together with a form of speed control rather than peddle cadence controlling the speed.

What I can't quite work out is if the end user wants to have definitive gearing, is why designers have not gone to the same way old British motor bikes used to operate, that of having the motor and gearbox as seperate units, connected via a drive belt. If we were to say have a very compact motor such as the TQ with only a minimum reduction of motor speed at the drive shaft, feeding via a belt on the left side of the bike to say a version of the latest Shimano 10 speed hub gear all mid mounted, feeding the drive belt to the rear. If you look at the TQ motor and Shimano units together, they take up a similar space to the latest motors.

 

[slickrock]

If you don't have any gearing, how can you climb in a situation where motor + human are insufficient torque? Maybe I misunderstood something.

It can be a bit hard to grasp that with a planetary E-CVT that there is actually "gearing and gear changing" happening without the usual mechanical mechanisms of having and activating/deactivating multiple sized gears for each gear ratio you want the gearbox aspect of the MGU to produce. Do take a look at the vlogger video at the 2:25 to 4:30 time mark, as I mentioned in the OP to get a better picture. There is a dedicated second motor that controls the gearing by activating the sun gear of the planetary gearing mechanism when you want to increase the gear ratio. It's effectively like increasing the diameter of the sun gear in relation to the ring gear: the larger the sun gear, the higher the gear ratio for planetary gear setups. Luckily, that control motor provides assist to the drive side as well, so along with the assist from the main motor, provides in excess of 100NM for this design.

 

[slickrock]

A Planetry gear system is not really a CVT gearbox. All your murmurings of complexity of say the Pinion gearbox in truth is probably no more complex than the Revonte system where gears are still used but just in a different manner. Where the Planetry sytem wins is compactness but that same planetry system is quite poor in its ability to choose gear ratios compared to a conventional geared gearbox. ...

Agree in the sense that typical planetary gear systems is not known for changing gear ratios; rather just for maintaining one set gear ratio. They are in important way to provide a gear ratio when radial weight concentricity is required, like in a geared hub in a Rohloff. So how does a Rohloff provide many different gear ratios? Well it has a bunch of separate planetary gear configurations for each gear ratio and uses pawls and clutches to activate and deactivate each entire planetary gear configuration:

Take one look at this and ask yourself how a single planetary gear setup of E-CVT is as complex as the orchestration of many planetary gears in a Rohloff, or for that matter the multi-axle gear network of a Pinion (which also has a planetary gear setup inside the counter axle):

Another way to look at it is to take standard e-bike motor that is only provides a single gear ratio and only controls power to the motor to control torque. Some of these motors already have planetary gears in them. By effectively adding one additional motor to drive the sun gear, you transform you e-bike motor into an E-CVT MGU (an oversimplification but you get the idea). To me this seems a more appealing design approach for motor-first companies like Bosch, Brose, and DJI than to companies starting from gear-side like SRAM and Shimano. Heck, DJI design is already halfway there with a planetary config:

...
Customer resistance to new ideas in bikes ( who wants to outlay say £ 10K as a guinea pig ) is probably holding us up at the moment more than design, particularly where R&D funding is dependant on buyers wanting to buy your product.
...

Agreed. That existing MGU's like the Pinon are so complex, it will be harder to get the cost of this technology down. That's why a big player needs to join the fray, and hopefully approach it from a different way.

...
At the end of the day a true belt driven CVT would be optimal in our use. A lot of small cars are now using it to great effect, a lot of small scooters have a very simplified version. But we as riders would have to allow the feel of constant cadence peddling together with a form of speed control rather than peddle cadence controlling the speed.
...

I would disagree a number of fronts about using belt-driven CVT, but one that stands out is that the EMTB community will have a hard time embracing any MGU that doesn't allow manual shifting, as in your words as giving up pedal cadence to control speed. E-CVT does not preclude this as the choice is yours to go automatic or manual. And it's conceptually more flexible than even that because gearing and motor power can be controlled electronically in concert in a nuanced way. Much like the various profiles of a Bosch motor, you could have profiles that are cadence-biased, work-biased (to maintain constant heart rate), torque-biased, battery-optimized biased, and power-biased.

