A different kind of E-bike

[JCG]

Hi all,

I'm pleased to have found this place. I'm in the stage of figuring out what is needed for an e-bike conversion project that I'll be working on with my students, and since e-bike conversion kit retailers provide very little in the way of electrical specifics, I'm turning to the experts... and those with valuable experience.

First off, this project is a precursor to a full electric vehicle conversion (of the old GM EV-1 which was donated to my University - from purely battery power to a series hybrid). To test the drivetrain ideas in a low-cost way, I decided that an e-bike might make a good platform for learning and experimenting. Here is the basic idea.

Main source of power/main energy storage: human muscle.
Secondary source of power/secondary energy storage: ultracapacitor modules.

That's right, no batteries are planned for use in propulsion. I live in Washington, DC and commute about 3 miles each way to work on my bike. There is a lot of stop-and-go, and a lot of close calls when you don't accelerate fast enough for the guy in the car behind you. With the Ultracap Bike, I plan on using the limited capacitor storage to provide quick starts from a rest (or very low speed) and will require an electric motor with good low-end torque. It won't be used at higher speeds, but I wouldn't say no to 20+ mph while the power lasts. I believe that a 500 W motor would suffice.

The ultracaps will be two 16 V, 330 F modules connected in series, for a total bank of 32 V and 165 F. I have already received them as a donation from Maxwell Technologies. I've tested them, and they are a lot of fun.

I foresee the following issues:

1) Regen braking is a must. Not just that, but STRONG regen. I need a controller that can send up to 30 A or so back to the ultracaps when I ask it to (gradually at first, of course! Charging a capacitor is like a short circuit).
2) I also hope to be able to use regen on the fly. That is, to be able engage it while cruising on a flat slope, and pedal against the generator resistance to charge the caps slowly if they need it. I envision using a front hub motor, separate from the normal rear wheel and cassette, to do this.
3) Ultracaps, like any caps, will drop in voltage as they discharge. I'm planning on draining them to no more than half of nominal potential at any time (16 V). This would be accomplished by using 75% of the available energy (63 kJ). If my motor is cranking out 500 W, that energy should last up to two minutes in a single burst, but as the voltage drops from 32 to 16 V, the current would jump from ~15 to ~30 A, for example.

I'll need a controller that can accept that input voltage range, and a motor that can take that voltage swing as well (without using a DC-DC booster, I hope). Would a range of 16-32 V input require a 36 V motor? Should I overvoltage a 24 V motor for part of the time? Will the controller care? I've heard that the Kelly controllers can be programmed...

Asking a motor and controller to deal with such a large voltage swing, and to send large currents back in regen, is not something I've seen discussed with relation to e-bikes yet. Thanks for taking the time to read, and thanks in advance for any help or suggestions as to what to buy!

J. Ganley
Howard University

[Ypedal]

Welcome to ES !

Wow !! that is one ambitious and interesting project !!!

I will say that it's above and beyond my skill level and understanding but i get a feel for what you are after..

A very small pack of A123 cells ( Dewalt 36v lithium power tools ) would make this a whole lot simpler but what's the challenge in that ? !. hehe..

It's a candidate for a direct drive brushless hub motor, you might want to contact Justin L. at http://www.ebikes.ca regarding controllers.. the 2 of you would have one of those conversations that would make even 50 % of the members on this forum glaze over .. The unit he drove across canada may work along the lines of what you want where the throttle becomes the brake level when the brake lever is applies( variable regen using the same control as the throttle, very smooth !! )

I encourage you to fire up the digital camera and start a build thread !!! i'd love to see this progress !!!

Cheers !

[gmouchawar]

Very cool project.
I agree that a small battery would make the project much simpler.
The controller wont work from zero volts. Its regulator need to be fed for the logic part to work properly.
So from a discharged state of the caps, no regeneration is possible. You can feed the logic seperately
or put a very small AH rechargable accross the caps. The SubC RC batteries are pretty small.
That way the Low Voltage cut off would keep the logic fed and regen possible.

