Using RC motors on E-bikes [Archive]

[recumpence]

Thanks for the input.

I do not mind spending upwards of $1,000 on a motor if it will do what I want.

The Lehner inrunner would only be run at far below its design output (maybe an average of 600 watts for this 11,000 watt mootor) with the occasional 18,000 RPM, 6,000 watt blast. So, I am not too worried about heat build-up. (Oh, I am not sure you can get the 3080 with a fan, I think the fan is only for the 22 series motors).

My AXI is a 63mm wide motor. I only have about 70mm clearance total. the Terminator is 77.5mm diameter. I would have to do alot of machining and modding to make it fit, but I can do it. The lehner is 60mm wide. It will fit right in. Even the bolt holes are the same location as my AXI. Besides, I would be running the Lehner at far below it's rated capacity. However, that being said, even with a low KV, I would be running it at an average RPM of about 7,000. It would be run as low as 4,000 RPM at times and as high as 15,000 to maybe 17,000 RPM for short runs at top speed.

However, that being said, I do not mind doing the machining deeded to make the terminator fit if that is the best option.

Matt

 

[Jeremy Harris]

Those Lehner motors look interesting, especially as they have a water cooling option. In many ways, a low rpm/volt, water cooled, in-runner like this is well-suited to a high-power bike application. The ability to deliver up to 11kW and keep the motor cool for decent reliability and efficiency sounds good.

All we need now is a controller that will drive it, for a reasonable price.

Jeremy

 

[Miles]

Found the data on the Lehner 3080 / 15

Lehner 3080 / 15
Kv 356 rpm per V
No-load current 4 A (@40 V)
Winding resistance 0.009 Ohms

Plettenberg Terminator:
Kv 215 rpm per V
No load current 3.7 A (@11 V)
Winding resistance ???

 

[safe]

Multiple Gears... the best option?

I will state without reservation that since I own a multiple speed bike with a small electric motor that in my own case I just can't imagine trying to get by with any type of one speed whether it's a hub motor, geared hub motor or geared down RC motor. The benefits of gearing are huge in that it allows the motor (whatever size it is) to operate in it's preferred part of the powerband and this definitely reduces heat, increases overall power and efficiency.

For some reason there is lack of "linkage" in these discussions about the relationship between the motor and the multispeed gearing. It's the synthesis of both ideas (small RC motor and multiple speed gearing) that makes for the maximum design. If you subtract one element then you are going to get less from this.

Does anyone seriously argue for "one speed" ebikes of any variety?

 

[Miles]

Well, when you're talking about 2kW plus motors, multi-speed gears might be difficult to handle

 

[docnjoj]

Maybe time for Nuvinci! Gotta cut the weight though!
otherDoc

 

[TylerDurden]

Maybe time for Nuvinci! Gotta cut the weight though!

+1 Too heavy.

 

[docnjoj]

With 2KW, who cares about bike weight. I gotta eat some more fried eggplant!
otherDoc

 

[John in CR]

Maybe time for Nuvinci! Gotta cut the weight though!
otherDoc

Can the Nuvinci handle the kind of torque involved? I thought it was a pedal power transmission. I have to assume that standard bike chain, derailleurs, and geared hubs aren't up to the task, so what are the other options? Maybe a motorcycle gearbox could be converted and remain compact. Next week I hope to have my electric motorcycle running, on which I'm using a Comet torque converter for a 3:1 CVT gearing range, but it's heavy and bulky for use on a bike. It's seems that a simple, light, and small hi/low transmission is all that's needed. All ingenious ideas are welcome.

John

 

[John in CR]

Maybe some type of freewheel mechanism that can be locked and unlocked at will. Then run 2 chains at different gearing ratios and switch back and forth between the two as needed. The 2kw scooter hub motor that has a high and low set of windings is looking better and better all the time. If I could just get it running.

