drone pancake motor low reduction mid drive with regen

mechhead

10 W
Joined
Mar 10, 2017
Messages
90
Location
adelaide south australia
H all im new here but have been avidly reading endless sphere stuff for a long time

I would like to start a thread for a community developed opensource high quality mid drive kit that may incorporate regenerative breaking and is designed around the newly available agricultural drone pancake motors

I will start with a problem statement that outlines the short comings of many of the current mid drive units.

problems with currant mid drive setups; weight and quality of existing kits, motors need to be large and heavy to deal with low reduction ratios, or for the rc motors and astros, need high reduction ratios to make effective torque at crank speeds, or direct drive through seperate chain system lacking gears and compromising efficiency at low or high speeds and also lack implimentation of regenerative breaks

I like the idea of very light motors and drive trains that see the existing bike gears, it saves weight and increases efficency and for the real world makes towing cargo or children and dogs easy and still has long legs for solo missions

Enter the new class of motor that no one on here is talking about yet that IMHO holds the most promise of efficency/light weight and needing minimal reduction.
the drone pancake motors, i have ordered a smaller one to play with and test out, specs are 100kv 83/18mm stator 2.7kw 180second rating 48 poles efficient rpm band 500 to 7000 T-motor do some nice ones a u12 and u13 109mm dia by 40mm long and 118mm dia by 56mm long with the u13 rated at 3100watts 180 second . the power ratings are probably on the high side for a bike motor as the airflow cooling they experience turning a 25 to 30 inch props on a large drone is would mean very large cooling airflows. With more conservative power levels should be good for a bike drive. the best reduction would seem to be small chain (efficent, strong but noisey with lots of sprocket options) belt drive (quiet and efficent but less ratio options or precision planetary gearhead ie neugart parker apex etc (quiet, strong and efficent, fixed ratio, more compact than either belt or chain but possibly heavier) i like the planetary solution because it is so compact and potentially long lasting and with one of these axially short drone motors easy to fit between crank arms.

The pancke motors main advantage over other rc motors or astros is their low rpm efficency from the huge pole count and large diameter, if you go look on alienexpress there are heaps of different kinds to choose from.

Implimenting regen

This could be done using a friction roller on an arm that pivots about the motor axis and contacts the tyre spinning the motor backwards. the rotation being opposite when in breaking mode to driving mode allowing the reduction stages to be bypassed this would be easy to impliment with either a belt or chain drive but more difficult with a planetary setup but still could be done by diving the other end of the motor shaft. this system would allow a regenerative break with a mid drive as all reduction would freewheel in reverse.

for smaller power levels the reduction could be done with 1/4" chain driving a large sproket that mounts inside of the triple chain rings it would be similar size as the smallest 28 to 34 tooth sproket on the inside of the chain ring setand would need to be cantilever mounted to fit around the BB there is room for a sprocket for this chain size between the front derailer and the frame on many bikes and it would look stealthy hiding behind the chain wheels just above the bottom bracket where on the other side there is room for the other stage of reduction and a regen arm and roller that engages the rear wheel with a cable paired to the rear brake.

I can picture an extended shaft on the pancake motor that drives a planetary gb through the motor base and a small tooth belt or chain sproket on the other side(left) of the motor and planetary box that goes to the arm that carries the wheel breaking roller.

for minimal power levels 200 to 750watts a 50mm planetary gb could be used, for higher power levels a larger 60 or 80mm planertary box would be needed
due to the low reduction requirements 10:1 planetary single stage and 2:1 to 4:1 final small chain drive of chain rings. there are several neugart threads already but there are many other high strength precision planetaries to choose from although none of them are cheap although there are many choices of used ones on ebay .

A 10:1 planetary and 3:1 final drive would give 30 :1 total reduction to the chain rings making 2700rpm motor to 90 rpm pedals

any feedback would be welcome
 
Welcome to the forum. I am all for choices and moving forwards.
It seems you might try to reinvent things here. Your motor of choice will still need reduction as far as I can tell. Well if reduction is needed why would neugart gearboxes be a better choice then say Astro 32xx in a Davinci drive? Because Matt uses left side drive on the Davinci's? Do you think that will render drive and motor "less usable" somehow? If you need more, just add another motor. Correct gearing in the first place will make the need for a crank drive obsolete. And that is a good thing, as you can pick and choose the correct chain and sprockets quality to fit your power output. That brings longevity and durability.
 
What about belt reduction ? That would be more quiet and leave out the need for oil inside of a gear box ... Yes ?

>
" Correct gearing in the first place will make the need for a crank drive obsolete "
>
? are you talking about a separate drive , sprocket on the disc brake side behind or in place of the disc brake rotor on the back wheel ?

do you have any link to someone using RC motors that do not use crank drive ? ( not including tire friction drive )



macribs said:
Your motor of choice will still need reduction as far as I can tell. Well if reduction is needed why would neugart gearboxes be a better choice then say Astro 32xx in a Davinci drive? Because Matt uses left side drive on the Davinci's? Do you think that will render drive and motor "less usable" somehow? If you need more, just add another motor. Correct gearing in the first place will make the need for a crank drive obsolete. And that is a good thing, as you can pick and choose the correct chain and sprockets quality to fit your power output. That brings longevity and durability.
 
