Amberwolf's other "Bottom Bracket"-type Drivetrain Idea

I've been pondering this idea for a while, but never did more than sketch it up. I was looking for them earlier today, to post them up for [Team1]'s thread about the drive they're working on, but I can't find them. I'll probably resketch them at lunch today and scan them in here then. Until I do, you'll have to get the idea from the text instead.

It's kinda like a cyclone or stokemonkey setup, but drives in a different place.

I haven't seen this done yet, probably for good reasons, but:

How about driving the bottom of the chainline coming out of the cranks, so that it pulls the chain along? The drive sprocket on the motor shaft would slide along it's shaft to stay aligned with the chain as you shift gears front and back. The motor itself could be mounted to the kickstand plate area on many bikes.

A disadvantage of this idea is that it could require adding some links to some (many?) bike chains, depending on how the chain would have to wrap around the sprocket to provide drive force.

Also would still need a freewheeling crank, if you don't wanna pedal along with the motor.

It's main advantage is that it allows all the chainrings to stay on the front and be usable for the pedalling, as none are taken up by the motor chain. (assuming you don't have to change them all out because of the freewheeling cranks).

I'm not sure yet if that advantage is enough to make all the other complications worth it, but if you only had a low-power motor, and needed to be able to pedal to help up steep hills or with lots of cargo, then it would definitely be useful.

amber,

Isn't this how the normal Cyclone kit is?

It's also the way the Panasonic and Yamaha drives link up.

Miles said:

amber,

Isn't this how the normal Cyclone kit is?

It's also the way the Panasonic and Yamaha drives link up.

Click to expand...

Not quite. "The drive sprocket on the motor shaft would slide along it's shaft to stay aligned with the chain as you shift gears front and back"

I think it would work, but only if the motor wasn't powering the splined/keyed freewheel sprocket that slides on the shaft during shifting. If power was maintained during shifting i think it would grip the shaft and break/bend the chain.

Remember that there is also a chain height difference when switching between the big and small chainrings. How would you get around that? Vertical and lateral movement would be quite the engineering task for a driveshaft.

I feel like the best way to drive the cranks would to make a custom #25 chain sprocket, like 150 teeth, the size of a 60+ bike chainring, but it would be able to be correctly fitted to the normal large bike chainring. It would have to be right next to the crank arm to provide enough clearance between chains. You would then use the 15 tooth roller bearing clutch #25 chain on the output shaft. 10:1 reduction, so you'd have the shaft at 800rpms. From there, you do a belt reduction to the motor, which would be quite doable at only about 5:1, for a 4000rpm motor.

I wish I had better tools and resources to built that.

amberwolf said:

It's kinda like a cyclone or stokemonkey setup, but drives in a different place.

Click to expand...

boostjuice said:

Not quite. "The drive sprocket on the motor shaft would slide along it's shaft to stay aligned with the chain as you shift gears front and back"

Click to expand...

Obviously, I was referring to the layout.

The floating sprocket is a nice idea, in principle. It might be worth trying by just using a roller clutch - you'd need a freewheel anyway....

As for this being perhaps less practical than existing types of BB drives, well, that doesn't really matter--as an idea, I just wanted to see if I could try something different. Just because. :lol:

I'm also pretty sure power would have to be interrupted during shifting, but that's not a big deal; I do that when pedalling too, for a different reason--stop applying force but still keep the pedals rotating, so I don't jam the chain.

However, if I could integrate lateral-movement bearings into the keyway tab on the drive sprocket, I could probably even shift under power. More complicated than I could likely manage to machine, but an idea, anyway.

I know too little about roller clutches to know if they would do what I want, but I expect that's how I'd get a motor-side freewheel in there.

I likely would not use a crank freewheel with this design; I'd just have to pedal along. Otherwise it'd require a custom-made crank spider that had a freewheel at the root of the spider (unless someone makes such a thing already?), allowing a normal 3-ring set to be bolted right on.

There is no need to deal with the height change; once I get the sketches scanned in hopefully it'll be obvious (or you can point out whatever I didn't think of, which is more likely :lol. The chain passes under (rather than over) the drive sprocket, and then up and over an idler/tensioner *behind* the drive sprocket, which keeps more chain wrapped around the drive sprocket than otherwise--since it's such a small sprocket then this is the only way I could think of to keep more chain and teeth engaged for better grip and power transfer, with less tooth wear and slippage. Also keeps it from coming off the drive sprocket during bumps/etc.

