Rough draft of a MelonDrive

ARod1993

100 W
Joined
Feb 24, 2014
Messages
249
Location
Cambridge, MA
Hey guys! I've started messing around with SolidWorks again, and while home sick from work I bashed out a rough draft of a drive unit designed to work with the Turnigy C80100 "melon" motor. It uses two internal jackshafts to achieve a 6:1 reduction internally, and then as of now I have it finishing with a 30t sprocket for #25 chain; pairing it with a 60t sprocket attached to the rear wheel would give you 12:1 reduction (and thus a top speed of about 50mph on a 26" bike). The gears and sprockets are all straight off McMaster, and the side plates would be custom machined (with 4.875" standoffs on the corners to hold it all together). I'd love to get feedback and user thoughts on efficiency, ease of assembly, and anything else you guys can think of!

mZ0Cz39.png

nc4cwdc.png

8lFHSq2.png


If the images aren't showing up then you can see them at https://imgur.com/gallery/4s1HTEt
 
markz said:
Looking good!

How do you mount a gear onto the motor?
Where do you purchase the gears?

Thanks; the gear in this case is a 1/2" shaft while the motor has 12mm, so I figured either a set screw or machining a keyway would do it. The gears, sprockets and shafts are all from McMaster-Carr :)

Here are a couple of images of the MelonDrive mounted in a scooter chassis alongside a Kelly KLS7218 controller and a 20s8p pack of 21700 cells:

u3TGIVH.png

S0J11Op.png


The two little boxes on the side of the pack are 20s BMSes from SuPower Battery, and the box behind the battery pack is a KLS7218 (I haven't added mounting points for it yet, though).
 
Its a nice model but since you are asking I see a few obvious areas of concern.

Efficiency:
Every reduction is another efficiency loss. You have more stuff creating friction and adding weight at every stage. Basically you have to have a really compelling reason to use a small motor and lots of heavy steel gears and shafts. Otherwise you are almost always better off adding that weight back in the form of copper magnets and battery. Those are the real heroes and all other weight is costing you.

I had an all aluminum 80100 drive setup with a gates belt primary reduction. I swapped it out for a 30mm lmx motor and the weight is exactly the same but now its just one stage to the rear wheel and only 2 bearings in the motor. Before there were 3 extra bearings not including the motor. I'm now totally a believer in the theory of get it done in the least amount of reductions possible.

Lubrication:
Are you planning on running those gears dry? If not you need the weight of an enclosure and the complexity and drag of seals on all the rotating parts.

Too much chain load:
Generally you want to get as much reduction as possible in the final stage unless you have a massive chain. A tiny final sprocket with a large rear wheel sprocket is the best way to get a reasonable bike size chain to hold up. #25 chain only belongs on the primary at high rpm to keep the load down. The slower the chain is moving to do a given job, the higher the load on it will be.

Noise:
Straight cut gears are going to make that thing sound like a chainsaw. Might actually be kinda cool.

I do like drives like this. There have been some really neat ones built here on ES in a similar style. You could end up with something pretty cool with some tweaks to your plan. Looks like you have the cad skills.
 
Are you interested in working on a 2-speed retro-direct transmission? The output shaft always spins in one direction, but by having the motor reverse, a second gear is engaged.

If low gear is only used for the rare steep hill, perhaps the gear with the least amount of friction and stages should be the second gear, for daily cruising...
 
Thanks
DanGT86 said:
Its a nice model but since you are asking I see a few obvious areas of concern.

Efficiency:
Every reduction is another efficiency loss. You have more stuff creating friction and adding weight at every stage. Basically you have to have a really compelling reason to use a small motor and lots of heavy steel gears and shafts. Otherwise you are almost always better off adding that weight back in the form of copper magnets and battery. Those are the real heroes and all other weight is costing you.

I had an all aluminum 80100 drive setup with a gates belt primary reduction. I swapped it out for a 30mm lmx motor and the weight is exactly the same but now its just one stage to the rear wheel and only 2 bearings in the motor. Before there were 3 extra bearings not including the motor. I'm now totally a believer in the theory of get it done in the least amount of reductions possible.

Lubrication:
Are you planning on running those gears dry? If not you need the weight of an enclosure and the complexity and drag of seals on all the rotating parts.

Too much chain load:
Generally you want to get as much reduction as possible in the final stage unless you have a massive chain. A tiny final sprocket with a large rear wheel sprocket is the best way to get a reasonable bike size chain to hold up. #25 chain only belongs on the primary at high rpm to keep the load down. The slower the chain is moving to do a given job, the higher the load on it will be.

Noise:
Straight cut gears are going to make that thing sound like a chainsaw. Might actually be kinda cool.

I do like drives like this. There have been some really neat ones built here on ES in a similar style. You could end up with something pretty cool with some tweaks to your plan. Looks like you have the cad skills.

That makes a lot of sense; the big restriction I imposed on myself here was on using McMaster-sourced parts and trying to keep the whole thing compact. Taking a second look at the McMaster gears available it looks like there's a 16t and an 80t gear both available in 16 pitch (which means I could run a 5:1 primary reduction (which is almost enough to get me down to where I want to be running a 12" wheel at 20s; for a 26" wheel I'd need to go down by another 2:1 or 2.4:1. Alternately an 8057/75 from Neumotor would put me just right for a 26" wheel at 5:1, and a whole lot less on a scooter.

Re: lubrication, you're right; I talked to a MechE colleague at my job and he said that if the small gear is spinning at ~9k RPM (20s at 130 rpm/V gives 9360RPM nominal) then it would need to be sealed and lubricated, which does add a bunch of drag depending on how it's done. I was wondering about maximum torque loads on a given chain size; 130rpm/V comes out to about 13.6 rad/s/V, which would give us 13.6A/Nm, so at 120ish A we'd have about 8.82Nm on the output shaft. Gear that down 6:1 and now you have about 53-54Nm on the output drive maximum. Is 48N*m input up to 96-100N*m output a reasonable load for bike chain?

I have the same question for the Neumotor; at 0.127 Nm/A peak torque is likely 25-30 Nm out of the motor (which becomes about 150-180 Nm at the wheel if geared down by 5-6:1, or about twice the torque output of the SmartCar engine). What kinds of powertrain components can I expect to handle that gracefully?

My other concern with chain/belt vs gears has to do primarily with making sure I can get decent chain wrap on the small sprocket so I don't chew through them or slip teeth; do you have advice about how best to do that when the sprocket sizes are wildly disparate?
 
Back
Top