Leaf / leafmotor / leafbike high efficiency 1500w motor

[neptronix]

It depends on how much torque you want. By having your controller outside of the hub your phase wires can become a limiting factor forcing you to choose a slower winding.

If the winding gets too slow the "full thermal headroom" won't matter anyway because rpm has become too slow. Adding in field weakening helps regains top speed but also adds a lot of heat.

Never been the case for me or anyone on this forum i've heard of.

Cut the wires an inch out the axle, enlargen them.
No problem running 2-4x the rated power this way, only a single digit % of the original wire remains in this case and isn't a significant choke point compared what conducts exceptionally better down the length of the wire.

Can't run 2x-4x the rated power on some internal controller. You'd need to replace it with a bigger controller. You won't find one. They're model specific. You also have a bigger heat problem when doing that, limiting the maximum power you can achieve on that motor. You've handicapped the power density of that motor even more than it was intiially.

That's why nobody pushing significant power takes internal controllered hub motors seriously. Lower power to weight ratio than external controller motors, less moddability, and historically, much lower reliability.

PS - i shoved 250 phase amps into a 4T leaf. I had either 12 or 10 gauge wires extending an inch out the axle. It could make more torque than physics would allow.. that's the only reason i never went higher than that!

 

[ebike4healthandfitness]

PS - i shoved 250 phase amps into a 4T leaf. I had either 12 or 10 gauge wires extending an inch out the axle. It could make more torque than physics would allow.. that's the only reason i never went higher than that!

According to ebikes.ca calculator 4T (notice Kv changed to 13.11) leaf 1500w using a 52v 80 amp controller set to 250 phase amps only makes 125 Nm of torque:

Motor Simulator - Tools

Our ebike motor simulator allows you to easily simulate the different performance characteristics of different ebike setups - with a wide selection of hub motors modeled, and the ability to add custom batteries and controllers and set a wide variety of vehicle parameters you'll be able to see...

ebikes.ca

That's really not much when looking at how other EV and motor systems are evolving.

Take for example, the Bafang M630 mid drive. That one has 250 Nm of torque:



Obviously when using a chainring smaller than the rear cog that 250Nm torque will be much higher when measured at the rear wheel. Then there are, of course, the Surron's and the Talaria's that have anywhere from 250Nm to 520Nm of torque factory stock. In contrast, the QS273 is 200Nm but the phase wires are still absolutely massive despite the 273mm stator diameter.

 

[neptronix]

According to ebikes.ca calculator 4T (notice Kv changed to 13.11) leaf 1500w using a 52v 80 amp controller set to 250 phase amps only makes 125 Nm of torque:

Motor Simulator - Tools

Our ebike motor simulator allows you to easily simulate the different performance characteristics of different ebike setups - with a wide selection of hub motors modeled, and the ability to add custom batteries and controllers and set a wide variety of vehicle parameters you'll be able to see...

ebikes.ca

View attachment 363746

A couple problems with using that graph to compare to my situation.
1. a controller capable of pushing that has a much lower resistance.
2. this battery is sagging a ton because it's not capable of the output. I used RC Lipos and had nothing like that sag.
3. That's not 6kw.What you're looking at is barely pushing it.

This is closer, except we're not accounting for dramatically lower resistance, really huge connectors, and also shorter phase wires. Those changes might be responsible for an additional ~1% of efficiency, but this is close enough.



..and it can produce more power than this, as others have shown.

That's really not much when looking at how other EV and motor systems are evolving.
Take for example, the Bafang M630 mid drive. That one has 250 Nm of torque:

View attachment 363747

That 250nm of torque is rated at crank speeds, like all mid drives are ( ~80 RPM ) versus that motor graph above, spinning at 489rpm.

489 / 80 = 6.11
250nm / 6.11 = 40nm at rear wheel at top speed

Let's say the pedal cadence is actually 90rpm.

489 / 90 = 5.43
250nm / 5.43 = 46nm at rear wheel at top speed

If you wanted to beat a hot rodded leaf with a mid drive, you'd need something more like this:
https://www.cycmotor.com/product-page/x1-pro-gen-4

 

[ebike4healthandfitness]

That 250nm of torque is rated at crank speeds, like all mid drives are ( ~80 RPM ) versus that motor graph above, spinning at 489rpm.