As the chain shifts left to smaller rings, then since the drive sprocket stays at the same height (about even with the bottom of the largest chainring, or even lower), it gets more of an angle from it to the chainring it's pulling from, and a tiny bit more chain wrap, too, although probably not enough to even add another whole tooth's worth of wrap.

So lateral movement is all it has to do, and that's theoretically easy. The main thing I'd worry about with it is road gunk ending up stuck in the lateral keyways, which would end up like a lot of sliding door tracks do--gummed up and unable to easily slide the door in the tracks at the bottom.

I'd like to at least put a ball bearing in a pocket machined into the inside end of the key on the sprocket. Two keyways, 180 degrees apart on teh shaft, and two keys, or better three keyways at 120 degrees, with three sprocket keys and a bearing in each one. That would allow much smoother movement laterally than just metal to metal. The bearings would just be pinned in place by the sprocket/keys against the shaft. I could probably manage to hand-file or dremel the pockets in the sprockets, then harden the keys a little if I can get a hot enough torch.


The shaft itself probably would not be the actual motor shaft, as I expect it would require at least one reduction stage from motor to chain, for efficiency of operation/current draw on the motor/batteries. So this would actually be a jackshaft between the actual motor shaft and the bike chain. I expect a belt reduction is probably the most suited to such a small space, though perhaps an actual toothed-gear reduction could be even smaller (it'd probably be heavier).


Anyway, gotta get the sketches scanned in and posted to help clarify what I'm talking about.

A roller clutch might self-align well enough for small movements.

Looking forward to your experiments with this idea, amber.

Personally, I couldn't leave front dérailleurs behind quick enough

amberwolf said:

As for this being perhaps less practical than existing types of BB drives, well, that doesn't really matter--as an idea, I just wanted to see if I could try something different. Just because. :lol:

Click to expand...

That's fine, but I hope you are aware that it will not give you the same range as driving a chainwheel would. Rather, it will give you the same range, but the motor pulls chain at a constant rate, so it does not "see" which chainwheel is being used. :?

Functionally, it is the same as two chains, (one for the pedals and one for the motor) driving a hub gear.

It's still quite useful to be able to vary the ratio between the pedals and motor drive, to accommodate different cadences, though. The rear dérailleur on my Moulton set-up gives me this.

Miles said:

It's still quite useful to be able to vary the ratio between the pedals and motor drive, to accommodate different cadences, though. The rear dérailleur on my Moulton set-up gives me this.

Click to expand...

Understood. What spread of sprockets are you using there now?

JennyB said:

Understood. What spread of sprockets are you using there now?

Click to expand...

14t;16t;19t;24t That fills in the gaps of the hub gear, well. I got them as individual Marchisio sprockets from Highpath.

Hi amberwolf.

Thanks for your friendly input the other day. Hopefully I can return the favour:

I didn't understand everything you guys spoke of (my technical bicycle vocabulary needs a bit of a polish), but I think I came up with an idea which can save you all the trouble of a lateral-movment bearing for a drive sprocket getting dirty. It also eliminates the problem of the "chain height", gives a bit of a gear reduction and solves (a little bit) the "bending chain" problem too (at least I think it would ).
Though - the drive's location is a little different.
So what if you dont't make the drive sprocket following the chain movement but rather make the drive sprocket move the chain itshelf?
For that you have get rid of the original derailleur - and place your Motor there. BUT - there you don't move the drive sprocket laterally on the motor shaft - you move the whole motor laterally.
In that way you have no movable and complicated parts on the shaft and also more stability, because the bearings to move the whole motor can be much bigger and are less prone to dirt!

I made a draft which should be easy to understand, please look at the attachment.

greetings,
Gabriel

Miles said:

JennyB said:

Understood. What spread of sprockets are you using there now?

Click to expand...

14t;16t;19t;24t That fills in the gaps of the hub gear, well. I got them as individual Marchisio sprockets from Highpath.

Click to expand...

