GOBAO ECVT EMTB Mid-Drive Motor

[emtbeast]

I have wondered why Avinox doesn't adopt a higher voltage . Higher voltage equals smaller copper wire throughout the system.

Voltage vs motor size

In electrical engineering, higher voltage 52 vs. 36 does not automatically mean a motor must be larger. In fact, the opposite is often true:

Higher voltage allows a system to achieve the same amount of power (watts) using less current.

Less current means the motor can use thinner copper wire, with its windings while generating less heat, which often allows engineers to design a smaller or more lightweight motor for the same power output.

Yes I agree, should have been more specific, the other components around the motor itself are usually larger, insulation, controlers, magnet width,...

 

[emtbeast]

The only way energy could be lost from the smaller motor is through heat (friction ) or for its energy to be absorbed in helping the main motor. In simplistic terms it’s far better to think that both motors are driving the bike forward as any excess energy beyond heat will be outputted to the drive train via the output shaft.

With modern plastic and highly accurate machine gears running on quality roller bearings there is minimal efficiency loss. I would say that in a way the planetary gears may have less meshing components ( friction ) than say the Pinion gearbox.

The smaller motor doesn't help the main one it's used for creating a the electrical virtual gear ratios. So depending on the input and the chosen mode it spins and so uses electricity also just like the main motor. My thoughts on efficiency were current ebike motors with one electric motor vs the e-CVTs with two.
There is also one thing to consider in favor of the e-CVT and that’s that we loose the mechanical looses of the derailleur and cassette. Maybe that's enough to counter the additional electric motor power draw.

 

[Plummet]

Can we wait until we get some actual trail time in before making those sort of assumptions. I would doubt there will be much ride time difference between a conventional FF and one of the latest designs.

Interesting to see Gaboa go down the 52V route which has long been the go to for the experimental lot who have now moved on again to even higher voltages. 50V is generally accepted as the highest safe DC voltage so going higher maybe a tad too far.

If we waited until all was revealed then this entire thread wouldn't exist.

However, Rob did say the motor cuts out at 10%. If that is true then Immediately I need 10% more battery to do the same assisted duration or vert of riding, that is assuming zero extra losses through the cvt.

That for me is the big question. How much extra battery and weight do I need to do the same vert/hours?

 

[ArmyOfEvilRobots]

I've seen a few people conjecture that the MGU is inoperable (or stuck in the lowest gear) when the battery is fully discharged. While that _might_ be true, it would only be the case if their controls system has design flaws. In a properly designed eCVT, one motor can act like a generator while the other provides motive power; In other words, the gear ratio motor can be driven by the power generated by the other. While this means that all the actual traction power comes from your legs, you could still shift and ride the bike.

I mean, can you imagine how many Priuses we would see stranded on the side of the road if their transmission failed as soon as the traction battery ran down? They use the exact same type of transmission. That's not to say I'm looking forward to pedaling harder to charge the battery, but if it just pulls enough to keep the planetary motor turning...

 

[Astro66]

The first release of any new technology while have compromises, that will be resolved as the technology develops. It's been that way since man lived in caves.

But compared with the compromise a derailleur, cassette and chain transmission creates on an EMTB. These unknown MGU E-CVT compromises will be insignificant.

I put it to you this way. If MGU E-CVT was the normal. Could you imagine someone trying to sell you a derailleur, cassette and chain transmission on an EMTB. Could you imagine your reaction if you were stupid enough to convert over to a derailleur, cassette and chain transmission. And on the first ride you bent the cage, rending the bike imobile, requiring you to walk it home.

This exactly happen to me, on my first ride of my brand new, over $10,000 EMTB, that had the latest SRAM AXS Transmission. Hence my absolute relief that an alternative is being developed.

 

[doochenozzle]

View attachment 187329

View attachment 187332

Getting a separate thread started for this new ECVT Motor to differentiate from the Avinox Gearbox announcement.

https://www.pinkbike.com/news/first...emtb-arms-race-with-a-1500-w-150-nm-ecvt.html

There's quite a bit to unpack here and not just the motor. Will give full take later.