[JCG]

Thanks very much for the replies. We decided to avoid large battery packs mostly due to the limited current you can draw per cell without seriously impacting lifetime, and the concept of not requiring an external charger was kind of nice. My ultracap pack will be able to deliver very high currents if needed, and can go through hundreds of thousands of more charge-discharge cycles than a battery pack without degrading. Granted, the amount of energy that these capacitors store is nothing compared to a large Li-ion battery pack, but 84 kJ is not too shabby; and will last plenty of time to accelerate up to cruising speed where power demand drops greatly. This will serve well as an urban commuter bike; it sure wouldn't be so great in a long-range cruiser where stop-and-go is minimal, and you'll just be hauling extra weight that doesn't help you most of the time.

Ypedal, I'll make sure to contact Justin ASAP. I remember seeing some twist throttles that would rest in a neutral position, could be twisted down for motoring, and twisted the other direction for regen, but I assume that these will be specifically designed for certain controllers (and I can't seem to find a place that will answer controller questions!).

Gmouchawar, your point is well taken about the logic voltage input requirement... though I don't plan on discharging the ultracaps past 50% (16 V), and they can hold their charge for months, there will be times when they are completely discharged. I like the idea of keeping a constant battery voltage fed into the controller, as long as the controller can give me a steady 32 V and plenty of current to do the charging whenever I ask for it, or as much current as I can pedal out.

Here's my rough plan so far:

1) Donor bike: check. I have an old steel-frame Rayleigh C-40 to use (my old bike)
2) Ultracaps: double check. If I could find the space I could use all three of my 16 V, 330 F modules to get 48 V tops but that might get heavy, they weigh 11 pounds (5 kg) each, and 33 lbs might be pushing it. For now I'll stick to 32 V
3) Motor: brushless DC direct drive should be a good choice for regen, thanks Ypedal. I assume I should find a 24 V BLDC front hub motor; as it should be able to handle the max 32 V and won't switch off at 16 V (as long as the controller doesn't either), I guess?
4) Controller: I was looking at this one: http://tinyurl.com/6bv4yu, which I guess can fit in the voltage range and passes plenty of current. These Kelly controllers are supposedly programmable too; which sounds nice.
5) Throttle: I would love something that could control motoring and regen in one twist grip, but I have no idea where these are sold anymore, and if they would be compatible with the Kelly controller. Any advice?

I'm looking to make some purchases in a few days... I'd be happy to post project pictures and updates once I have some of the stuff ready to go.

[paultrafalgar]

JCG, welcome to the Sphere.Have you searched for Ultracaps on this forum? There has been some discussion of them here. My memory says that the general feeling was that currently they don't offer enough energy/kg or /litre to warrant use on ebikes. If you study Justin's Trans-Canada data(attached) you will see that the overall regen is 2.4%. Of course, you can make whatever type of machine you choose, but the energy/weight ratio of ultracaps currently available and the regen energy available on bikes seems to make your project a little esoteric? I hope I'm not being unfair/ talking rubbish! When (should that be if?) EEStor every produce their wondercaps, it will be a different story.
Best wishes,
Paul

[JCG]

Thanks for the comments, Paul. No doubt these are some common ideas relating to the use of ultracapacitors in electric drive systems, I'll see if I can explain why the move from batteries to ultracaps made sense in the case of series hybrids.

You are right that there is a kind of esoteric character to this particular application; as I mentioned before it makes the most sense on an urban riding circuit. Long-distance cruising has almost no need to take advantage of regeneration energy capture, unless the terrain is ridiculously hilly. For long range cruising, it's desired to have an energy source that has a high amount of storage, or high [url]energy[/url] density. For high energy, low power storage devices, it's very hard to beat something with a fuel tank (or in the very small power range for bicycle cruising, the human body): http://tinyurl.com/5s5es3

The regeneration limits we see in battery-powered bicycles (5-20%) and battery-powered cars (5-30%) are largely due to the limitation of current that can be sent back to a battery, relating to the limited power density of those devices. Since an electrochemical battery is governed by the kinetics of a surface chemical reaction, there is a finite rate at which it can occur per available surface area. That is to say, a Li-ion battery cell, no matter how many cells are with it in series, can't charge faster than about 10 A without harming it and its electrolyte, which shows up as a loss of capacity over its lifetime: http://tinyurl.com/5g7kq9. Ultracapacitors can charge as fast as they can discharge - which is extraordinarily fast. The two modules I mentioned earlier can easily provide over 3000 A at short circuit... and will come back for more many hundreds of thousands of times. A data sheet that might prove interesting is here: http://tinyurl.com/6nlrw6

Once cruising speed is obtained, power requirements are quite low (even with a heavy bike) compared to those required during acceleration. I can provide some electric vehicle simulation results (in graph form) for comparison if anyone is interested. In short, the goal of the urban e-bike with ultracaps would be to give quick acceleration to cruising speed, would provide efficient (80% in most published estimates) recovery of energy through regenerative braking (believe me, braking happens ALL the time here in the city), and would be powered by muscle at cruising speeds (~125 W). Your muscles will use about 530 food calories to take you 15 miles at that power use rate, with a heavy bike. (http://tinyurl.com/66uva) And all you need to do is stop for lunch, quicker than recharging!

