Leaf / leafmotor / leafbike high efficiency 1500w motor

[Cowardlyduck]

Having tried both I can comment on a few aspects...the Leaf worked well, heavily modified with FF, heatsinks and internal fan cooling it was great at up to 8KW peaks, regularly sitting between 3-5KW sustained and kept reasonably cool. I also vented it, drilled out the axle and replaced the wires through the axle with 12AWG silicone wire, then to 8AWG immediate out of the axle.
I also did lots of minor things to improve cooling like painting the inside of the covers black, and keeping it reasonably clean.

On the other hand, the RH212 I've just installed on my recumbent in a 20" wheel is performing great at 3KW+ and I haven't been able to break 75C yet. It only has FF and nothing else. Performance is fantastic and it pulls really hard.
My Phaserunner overheats first at 60A which I need to do something about.


Cheers

[ebike4healthandfitness]

A huge limiting factor for the 212 is the L1019 connector which will severely limit the input amperage if you wanted to run high power but besides that...everything below still holds . The L1019 big pins (phase current) are rated for 30A...although they can handle a lot more than that for very short periods of time, I wouldn't suggest much over 30A steady state.

Are you sure the L1019 connector is rated for only 30 amps? That doesn't sound right as Phaserunner (which uses L1019) is a 45 to 50 amp controller without a heatsink (and higher than 50 amps with a heatsink).

[Bullfrog]

30A for the big pins in the connector and 2A for the little pins...I got the specs here: https://www.higoconnector.com/products/ ... 1000/DK99K

I would say you could double that for short periods of time.

Grin says 100A currents for short times here: https://ebikes.ca/learn/connectors.html.

I guess it all depends on how long .

[Bullfrog]

Having tried both I can comment on a few aspects...the Leaf worked well, heavily modified with FF, heatsinks and internal fan cooling it was great at up to 8KW peaks, regularly sitting between 3-5KW sustained and kept reasonably cool. I also vented it, drilled out the axle and replaced the wires through the axle with 12AWG silicone wire, then to 8AWG immediate out of the axle.
I also did lots of minor things to improve cooling like painting the inside of the covers black, and keeping it reasonably clean.

On the other hand, the RH212 I've just installed on my recumbent in a 20" wheel is performing great at 3KW+ and I haven't been able to break 75C yet. It only has FF and nothing else. Performance is fantastic and it pulls really hard.
My Phaserunner overheats first at 60A which I need to do something about.


Cheers

Thanks...your real world data is worth more than all of my hypothesizing .

I have been thinking about a Phaserunner and mounting it on a finned copper block....and maybe a small fan to blow air across them.

[Cowardlyduck]

Thanks...your real world data is worth more than all of my hypothesizing .

I have been thinking about a Phaserunner and mounting it on a finned copper block....and maybe a small fan to blow air across them.

Thanks. Yeah, I plan to put some thermal padding between mine and my frame. I've done that before with good success.

Also, I have the L1019 connector. It doesn't seem to have a problem with 60A and I do load it down for quite a while at that level.

Cheers

[ebike4healthandfitness]

30A for the big pins in the connector and 2A for the little pins...I got the specs here: https://www.higoconnector.com/products/ ... 1000/DK99K

I would say you could double that for short periods of time.

Grin says 100A currents for short times here: https://ebikes.ca/learn/connectors.html.

I guess it all depends on how long .

Here is what ebike.ca links as the equivalent from cusmade:

https://www.cusmade.biz/Product/Detail/9

It is rated at 45 amps.

This thread also mentions it in the last post:

viewtopic.php?t=107618

It's actually a L1910 connector (19mm connector diameter, 10mm cable diameter) and it is made by cusmade: https://www.cusmade.biz/ IIRC it's mentionned somewhere on ebike.ca.
Just ordered a bunch directly from the factory. $10/pc
Came at specified length; quality is awesome in comparison to julet/higo. Nearly as good as Rosenberger RoPd - my all time favourite connector.
Manufacturers like xiongda or mxus usually agree to fit motors (if the axle is wide enough) with custom connector if you ship them to the factory. Does not work with bafang though. Mac and 9c hubs come with L1910 already fitted if you ask for it.

Worth noting, the L10 connector will carry motor phase amps, not battery amps. Phase amps are often 2-3x higher than battery amps - Grin's 40A Grinfineon controller puts out up around 110A on the phase wires, for example, and their tiny Baserunner controller puts out 80 peak phase amps.

My understanding is that phase amps might be higher than batt amp, but also at very high frequency - which means the load keeps beeing turned on and off on each phase and is actually divided between the 3 wires - hence they can be smaller than batt cables but have to be optimized to reduce skin effect.
You can't squeeze out of the controller more energy than goes in!