...
What I can't quite work out is if the end user wants to have definitive gearing, is why designers have not gone to the same way old British motor bikes used to operate, that of having the motor and gearbox as seperate units, connected via a drive belt. If we were to say have a very compact motor such as the TQ with only a minimum reduction of motor speed at the drive shaft, feeding via a belt on the left side of the bike to say a version of the latest Shimano 10 speed hub gear all mid mounted, feeding the drive belt to the rear. If you look at the TQ motor and Shimano units together, they take up a similar space to the latest motors.

Shimano may actually be working on something similar to this. I guess we will see just how kludgy or elegant it will be.

 

[Waynemarlow]

I would disagree a number of fronts about using belt-driven CVT, but one that stands out is that the EMTB community will have a hard time embracing any MGU that doesn't allow manual shifting, as in your words as giving up pedal cadence to control speed. E-CVT does not preclude this as the choice is yours to go automatic or manual. And it's conceptually more flexible than even that because gearing and motor power can be controlled electronically in concert in a nuanced way. Much like the various profiles of a Bosch motor, you could have profiles that are cadence-biased, work-biased (to maintain constant heart rate), torque-biased, battery-optimized biased, and power-biased.

I agree largely but I think things maybe futher along than you think ( hopefully ). Decatholon are selling their Rockrider E-ACTV 900 which has the Owuru propulsion module for eBike by E2 Drives CVT. I think though that it is more a constant cadence and only 65Nm and 310% ratio. Its a start and it would be very interesting to take one for a spin.

 

[Straker]

Disagree on the preposition regarding the Nuvici hub re ‘feel’. Having done considerable distances on one of these in addition to multiple hub gears it’s ok to say int different….. however by not being a faux cvt it had smooth stepless performance.

Its only problem was that it was heavier than other hubs…. An issue that has dissipated with the rise of the e bike.

 

[bram.biesiekierski]

It would be interesting to see a design like Toyota's hybrid ecvt condensed down to ebike proportions. That system is ingenious the way it uses 2 electric motors and a planteary gear system to provide motor assist, EV mode, infinite gearing, reverse gear, starter motor, generator, park brake. All by simply controlling the speed and direction of 2 motors. None of that crazy belt drive nonsense of conventional cvt.

 

[slickrock]

I agree largely but I think things maybe futher along than you think ( hopefully ). Decatholon are selling their Rockrider E-ACTV 900 which has the Owuru propulsion module for eBike by E2 Drives CVT. I think though that it is more a constant cadence and only 65Nm and 310% ratio. Its a start and it would be very interesting to take one for a spin.

Yes, the Owuru is definitely in the E-CVT camp that uses the planetary/two-motor configuration as described in the OP and it was included in the CycleAbout video I linked to.

And as you said it's actually out there in the wild getting some real-world use. That said, the reasons I didn't include it in the short list:

1. It's appears intended as city bike motor instead of EMTB. Namely only 265% gear range and at best mid-power assist.
2. The motor uses 2 internal belts. This is because the motors are side-by-side instead of transverse mounted - probably to avoid patents. Belts have certain advantages (less weight instead of transfer gears, quieter, ), but 2 of them? I still have lingering reliability memories with Brose motors on my 2nd Gen Levo.
3. It uses a secondary drive pulley to drive the chain along with the crank as well. It's like both primary and secondary axles both provide output. If that's the case, why 2 belts that internally interconnect them. I'm OK with a secondary axle to provide full power with the pedal axle providing rider input, like the Effigear gearbox, which is great for high-pivot MTBS.

Curious of anyone on the forum has tried the Decathlon bike with this motor?

Disagree on the preposition regarding the Nuvici hub re ‘feel’. Having done considerable distances on one of these in addition to multiple hub gears it’s ok to say int different….. however by not being a faux cvt it had smooth stepless performance.

Its only problem was that it was heavier than other hubs…. An issue that has dissipated with the rise of the e bike.

Yes, I get this. I've used several internal gear hubs and they all feel a bit different. I recall the Alfine/Nexus hubs felt a bit vague during pedal engagement compared to the Rohloff (perhaps because of the silent hub sprag free wheel of the Shimano. And the Rohloff feeling bit vague in the 1st gear range compared to derailleur with a pawled freewheel. I tried the Nuvinci at a bike shop, but didn't use it long enough other than to develop an opinion that it had a different feel, and sensed that it was less efficient than a Rohloff. But I think all of this matters more with MTBs compared to EMTBS, where the motor mediates the pedal from wheel and so the "feel" of the motor plays more defining role on the ride of the bike than the transmission (and having an additional freewheel in the motor beside the hub freewheel doesn't help matters either).

