Leaf / leafmotor / leafbike high efficiency 1500w motor

amberwolf » Nov 19 2020 11:31pm

But...how are you going to get that thru the axle with the existing wires? (Assuming you're going to use the thermistor to monitor stator / winding temperature, like the existing sensor does, the sensor has to be inside the motor, and the wires have to pass thru the axle bearing somehow, to reach the monitoring device (CA, etc)).

Click to expand...

Sorry for not stating this, leafmotors can be ordered with a temperature sensor. It does not work with the CA3 but the wires are ran thru the axle and a person needs to change out the thermistor within the motor to the existing wires or wire as it does not have a ground. The ground off the new thermistor can be connected to the Halls ground wire within the motor.

ZeroEm said:

... needs to change out the thermistor within the motor to the existing wires or wire as it does not have a ground. The ground off the new thermistor can be connected to the Halls ground wire within the motor.

Click to expand...

That is the plan. The thermistor I ordered to replace the one in the motor has a 2-pin JST on it, and I will have to cut the wire, connect the thermistor to the ground wire and white wire slotted through the motor, and then connect the JST to those same wires outside of the axle.

As problems develop, I will start replacing connectors with more permanent solutions.

amberwolf said:

THere's other types I don't recall the names of that are small and waterproof like Higo, but are user-assemblable onto existing cabling. One of them is what Incememed uses on his SFOC controller units, and are named in either his Hall Sensors B'gone thread or the manual for his controller that's also in that thread. Both Higo and what he is using are small enough to pass thru axle nuts, too, which the 5/6-pin JST's aren't really.

Click to expand...

I did some searching. Would that be the Binder/99 series of connectors?

That thread contains valuable information. This controller may be a future upgrade for me:

https://endless-sphere.com/forums/viewtopic.php?f=2&t=30680&start=25#p1393153

I'll need something much more efficient than a Leafbike motor to take advantage of it though, and AFAIK, one is not available for my application.

Had some issues the first month of use then just wrapped the connections with silicon tape. Think this help hold the connectors together and it made a water resistant cover, many not hold up under water but have driven it in the rain. When I get around to it and pull the motor off I will see if there is any corrosion. The motor i'm putting on has grins new favorites L1019. Round plug with all the phase and halls wires in one connector.

ZeroEm said:

amberwolf » Nov 19 2020 11:31pm

But...how are you going to get that thru the axle with the existing wires? (Assuming you're going to use the thermistor to monitor stator / winding temperature, like the existing sensor does, the sensor has to be inside the motor, and the wires have to pass thru the axle bearing somehow, to reach the monitoring device (CA, etc)).

Click to expand...

Sorry for not stating this, leafmotors can be ordered with a temperature sensor. It does not work with the CA3 but the wires are ran thru the axle and a person needs to change out the thermistor within the motor to the existing wires or wire as it does not have a ground. The ground off the new thermistor can be connected to the Halls ground wire within the motor.

Click to expand...

Yes, but he had said he wasn't going to use the existing wires, and was going to use the wires and JST that came on the new thermistor, although it sounds like his plan has now changed, to use the existing wires between the thermistor and it's JST.

The Toecutter said:

I did some searching. Would that be the Binder/99 series of connectors?

Click to expand...

Looks like it based on this image search on that name. When I tested that controller, the connectors worked well and didnt' get wet even on the occasion of complete submersion of the whole bottom of the trike in flash flood waters.
https://www.google.com/search?q=Binder/99+series+of+connector&tbm=isch

That thread contains valuable information. This controller may be a future upgrade for me:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=30680&start=25#p1393153
I'll need something much more efficient than a Leafbike motor to take advantage of it though, and AFAIK, one is not available for my application.

Click to expand...

Depends on the "advantage" you want from it. If you're simply looking for sensorless FOC sinewave operation of the motor, it should work fine with any motor.

If you're wanting high efficiency, I'm not sure there's any off the shelf common 3-phase BLDC hubmotor that's really that much more efficient than any other, and it will depend on the specific loading and usage scenarios whether you would get the efficiency any manufacturer "rates" their motor for, anyway.