[Edit: Actual Review of the motor: Revolutionary Gobao X1P Gearbox Motor Review: Is This the Future of eMTB Drive Systems?]
[Edit: full take below:

The Gobao XP1 is perhaps as much a consequential introduction as the Avinox MG motor, but for different reasons. Most importantly though, both of these announcements not only inject excitement into the glacially slow uptake of mid-drive motor-gearbox bikes EMTBS, but they both cross-validate ECVT transmission technology as the best way forward for a life without derailleurs for the EMTB crowd.

I can’t stress enough how important it is for both of these companies to abandon typical complex spur gear-laden carriages of gearbox solutions like with Pinion and go instead with ECVT. There many, many advantages to using ECVT, which I’ll save exposition here since I’ve fully expounded and explored a year ago on this thread dedicated to ECVT tech: https://www.emtbforums.com/threads/could-e-cvt-save-the-motor-gearbox-unit-mgu.42671/

What’s interesting is other companies have announced ECVT motors, but for varying reasons they have yet to see daylight. Revonte, which is really the first progenitor of ECVT tech when bankrupt years ago, though sold its IP to some unknown company (hmm). Next is Owuru/E2, which came out with ECVT, but is only currently used on a Decathlon city bike and not really built for the gear range and torque needed for EMTB. Then came Villager Dynamic which announced a ECVT 2 years ago with a built-in tensioner, but they have since gone silent. Last year at Eurobike Owuru/E2 made similar waves with an update motor geared for EMTB, and generated interest but there were rumblings that their motor would not be available until 2028. And both Shimano and SRAM announced MGU patents in 2025, both of which are based on Pinion-like gearbox designs, so it was looking like ECVT tech would languish as well. Well, today changed everything and ECVT is now in the spotlight.

What makes the XP1 (ignoring the X1 version for now, since it seems more geared for city bike use) a compelling announcement would include the following:

1. Gobao, has been in the background making OEM-branded ebike motor systems for some years now, perhaps even longer than DJI itself. This not a fledgling company jumping into new motor tech, like Revonte did years ago.
2. The XP1 system appears more fleshed out compared to the Avinox MG and seems closer to production. More clear specifications are provided and influencer ride demos are not embargoed for Eurohik, unlike with the MG. It’s quite possible this system will hit the market before Avinox.
3. While internals have not been shown, this is being advertised as a true ECVT, with two internal motors and planetary gear system which are the hallmarks of ECVT design.
4. The battery system as advertised is super compelling on its own, but should be seen as level of sophistication for the overall Gobao platform rather than the motor tech, specifically.
5. Rob actually demoed the motor and pretty much went bonkers over it. In a way, I’m not surprised because older testaments from people who have tested Revonte, Owuru, etc. have posted impressive demo takes on those ECVT alternatives. But this is the first time I’ve heard how refined the performance from an ECVT motor.
6. There’s just something about the announcement that feels like a full-fleshed out solution that harks back to when DJI hit the market with their motor years ago.
7. Rumors about Specialized branding this motor and jumping straight into the MGU market with this system, since Gobao specializes in custom OEM branded solutions. If so, Gobao would act like a MGU-Brose for Specialized, instantly validating Gabao as a major player in the MGU space.

Some other subtle observations on the specs and photos:

1. The rear sprocket is pretty large, which means the internal gearing to step down the motor RPM is lower, so implicitly will be quieter. This is the same approach Rivian/ALSO TM-B took with their pedal-by-wire bike.
2. Gear ratio is 500%, which is less than the 525% ratio of the Avinox MG. However in the scheme of things with the more powerful motors and greater torque, the benefits of greater range diminish somewhat.
3. Power and torque parity with the Avinox MG, which in turn has parity with their powerful M2S.
4. Weight is lighter than the Pinion MGU and about 200g net more than similarly equipped Avinox M2S system with AXS transmission. Still waiting for a compared weight against the MG motor.
5. The battery energy/weight density is better than Avinox, so batteries have more capacity and weigh less comparatively.
6. Battery form factor a bit up I the air. Sample bikes have the chubby down tube, not knowing what size battery is in them. Aesthetics of the MG system appear better at this time.
7. Battery charging time is off the chain with a 30A charger. For many this would be extremely compelling on its own.
8. Spindle location a bit higher than expected in this motor. With the pictured 39T Gates front sprocket, it sets higher than the bottom of the motor. Then again gates sprockets are around 20% smaller than an equivalent chainring by tooth count, so this may be the best gauge. The MG looks tidier on the underside. Both still require bash guards. Avinox is notorious of not having support for them, unlike Bosch. Let’s hope Gobao has something in mind.
9. Belt tensioner looks decent so far: not too large and properly tucked behind the motor. Some of those MG tensioner prototypes look horrendous.
10. Controls, touch displays, and software look to approach the level seen with the Avinox platform.