I've read a few things about EEStor this year. Until they actually test a mass-producable product, I'll stick with Maxwell!

[Miles]

Hi JC,

You might like to have a look at these:

http://groups.google.com/group/nohassel?hl=en

http://endless-sphere.com/forums/viewto ... &sk=t&sd=a

[mcstar]

Here's some food for thought from someone that's been e-biking for about a year now with a realtime power meter in veiw. A good 10Ah LiFePo4 can absorb 5C recharges (50A for a 10Ah pack) for years and years with no problem at all. My current 48V pack weighs in at 22lbs and can deliver 500 watts for over 20 miles before needing a recharge on my ebike. Compared to pedalling, that's at least 2-3x's more power than I can put out with my legs alone. I can hit maybe 200 watts of pedal power for short bursts. Lance Armstrong was here in the area last week and we calculated his energy output to be about 400 watts for 30 minutes For my own riding, at a cruising speed of 24 - 30 mph, my constant power draw from the pack can easily be 10-12 amps (especially if it's windy or hilly) at 48V. Although it's definitely possible to pull more power than that on take off, in practice it rarely happens since I just don't take off that fast, and at low rpm's at my pedal input plays a larger role.

I think your idea is interesting in premise, but it may be that you'll get a much better experience using something that's lighter and better suited for the task at hand. For instance, 5AH of LiFePo4 at 36V would only weigh about 7lbs and would carry you at full speed both ways, no regen needed at all. You could get such a pack for about $200 from China.

[Miles]

I think your idea is interesting in premise, but it may be that you'll get a much better experience using something that's lighter and better suited for the task at hand. For instance, 5AH of LiFePo4 at 36V would only weigh about 7lbs and would carry you at full speed both ways, no regen needed at all. You could get such a pack for about $200 from China.

I think you're missing the point, somewhat.....

[Link]

That is to say, a Li-ion battery cell, no matter how many cells are with it in series, can't charge faster than about 10 A without harming it and its electrolyte, which shows up as a loss of capacity over its lifetime: http://tinyurl.com/5g7kq9. Ultracapacitors can charge as fast as they can discharge - which is extraordinarily fast. The two modules I mentioned earlier can easily provide over 3000 A at short circuit... and will come back for more many hundreds of thousands of times. A data sheet that might prove interesting is here: http://tinyurl.com/6nlrw6

True, caps are more power dense than batteries, but, FWIW, even the highest-powered bikes around here don't use more than 100A max. Plus with batteries you don't have to deal with the whole voltage/remaining capacity relationship.

84kJ might be decent for a cap, but it won't get you very far. 84kJ works out to about 23Wh, and that's only if you drain it all the way flat. The average ebike uses 25Wh to go one mile. I have a 72V 4.5Ah SLA pack that I used to use on one of my bikes. Accounting for Peukert and voltage sag, I can get about 500kJ out of it. This was just BARELY enough for me to make the 4 mile or so trip to work, and I have no hills, usually no wind, few stops, and was somewhat gentle on acceleration.

Might need a little more supercap, but I like the idea. I contemplated a small bank on my bike to store regen energy and use it either for acceleration or in conjunction with my battery pack for a short boost in top speed once, but tossed the idea since it wouldn't be worth the ~$200 it would have take me to do it.

1) Regen braking is a must. Not just that, but STRONG regen. I need a controller that can send up to 30 A or so back to the ultracaps when I ask it to (gradually at first, of course! Charging a capacitor is like a short circuit).
2) I also hope to be able to use regen on the fly. That is, to be able engage it while cruising on a flat slope, and pedal against the generator resistance to charge the caps slowly if they need it. I envision using a front hub motor, separate from the normal rear wheel and cassette, to do this.
3) Ultracaps, like any caps, will drop in voltage as they discharge. I'm planning on draining them to no more than half of nominal potential at any time (16 V). This would be accomplished by using 75% of the available energy (63 kJ). If my motor is cranking out 500 W, that energy should last up to two minutes in a single burst, but as the voltage drops from 32 to 16 V, the current would jump from ~15 to ~30 A, for example.