489 / 80 = 6.11
250nm / 6.11 = 40nm at rear wheel at top speed

Let's say the pedal cadence is actually 90rpm.

489 / 90 = 5.43
250nm / 5.43 = 46nm at rear wheel at top speed

250Nm is the 0 rpm torque output measured at the motor output shaft before any torque multiplication or torque reduction happens through the drive train.

And this discussion we are having is not about torque at top speed. We are talking torque from a standstill. The torque determined by phase amps.

So while you claim (using that chart you posted) leaf with RC lipo and 78v 80amp 250 phase amp is capable of just short of 180Nm standing start the Bafang M630 bike below using a Bafang 5 speed IGH and ordinary lithum ion battery would have more than that. Since that chainring to cog ratio looks pretty close to 1:1 I am assuming 1st gear is around 250Nm at the wheel.

Hot Selling Bafang M630 1000w Belt Drive Fat Tire Step Thru Hunting Electric Bike For Adults - Buy Eu Warehouse Electric Mountain Bike trek Electric Mountain Bike mid Drive Electric Mountain Bike 29" Electric Mountain Bike electric Mountain Bike Mid

Hot Selling Bafang M630 1000w Belt Drive Fat Tire Step Thru Hunting Electric Bike For Adults - Buy Eu Warehouse Electric Mountain Bike trek Electric Mountain Bike mid Drive Electric Mountain Bike 29" Electric Mountain Bike electric Mountain Bike Mid Drive electric Mountain Bikes For Adults...

www.alibaba.com

And this ~250Nm is from a cheap pre-built with nothing exotic.

Other electric bikes do better though. For example this $5000 E Ride Pro SS 3.0 (a Surron/Talaria clone) has 520 Nm wheel torque:

Pro SS 3.0: (19"/18") Electric Dirt Bike 2025 News | E Ride Pro

With a peak motor power of 16kW 72V 50Ah long range battery, the E Ride Pro SS 3.0 is built to dominate rugged terrains and provide the ultimate off-road riding experience.

eride.us.com

 

[Cowardlyduck]

Good luck pushing all that torque through a bicycle chain reliably.

This theoretical discussion is nice an all, but all I can say is...

This:

Like this:

Plus this:

Like this:

Together with this:


In this:


At 8KW was an absolute blast, rarely overheated and regularly did things like this....which is climbing a mountain over 1400M (~4600ft)


Cheers

 

[ebike4healthandfitness]

Good luck pushing all that torque through a bicycle chain reliably.

It's through a belt

 

[Cowardlyduck]

It's through a belt

That doesn't necessarily help much....jump over to the LMX 64 owners thread to see how we all kept breaking belts for a long time at roughly 2-2.5Kw before finally settling on expensive carbon belts.

That's not to mention the wear on sprockets, teeth slipping etc from running that kind of power. It's better now, but it took a long time to dial it in for most of us.

All that compared to the simplicity of a hub....unless you like to mess about with things (and I do ) most people would prefer something that just works and with the right torque arms, a hub motor like the leaf is exactly that.

Cheers

 

[ebike4healthandfitness]

That doesn't necessarily help much....jump over to the LMX 64 owners thread to see how we all kept breaking belts for a long time at roughly 2-2.5Kw before finally settling on expensive carbon belts.

It already comes with the expensive carbon belt:

(picture below from the M630 bike link I provided in post #3429)

 

[Cowardlyduck]

I don't see any mention about carbon there, but ok, if that's what it has then maybe it will work well.

You should choose/do something and let us know how it goes. Do you have a built thread yet?

Cheers

 

[ebike4healthandfitness]

I don't see any mention about carbon there, but ok, if that's what it has then maybe it will work well.

All the gates bicycle belts are carbon.

Anyway, my point (with that example and others I posted) is that factory verified high 0 rpm wheel torque is pretty common these days.

Unfortunately for hub motors even the mighty QS273 with its massive 273mm diameter stator requires phase wires that look like hoses just to hit a factory verified 200Nm torque. That is not good!

 

[neptronix]

250Nm is the 0 rpm torque output measured at the motor output shaft before any torque multiplication or torque reduction happens through the drive train.

Right, so it gets reduced like i said, and isn't 250nm anymore.

Let's do the reverse comparison.