Yes I do, look up ES member recumpence he has done several hi powered build even past 30 kw peak power and he does not use crank drive. He is the maker of the davinci drive, http://davincidrives.com

Several other ES members has done build with his davinci drive and astro motors. You should get the Astro motor(s!) from Matt (recumpnce) when ordering the drives. That way you get fan cooling, stronger axle etc. It also comes with a slipper clutch built in, to take away some stress on the whole drive train.

Davinci drive to the left and his v4 drive to the right. Notice how the davinci drive utilizes the whole drive as a heat sink. Nice detail.

nbt0iiG.png
 
Hi Mechhead,

Thanks for bringing these new motors to our attention.

While I agree that Matt / Recumpence makes excellent drives using Astro motors, that is no reason to ignore a new class of motor that could be worth attention.

I have looked up the T-motor website and researched the U13 particularly. It does seem to have three advantages over many other motors that have been used before, as you state:
1) The large diameter and high pole count should give inherently higher torque and lower kV (rpm per volt) than many other RC motors (having said that, my understanding of the fine points of brushless motors is only moderate, and I would like to see comments on this thread by some of the more expert members). This means less reduction between motor and wheel would be needed than with a higher revving motor;

2) The large diameter and short length, together with the very open base and rotor end cap would allow excellent cooling of both stator and rotor if air could be induced to flow through the motor. Sure, in RC drone use there is a propellor attached to the motor, but if you look at the props on the T-motor website, the central portion that is within the diameter of the motor has little if any pitch, so would not directly be causing significant airflow through the motor. Movement of the drone through the air would help of course.
In e-bike use, a machined centrifugal fan attached between the top of the rotor and the belt pulley / chain sprocket / gearbox input could create a lot of airflow. Alternatively a small electric fan could blow air through a duct to the base of the motor. This could be controlled by a temperature sensor in the windings so the fan would only run if the motor got hot; and

3) The short overall length of the motor and the existing drive flange on top of the rotor would make it very easy to attach transmission components to the motor and still retain a narrow overall width (maybe even two different diameter belt pulleys and two belts, allowing a two speed drive to a countershaft equipped with dog clutches??)

The power handling of the U13 motor is stated to be over 3kW for 180 seconds, presumably with good cooling airflow. Possibly not enough for massive hills at continuous high speed, but certainly enough for some of us..

Check out https://www.dronevibes.com/forums/threads/t-motor-u13-100kv-cf-29x9-5-3-blade-prop-review-thrust-test.31750/ where one of the videos shows one of these motors on a fixed test stand (no airflow except from the propellor and the profile of the endcap of the rotor) for 12 minutes at full throttle , which seems to indicate good power handling without excessive temp rise.

I'll be watching this thread and hoping someone takes the plunge and tries one of these motors. After getting a couple of other projects out of the way, I could be interested in trying one myself.
 
Each type of motor has its advantages and disadvantages. For different uses, the same person might select different motors of the same power.
For example if I was building a mid-drive cargo bike for use on smooth sealed roads I might choose a hub motor, or a lower kV motor like the BHT or similar.
But for a light, nimble full suspension offroad bike for singletrack and exploring off track, I would want the lightest, most compact drive system possible even if it meant two-stage reduction was needed.

So not saying one is better than the other, but for some situations I believe this motor would be better than a hub motor because it is lighter, smaller and has very much better heat dissipation potential.
 
Drum you get it ultra light weight and these flat pancake motors have excellent cooling ability curtesy of their very short stator laminate stack and their high pole count. These two things together make many short paths through the windings through which great volumes of air can move freely , with a temp sensing fan thats driven seperately to the motor i suspect they could be run stationary at their max power rate continuously even at low rpm. I have one on order ill post test specs if i can devise a way to accurately measure torque , possibly a string spool and an adjustable weight along with a watt meter could measure centimeter gram seconds or some such then the unit conversion fun begins newton/centimeters or in/lbs take your pick. I digress. there are other benefits they are axialy thin so fitting them across the frame is easier and more compact than other solutions, iI have a 8318 on order its specs are 3000w for 180seconds or 2300w continuous 100rpm/v 42 poles, lets do a bit of math here on a similar astro and compare

alien 8318 42 poles (630 grams) @22v =no load rpm 2200rpm reductionneeded to see wheel 323rpm(40kmh) is 6.8:1 this is easily achievable in a single stage with #35 chain to the rear wheel or dual or single stage reduction to crank drive.