1. 30A isn't much to ask from a motor being used as a brake. I think there's a Kelly or two that can handle it without issue.

2. You're probably not going to want to pedal against the motor while it's acting as a brake unless you're really itching for a workout. Most hub motors are noticeably more difficult to propel than a normal hub just on their own. Though probably worse than most, mine drags me down to less than 10mph if it doesn't have power.

3. A motor will generally draw whatever the controller allows it to. My "500W" Golden will pull 2kW peak. The Crystalyte 5-series motors (rated for 750W) will handle 5kW+ no problem. Doc's reaching the limit on his, but he's running 100V and 100A.

You'll also see less power as the voltage in the caps drops. Since the amperage is what the controller limits, you could potentially draw 3.2kW from a 100A controller at full charge, but that will drop to 1.6kW from the same controller further down the line.

3) Motor: brushless DC direct drive should be a good choice for regen, thanks Ypedal. I assume I should find a 24 V BLDC front hub motor; as it should be able to handle the max 32 V and won't switch off at 16 V (as long as the controller doesn't either), I guess?
4) Controller: I was looking at this one: http://tinyurl.com/6bv4yu, which I guess can fit in the voltage range and passes plenty of current. These Kelly controllers are supposedly programmable too; which sounds nice.
5) Throttle: I would love something that could control motoring and regen in one twist grip, but I have no idea where these are sold anymore, and if they would be compatible with the Kelly controller. Any advice?

3. Yah. But it doesn't have to be rated for 24V. Motors don't care what voltage they get as long as it's not enough to destroy something, so any one will work. The C'lyte 4XX motors are about right for what you're looking for. Maybe a BL-36, too.

4. That controller will work. Kelly makes good stuff.

5. Me neither. Only thing I know of that works like that is the one on the Vectrix. You might be able to get one from them, but I'm not sure of its technical details or compatibility with the Kelly.

[Miles]

Most hub motors are noticeably more difficult to propel than a normal hub just on their own. Though probably worse than most, mine drags me down to less than 10mph if it doesn't have power.

I think hub motors should be ruled out, for this reason.

[JCG]

Miles - thanks for the links. I'll check those out. Re: ruling out the hub motor, it would be better to use a chain-mounted motor instead? I can't say I'm looking forward to the added complexity...

McStar - I'd love to hear more about the power meter you're using; does it have the ability to record or log the data for a trip (peak power, avg. power, total Wh, etc.)? I will need to compile this data to make plots for publication later. I appreciate all the benefits of fast-charge batteries, but... my ultracaps are here now, and I got them for free

Link - Thanks for all this. I was reading your posts in other places and was hoping you would chime in. I agree, the largest problem with the caps is the voltage drop with discharge. If the controller is ok with it (as the BYU EV-1 drag racer was: http://tinyurl.com/5w7oa5), then great; but if not there is always room for a small DC boost converter (I have some small ones on the way).

In terms of range, no problem. The caps should only be delivering power during acceleration, which should be less than 10 seconds a pop. Regen can occur whenever convenient.

Thanks very much for the motor suggestion, and I'm really glad to hear that the controller sounds good. This is the kind of direction I'm most in need of right now. I'll check on the throttle, my hope is that the Kelly controller might be adaptable to several different kinds of inputs. I'll try to find out from Kelly the usual way that regen is engaged using their controllers (unless someone here already knows). I'm starting to regret my chemical engineering training, only one circuits course...

The main idea is that the electric motor will be in motoring mode only very sparingly - the system is not designed to be an electric cruiser, just an acceleration assistor and energy capture method. I hope it doesn't sound like I'm out of shape, but biking in the city is a real pain (even when you DON'T get hit by a taxi cab). The reason you see bikers in the city running stop signs and red lights is that they just don't want to waste all the momentum they've got going. If this ultracap bike works, I might even start obeying the traffic laws myself...