If you're wanting higher power from it, then if it's just bursts, the leafmotor would probably handle those, as long as the axle is capable of handling it (the MXUS3k axles apparently aren't, but they're poorly built motors)--higher power for longer periods might need a bigger motor (something like the QS205 50, perhaps, which would easily handle a few kw, and some have run them much higher).

amberwolf said:

If you're wanting higher power from it, then if it's just bursts, the leafmotor would probably handle those, as long as the axle is capable of handling it (the MXUS3k axles apparently aren't, but they're poorly built motors)--higher power for longer periods might need a bigger motor (something like the QS205 50, perhaps, which would easily handle a few kw, and some have run them much higher).

Click to expand...

In the long term, I'm going to want more power than the Leafbike motor can provide. A similar sized hub motor with a higher peak efficiency and good mechanical robustness would allow that, but there is nothing of that sort on the market that can accept a bicycle cassette or freewheel. I need the motor to remain relatively lightweight, as I'm trying to keep the whole build under 100 lbs, in the interest of keeping it pedalable with the motor shut off. But at some point in the future, I'd like to do 10 kW peak or more, and I don't think the Leafbike can handle that on any regular basis reliably, even if for only a few seconds at a time... The motor I need doesn't seem to exist and I might have to design and build it unless something that meets my requirements comes onto the market. Or go to a mid drive and all of the issues that it entails(I would have to keep it separate from the human power drivetrain in the interest of pedal drivetrain longevity).

The Leafbike will satisfy me in the near term though, starting at 1.5 kW with the 46.8V pack, then when I build my 20S6P 72V pack, maybe 4-5 kW peak... perhaps 7 kW with an A123 pack... 4 kW or more will allow for car-like performance in a velomobile, which is the goal. With the Leafbike at 4-7 kW, it will perform more like a worn-out decades-old economy car than a new sports car though, the latter being the ultimate goal.

The Toecutter said:

In the long term, I'm going to want more power than the Leafbike motor can provide. A similar sized hub motor with a higher peak efficiency and good mechanical robustness would allow that, but there is nothing of that sort on the market that can accept a bicycle cassette or freewheel. I need the motor to remain relatively lightweight, as I'm trying to keep the whole build under 100 lbs, in the interest of keeping it pedalable with the motor shut off.

Click to expand...

That sounds a little like wanting a 10,000 TEU container ship that can go by sail in case of engine problems. I mean, yeah... but no.

Pedal power works for vehicles that require 100-200W continuous, 300-1000W peak. If your performance requirements don't fit in this range, don't bother.

by Balmorhea » Nov 21 2020 1:20am

The Toecutter wrote: ↑Nov 20 2020 11:58pm
In the long term, I'm going to want more power than the Leafbike motor can provide. A similar sized hub motor with a higher peak efficiency and good mechanical robustness would allow that, but there is nothing of that sort on the market that can accept a bicycle cassette or freewheel. I need the motor to remain relatively lightweight, as I'm trying to keep the whole build under 100 lbs, in the interest of keeping it pedalable with the motor shut off.

Click to expand...

That sounds a little like wanting a 10,000 TEU container ship that can go by sail in case of engine problems. I mean, yeah... but no.

Pedal power works for vehicles that require 100-200W continuous, 300-1000W peak. If your performance requirements don't fit in this range, don't bother.

Click to expand...

When Luke tested the leafmotor it look like 2000w continuous was doable. I'm sure many things would factor in, I'm thinking that is a lot for a Velomobile. I"m sure to get higher power with a weight limited motor mid drive will be needed.

I think a lot about more power, less drag, with regen. I keep coming up with something like Grin's front DD or a leafmotor and a geared or midrive that would disengage to keep drag down. Most geared motors just do not put out much power. The trouble of a high power mid motor is that trouble.

Think about AWD on my trike, two of Grin's front DD and a leaf motor in the rear, that should net 7K and more speed than needed. Not to say how heavy and costly.

Someone may come up with something someday. At the moment the leafmotor is one of the better options. I can go without power 15 mph/24 kph and don't really feel the drag, not that it is not there. It is the hills where I feel how heavy my trike is with me on it.