So, four direct drive gears, two overdrives and two grannies (19 direct = 14 low, 14 direct = 19 high)? Which of the duplicate gears do you favour?

To bring this back to the the original chain-pulling scheme: the only gearing that makes sense to me would be an old-fashioned half-step with a super-wide 4 or 5 cog block. For example, either 52 and 42 rings with 12, 17, 24, 34 or 48, 42 with 11, 15, 20, 26, 34. "Half-step" because the ratio between the chainrings is half that of the gap. It worked very nicely on my old Cannondale, but these days an Alfine or a wide-ratio 8-speed cassette would give nearly the same spread on a single chainring.

JennyB said:

So, four direct drive gears, two overdrives and two grannies (19 direct = 14 low, 14 direct = 19 high)? Which of the duplicate gears do you favour?

Click to expand...

If I'm using the motor, I stick with the 14t or 16t and use the hub gear. When I'm not using the motor, I change with the derailleur and keep the hub in direct, mostly. Does that answer the question?

@[Team1] Gabriel: That looks like an interesting idea; the disadvantage of it is that it would require significant space within the triangle, which I would hope to use for batteries instead. Also, some bikes don't have very much space between the toptube and the downtube, and I might not be able to fit the motor unit there (depending on it's design). My DayGlo Avenger is one of those.

But assuming I have the space, it might be a good way to simplify it, as long as a setup can be made that will be easily attachable to a regular unmodified front derailer. If the unit is placed so that it has an idler pressing the chain down below the level of the top of the derailer, so that it can still pass within the derailer, and the motor drive sprocket is forward of that with teh chain passing over it and then down onto the pedal chainrings, it would allow operation without modification to the derailer itself.

All the derailer has to do is slide the motor back and forth on a lateral bearing of some type, on the separate motor mount to the seat tube and/or downtube.

It does however have some disadvantages (besides taking up space in the triangle) I can see right now:

1--the pedal chainrings will have a lot less chainwrap than they used to, so there could be slippage under high-torque situations, on the granny gear (smallest ring) especially.

2--the motor assembly would be more obvious than if it were in the kickstand location.

3--there is slightly more chance of getting pants/etc stuck in the chain loop than before.

JennyB said:

That's fine, but I hope you are aware that it will not give you the same range as driving a chainwheel would. Rather, it will give you the same range, but the motor pulls chain at a constant rate, so it does not "see" which chainwheel is being used. :?

Click to expand...

Yes; it is still not as good as my CrazyBike2's drive is at fully utilizing the drivetrain for the motor, but at least could do as Miles says, and allows different pedal cadence to motor speed, to allow me to assist the motor in my least painful way and also to assist more on hills and such than I could with a regular Cyclone type drive.

A motor driving a completely separate chainring on the *left* side of the BB would give the same as I have with CrazyBike2: full use of the drivetrain. But it would also require modification of the existing pedal setup.

What I would like to do is get this working without any modification of an existing bike other than perhaps removal of the kickstand to allow bolting the motor unit in it's place. (I generally can't use a kickstand anyway, as my bikes end up too heavy when loaded with cargo for any normal kickstand to handle, and they either bend or break, or the load is high enough up that it just tips over anyway. ).

Miles said:

A roller clutch might self-align well enough for small movements.

Looking forward to your experiments with this idea, amber.

Personally, I couldn't leave front dérailleurs behind quick enough

Click to expand...

I'm not sure how long it will be before I can actually do any physical work on the idea, but at least it can be worked out in theory for now.

I don't use the front derailer very often, but when I need it, I *really* need it, even with the motor assist, or most especially, when the motor assist breaks down and I've got a cargo load, or something else. Or simply when I have a long way to go so I'm going to need to pedal assist the whole way to be sure I don't have to *just* pedal the last leg of it (due to running out of battery), and my knees aren't going to take that under load for long, so shifting down to a lower gear for me but not necessarily for the motor would be nice.

I realize that simply using a left side rear wheel drive plus something like Thud's 2-speed would accomplish something similar, and allow the motor to stay more efficient than with only one ratio, but I like to try different things, especially when I think I'm the first to think of whatever dumb idea it is. :lol: I haven't really searched very hard, so it's likely someone else has already done this and proved it unworkable. Until I see that, though, and why, there's no reason not to ponder yet another way to drive a bike, if it gives some advantage (however slight) or combination of features that another one does not give.