[ebike4healthandfitness]

30A for the big pins in the connector and 2A for the little pins...I got the specs here: https://www.higoconnector.com/products/ ... 1000/DK99K

I would say you could double that for short periods of time.

Grin says 100A currents for short times here: https://ebikes.ca/learn/connectors.html.

I guess it all depends on how long .

Here is another higo L1019 listed at 45 amps, but I think it is only for 2 pins. For 3 pins it says 40 amps:

https://www.higoconnector.com/products/ ... 099K#title

(Notice higo lists a thicker wire at 5.3mm for this one vs 4.0mm for the L10 rated at 30 amps. The cable diameter is also thicker at 11mm vs. 8.5mm for the other L10)

It is also listed as NEW under their motor connectors section.

https://www.higoconnector.com/products/ ... otor/w999K

[ebike4healthandfitness]

By the way, that 16ah battery in the simulator has 3v of voltage drop on that 25A load. that's a very wimpy battery. Select one with much bigger capacity and a lot less resistance, because this is reducing the efficiency of both motors and thus clouding the picture.

Also, try the test on a 72v battery with 0.01 mOhms resistance and a 80A batt / 250A phase. Then you will really see the differences of both motors. It's much larger than you think it is.

How much resistance for the custom controller in the second test? Also what throttle type? (Voltage, amps or torque)

[ebike4healthandfitness]

There's a big correlation in hub motors with 'square shaped-ness' of the stator and efficiency, and that has to do with how much extra end turn losses there are ( extra copper on the end of the stator that isn't doing much useful work ). The more poles, the taller the stator, the more end-turn losses you have ( compared to a totally square, low pole count motor ), and that will always show up in a dyno graph as a loss of continuous power and peak efficiency.

What about the use of stronger permanent magnets in the motor? Would that help reduce end-turn losses in motors with tall and narrow stators?

[neptronix]

How much resistance for the custom controller in the second test? Also what throttle type? (Voltage, amps or torque)

Same exact settings on both, and default on the throttle type - i didn't want to bias the test.

What about the use of stronger permanent magnets in the motor? Would that help reduce end-turn losses in motors with tall and narrow stators?

Justin has experimented with that on the all axle motor and seen that efficiency goes down a tiny bit ( ~0.5% ), but power potential increases by a higher amount..

Means that the all axle motor uses a tiny bit of extra power during cruise, but during mid-high power levels, the efficiency is notably increased because the stronger magnet potentiates what the copper in the motor is able to dish out.

[E-HP]

Justin has experimented with that on the all axle motor and seen that efficiency goes down a tiny bit ( ~0.5% ), but power potential increases by a higher amount.

I think going the other direction has a bigger effect, when a motor melts down. The motor heats up, the magnets demagnetize, the motor KV goes up, and as the rpm tries to go up, it draws more current, which makes even more heat.

[Bullfrog]

You guys realize something we have not discussed...at least I don't think we have, is magnetic saturation. And the Grin Tech Motor Simulator does not take it into account.

If anybody could get this IEEE paper I sure would appreciate a copy either electronically or I can send you my address:

https://ieeexplore.ieee.org/abstract/document/5947918

It is specifically about magnetic saturation for a BLDC ebike motor.

I know magnetic saturation occurs when the amperage you are supplying gets relatively high and the torque output no longer has a linear relationship with the amperage but I don't fully understand the parameters that lead to magnetic saturation or how to know when it is going to occur. I know how to tell it HAS occurred and its effects, but I want to understand when it is going to occur before it happens .

Thanks

[ebike4healthandfitness]

How much resistance for the custom controller in the second test? Also what throttle type? (Voltage, amps or torque)

Same exact settings on both, and default on the throttle type - i didn't want to bias the test.

Here is what I get with Leaf bike 1500 Kv set at 9.6:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.6

(Peak power for RH212 occurs at 378.5 rpm with 70.7% efficiency for 4032 watts. Peak power for Leaf bike 1500 (Kv 9.6) occurs at 426 rpm with 71.7% efficiency for 4066 watts)

Here is what I get with Leaf bike 1500 Kv set at 9.0:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.0

(Peak power for RH212 occurs at 378.5 rpm with70.7% efficiency for 4032 watts. Peak power for Leaf bike 1500 (Kv 9.0) occurs at 378.5 rpm with 68.2% efficiency for 3891 watts)

[ebike4healthandfitness]

What about the use of stronger permanent magnets in the motor? Would that help reduce end-turn losses in motors with tall and narrow stators?

Justin has experimented with that on the all axle motor and seen that efficiency goes down a tiny bit ( ~0.5% ), but power potential increases by a higher amount..

Means that the all axle motor uses a tiny bit of extra power during cruise, but during mid-high power levels, the efficiency is notably increased because the stronger magnet potentiates what the copper in the motor is able to dish out.