Astro of similar power rating 3210 8 poles (963 grams) 2000w to 3000w 275 rpm/v @22v = 6050rpm no load reduction required to see 323rpm(40kmh) would be 18.7:1

there is one other side to this equation the erpm figure for each motor at the above mentioned speeds
astro 3210, 6050 rpm x 8 poles = 48400erpm
alien 8318, 2200 rpm x 42 poles=92400 erpm

at this point i would love for someone else who knows better to point out any flaws in my rekoning and math.

this raises another interesting factor take a vesc for example it recomends not exceeding 100,000erpm becaus of intricacies like rise and fall times of fets even when driven very hard with very fast switching darlington transistors .Tthe more time on the rise or pulldown a fet junction spends the more heat it makes and the less efficient it becomes.

this would mean that the alien 8318 with the huge pole count would be capable of similar efficiency to the astro at 5.25 times its speed and if an astro gives best efficiency betweem 6000 and 10,000rpm the alien would give similar efficiency at 1142rpm to 1900 and also mean that the alien would have 5.25 times more torque at a given rpm than the astro when eating the same watts

so 18.7:1 astro or the pancake @ 6.7:1. a heavier motor needing two stages of reduction to see the wheel or three for the crank ie a davinci style setup, they are drool worthy. but one less reduction stage would save more weight


An 11 tooth to 76 tooth #35 chain to drive the rear wheel or two stages to drive the crank. IMHO the lighter motor with better cooling and one stage less reduction would yield a good weight saving and efficiency gain from the loss of one reduction stage, and it makes for esier packaging. With adequate cooling air supply may out live a similarly loaded astro. Now dont get me wrong im not dissing the quality of astros, its just they are not exactly ideal for pushbike propulsion requiring such high rpms to work efficiently. now here is a thought how about getting astro to make a 150mm dia by 25mm or 50mm along axis motor with 84 poles and an out runner that can mount a sproket on the end of the spindle there are some nice flat flange mount ones available for kart motors that would do then only one stage of reduction to crank or wheel would be required add a controller with built in delta star switching and away you go with a two speed drive without the two speeds. im sure if enough people asked for these motors astro would make them to their exacting standards.

maby spinning magnets would have somthing to say about large dia pie motors with astronomical pole counts.

the other option is to gut a suitable hub motor like the stoke monkey and lighten it by cutting off all that excess metal thats not needed to suppot a wheel and see what kind of narrow low redustion middrive can be had like that but most of then are skimping on the pole count aswell.
 
Here is a new motor I spotted that direct left side drive users may like

https://wholesaler.alibaba.com/product-detail/2016-hot-sale-MP10850-50KV-7kw_60536628708.html?spm=a2700.7782932.1998701000.30.LWZT5T

what do you lot think ?
 
It may work, my biggest concern is how much magnetic repulsion you can generate. Even if you ran all 2.7 kilowatts through the thing it might not be enough gauss to get you anywhere.

I am not sure what you're responding to? There's enough data on the alibaba motor to judge the torque.

3000 rpm 7kW = 22Nm (max?)
3000rpm 15Nm (continuous?) = 4.7 kW
Figures don't make sense like they never do on Alibaba. only way to know is to buy it.
 
Drum said:
I'll be watching this thread and hoping someone takes the plunge and tries one of these motors. After getting a couple of other projects out of the way, I could be interested in trying one myself.

Hi Drum. Did you or anyone else try out one of these T-Motors? I'm looking at the U15II KV80 right now. The numbers that I understand seem incredible. It makes me suspicious. Perhaps I'm missing something.
 
Numbers give 76% efficiency at 8550w input - horrible.
To get 3 minutes of 9.9kW (which they state as max) will result in even lower efficiency: at an assumption of 70% efficiency about 3kW of heat has to be removed or it will be put into motor temp increase.

It’s too high for such a small motor, even with a propeller creating 6.5kW of airflow over it.

—>the spec is false, it’s as simple as that.
 
There's hardly anything that's lighter and more efficient than a single stage reduction with belt or chain. Unless you have special space configuration issues, or need to slave multiple motors together, it's not worth the weight and efficiency hits.

Most non crank drive, non friction rc drives basically are some variant of something like these... ( Thank you uncredited ES hivemind )

MattyCiii1-300x239.jpgdownload (5).jpeg

There was a Euro one that had twin Hacker pancake motors (maybe), with a large diameter sprocket bolted to each side of the back wheel with belt drive and a big plastic circular mud guard over everything, but couldn't find an image.
 
larsb said:
Numbers give 76% efficiency at 8550w input - horrible.
To get 3 minutes of 9.9kW (which they state as max) will result in even lower efficiency: at an assumption of 70% efficiency about 3kW of heat has to be removed or it will be put into motor temp increase.

It’s too high for such a small motor, even with a propeller creating 6.5kW of airflow over it.

—>the spec is false, it’s as simple as that.

Good to know. Thank you!
 
Voltron said:
There was a Euro one that had twin Hacker pancake motors (maybe), with a large diameter sprocket bolted to each side of the back wheel with belt drive and a big plastic circular mud guard over everything, but couldn't find an image.

Are you referring to the Aden?ADEN.JPG
 
Back
Top