Balmorhea said:

Pedal power works for vehicles that require 100-200W continuous, 300-1000W peak. If your performance requirements don't fit in this range, don't bother.

Click to expand...

The thing is, I'm seeking to build a vehicle that can do 20-25 mph on 100-200W continuous and maintain highway speeds at 1,000W peak. With that vehicle basically built, I will then shove as much power as I can into it in the interest of automobile-like acceleration performance and top speed, criteria that 100-200W continuous and 1,000W peak is generally a poor fit for. And if it can still do 20-25 mph on 100-200W continuous and maintain highway speeds at 1,000W peak with the motor installed, that means that human power is still a good fit for that speed range, and can also be worthy of supplementing the electric motor at all automobile-legal speeds in the interest of providing a significant efficiency increase and range extension. It also fits the design philosophy of building a vehicle that is not grid-reliant by many meanings of the word "grid", and also renders moot the concept of range anxiety.

It's more like wanting a small 100+ mph capable speed boat, with a deployable sail as backup and designed in such a way so the sail can supplement the vehicle's performance at low power demand and/or speed but for the sail's presence not to interfere with the vehicle's maximum performance potential when the sail is not in use, than it is a 10,000 TEU container ship.

ZeroEm said:

When Luke tested the leafmotor it look like 2000w continuous was doable. I'm sure many things would factor in, I'm thinking that is a lot for a Velomobile.

Click to expand...

2,000W continuous would get a well-designed body shell to maintain 70 mph. Getting that design is easier said than done though. A lot of aerodynamic work will be necessary for it to be stable with heavier vehicles passing it at higher speeds and with strong cross winds, which is going to work against the goal of keeping drag low. The center of pressure is going to need to remain behind the center of gravity in as many operating conditions as possible to keep the thing straight, and it's going to need net downforce over the entire vehicle, and to retain these conditions over uneven road surfaces that can introduce unanticipated yaw.

Sustained 70 mph is definitely not happening with my zip-tied, taped-up, coroplast rolling coffin! It's been over 50 mph downhill for short periods, and was stable, could still safely make lane changes, and felt well-planted, but I wouldn't trust it at anything more. I wouldn't trust the Schwalbe Marathon Greenguard tires and Velocity USA wheels for long-distance cruising at anything more than 35 mph, really. 70 mph might just rip the body apart!

I think a lot about more power, less drag, with regen. I keep coming up with something like Grin's front DD or a leafmotor and a geared or midrive that would disengage to keep drag down. Most geared motors just do not put out much power. The trouble of a high power mid motor is that trouble.

Click to expand...

Those front drive all-axle motors are still quite lossy. The published Hysteresis drag torque is 0.45 Nm, and the Eddy Current Drag Torque is 0.0005 Nm/rpm. A motor in each front wheel, assuming a 20" bike wheel, for the pair is going to require 59W at 20 mph, 79W at 25 mph, 101W at 30 mph! In something as efficient as a velomobile, those cogging losses could easily be costing an extra 5-7 mph of speed when operated unpowered versus not having a motor at all. If they had thinner 0.2mm or smaller laminations instead of the 0.35mm they currently have, those losses could be cut significantly, but to me, the ideal is just sticking with one motor that is of low losses and of sufficient peak efficiency that it can make high peak power without melting so that when operating solely under human power with a dead battery, the motor is only costing the rider 2-3 mph at 25 mph on flat ground.

I'd like to see an affordable 7-lb rear-drive synchronous reluctance hub motor of 97-98% peak efficiency that can fit a 9-speed cluster in a 135mm dropout set, that is designed for mechanical reliability(especially on the human-powered side), and can perhaps handle short bursts of 15 kW peak for 30 seconds at a time and do about 2 kW continuous.

Peak means "for a few seconds" at most a few minutes.

So not available for "maintaining" anything

the "top highway speed" needs to be within the continuous output capacity of the system

even taking headwinds and long climbs up gentle slopes into account

The top speed doesn't need to be so conservative that it is always within the continuous rating. It needs to be set at an amount that once reached, the battery will run out of charge before the motor overheats.