If you plug in those same battery and controller settings used in my previous example here is GRIN V1 5T winding vs. GRIN V2 fast winding:

https://ebikes.ca/tools/simulator.html? ... _40&hp_b=0

[ebike4healthandfitness]

You might complain about the leaf's wimpy phase wires but it's the copper wrapped around the stator that is responsible for >90% of a motor's 'power bus'. After cutting my wires, and replacing most of their length with girthy copper ( meaning that i only had a few inches of restrictive wire ), i was pushing 6kw regularly and never noticing extra heat in the thinner wire section.

If the leafmotor came with this phase leads from the fait'd be good for an additional 1-50w power at best VS us fixing the problem after the fact. Not a deal breaker for me. Also, absolutely not a main deal-maker for any motor, unless the internal wiring is horrifically undersized.

At max power I would expect phase amps (aka motor amps) to be roughly equal to battery amps. It's only at low rpm that phase amps can greatly exceed battery amps.

So the thin wires are mainly only a problem when spending a lot of time at low rpm and full throttle. (e.g. heavily loaded and/or going up a steep hill).

As long as the system (controller and battery) is strong enough and load (steepness of hill and total weight of rider(s) + bike) is light enough the phase wires shouldn't be loaded for very long.

[ebike4healthandfitness]

You guys realize something we have not discussed...at least I don't think we have, is magnetic saturation. And the Grin Tech Motor Simulator does not take it into account.

If I increase amps for either leaf bike 1500 or RH212 there does come a point in the Grin Tech motor simulator where power doesn't increase anymore. It appears that Leaf bike 1500 and RH212 are about the same in this respect. It was only until I loaded a motor into the simulator with a wider stator that I could finally get power to go up.

Volts is a different story and both leaf bike 1500 and RH212 could still increase power by increasing voltage even when increasing power via amps failed.

[john61ct]

That comes back to winding count / Kv

Which apparently is only selectable when buying a Leaf?

[ebike4healthandfitness]

That comes back to winding count / Kv

It's related to magnet and stator width. And it makes sense because bigger motors should make more power.

The outlier, of course, being the RH212 which compensates for it's narrower stator by having a taller stator and more stators ( as well as possibly stronger magnets).

[john61ct]

> increasing power via amps

is all I was talking about there, not overall power capacity

[neptronix]

Here is what I get with Leaf bike 1500 Kv set at 9.6:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.6

(Peak power for RH212 occurs at 378.5 rpm with 70.7% efficiency for 4032 watts. Peak power for Leaf bike 1500 (Kv 9.6) occurs at 426 rpm with 71.7% efficiency for 4066 watts)

Here is what I get with Leaf bike 1500 Kv set at 9.0:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.0

In both of your simulations, you have the battery and phase resistance down to nothing, and the power jacked up so we can see the true but you haven't adjusted for RPM. Watch what happens when we do this:

Upright bike ( your scenario, just adjusted for RPM differences )

https://ebikes.ca/tools/simulator.html? ... e=mountain

2021-12-20 05_51_41-Motor Simulator - Tools.png (97.39 KiB) Viewed 366 times

Now if you take that simulation and put the stator in more of a low-load, high RPM situation, IE in a semi-recumbent with 20" wheels ( better aerodynamics, very different load point on the motor ) we start seeing the achilles heel of the RH212 - the lower efficiency during cruise conditions is noticeable:

https://ebikes.ca/tools/simulator.html? ... &kv_b=9.71

20 incher.png (101.04 KiB) Viewed 364 times

This is the effect of the extra poles and higher iron losses here.

You sometimes see the opposite with a 29" wheel. The RH212 pulls ahead in efficiency in mid-high power situations. In low power situations, the leaf can often the efficiency race by ~0.1%.

https://ebikes.ca/tools/simulator.html? ... &grade_b=0

The RH212 performs admirably for it's weight in 26/29 wheels... but the smaller the load, the smaller the wheel ( when you adjust for the RPM difference ), the more the leaf shines.

You might think this is splitting hairs, but when you're pushing serious power for a long time, efficiency matters a lot. A 2% loss of 2000W creates 40W of heat. A hub motor of these sizes can shed around 200W of heat continuously without any cooling mods - we have a very small 'heat budget'... so even 1% matters.

Here's the thing that really disappoints me about the RH212.. it is a bit smaller motor, and therefore should have less iron losses than the leaf.. but instead, it has a bit higher iron losses..

[ebike4healthandfitness]

> increasing power via amps

That is what Bullfrog and I am referring to as well but in the context of magnetic saturation.

You guys realize something we have not discussed...at least I don't think we have, is magnetic saturation. And the Grin Tech Motor Simulator does not take it into account.