Let's gear the 180nm leaf motor down so that it's running at an equivalent crank speed. Let's do a 6:1 reduction, so the leaf is running 180nm x 6 = 1080nm of peak torque from a stall.

So while you claim (using that chart you posted) leaf with RC lipo and 78v 80amp 250 phase amp is capable of just short of 180Nm standing start the Bafang M630 bike below using a Bafang 5 speed IGH and ordinary lithum ion battery would have more than that. Since that chainring to cog ratio looks pretty close to 1:1 I am assuming 1st gear is around 250Nm at the wheel.

Right.. in a 1:1 gear that goes a few mph.. you will get that 250nm torque for a short period of time, at low RPM.
In real world usage you'll never use that gear because it's speed range is too low.

In the real world, when you're doing hill climbing on say, something power equivalent to a BBSHD ( 1-1.5kw ).. you're doing 20-30mph and pedaling furiously to keep pace, whereas my leaf configuration is doin' 35-55mph.

Here's another thing..
Mid drive peak torque ratings are essentially useless as a means for assessing performance.
Mid drives don't operate at 0RPM continuously, so a 0RPM measurement is not useful. What's useful is torque at mid to high RPM.

Here's an example of how dumb this rating is.. CYC Pro X1 ( 5kw rated ) and this Bafang ( 1-1.5kw rated ) are both rated 250nm.
One is rated at absolute peak, the other is rated somewhere mid RPM.

It's a marketing trick by Bafang that's intended to fool first time ebikers who don't understand mechanics/electrics

 

[scianiac]

All the gates bicycle belts are carbon.

Anyway, my point (with that example and others I posted) is that factory verified high 0 rpm wheel torque is pretty common these days.

Unfortunately for hub motors even the mighty QS273 with its massive 273mm diameter stator requires phase wires that look like hoses just to hit a factory verified 200Nm torque. That is not good!

I don't even understand what argument you are trying to make, first you say you want a controller inside the motor for phase wire reasons, then we tell you there is no point because it would be better to just use larger wires and you can easily fit large enough wires to a Leaf that will handle all the power the motor can really handle. So then you compare it to geared mid drives and claim you want more torque than the leaf can provide which yeah you can gear motors and achieve more torque that's how gears work and having no gears is sorta the whole point of a hub motor. You also make comparisons to a Sur-ron which is just an entirely different thing, yes a normal bike frame with a Leaf won't be powerful like a lightweight E-dirt bike. Then you talk about how the QS273 which is a more powerful hub motor, which seems to be what you want and talk about how it has much larger phase wires which is exactly what we were saying, it's better to use larger phase wires. And again comparing the rated torque of the QS273 which lets face it is way below what most people actually run those motors at vs the Bafang mid drive which is probably highly optimistic and only works at a very limited speed.

If your argument is hub motors are not as good in some situations then yes we all already know that but picking very specific points of comparison doesn't really provide any useful real world information unless you only use the bike at 0-5mph or something.

 

[Bullfrog]

I still wouldn't want them. I want the full thermal headroom of a motor without additional things inside it. I want standardized, swappable parts that aren't specific to one motor. I want as absolutely low of unsprung weight as i can get.

No such thing exists on the market today anyhow.

I completely agree...heat no matter the source is your enemy and combining things that produce heat only makes the situation worse.

 

[Bullfrog]

Phase amps are not continuous like battery amps so 100 phase amps does not require the same size wiring size as 100A of battery current.

The difference in the phase amps required for a faster wound motor to produce the same torque as a slower wound motor is not that much different but they are a little higher so your controller has the potential to be the limiting factor. IMO the difference is not enough to worry about upgrading your phase wires in most applications.

The only point I wanted to make earlier is all windings of a given motor design output the same torque per per BATTERY amp so the terms "torque winding" is a misnomer.

 

[Bullfrog]

As far as bicycle chains and mid drives go....my BBSHD puts out over 300 NM at the output shaft and I have never had any problems with my bicycle chain handling the torque.

 

[ebike4healthandfitness]

The only point I wanted to make earlier is all windings of a given motor design output the same torque per per BATTERY amp so the terms "torque winding" is a misnomer.

Yes, you said that but you also added the fast winding needs more phase amps:

Post #3418:

Just a couple notes....