 

[Kraeuterbutter]

@Jeremy Harris:
hmm.. if watercooling is THAT cool ?
i remember the tests of a rc-boat-guy (i think it was with Lehner or Kontronik-motors)
he found, that a good aircooling (running THROUGH the motor, cooling windings and magnets, bringing the heat out)

worked better for him than water cooling were only the case of the motor is cooled, and the heat has to work through the windings to the outside case...

so watercooling seems to be not always necessarely the better solution (on BL-Motors)

====================================================================

@Matt:
i would like to see your kind of mod with european legal paremeters (250Watt, 500Watt peak, 25km/h)
with 10times less peak power the whole thing (motor, controller, and most notably the weight of the gearing/belts (less torque) could be lighter..

would be interesting how light a european street-legal version could be made.. and how small (maybe not much smaller, but for sure lighter)

i mean: 250Watt input power -> that can be done by an 200g light outrunner easily
i personally would be pleased by a assistance as strong as iam (equal to my poweroutput, so the electric motor would be like a partner to me)..
not 10times stronger, that would mean: my afford on pedaling is worth only 10%, like nothing


=====================================================================

@SAFE:
yes, multiple gearing
i also have often heard (in RC forums) that electric motors does not need multiple gearing and even do not benefit from it..
thats nonsense..
in rc-hobby they often put - for my taste - too strong motors into cars to get the speed and torque they want
result: expensive controller, expensive batteries, and often too much power and so broken drivetrain

i have put in my rc-car with 2 gears a small motor.. i have enough torque for wheelis in 1st gear and nevertheless the same topspeed in 2nd gear like one of the over-motorrized vehicles..
so the little motor greatly benefits from the gearing..
with no gearing it would be slow with torque or fast but no torque

i think on slow bikes (the street legal european bikes up to 25km/h) gearing is not that neccessary
you don´t drive unter 7km/h, more like 10km/h and at 16km/h is peak performance at than assistance gets less and less until 25km/h

on your faster bikes (40km/h ? or more) i think gearing becomes more and more important
a hubmotor going 40km/h flat -> how well should this performe with 10km/h on a steep hill ???

a hub-motor like the SRAM sparc would be cool (integrated hubgear for the biker) but the motor should use that gears as well

 

[Kraeuterbutter]

wah..
so fast - while iam typing several new postings..

The 2kw scooter hub motor that has a high and low set of windings is looking better and better all the time. If I could just get it running.

the Lehner Motors have also a "electric gearing" included..
you can use them in triangle or in star-configuration

 

[docnjoj]

Maybe time for Nuvinci! Gotta cut the weight though!
otherDoc

Can the Nuvinci handle the kind of torque involved? I thought it was a pedal power transmission. IJohn

Dunno! They say 130 ft/lbs. Thats a lot!
otherDoc
Whoops, sorry! thats Nm! 100 lb/feet, still a bunch of torque!

 

[dirty_d]

what if you make the chain go straight from the chainring to one of sprockets in the back thats out of the way, like the biggest one, then run the motor chain though the derailleur, so that way you would at least have 8 gears, you would still need some kind of freewheel for the pedals though. pedals and a motor a such a pain in the ass to get to work together with gears for both.

 

[safe]

My logic goes something like this...

If you have a motor of a given power output if you add multispeed gearing to it so that it's low ratio to high ratio is roughly two to one you improve the power across the speed range by something like 25% and the efficiency about 25% while reducing the needed peak torque and therefore the heat production by 33%.

These numbers could go higher and lower depending on specific circumstances (obviously) but as a rough estimate I think they are going to fit pretty well. It's the "ballpark estimate" idea... not exact, but good to think along the lines of.

So if we can reduce the motor size by increasing the rpm using the RC motor and then use multiple speed gearing to improve the situation further still we start to get a motor that is extremely light weight because the instantaneous demands on it get to be so low. For each revolution of the motor it's being geared to a speed that is "just right" to deliver the optimal result. If you can further connect the motor rpm so that it's in perfect harmony with the pedal rpm and integrate those two power sources you then approach the perfect system.

I know I'm setting a very high bar to clear, but it's good to imagine perfection first and then fall short of it in practice as practical issues surface. In practice I'll tend to go for the cheapest solution to try to implement the abstract ideal of perfection.