I’m sure more details will come out soon as the system is showcased at Eurobike. I do encourage anyone planning attend to get a demo ride on the bike and really put through its paces, albeit on flat asphalt. Do focus on manual shifting, initial uptake, auto mode profiles (whether biased toward cadence, power, or motor efficiency), sprinting from standstill, noise, feel at the pedals, any noticeable slop, and any other intangibles worth noting. ]

Gear range is the stupidest metric to focus on. It needs to be greater than 0 but I would think 300% would be adequate with modern torque values.
Focus on weight.

 

[Gauss Guzzler]

I have wondered why Avinox doesn't adopt a higher voltage . Higher voltage equals smaller copper wire throughout the system.

This is a common misconception but in practice, voltage is completely irrelevant to the consumer as it makes no meaningful difference in size, weight, power, efficiency, or charge speed. Manufacturers pick a voltage based entirely on the way they want to stack the cells in the frame - that's it.

Yes, Ohm's Law is correct in that a higher voltage allows the same power to be delivered thru smaller wires, but it's only that simple for the wires between the battery and motor (or charger). Both the battery and the motor end up needing the same amount of copper regardless, and the trivial weight savings that results from smaller interconnecting wiring is typically offset by the weight gain of a proportionally more complex battery cell balancing circuit.

 

[Diby2000]

I have wondered why Avinox doesn't adopt a higher voltage . Higher voltage equals smaller copper wire throughout the system.

Voltage vs motor size

In electrical engineering, higher voltage 52 vs. 36 does not automatically mean a motor must be larger. In fact, the opposite is often true:

Higher voltage allows a system to achieve the same amount of power (watts) using less current.

Less current means the motor can use thinner copper wire, with its windings while generating less heat, which often allows engineers to design a smaller or more lightweight motor for the same power output.

This is a common misconception but in practice, voltage is completely irrelevant to the consumer as it makes no meaningful difference in size, weight, power, efficiency, or charge speed. Manufacturers pick a voltage based entirely on the way they want to stack the cells in the frame - that's it.

Yes, Ohm's Law is correct in that a higher voltage allows the same power to be delivered thru smaller wires, but it's only that simple for the wires between the battery and motor (or charger). Both the battery and the motor end up needing the same amount of copper regardless, and the trivial weight savings that results from smaller interconnecting wiring is typically offset by the weight gain of a proportionally more complex battery cell balancing circuit.

I agree that the user does not need to know the difference but there are many benifits to using a higher voltage stem.
Generally speaking, almost all inefficiencies in electronics come from heat losses one way or another. Heat losses increase across a system as current increases. So less current generally translates to less heat losses and better efficiencies. But for this system and it's intended use case, I believe the user will see the biggest advantage at the charger. Most charging systems are current limited, so be increasing the voltage, you can increase the power throughout without without significant changes elsewhere.
As stated elsewhere, 50V is the extra low voltage threshold for DC. So I think the closer to this we can get, the better for us consumers.

 

[Gauss Guzzler]

Even for the charger voltage doesn't end up mattering. The FETs and capacitors inside the charger are sized by either voltage or current so the circuitry ends up being about the same size regardless. The wire and connector going to the bike could conceivably get smaller with higher voltages but those are often sized for handling loads, not electrical loads. A high voltage charger might end up being a little more efficient and cooler running, but the overall size/weight/performance inherently remains similar to that of a lower voltage system.

 

[Diby2000]

Again I agree with everything you have said. I don't work in this area, but experience had told me on the post that higher voltage charging (even at the same power) is both cheaper and faster, even thought both are possible and even done commonly.
One simple example is charging 4S2P (4 banks of cells in series that each have 2 cells in parallel) will charge faster (I can go into the detail of why but I don't know if everyone here want to hear that) at I higher C rate (technically more stress on the battery cells) at the same max power rating, than a same capacity 2S4P arrangement. And the components in the charger will also cost less, probably have lighter mass and more efficient.