That said, I'm feeling slightly trepidatious about potential pedal feel of a E-CVT. With 2 motors involved, and one driving gear ratios, I'm just hoping that the torque, cadence, and speed sensors are refined enough and control systems clocked fast enough to instantaneously response to pedal input so rider feels "connected" to the wheel. Normal e-bike motors of the past I feel struggled with this but with more recent editions, the ebike motor of today feels far better in this way. I would be cool if E-CVT had this out of the starting gate, but it could take a few iterations to get the pedal refinement down.

It would be interesting to see a design like Toyota's hybrid ecvt condensed down to ebike proportions. That system is ingenious the way it uses 2 electric motors and a planteary gear system to provide motor assist, EV mode, infinite gearing, reverse gear, starter motor, generator, park brake. All by simply controlling the speed and direction of 2 motors. None of that crazy belt drive nonsense of conventional cvt.

To quote the Wolf in Pulp Fiction: "Spoken like a true prodigy."

 

[BillyBob]

Enjoying this thread. Been watching the evolving MGU situation since Pinions launch at Eurobike. It did seem to me that the 2 part criticism of the Pinion bikes was a bit unfair; noisy in lowest 4 gears, and additional weight, because although valid points, in the bigger picture they seemed to take over the conversation and ignore what a big tech breakthrough this was. Been waiting to see if a big beast brand would come on board or release their own MGU and with exception of Haibike's recent pinion MGU - and that seemed a very low profile release, guess Haibike just don't get the headlines anymore in this DJI world - there hasn't been any big release. The haibike version being carbon seems to address the weight issue a bit, being lighter then the Nicolai for instance, and Rob tempted us with some pics on Insta of him testing the Haibike though no review yet.

Back at the same Eurobike that Pinion launched I thought that Brose had been suggesting they had an MGU up their sleeve, but nothing since. Perhaps a casualty of this new 'Power arms race' set in motion by DJI.

 

[BillyBob]

Also, the bike industry 'boom then bust' cycle since Covid perhaps has meant that R&D into gearboxes was abandoned or put on hold as bike brands scrambled just to survive, with warehouses filled with unsold models. The Spesh Levo Gen 4 pushed back a full year etc. So we might look back and see that the volatile bike market killed off MGU tech development for now at least.

 

[WatermanJim]

Cool thread.
More than 20 years ago I had a rolloff and set my kids bikes up with nexus 7 so they could ride hills at 5 years old.
After thinking about it for all this time, the main reasons we don't have fully integrated ebike systems is that they are expensive, hard to fix and inefficient. Even if all the other problems like weight and noise and feel were solved, you would still have a very complex and delicate mechanism that will be unfixable on the trail and cause more problems than they solve.

Also cvts are for crusing not racing. I don't think there is a top race vehicle that uses a cvt. If they were better than a gearbox then they would be used in F1, nascar, motogp, motocross, etc.

 

[bram.biesiekierski]

Cool thread.
More than 20 years ago I had a rolloff and set my kids bikes up with nexus 7 so they could ride hills at 5 years old.
After thinking about it for all this time, the main reasons we don't have fully integrated ebike systems is that they are expensive, hard to fix and inefficient. Even if all the other problems like weight and noise and feel were solved, you would still have a very complex and delicate mechanism that will be unfixable on the trail and cause more problems than they solve.

Also cvts are for crusing not racing. I don't think there is a top race vehicle that uses a cvt. If they were better than a gearbox then they would be used in F1, nascar, motogp, motocross, etetc.

CVT was used/tried in F1 during the 90s. But was banned by the FIA as it was thought it could be too competitive.

 

[Waynemarlow]

The other issue is that just how much does the extra weight of say a 10 speed wheel mounted hub actually affect the handling of the bike for 95% of riders who will never ever exploit the benefits of a mid mounted motor and gearbox ?

 

[slickrock]

Enjoying this thread. Been watching the evolving MGU situation since Pinions launch at Eurobike. It did seem to me that the 2 part criticism of the Pinion bikes was a bit unfair; noisy in lowest 4 gears, and additional weight, because although valid points, in the bigger picture they seemed to take over the conversation and ignore what a big tech breakthrough this was. Been waiting to see if a big beast brand would come on board or release their own MGU and with exception of Haibike's recent pinion MGU - and that seemed a very low profile release, guess Haibike just don't get the headlines anymore in this DJI world - there hasn't been any big release. The haibike version being carbon seems to address the weight issue a bit, being lighter then the Nicolai for instance, and Rob tempted us with some pics on Insta of him testing the Haibike though no review yet.