The dynamic tension of ideas ("cheap" verses "perfect") is the foundation of the things that succeed. If you ignore the bottom line you fail because your product ends up unaffordable. If you ignore the technical search for perfection you end up with a boring product that can lose in the market to something more attractive.

So that's my two cents worth anyway...

 

[dirty_d]

oh how i would love a 95% efficient CVT with a 500% gear range thats the same size as the motor. wire that bad boy up to an automatic electronic shifter that monitors the motor voltage and rpm to keep the motor running at peak power or efficiency for whatever speed/load you are at... maybe in 2020. does anyone think something like that could ever be built? how efficiently do those toroidal CVTs run? they use some kind of traction oil to transfer torque instead of metal-metal contact.

it would probably be a lot simpler to design a motor with reconfigurable windings, i.e they can switch from being a huge low resistance single turn around the poles to many turns of thinner wire or anywhere inbetween, its definitely doable, you can wrap a bunch of single strands around the poles, and terminate them on a circuit board where they are reconfigured with FETs depending on how you need the motor to run, im just not sure if all that extra resistance from the FETs would eliminate any gain you get from being able to "shift" the motor.

 

[Jeremy Harris]

I think that the reason water cooling MAY have a slight edge is primarily to do with maintaining motor efficiency at high power levels. If using a simple water cooling system (and I'm thinking along the lines of an adapted PC over-clockers type water cooling kit) halves the winding temperature at maximum load it might give an appreciable increase in motor efficiency, by reducing copper losses.

With an in-runner, like the Lehner, the water cooling pipes can be closer to the windings, with an out-runner this is much harder to engineer. Water has a very much high specific heat capacity than air, so will be much more effective at getting heat out of a small space.

Whether it would be worth the added complexity and energy penalty of the cooling system is debatable, but it might still make for a more robust overall system.

Jeremy

 

[Kraeuterbutter]

hmm.. maybe is my English to bad..
i didn´t get this with the two windings..

do you mean that there is a winding for fast spinning (high power) with thick, few windings

and then there are for slow spinning (less power at given voltage but more torque/volt) with thinner, more windings

so two seperate windings, with two different wires..
but that would mean that only always half of the copper in the motor is used = bad efficience..
but you said something with FET.. that sounds mystical to me, how you want to make this ??
(simulate a thick wire by paralleling the thin wires from the slow spinning mode per fets to one thick wire??)

===========================

for the shifting: i think all technics and knowledge are already developed and here
there are some people that are building there own brushless-controllers..
for example: http://www.aerodesign.de/peter/2001/LRK350/SPEEDY-BL_eng.html

i think people, who can build a brushless-controller would also be able to add features to it, like:
when the rpm of the motor goes down and the ampers up, that the controller at a certain point/boundary
gives a signal to a Servo (from the rc-hobby-world) that shifts a modified 3 speed hub-gear into a lower gear...
the only problem would be: shifting under (heavy) load..
maybe the motor should go of power 1/10sec before shifting, and after shifting - the controller measures motor rpm and wheel-rpm - it softly powers up again, so that the shifting is smooth...

i think if somebody with the knowledge would put some effort on this that should be not that big of an problem...

when the rpm goes up an up and the load (amperes) becomes lower and lower, on a certain point:
motor power down (so that no load on the gears)
shift up to next higher gear
motor power up to the point were rpm is slightly lower than rpm of wheels given, and than softly "fade in"
so you don´t feel a hard twitch

 

[Kraeuterbutter]

@Jeremy Harris:

i see what you mean..

but i didn´t want to compare a outrunner with an innrunner and water (or air) cooling

i meant: a outrunner has benefits when it comes to temp compared to an inrunner without cooling (like most inrunners have been the last years)

NOW there come inrunner with aircooling (Lehner has introduced them this year,
Kontronik has introduced them this year, see: http://www.kontronik.com/Grafik2008/Kira600beidegross.jpg)

i also said: maybe this will change something in the heli-scene.. over the last years the outrunners have taken over completley the market in rc-helis, you don´t see any Inrunner (lehner, Kontronik, hacker) in an heli
maybe this will change know with the cooling..