 

[Gauss Guzzler]

The "C-rate" term is specifically intended to bypass this myth. From the cell's perspective a 1 hour charge is a 1 hour charge no matter what the S/P arrangement. The only meaningful difference is the size of the wiring between the battery and the charger. And at the limit a 4V battery will charge faster and more efficiently than a 4 million volt battery because it avoids the balancing losses.

 

[Ndanger]

Pretty wild you guys still doubt the speed at which Avinox can put out product. I did get the gist that it’s well behind the gobao, but I have no doubts it’ll be refined and ready by early next year. Will it have first gen issues, possibly. If that risk concerns you wait for second gen and enjoy the early adopter reviews.

As I wait for availability of the new Avinox powered bikes espically the Amflow PR, I am starting to notice the "new guy in town" especially some of the technology which appears to be "out innovating the innovator". Their decision to go with a higher voltage standard is a prime example of potential advantages.

 

[patdam]

Everyone is weighing in with their speculations. Since I disagree with many of them—and feel it’s worth correcting ideas that seem wrong to me (until proven otherwise)—I’m going to share my own take:
"Higher voltage isn't better": I disagree; lower amperage generates less heat in the components, which benefits reliability.
"The second motor is just for gear shifting and doesn't need to be powerful": That’s not my understanding of the ECVT. I view it as a three-stage gearing system, with a specific drive source for each stage (human power + 2 electric motors) that work in combination. Both motors need to be capable of driving the bike: one is geared for torque (low gear ratio = starting off), while the other is geared for speed (high gear ratio). Consequently, the ECVT requires more energy to operate and a larger battery to match the performance of a mechanical derailleur system.
"The system can act as a brake": Yes, on a hardtail. But in my view, no—not with rear suspension, which requires a final drive (chain or belt) with variable length (using a tensioner).

 

[cream]

Was that bike you tested the Decathlon with Owuru motor?

No, it was an Ellio max, a belgian brand. It’s a hub motor based coupled with an e-cvt like transmission. Of course, with regen.

 

[Diby2000]

The "C-rate" term is specifically intended to bypass this myth. From the cell's perspective a 1 hour charge is a 1 hour charge no matter what the S/P arrangement. The only meaningful difference is the size of the wiring between the battery and the charger. And at the limit a 4V battery will charge faster and more efficiently than a 4 million volt battery because it avoids the balancing losses.

One again I agree with most of what you are saying. But the details matter here.
When charging at the chargers output limit, there is no difference. I should have been more specific for my use of the C rate term here. At the pack level, you are right, it makes little to no difference. But once the cells are at their max voltage and the charge cycle goes into constant voltage mode, the available charge energy becomes limited (below the chargers max output power). This happens later (at a higher state of charge at both the pack and cell level) in a series connected battery.
Again, technically you are right about cell balancing. A parallel connected battery does not balance the cells (because it can't). This is another advantage for the series pack and actually another major disadvantage for the parallel pack. Balancing of the cells does not need to be done every charge cycle, and can be done at the very end of the charge anyway. So you can fast charge the majority of pack in both configurations (again the series pack will be slightly faster than the parallel pack), then disconnect and joy, or at this point of still connected the series pack has the ability to balance the cells as required. Sounds like you know a bit about this stuff, so you will know The importance of equal cell balance.

 

[Diby2000]

Thank you for this, it has been on my mind also.

Everyone is weighing in with their speculations. Since I disagree with many of them—and feel it’s worth correcting ideas that seem wrong to me (until proven otherwise)—I’m going to share my own take:
"Higher voltage isn't better": I disagree; lower amperage generates less heat in the components, which benefits reliability.
"The second motor is just for gear shifting and doesn't need to be powerful": That’s not my understanding of the ECVT. I view it as a three-stage gearing system, with a specific drive source for each stage (human power + 2 electric motors) that work in combination. Both motors need to be capable of driving the bike: one is geared for torque (low gear ratio = starting off), while the other is geared for speed (high gear ratio). Consequently, the ECVT requires more energy to operate and a larger battery to match the performance of a mechanical derailleur system.
"The system can act as a brake":

I would like to push back a bit here also.
I don't think the ecvt will be replacing a chain and deralleur on a non electric assisted bike. On an ebike, different question since we already have one electric motor.
Yes the ecvt requires 3 inputs, one human powered and 2 electric motors. Three motors can also be different configurations for different requirements (speed or torque) but these two motors will not be both as powerful as the output rating. I'm fact neither of them will be this powerful. The rated output will be a combination of the 2 motors together, with the gearing multiplication. So your statement that the ecvt will inherently require more energy and a larger battery may not be true. Time will tell, but this is not a smoking gun as single large motor + chain + detailer still has efficiency losses.