Back at the same Eurobike that Pinion launched I thought that Brose had been suggesting they had an MGU up their sleeve, but nothing since. Perhaps a casualty of this new 'Power arms race' set in motion by DJI.

This thread on EMTB tries to take stock of other bikes with the Pinion MGU, and Haibike is there as well, but the lack of activity on it is partially why I started this topic in the first place - to address the kind of lack of enthusiasm I would have expected otherwise with what MGU's bring to the table. The DJI product disruption sucking up all the oxygen didn't help matters either. This is why I hope DJI is following this thread - they seem unbridled enough to take the risk on the E-CVT space.

Also, the bike industry 'boom then bust' cycle since Covid perhaps has meant that R&D into gearboxes was abandoned or put on hold as bike brands scrambled just to survive, with warehouses filled with unsold models. The Spesh Levo Gen 4 pushed back a full year etc. So we might look back and see that the volatile bike market killed off MGU tech development for now at least.

Agreed. I was expecting a Cambrian Explosion of ebike tech but, but turned out to be more of implosion instead.

Cool thread.
More than 20 years ago I had a rolloff and set my kids bikes up with nexus 7 so they could ride hills at 5 years old.
After thinking about it for all this time, the main reasons we don't have fully integrated ebike systems is that they are expensive, hard to fix and inefficient. Even if all the other problems like weight and noise and feel were solved, you would still have a very complex and delicate mechanism that will be unfixable on the trail and cause more problems than they solve......

This is consistent set of arguments that support why internal-gear hubs and gearboxes have not taken hold for the last 15 years with MTB. Touring bikes?, yes. City, bikes?, yes. MTB?, no, but here come the downhill rigs now. Suddenly those gearbox issues turn into advantages.

EMTB? I don't see why not (especially with an MGU). I mean we already crossed the rubicon with putting complex, electro-mechanical systems that are "expensive and hard to fix" on our bikes. LBS's don't even try to fix ebike motors - they just swap them out under warranty. The same would be the case with an MGU. And that's why it needs to as reliable as a straight ebike motor and I feel E-CVT is closest to that.

The other issue is that just how much does the extra weight of say a 10 speed wheel mounted hub actually affect the handling of the bike for 95% of riders who will never ever exploit the benefits of a mid mounted motor and gearbox ?

Unweighting the unsprung rear-suspension and also getting rid of the chain is profound for EMTB performance. This is a huge reason why downhillers are taking on gearboxes. The suspension feels unbridled and free of kickback, chain whip, and other deleterious things that hamper suspension performance. You won't ever need an Ochain or SideKick on a belted MGU bike.

CVT was used/tried in F1 during the 90s. But was banned by the FIA as it was thought it could be too competitive.

Here's to hoping that E-CVT eventually does the same for the World E-Bike Series and UCI tourneys.

 

[BillyBob]

This thread on EMTB tries to take stock of other bikes with the Pinion MGU, and Haibike is there as well, but the lack of activity on it is partially why I started this topic in the first place - to address the kind of lack of enthusiasm I would have expected otherwise with what MGU's bring to the table. The DJI product disruption sucking up all the oxygen didn't help matters either. This is why I hope DJI is following this thread - they seem unbridled enough to take the risk on the E-CVT space.


Agreed. I was expecting a Cambrian Explosion of ebike tech but, but turned out to be more of implosion instead.


This is consistent set of arguments that support why internal-gear hubs and gearboxes have not taken hold for the last 15 years with MTB. Touring bikes?, yes. City, bikes?, yes. MTB?, no, but here come the downhill rigs now. Suddenly those gearbox issues turn into advantages.

EMTB? I don't see why not (especially with an MGU). I mean we already crossed the rubicon with putting complex, electro-mechanical systems that are "expensive and hard to fix" on or bikes. LBS's don't even try to fix ebike motors - they just swap them out under warranty. The same would be the case with an MGU. And that's why it needs to as reliable as a straight ebike motor and I feel E-CVT is closest to that.