[Malcolm]

Sounds like a fascinating project to me. If you ride a bike in the city most of your effort goes into getting up to speed and it gets to be a pain when you have to dump all that hard-earned momentum into the brakes every 50 yards for another set of lights. This approach is rather like having an electronic flywheel, but one that can spin up and down in milliseconds.

Miles, I reckon that Split-Pi buck boost controller you linked to would be close to ideal, definitely worth getting touch with Andy at Green Energy Technologies.

[Miles]

Re: ruling out the hub motor, it would be better to use a chain-mounted motor instead? I can't say I'm looking forward to the added complexity...

Chain or synchro belt - then you'll need a locking freewheel or a clutch.

Otherwise, the drag from a direct-drive hub motor will spoil your cruising.....

[JCG]

Malcom: right on. I hate stop signs with the fire of a thousand suns.

Miles: The motor drag got me thinking. Is there a rough estimate of the amount of force to be overcome by the hub motor drag at cruising velocity, or a guess at how much power is sapped by overcoming it? It may be possible to pedal the bike normally while drawing only a very small amount of power from the caps.

Link, if you pedaled your bike, and (assuming you had the ability to run at a very low throttle) supplied just enough juice from your battery to make it feel like you were riding a normal, unmodified bike, can you guage the amount of current you'd be using for a given pack voltage? We can use that to figure out a value for the parasitic power loss.

[Miles]

Miles: The motor drag got me thinking. Is there a rough estimate of the amount of force to be overcome by the hub motor drag at cruising velocity, or a guess at how much power is sapped by overcoming it? It may be possible to pedal the bike normally while drawing only a very small amount of power from the caps.

Maybe this will be a guide: viewtopic.php?f=3&t=7223

[JCG]

Miles: The motor drag got me thinking. Is there a rough estimate of the amount of force to be overcome by the hub motor drag at cruising velocity, or a guess at how much power is sapped by overcoming it? It may be possible to pedal the bike normally while drawing only a very small amount of power from the caps.

Maybe this will be a guide: viewtopic.php?f=3&t=7223

This is good stuff. Thanks Miles.

[dogman]

How can you get regen if you have a freewheel type motor? Won't the motor just freewheel instead of generating?

I also find the motor ( brushed) drags me down a lot if not powered up. I can peadle maybe 10 mph with no power, but a tiny bit of throttle, enough to go 3 mph or so, and I can peadle much easier to about 15mph.

If you want a 24v brushed hub motor, they are getting rare, everybody is going brushless, but they still show a 24v brushed kit on the Werelectrified site, the Wilderness energy supplier. It may be a good idea for you to buy a kit, from them or Justin, and fool with it some to get an idea what a battery powered bike does, and then start modding. Cruising at 24v would be pretty efficient, but a cap to give you a burst of 60v at times would be cool.

[Miles]

How can you get regen if you have a freewheel type motor?

You can't.

[johnb]

Miles: The motor drag got me thinking. Is there a rough estimate of the amount of force to be overcome by the hub motor drag at cruising velocity, or a guess at how much power is sapped by overcoming it? It may be possible to pedal the bike normally while drawing only a very small amount of power from the caps.

Maybe this will be a guide: http://endless-sphere.com/forums/viewto ... f=3&t=7223

This is good stuff. Thanks Miles.

Those figures are not the same as the pedalling input you would need to overcome drag though (much lower speed etc)

I find that I can pedal with my 409 and hardly notice any drag.

[Miles]

Maybe one could get an accurate value by driving it with another motor, or weights and string.....

"I find that I can pedal with my 409 and hardly notice any drag." What figure do you put on that?

Anyway, it's the input power needed that we're concerned with, in this case.......

[JCG]

I think I see what everyone is talking about now after reading some stuff on ebikes.ca:

Common to all direct drive hub motors is that they are always mechanically engaged. This is both a good thing and a bad thing. It is good because it makes regenerative braking possible. All direct drive hub motors can be made to do regen whether they are equiped to do so or not. Both the BionX and TidalForce both have sophisticated controllers which allow varied braking energy to be stored back in the battery, while most of the Chinese kits like Crystalyte, Golden Island, and Wilderness Energy need 3rd party controllers to do so. Always engaged is a bad because it means that you are always overcoming the rolling resistance of the motor even when you are not using it. Depending on the symmetry and quality of the motor, this additional drag torque can range from imperceptible to feeling like you are always riding with a flat.