Assume a Leafbike 3T wind running at 96V in a 16" moto wheel(equivalent size to 20" BMX wheel), in a velomobile with a laden mass of around 130 kg(including rider, luggage, tools, ect.) and a CdA around 0.08 m^2 with DOT tires that have a Crr around 0.008(tires with this low of a Crr may not currently exist in a format for motorcycles/scooters, but for cars, they already do, and it is only a matter of time before someone decides there's a market for such a thing). The motor would be capable of about 2-2.5 kW continuous at that voltage, which would be sufficient for 70-75 mph on the flat in such a vehicle, indefinitely, but the battery isn't going to be able to power the thing an infinite amount of time. For flat ground, it is likely that the motor would be able to do 90 mph without overheating while drawing down an entire 1.5-2 kWh pack, including a few hard accelerations to get there, because it would be using 4 kW of power to maintain 90 mph but drain the pack in under 30 minutes.

That's what an off the shelf $300 Chinese motor is theoretically capable of, in the right vehicle. It gets just over 90% efficiency peak and 75-85% over most of its usable range. In practice, some of the builders in this topic have mentioned flogging their ebikes with 4-7 kW until the battery dies and not destroying their motors. For those who worry about what happens trying to go up mountains at those speeds and power levels, a thermistor and a Cycle Analyst computer can prevent the user from melting it by cutting back the phase current when the temperature becomes too much.

The state of the hub motor technology available on the market today is a step backwards from the hubmonsters/mid monsters/mini monsters of almost a decade ago, but the actual technology that has been developed has improved greatly since the hubmonsters/mid monsters/mini monsters entered the market.

Imagine what happens if a motor is made with technology that isn't as antiquated using more robust components in a manufacturing process with better quality control. A motor with 97% peak efficiency and 90-95% efficiency over most of its usable operating range is possible today without exotic materials or manufacturing processes. It's an unexploited market niche regarding the ebike market. Someone with sufficient capital to make it happen is likely to eventually realize this.

> 2,000W continuous would get a well-designed body shell to maintain 70 mph

That 70mph is not a peak top speed is it?

by The Toecutter » Nov 21 2020 3:30pm

The top speed doesn't need to be so conservative that it is always within the continuous rating. It needs to be set at an amount that once reached, the battery will run out of charge before the motor overheats.

Assume a Leafbike 3T wind running at 96V in a 16" moto wheel(equivalent size to 20" BMX wheel), in a velomobile with a laden mass of around 130 kg(including rider, luggage, tools, ect.) and a CdA around 0.08 m^2 with DOT tires that have a Crr around 0.008(tires with this low of a Crr may not currently exist in a format for motorcycles/scooters, but for cars, they already do, and it is only a matter of time before someone decides there's a market for such a thing). The motor would be capable of about 2-2.5 kW continuous at that voltage, which would be sufficient for 70-75 mph on the flat in such a vehicle, indefinitely, but the battery isn't going to be able to power the thing an infinite amount of time. For flat ground, it is likely that the motor would be able to do 90 mph without overheating while drawing down an entire 1.5-2 kWh pack, including a few hard accelerations to get there, because it would be using 4 kW of power to maintain 90 mph but drain the pack in under 30 minutes.

That's what an off the shelf $300 Chinese motor is theoretically capable of, in the right vehicle. It gets just over 90% efficiency peak and 75-85% over most of its usable range. In practice, some of the builders in this topic have mentioned flogging their ebikes with 4-7 kW until the battery dies and not destroying their motors. For those who worry about what happens trying to go up mountains at those speeds and power levels, a thermistor and a Cycle Analyst computer can prevent the user from melting it by cutting back the phase current when the temperature becomes too much.

The state of the hub motor technology available on the market today is a step backwards from the hubmonsters/mid monsters/mini monsters of almost a decade ago, but the actual technology that has been developed has improved greatly since the hubmonsters/mid monsters/mini monsters entered the market.