Unweighting the unsprung rear-suspension and also getting rid of the chain is profound for EMTB performance. This is a huge reason why downhillers are taking on gearboxes. The suspension feels unbridled and free of kickback, chain whip, and other deleterious things that hamper suspension performance. You won't ever need an Ochain or SideKick on a belted MGU bike.


Here's to hoping that E-CVT eventually does the same for the World E-Bike Series and UCI tourneys.

Eurobike starts today I think. Interested to see if there are new developments in this area.

 

[DirkWisely]

Day 1 coverage video I saw only had one MGU, but it looked nice.

 

[slickrock]

Why would there be no kickback with belt + MGU?

I think it's best to distinguish between true pedal kickback and what I would call Kickback Effects or kickback sensation. My statement above was about the latter and I cover it in detail in the Anti-Pedal Kickback Tech thread. The proof is if you see any of the current downhill racers with Zerodes, Athertons, AONs, etc. that are belt and gearbox equipt rigs also use Ochains. So far, I haven't seen this, because Ochains are meant to deal with... chains. One of the greatest strengths of MGU's and gearboxes is to rid yourself of the chain (let alone the derailleur as well).

Day 1 coverage video I saw only had one MGU, but it looked nice.

Yes, the Interdrive will be there. What's interesting about that one is that it fits bikes with Shimano EP08 motors, but I'm not sure if it also uses the Shimano battery infrastructure, which if true would be cool indeed. That said, it's a multi-geared MGU, not of the E-CVT variety, but all new entries are welcome in my book. Owuru will be there as well and that's a true E-CVT. What I really want is Villeger to show up with their E-CVT with integrated belt tensioner, but not holding my breath.

 

[slickrock]

OK, so good news from Eurobike on E-CVT front: Owuru/E2 is showing off a new EMTB version of their E-CVT city motor. Other than the announcement there's not much else revealed in the segment, and it's begging for follow up segment that needs to go into its target specs, like power, torque and gear range. The bad news is that it probably won't get to market until Q4 2026.

Spoke too soon, there's a new segment that covers Owuru/E2, but still light on the details other than it looks like they are giving demo rides, which is important because pedal response/feel is really important for this emerging tech. Curious as to what version of the motor they are demoing. If anyone on this forum is also attending Eurobike, do check out these demos and report back on your experience.

 

[DirkWisely]

Their tech looks cool, but man is it bulky.

 

[slickrock]

Their tech looks cool, but man is it bulky.

Some more info on the Owuru/E2 EMTB E-CVT arriving in this YT Short. First, it looks like they ARE demoing the EMTB motor on bike, which is huge! Specs sound great with 90-150NM torque and 1KW of power. At 3.9Kg it still seems heavy, but that's 200g lighter than the Pinion, and don't forget that this will displace 470g of derailleur, 480g of cassette and 200g of chain-to-belt weight difference.

I agree with you that on center the motor looks bulky on the bike. It would be more compact if one the internal motors was transverse mounted like on the Revonte, instead of side-by-side with belts interconnecting them (this is an assumption based on the design of the city motor version). At this point though its all about exposing the tech and letting reviewers ride the demo bike is probably the best way to go about doing this.

Again, if anyone is at Eurobike, please throw a leg over this demo and report back here.

 

[slickrock]

OK, more posts from is coming through from Eurobike on Owuru Ride E-CVT. The Pinkbike article goes into detail on the operation of the of the transmission and uses the concept of a differential as an alternate way of describing how the planetary E-CVT operates, which I think is a good way to wrap your mind around how it works. The Ride will have 500% gear range which is the tidbit of info that I was waiting for, albeit I was hoping for more, but anything in the 500 range or above is good enough for EMTB.

Also the article suggests that planetary gear setup is directly driven by the motor so no belts are required, unlike the city motor version of Owuru which has 2 belts! And not to mention that the demo bike is an UNNO! I would be all in on that combination.

 

[TorAtle]

I'm not too concerned about weight or bulkyness, the proposed Owuru unit is already acceptable given what it offers.

Some of the modes it will (assumingly) offer will take getting some used to. A fixed selectable cadence will probably be nice for commuting, but the revs from the 2 motors may jump up and down depending on how much power you're asking of the system. I've always found this pretty annoying in cars like Toyota Prius, but at least here we're not dealing with a combustion engine so the noise will be lower.

Then there's the artificial gears-mode. In theory you could have as many gears as you like. For "sporty" riding, I guess this is the mode we think we will be using the most. I'm pretty sure there will be some rubber-band effect when sharply putting down power on the pedals, while the system is trying to adjust to the selected gear ratio. There will be some disconnect between what you expect the system to do and what it does.