for water vs. air:
here i meant:
a water-cooled inrunner (cooling from the outside) vs. a air-cooled inrunner (air flows through the motor..)
here has a friend noticed, that the new Kontroniks with aircooling stay cooler in his boat, than he has managed before with watercooling (from the outside of the motor)
(Best would be a combination of both, cooling of the windings from the outside with watercooling, and from the inside with aircooling)

besides: windings hold 200°C and more before they burn, the magnets (neodym) would get ruined when they exceed 110°C
so cooling with air through the motor cools also the magnets.. cooling the windings from outside with water will not have that much effect on the magnets which are the fragile part in the motor

 

[John in CR]

hmm.. maybe is my English to bad..
i didn´t get this with the two windings...

Here's a pic of the inside of the Erato hub motor I have. It has 18 coils and 20 magnets. The 2 pics are with the hi/lo switch in it's 2 different positions. The yellow plate that rotates slightly is connected to a rod that comes out of the opposite end of the shaft from the hall and phase wire duct. It seems like an elegant solution to achieve gearing via a manual switch. In low mode, each of the phase wires connects to 6 coils, giving it the most windings and highest torque. Then in high mode, each phase wire appears to connect to only 2 coils enabling high rpms.

If all coils have connections running to the outside of the motor, I'm sure a similar result can be achieved just by a means of switches or a specialized controller. With 18 coils it seems that a 1st, 2nd and 3rd gear is possible, but a 2 speed would only need 12.

 

[dirty_d]

Kraeuterbutter, you know how some motors use multiple strands of thin wire in parallel to get the same effect as one strand of thicker wire, but being able to fit more wire around the stator poles. instead of just joining the ends together in parallel permanently, you could have them go to a controller board. the board could switch them all into parallel so they act all as one wire, giving very low resistance and very low Kv for "low gear", and it could progressively switch them into more turns of effectively thinner wire. for example if there were 8 separate wires, in the lowest gear, they are all in parallel, so its like 1 turn of very thick wire, 2nd gear would have 2 sets of 4 in parallel with the ends in series effectively giving you 2 turns of thinner wire, 3rd gear would be 4 sets of 2 in parallel, giving you 4 turns of yet thinner wire, 4th gear would be 8 sets of 1 wire, giving you 8 turns of the thinnest wire, this would be the highest Kv configuration. im not exactly sure how you would do this electronically, it seems pretty difficult managing all those connections.

 

[Miles]

It would be really cool, if this were possible. It still wouldn't be a substitute for mechanical gears but it would be much better than field weakening.

On multi-stranding for packing density:
http://www.rcgroups.com/forums/showthread.php?p=9658190

 

[eP]

It would be really cool, if this were possible. It still wouldn't be a substitute for mechanical gears but it would be much better than field weakening.

On multi-stranding for packing density:
http://www.rcgroups.com/forums/showthread.php?p=9658190

You are wrong Miles. I wouldn't be cool anyway. It is a crap solution.
Iron loss depend on flux density and flux frequency (poles count and rpm).
So you are unable to lower iron loss this way.
The only one thing you are able to do that way is increase the copper loss, but i'm sure you don't want do that.
That is the reason this is a crap and cannot help anyway.

Best regards

 

[Miles]

Piotrek,

Are you sure you understand what dirty-d was proposing. It seems pretty unlikely that it's feasible but, if it was, I don't understand your (non) argument. Could you explain for us?

 

[John in CR]

It would be really cool, if this were possible. It still wouldn't be a substitute for mechanical gears but it would be much better than field weakening.

On multi-stranding for packing density:
http://www.rcgroups.com/forums/showthread.php?p=9658190

You are wrong Miles. I wouldn't be cool anyway. It is a crap solution.
Iron loss depend on flux density and flux frequency (poles count and rpm).
So you are unable to lower iron loss this way.
The only one thing you are able to do that way is increase the copper loss, but i'm sure you don't want do that.
That is the reason this is a crap and cannot help anyway.

Best regards

What about the way Erato handles it? Is the iron loss reduced by 2/3rds by reducing the poles count by 2/3rd for the high speed setting to enable much higher rpms?