A. You can run a BBSHD on 72v with an aftermarket controller...I did it.

B. Torque output for different motor windings...for the same BATTERY amperage, all windings of a given motor design (Leaf motor for example) will produce the same torque. The difference is in the phase amperage required...a fast wound motor (Leaf 3T for example) will require more PHASE amperage to output the same torque as a slow wound motor (Leaf 6T for example).

To be a little clearer...if your controller is capable of putting out enough PHASE amperage, the torque output of a given motor design like the Leaf motor for example is the same for all windings i.e. 4T, 5T, 6T, etc. assuming you feed them the same power i.e. Volts X BATTERY amperage.

The terms "Torque Winding" and "Speed Winding" for a given motor design are misnomers.

But as you know there is a limit to the amount of phase amps that can be supplied. So while what you say is true in theory it is not true in practice because of controller and phase wire limitations.

Speaking of aftermarket BBDHD with 72v which you also mention in your post above I see Luna cycles made a tiny 72v 100 amp controller fit inside the little BBSHD shell using nothing more than silicon based fets:

Luna BBSHD Ludicrous V2 Released!

The Ludicrous V2 is the result of many months of R&D designing a motor controller from scratch with the latest innovations in motor control, spanning over 100,000 lines of code, the best automotive qualified materials, and vastly proven building blocks; all tightly integrated into the stock...

lunacycle.com

So yeah, I don't see what the big deal is for putting a controller inside a much bigger DD hub motor is. This especially with statorade cooling and GaN FETs. Anyone building a high end DD hub motor should be able to do this.

 

[ebike4healthandfitness]

The difference in the phase amps required for a faster wound motor to produce the same torque as a slower wound motor is not that much different but they are a little higher so your controller has the potential to be the limiting factor. IMO the difference is not enough to worry about upgrading your phase wires in most applications.

It is a pretty big difference actually.

 

[99t4]

Other electric bikes do better though. For example this $5000 E Ride Pro SS 3.0 (a Surron/Talaria clone) has 520 Nm wheel torque:

That's only an "electric bike" in the most forgiving sense of the term. It's an e-motorcycle, come on. What's your point and what does it have to do with this thread? (Leaf / leafmotor / leafbike high efficiency 1500w motor)

 

[neptronix]

Notice that controller has an optimized heat path to the atmosphere and hub motors with internal controllers don't.
Notice that the controller also doesn't produce heat inside the stator casing, the motor and controller both have their own separate heat zones.

The motion of heat from one heat zone to another is miniscule compared to the internal controller in a hub motor scenario.

..this is why after many iterations of motors that continually blew their internal controllers ( didn't dock the motor/controller power enough to account for thermals ), golden motor started making their internal controllers much more external... at the expense of motor width.. which is a negative for efficiency and power, because the wider the stator, the lower the proportion of end turn losses.

Let's look at the magic pie V2.. all the controller heat ends up in the motor casing.. the most effective path from controller heat to atmosphere is through the axle.. which the stator is also trying to utilize.



Sure, can can improve the efficiency of the controller a bit; now you have less of a problem that you created for yourself. Ideally, you wouldn't create the problem in the first place.

 

[ebike4healthandfitness]

That's only an "electric bike" in the most forgiving sense of the term. It's an e-motorcycle, come on. What's your point and what does it have to do with this thread? (Leaf / leafmotor / leafbike high efficiency 1500w motor)

People have no problem using that level of torque and often complain of wanting more.

And often times our ebikes use larger diameter tires which need more torque than a smaller diameter tire would.

 

[scianiac]

So yeah, I don't see what the big deal is for putting a controller inside a much bigger DD hub motor is. This especially with statorade cooling and GaN FETs. Anyone building a high end DD hub motor should be able to do this.

I don't think you actually understand anything about the thermal pathways here, that Luna controller or actually any BBSHD controller is mounted directly to the case which is covered in heatsink fins, probably the FETs are either on a MPCB or on a dense array of vias and that is mounted directly to the case with some thermal paste.

Now in a hub motor, you can't connect anything to the case, that statorade works great but only compared to air and you don't understand how it works at all. It only connects the stator and magnets with something more conductive than air but still not great. How exactly are you going to cool the FETs like that? You going to mount them to the stator and try and conduct the heat through 2 inches of steel that is also being warmed by the windings and then through the statorade and through the magnets and then to the housing. Unless you are venting the case or pumping oil through it there is just no way this makes any sense. All to save what? The shaft of the Leaf will already fit 10GA wire which is well more than enough to run it well into staturation. Not to mention controllers don't like it when you start making the battery leads too long, the inductance starts to become an issue.