Imagine what happens if a motor is made with technology that isn't as antiquated using more robust components in a manufacturing process with better quality control. A motor with 97% peak efficiency and 90-95% efficiency over most of its usable operating range is possible today without exotic materials or manufacturing processes. It's an unexploited market niche regarding the ebike market. Someone with sufficient capital to make it happen is likely to eventually realize this.

Click to expand...

I agree with you. So far the leafmotor has been the best option and it's hard to understand why someone has not made something better. The top speed I try to force on myself is 28mph/46kph but the top speed for this 7T motor is more like 33 mph. This is well with in range of this motor in my area with low rolling hills. Did a 55 mi run between 25-35 mph used about 900w from my 1.8kw pack. Half way took a break and the motor was only warm to the touch not hot so not saying much. I know others want to go faster but this is fast enough. Just saying the leafmotor works well for this.

This setup lets me peddle unpowered 15-20mph if any thing happens. I look for better options but want lots of gears and low drag and big wheels. Don't bigger wheels cause the motor to turn slower/less drag. Not that for high speed/power a smaller wheel is better. But i'm split between these two issues. for long riding i'm sticking with bigger wheels and gears.

john61ct said:

That 70mph is not a peak top speed is it?

Click to expand...

No. But if there were a battery pack with an infinite supply of energy, the motor would never overheat at 70 mph, at least on flat ground, in the hypothetical velomobile that I proposed. That is what a continuous rating 2 kW means, that the motor can withstand 2 kW indefinitely.

My first build won't even be able to reach 70 mph, this being said. The motor in my possession is a 5T wind and not the 3T wind which I would need, and it's built into a larger 26" bicycle wheel instead of the smaller 16" motorcycle wheel. 55 mph as a top speed is more likely for my first build once I have a 72V pack(which is about as much as my Phaserunner controller can handle), as opposed to the 96V pack proposed, and while the motor would never overheat at 55 mph, I won't have the wheels/hubs/tires/axles/brakes or full suspension to handle 55 mph as a cruising speed, and 35 mph would be more realistic for a cruising speed given that I am using bicycle grade components and only have a front suspension.

But for a future build with a 3T wind in a 16" moto wheel(equivalent to a 20" bicycle wheel in diameter) running a 96V pack, cruising at 70 mph and reaching a top speed of 90+ mph is a possibility with just one of these Leafbike motors without a major risk of overheating it, if and only if I get the aerodynamics right for it and if I have the proper components to handle the stresses that such speeds will entail. And that's all a big "if". With 7 kW peak thrown to it with a maximum phase current of 150A, 0-60 mph acceleration in around 10-12 seconds is theoretically possible depending upon laden vehicle weight.

This is an impressive little motor when you look at it in the above context. And it still falls woefully short of what is possible with the current state of the technology developed.

12mph = average about 100w
20mph = average about 350w
25mph = average about 500w
30mph = average about 900-1000w
37mph = about ~1650w ( not really confirmed in this video as she was still accelerating a bit )


Is this on 48v?

momo said:

12mph = average about 100w
20mph = average about 350w
25mph = average about 500w
30mph = average about 900-1000w
37mph = about ~1650w ( not really confirmed in this video as she was still accelerating a bit )


Is this on 48v?

Click to expand...

Those wattage figures are about accurate for an unfaired mountain bike. Power demand at a given speed is going to depend upon the vehicle's aerodynamic drag, rolling resistance, and stray mechanical losses. Voltage is independent of power requirement, although depending upon motor wind, 48V may or may not be enough to reach the speeds listed. A "48V nominal" pack, which if a 13S LiIon type battery is actually 46.8V nominal, would yield about 52V at full charge, which in the default 5T wind Leafbike motor in a 26" wheel, should give a top speed of 37 mph on a mountain bike at a load of about 1650W.

I see, I'm trying to reach similar speeds as the tidalforce 750x, which was 30mph on 36v and a 26" wheel. As far as windings, I wouldn't know. Wonder if you can gut the Factory controller and use an external controller on a tidalforce motor.

momo said:

I see, I'm trying to reach similar speeds as the tidalforce 750x, which was 30mph on 36v and a 26" wheel. As far as windings, I wouldn't know. Wonder if you can gut the Factory controller and use an external controller on a tidalforce motor.