Why is there always a good thing with a bad thing? It looks like I need to poll to see which hub motor has the lowest rolling resistance. Is there a favorite out there?

[Link]

Link - Thanks for all this. I was reading your posts in other places and was hoping you would chime in. I agree, the largest problem with the caps is the voltage drop with discharge. If the controller is ok with it (as the BYU EV-1 drag racer was: http://tinyurl.com/5w7oa5), then great; but if not there is always room for a small DC boost converter (I have some small ones on the way).

In terms of range, no problem. The caps should only be delivering power during acceleration, which should be less than 10 seconds a pop. Regen can occur whenever convenient.

Thanks very much for the motor suggestion, and I'm really glad to hear that the controller sounds good. This is the kind of direction I'm most in need of right now. I'll check on the throttle, my hope is that the Kelly controller might be adaptable to several different kinds of inputs. I'll try to find out from Kelly the usual way that regen is engaged using their controllers (unless someone here already knows). I'm starting to regret my chemical engineering training, only one circuits course...

Nah, it's not the controller that's the problem since it can run even if it hits LVC, it's that a cap's discharge curve is kind of inconvenient for driving a motor. At low speed when the cap's voltage is high, it's going to be very inefficient and waste a lot of power. After it eats up some of the charge and the cap's voltage drops, that's going to kill your potential top speed. Other than a high-powered DC/DC converter, there's no way around this. I guess it's fine if you're only using it to help acceleration, though, but don't expect it to be able to assist in cruising for long.

IIRC, there's two ways to engage regen on the Kelly, both of which are available through the settings. One simply activates regen whenever the throttle is set lower than what the motor's currently drawing. The other, I believe, relies on a switch/button that turns the throttle into a brake. Key phrase being "IIRC", though.

I know what you mean about the city. I have it fairly lucky, since my bike is about fast enough to keep cars happy on most roads (although another 5mph wouldn't hurt) and I don't have all that many stoplights. My 40A controller keeps my acceleration roughly equal to what a car will usually do from a stoplight.

Link, if you pedaled your bike, and (assuming you had the ability to run at a very low throttle) supplied just enough juice from your battery to make it feel like you were riding a normal, unmodified bike, can you guage the amount of current you'd be using for a given pack voltage? We can use that to figure out a value for the parasitic power loss.

I won't have any meters on my bike until I free up a few pins on the D-sub connector by converting to a sensorless controller, so not really. However, the primary cause of drag on a direct-drive motor is eddy currents in the stator. This can be taken care of with minimal power.

But this reminds me of another problem with a capacitor's discharge curve: A motor will not want to spin faster than the voltage with which it is fed, meaning that the speed you can pedal at (powered or not) is directly proportional to your remaining capacity. If, say, the motor would top out at 25mph at the capacitor's full voltage, it's going to try to recharge the caps if you try to pedal faster than about 13mph. Only way around this is a big diode preventing power from flowing backwards into the battery, but that will kill regen, too, so you'd also have to have a relay to provide a pathway around that diode that's activated when regen is.

Something to think about.

Why is there always a good thing with a bad thing? It looks like I need to poll to see which hub motor has the lowest rolling resistance. Is there a favorite out there?

The 4XX motors aren't too hard to pedal without power. Same for the BD/BL-36 ones for Wilderness energy.

The model Golden Motor sells from their site is probably the worst, but there's another you can get from e-bikekit.com that I'm going to try out soon that seems to spin more freely. An X5 would be serious overkill for a capacitor powered bike and probably be hard to pedal, so I wouldn't go with one of them.

[johnb]

Maybe one could get an accurate value by driving it with another motor, or weights and string.....

"I find that I can pedal with my 409 and hardly notice any drag." What figure do you put on that?

Anyway, it's the input power needed that we're concerned with, in this case.......

The 40x motors have a cogging torque of about 0.5 to 1 N-m which translates to about 10-20 watts at 200 rpm. The 409 has the lowest cogging torque.

[Miles]

The 40x motors have a cogging torque of about 0.5 to 1 N-m which translates to about 10-20 watts at 200 rpm. The 409 has the lowest cogging torque.

Thanks, John

If you're normally using between 50 and 100 watts to cruise along under pedal power, that's not insignificant...

Running the motor to neutralise any drag would use more power than that, too, hence the reference to the no-load figures.

I think, if it was my project, I would use an external motor and a clutch.