In my dream motor-gearbox combo, I would like a dual clutch type. Instant gear changes, actual mechanical gearing. The pitch of the motor will match your cadence like we're used to with derailleurs. I'm not sure this is feasible on an emtb, though.

 

[Waynemarlow]

Quite a lot of video on the Emtb unit from about 2 min in

 

[slickrock]

Hmm I thought pedal kickback was caused by the suspension increasing the length between cassette and chain ring, yanking the chain ring backwards and kicking the pedals. And thats why something like a low engagement hub would decrease kick back. And why the ochain allows a "bending" in the chain ring backwards direction. If it is indeed caused by how I described it above, I fail to see how that would disappear by replacing the chain with a belt

You seem to say its another effect entirely which cause the kickback problem, involving the sheer amount of chain slap going on under the chain stay / in the derailleur?

To be clear, I'm not saying that rotational kickback or suspension kickback doesn't exist or that pedal kickback tech (Ochain, Sidekick) are not effective to alleviate these phenomena. What I'm saying is the other effects of a long chain, derailleur circuitousness , and loose chain tensioning contribute to chain whiplash that far outweighs the strict suspension/rotational kickback effects and that moving to belt (in that it's lighter, shorter, tighter) knocks out the stronger underlying reason why folks use Ochains in the first place (i.e. not just to alleviate strict pedal kickback). Add to this the huge suspension improvement due to the reduction of unsprung weight loss going to MGU/E-CVT ( or gearbox for that matter), there's diminishing returns using such devices on belted drivetrains.

No, back to our regularly scheduled E-CVT program.

 

[Waynemarlow]

If frame designers and engine manufacturers got their design teams together and placed the lower pivot point on or as close to the crank centre point, then the whole chain kick back and chain / rear suspension elongation disappears, losing the derailleur and sundries along the way.

Way back in the early 90’s I built a frame using the crank as the shaft for the rear suspension mounting. Back then using the first Marzocchi coil shocks which on todays terms are pretty crude shock absorbers, the suspension was sweet as. You may notice more and more EBike frames are going to a high pivot point, it’s a compromise to get nearer to that ideal suspension pivot point.

It’s only a matter of time before the motor and gearbox are used as part of the frame, much like modern cars, for mounting the suspension onto.

 

[WatermanJim]

If frame designers and engine manufacturers got their design teams together and placed the lower pivot point on or as close to the crank centre point, then the whole chain kick back and chain / rear suspension elongation disappears, losing the derailleur and sundries along the way.

Way back in the early 90’s I built a frame using the crank as the shaft for the rear suspension mounting. Back then using the first Marzocchi coil shocks which on todays terms are pretty crude shock absorbers, the suspension was sweet as. You may notice more and more EBike frames are going to a high pivot point, it’s a compromise to get nearer to that ideal suspension pivot point.

It’s only a matter of time before the motor and gearbox are used as part of the frame, much like modern cars, for mounting the suspension onto.

I think the 1999 Cove G-spot was a game changer with the concentric pivot. I had the rolloff on there for a couple years and its still set up as a single speed. A rear concentric pivot is smooth and active but does not elimate chain feedback.

I like the e-cvt idea but it sounds like in reality it doesn't feel or sound normal/natural.
Also, every automated system needs a manual backup, so unless i can pedal it when its off, i dont want it.

Lastly, I want the motor to be absolutely silent. I think its terrible that they have not made this a priority. I have 2 gen 1 levos and they are almost silent, so anything louder is a step backward imo.

 

[slickrock]

...
Some of the modes it will (assumingly) offer will take getting some used to. A fixed selectable cadence will probably be nice for commuting, but the revs from the 2 motors may jump up and down depending on how much power you're asking of the system. I've always found this pretty annoying in cars like Toyota Prius, but at least here we're not dealing with a combustion engine so the noise will be lower.

Then there's the artificial gears-mode. In theory you could have as many gears as you like. For "sporty" riding, I guess this is the mode we think we will be using the most. I'm pretty sure there will be some rubber-band effect when sharply putting down power on the pedals, while the system is trying to adjust to the selected gear ratio. There will be some disconnect between what you expect the system to do and what it does.
...

All of this is hugely important to address for the E-CVT space to succeed, and I’ve been meditating on this, so please bear with this exposition.