Click to expand...

I don't know about the Tidalforce motor. Regarding the Leafbike motor using a 36V nominal pack with a 26" wheel, you will be able to do 35 mph with a 4T wind, 28 mph with the default 5T wind. That's on a fully charged pack. Top speed should drop slightly as battery pack voltage drops.

momo said:

Wonder if you can gut the Factory controller and use an external controller on a tidalforce motor.

Click to expand...

Not unless you custom-design one for it--it's a 5-phase motor, so AFAIK none of the available common controllers can run it (they're only three phase).

Hey fellow leafheads,
While I've done my own on road testing to tune my controller batt and phase amps for best performance without unreasonable temperature climbs (running FF with hubsinks), I've recently been playing around with the ebikes.ca sim and have a question. The highest I can get phase amps to go on the sim, at WOT from a stall, is 151.9a using 80a batt, and 137.2a on 65a batt. My controller is a YKZ7280JA bluetooth with 80a batt 230a phase maximums. I think the best performance I've gotten out of it so far (seat of the pants measurement) was at something like 80% batt 80-85% phase. Is the leaf simply approaching saturation at thise levels and therefore any higher is just creating heat? Or am I missing something or what?
Thanks all!

What Battery do you have? 13s 5p, I read some where and what motor a 4T or 5T?

I bought a used leaf laced in a 26 inch rim. I opened the motor to check the condition. I think it looks good. Can somebody tell me which version it is looking at the windings? I've read somewhere Leaf recommends a 5T for 26 inches.

I am on the waiting list for a Nucular 12F. In the meantime, I want to buy a cheap and powerful (programmable?) controller for my setup with an EM3ev 52V (14S5P) Jumbo Shark Battery with 25R cells rated ~ 40A Continuous, 55A Max Burst Current. Recommendations anyone?

I just got the 10k NTC thermistor installed today and used JB weld to adhere it to the lip of the case and up against one of the copper wire loops, since it wouldn't fit inside one of the loops like the stock thermistor that came in the motor did.

Anyhow, I have soldered pins onto the black, yellow, green, blue, and red wires coming off of the motor for the 5-pin JST connector that goes to the Phaserunner controller, only to find that the pins now won't fit through the holes in the plastic housing of the JST connector. The solder holding the pins to the wires is getting in the way. Is there any technique one uses to get the pins soldered to the wires where the pins can still slot through the JST ports? It's very frustrating working with tiny, cheap wires selected to save a fraction of a cent.

The Toecutter said:

Anyhow, I have soldered pins onto the black, yellow, green, blue, and red wires coming off of the motor for the 5-pin JST connector that goes to the Phaserunner controller, only to find that the pins now won't fit through the holes in the plastic housing of the JST connector. The solder holding the pins to the wires is getting in the way. Is there any technique one uses to get the pins soldered to the wires where the pins can still slot through the JST ports?

Click to expand...

The "quick" solutions other than crimping (which is the only way JST pins are designed to be used) new pins on (rather than reusing old ones) are, in no particular order:

--file the solder down till it does fit (may not be possible)

--cut the JST housing away enough to allow them to fit, and glue the pins in place (since you may not be able to lock them in), and use tape or heatshrink (applied over exposed conductor/contact areas that won't be inside the housing anymore, before insertion) to guarantee no shorts to anything.

--unfold the crimp tabs of each contact carefully, removing the original wire that had been crimped in them, and place the new wire down in the U channel, then solder that in place, and file any bits that won't fit in the housing carefully off.

--use an existing prewired JST (even one cut off some other device), and splice the new wires to the prewired wires some distance from the connector, preferably staggering the connections so that even if the splice insulation fails, they can't short to each other.

If you *are* using new pins already, but not crimping, then you must remove all of the crimp foldover tabs either before or after you solder, such that the pin ends up the same shape as a crimped pin. Otherwise it is not going to be able to fit into the housing, as it's only designed to be crimped before insertion.