People have no problem using that level of torque and often complain of wanting more.

And often times our ebikes use larger diameter tires which need more torque than a smaller diameter tire would.

We have no problem with bikes that have that much torque. Indeed I expect that a Wärtsilä RT-flex96C also has more torque than a Leaf motor but we aren't talking about container ships here, we're talking about applications that it would be reasonable to use a Leaf motor for. If you want more torque than a Leaf get a larger motor or one more suited for your application.

To idk possibly steer this ship back on course here, I guess I have ships on the brain. I've been pondering ways to add an encoder to a hub motor, I think with an encoder even the shity VESC on my Leaf bike could perform much better and solves any hall noise issues (not that those seem to be the issue at all). You can get magnet encoder rings but that would mean finding a VESC compatible encoder that will read one of those or adding that to VESC. I think a sin/cos encoder would do this, that is read a ring and not a disc.

 

[Bengy22]

Good luck pushing all that torque through a bicycle chain reliably.

I wanted to chime in here, I don't know where this comes from but people way over estimate how delicate mtb chains are. Either that or it comes from people using those super thin 12 speed chains. Run less gears to get the thicker chains, plus the more power you have the less gears you end up wanting anyway. I've been using a 8 speed with my bbshd since day once and never wanted more and never had a chain snap (9k miles). I'm now running roughly 6.5kw and I still get roughly 1.5k miles on my $8 kmc chain, every 3 swaps do the cassette and chainring. Only issue I had was moving from 4kw to 6.5kw killed my hub in less than 1k miles, had to upgrade to a star ratchet one but been good since. I've included the Ebike.ca sim for the gear I use 95% of the time 3rd (also note the top speed is off by a bunch due to no FW or MTPA). My highest gear is 1:1, but I've only used it a handful of times for fun (I was worried I was going to snap something lol) but to be honest even 3rd will climb anything.

Speaking of aftermarket BBDHD with 72v which you also mention in your post above I see Luna cycles made a tiny 72v 100 amp controller fit inside the little BBSHD shell using nothing more than silicon based fets:

Luna BBSHD Ludicrous V2 Released!

The Ludicrous V2 is the result of many months of R&D designing a motor controller from scratch with the latest innovations in motor control, spanning over 100,000 lines of code, the best automotive qualified materials, and vastly proven building blocks; all tightly integrated into the stock...

lunacycle.com

This is the controller I run, in my opinion it should be on every bbshd they sell.

 

[ebike4healthandfitness]

That Luna 72v 100amp controller is tiny.....with silicon FETs.

So with GaN FETs it should be even smaller.

As far as temperature goes it shouldn't even matter as far as I understand things (remember this post ---> Leaf / leafmotor / leafbike high efficiency 1500w motor)

P S. Apparently there is some real potential for GaN in extremely hot environments well beyond that of any hub motor ---> Turning up the heat on next-generation semiconductors

One reason scientists have not yet been able to send a rover to the planet’s surface is because silicon-based electronics can’t operate in such extreme temperatures for an extended period of time.

For high-temperature applications like Venus exploration, researchers have recently turned to gallium nitride, a unique material that can withstand temperatures of 500 degrees or more.

 

[scianiac]

Nah GaN fets are lame, what you want for really high temps is SiC FETs. But why? First those FETs are not just slap them in anything, they are often much harder to design controllers with in addition to being more expensive and more delicate. All for what exactly? What is this doing better? I would rather have the GaN FETs in a well cooled external controller where they will perform even better. Any combination of variables here will always yield the external controller working better. You use exotic FETs, they still perform better when not screaming hot, let alone you have to make everything else on the controller not overheat.

 

[neptronix]

Nah GaN fets are lame, what you want for really high temps is SiC FETs. But why? First those FETs are not just slap them in anything, they are often much harder to design controllers with in addition to being more expensive and more delicate. All for what exactly? What is this doing better? I would rather have the GaN FETs in a well cooled external controller where they will perform even better. Any combination of variables here will always yield the external controller working better. You use exotic FETs, they still perform better when not screaming hot, let alone you have to make everything else on the controller not overheat.

Exactly!