 

[patdam]

Thank you for this, it has been on my mind also.

I would like to push back a bit here also.
I don't think the ecvt will be replacing a chain and deralleur on a non electric assisted bike. On an ebike, different question since we already have one electric motor.
Yes the ecvt requires 3 inputs, one human powered and 2 electric motors. Three motors can also be different configurations for different requirements (speed or torque) but these two motors will not be both as powerful as the output rating. I'm fact neither of them will be this powerful. The rated output will be a combination of the 2 motors together, with the gearing multiplication. So your statement that the ecvt will inherently require more energy and a larger battery may not be true. Time will tell, but this is not a smoking gun as single large motor + chain + detailer still has efficiency losses.

In the course of my research to understand this, I seem to recall reading that the e-CVT has lower efficiency than a manual transmission . Don't you think so ?

 

[Rob Rides EMTB]

@Greg Watts

Gobao eCVT. Will it be less efficient than normal EMTB motor? And anything else you think that it brings to the table.

 

[Greg Watts]

@Greg Watts Gobao eCVT. Will it be less efficient than normal EMTB motor? And anything else you think that it brings to the table.

@Rob Rides EMTB - good timing, this is fresh off Eurobike 2026 so there's genuine uncertainty baked in.

On efficiency: genuinely unknown - and that's the honest answer, not a cop-out. eCVT systems have historically carried an efficiency penalty (two motors running where one used to), but the Gobao design is specifically intended to offset that by keeping the primary motor in its optimal efficiency window continuously, adjusting ratio in real time based on cadence, torque, speed and gradient. Whether the AI-managed ratio system saves more than the second motor costs is the central unanswered question. It'll need independent dyno testing to settle - early impressions suggest smooth, consistent power delivery, but "smooth" and "efficient" aren't the same thing.

What it genuinely brings: no derailleur, no cassette, no mechanical shifter - the transmission and the motor are one unit. Continuously variable ratio rather than fixed steps, so theoretically no inefficient cross-chaining or mistimed shifts on steep punchy climbs. Rider sets preferred cadence, system maintains it automatically. There's also claimed regenerative braking capability. The @slickrock's post on eCVT from the "Could E-CVT Save the MGU?" thread covers the planetary differential concept well if you want the mechanical detail.

Biggest unanswered caveat beyond efficiency: torque slip resistance - when a rider hammers the pedals suddenly, the ratio motor has to react instantly with significant counter-torque. That's exactly the kind of violent input eMTB trail riding produces constantly. Mass production isn't imminent either - Gobao's existing P100 is already in OEM bikes, but the eCVT unit is a different beast entirely. Worth watching closely, but "wait for the first real trail tests" is where this sits right now.

https://www.emtbforums.com/bike-finder/motors/5/avinox-avinox-m1

 

[Diby2000]

In the course of my research to understand this, I seem to recall reading that the e-CVT has lower efficiency than a manual transmission . Don't you think so ?

I honestly don't know.
There are less moving parts, less gear teeth interactions (more teeth that are actually transmitting torque and this is where most efficientcy losses happen, but less overall), maybe less rotating inertia? In the single planetary gearbox, but you also have 2 electric motors that the manual gearbox does not have.
I think this is not so straight forward a question/statement.

I think more relevant is:
Is a single large powerful motor with a chain and deralleur more efficient than an ecvt and chain/belt/whatever drive system you use to get the power to the rear wheel?
If so, what is the deference and do you think this trade-off is worth the potential benifits of the system?

 

[patdam]

https://www.emtbforums.com/members/1/



Are you able to navigate inside engine's software to analyze its behavior, or can you only collect online reports?




Are you able to navigate inside engine's software to analyze its behavior, or can you only collect online reports?