In the past, crude e-bike motors without torque or cadence sensors just applied power at a constant rate, even when you didn't need it, creating a huge disconnect between the rider's own sense of power application and the bike's speed. Nowadays with advanced motors, the dance between motor torque/speed and pedal power application is properly mediated to create a more natural sense of connection between the rider and bike, and the control system of the motor is responsible for this. For me, the Gen 2 Levo's power settings were more of a linear application of power enhancement to rider, which is a good example of this. The more clocked the control system and sensors, the more natural the connection to the rider senses - it feels largely like your just a much stronger rider.* That said, “you” are the control system that still manages the transmission, with shifting gears and controlling pedal cadence as you did before and so the transition to e-bikes has so far been straight forward.

However, with E-CVT this dance is more complicated because it throws in continuous gear control as well as motor speed/power control. Here again, the control system will be responsible for mediating power delivery from the rider to the bike, but this time “you” are no longer the control system that manages the transmission. What's left for the rider to control?: just human power delivery, primarily in terms of applying torque on the pedals and with some cadence preference, much like when we rely on our own pedal torque/cadence in MTB situations where there is not enough reason to change gears because your own natural torque/cadence preference is both balanced and optimized for yourself. Changing gears is just the act to allow yourself to get back to that your own torque/cadence equilibrium. Standard e-bike motors expand the range where you don't need to change gears as much because they augment the torque side of the equation of torque/speed equilibrium. This is the reason why 12 gears on your cassette is overkill on your EMTB. So E-CVT in a sense just expands this range even further, but in this case to the point where you don't need to change gears all. From this point of view, changing gears is just getting in the way from the optimal power delivery of yourself and the optimal power delivery from the motor to move your bike. No doubt the experience will feel strange at first, but going back to a standard e-bike motors may be stranger still once you are freed of gearing responsibility.

What's important is again designing the control system to be quickly and accurately responsive to keep the sense of rider input connected to bikes speed in all conditions by allowing the rider to actually participate in doing work to propel the bike. The rider needs to feel that he/she is still in control, and some secret sauce will be required here. I think this where manual shifting option could help the rider with this transition. If E-CVT were to provide a set of fixed gear ratios, being in a particular gear would just be degenerative case of what we experience now with our EMTBS: increasing pedal cadence is a consequence of keeping applied pedal torque from running out, only to be rescued by an uphill grade or shifting up gears to get that pedal torque back. But just hearing myself write this makes me want something more, like making each “gear” setting provide a level of pedal torque/cadence preference to the level of power/transmission assist, like a “training wheels” mode that gets me more comfortable with giving up gear control responsibilities.

* Bosch (and Shimano, I gather) I feel goes a step further in that it adds riding "profiles" for each power setting (rather than power levels), basically understanding that different riding conditions require different power curves, input torque/cadence, and overruns for an optimal power-assisted riding experience. Tour mode really feels different than EMTB mode which feels different than Race mode for a reason and I now prefer this over say my older Levo with 3 levels of assist. Similarly, I would expect E-CVTs to provide riding profiles, but also including gear ramping profiles that are are best suited for the part of the ride your’e on. Like hammering on the pedals off-saddle on step technical uphills? - there’s profile for that.

 

[Waynemarlow]

I think this where manual shifting option could help the rider with this transition. If E-CVT were to provide a set of fixed gear ratios, being in a particular gear would just be degenerative case of what we experience now with our EMTBS: increasing pedal cadence is a consequence of keeping applied pedal torque from running out, only to be rescued by an uphill grade or shifting up gears to get that pedal torque back

You have to be a bit careful in your use of the term " CVT" here. Is a two electric engined planetry gear system a true CVT sytem, yes and no. The advantage of using planetry gears rather than belt driven CVT technolgy, is that you can lock the gear ratio and use cadence as your control of speed with torque within that cadence as your energy input for say climbing. It maybe subtle but belt driven CVT, the simplest least mechanically advanced, will not offer that same scenario.

 

[Waynemarlow]

Owuru's CVT is real and will shake the market up a bit it looks like.

Owuru’s gearbox unit – the Pinion MGU killer?

With their new gearbox unit, Owuru unveil a stepless, fully automatic motor/gearbox unit delivering up to 150 Nm of torque and 1,000 W peak power. Here's what we thought of it on the trail.

ebike-mtb.com