If anybody could get this IEEE paper I sure would appreciate a copy either electronically or I can send you my address:

https://ieeexplore.ieee.org/abstract/document/5947918

It is specifically about magnetic saturation for a BLDC ebike motor.

I know magnetic saturation occurs when the amperage you are supplying gets relatively high and the torque output no longer has a linear relationship with the amperage but I don't fully understand the parameters that lead to magnetic saturation or how to know when it is going to occur. I know how to tell it HAS occurred and its effects, but I want to understand when it is going to occur before it happens .
Thanks

Normally torque at any given rpm increases linearly with amps and therefore power at any given rpm also increases linearly.

But when magnetic saturation this doesn't happen. The only fix at that point is more powerful magnets (which can come in various forms).

[neptronix]

It's not just magnetic saturation, but also the resistance of the copper in the windings increases as heat increases. As that resistance climbs, more heat is made.. and you are well on your way to a not-so-linear death spiral.

The biggest improvement to the RH212 you could make is using 0.27mm lams instead of 0.35mm... this would make the RH212 a quantifiably better motor because it'd have lower iron/magnet losses than the leaf, and higher continuous power.. and at lower weight!

Different magnets won't move the needle as much.

[Bullfrog]

Here is what I get with Leaf bike 1500 Kv set at 9.6:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.6

(Peak power for RH212 occurs at 378.5 rpm with 70.7% efficiency for 4032 watts. Peak power for Leaf bike 1500 (Kv 9.6) occurs at 426 rpm with 71.7% efficiency for 4066 watts)

Here is what I get with Leaf bike 1500 Kv set at 9.0:

https://ebikes.ca/tools/simulator.html? ... 0&kv_b=9.0

In both of your simulations, you have the battery and phase resistance down to nothing, and the power jacked up so we can see the true but you haven't adjusted for RPM. Watch what happens when we do this:

Upright bike ( your scenario, just adjusted for RPM differences )

https://ebikes.ca/tools/simulator.html? ... e=mountain

2021-12-20 05_51_41-Motor Simulator - Tools.png

Now if you take that simulation and put the stator in more of a low-load, high RPM situation, IE in a semi-recumbent with 20" wheels ( better aerodynamics, very different load point on the motor ) we start seeing the achilles heel of the RH212 - the lower efficiency during cruise conditions is noticeable:

https://ebikes.ca/tools/simulator.html? ... &kv_b=9.71

20 incher.png

This is the effect of the extra poles and higher iron losses here.

You sometimes see the opposite with a 29" wheel. The RH212 pulls ahead in efficiency in mid-high power situations. In low power situations, the leaf can often the efficiency race by ~0.1%.

https://ebikes.ca/tools/simulator.html? ... &grade_b=0

The RH212 performs admirably for it's weight in 26/29 wheels... but the smaller the load, the smaller the wheel ( when you adjust for the RPM difference ), the more the leaf shines.

You might think this is splitting hairs, but when you're pushing serious power for a long time, efficiency matters a lot. A 2% loss of 2000W creates 40W of heat. A hub motor of these sizes can shed around 200W of heat continuously without any cooling mods - we have a very small 'heat budget'... so even 1% matters.

Here's the thing that really disappoints me about the RH212.. it is a bit smaller motor, and therefore should have less iron losses than the leaf.. but instead, it has a bit higher iron losses..

Neptronix....NOW I understand and Thank You for posting the Motor Simulator runs .

[Bullfrog]

Regarding magnetic saturation....

My limited understanding is that....When Magnetic Saturation occurs, it is a function of the material used for the poles i.e. the piece that the stator coils wrap around. And it is nothing more than the magnetic strength of the poles is no longer linear with regard to the amperage circulating in the stator windings. It is basically defined by two lines...one with a fairly steep slope where the Magnetic strength increases with increasing amperage and then a relatively sharp knee in the curve and a somewhat horizontal line where magnetic field strength doesn't increase much with increasing amperage. But of course the I squared R heat generated in the stator windings goes up a LOT hence the death spiral that Neptronix mentioned. AND lets not even talk about having a geared hub motor like my beloved MAC because it just melts down even quicker .

If somebody understands Magnetic Saturation...please correct me if there are additional significant factors besides the pole material.

[neptronix]

When people are talking about a motor hitting saturation on this forum, they're usually meaning it's in the 50% efficiency range, ie as you progressively feed it more power, it makes more heat and less power.

It describes a state where any motor is making more heat than it is making more torque, and that may be influenced by temperature ( everything's performing worse ) and/or instantaneous power ( more power than the copper can do anything with, IE more than 100% of the magnet and copper's capacity to make useful motion ).

Maybe there's a more formal definition but that's how i understand it.