Sketches in next post.

Hi,

These kits accommodate multiple chain rings and it seems like a simpler design:

http://www.boosty.ch/en/conversatio...page=flypage.tpl&product_id=237&category_id=2

http://www.boosty.ch/en/conversation-kits/crank.html

amberwolf said:

What I would like to do is get this working without any modification of an existing bike other than perhaps removal of the kickstand to allow bolting the motor unit in it's place. (I generally can't use a kickstand anyway, as my bikes end up too heavy when loaded with cargo for any normal kickstand to handle, and they either bend or break, or the load is high enough up that it just tips over anyway. ).

Click to expand...

A quick way to test it would be to mount a Cyclone unit far enough back on a cargo bike that the chain does not jump when you change chain rings. That might even be a good idea for the long unsupported chain run on an Xtracycle! :wink:

I don't use the front derailer very often, but when I need it, I *really* need it, even with the motor assist, or most especially, when the motor assist breaks down and I've got a cargo load, or something else. Or simply when I have a long way to go so I'm going to need to pedal assist the whole way to be sure I don't have to *just* pedal the last leg of it (due to running out of battery), and my knees aren't going to take that under load for long, so shifting down to a lower gear for me but not necessarily for the motor would be nice.

Click to expand...

Been there, know what you mean! :lol:

JennyB said:

quick way to test it would be to mount a Cyclone unit far enough back on a cargo bike that the chain does not jump when you change chain rings. That might even be a good idea for the long unsupported chain run on an Xtracycle! :wink:

Click to expand...

Well, it would probably work fine without the lateral movement of the drive sprocket on a long chainrun, like an Xtracycle or my in-planning cargo longbike *might* have, but I'd like to try to use it on standard-length bikes, too. DayGlo Avenger, for instance, has been a heck of a workhorse even as a standard frame with just a single box and a rack! (but I can't count how many kickstands, including custom designs I've tried, that I've destroyed on it; gotta eventually make one that goes out from the seatpost at the top, instead).



@MitchJi: It is a simpler design, but it doesn't really accomplish what this idea is trying to do (whether it succeeds or not will be another story).


It has a few possible problems:

1--Unfortunatley it looks like it's ISIS only, which means all the older bikes with just square-taper (or even older) crankshafts would require using their cartridge and shaft. Bikes using one-piece cranks would require an adapter kit for the BB itself, too.

2--if I already had cranks I like, I have to use their freewheeling cranks, which might be longer or shorter than the ones I have (on CrazyBike2 I use 150mm cranks, rather than 170, for instance). This would also be a problem for people with shorter or longer legs than normal who use non-170mm cranks like they should.

3--If I use a wider (or narrower, though unlikely) non-ISIS crankshaft than the one it comes with, I'd have to go find a wider ISIS one, for yet more money (assuming one is made in sufficient width).

4--AFAICT, the largest chainring is used for hte motor, so it is unavialable for pedalling at higher speeds. It does show an option for 4 chainrings, so that might allow a larger chainring to be used for pedalling, but I'm unsure if it is a standard spider that I could use my ovoid (or any other) 5-arm chainrings with. It states it's a 5-arm, but clearly shows 4-arm in all the stuff I see in images/etc.

5--It still allows for catching pants/etc in the motor chain forward of the chainrings (unless a guard is built).

These problems are also problems with other "cyclone"/etc types of drive, part of what I'm hoping to workaround with this particular version, without adding too many other disadvantages to outweigh the fixes.

I wonder if the gearbox (or whatever it has) could be adapted for use with an RC motor? That's not specifically part of my requirements for this particular drivetrain, but I would like the possibility, since they seem to be quite popular and successful as ebike drive motors, once you get past the controller issues.

amberwolf said:

I wonder if the gearbox (or whatever it has) could be adapted for use with an RC motor?

Click to expand...

It's just one of these: http://szunitemotor.en.made-in-china.com/product/cMKQfWyjhnVx/China-Gear-Motor.html dolled up to make it presentable at $1100 kit price

Ah, so it isnt' even brushless. Well, that doesnt' really matter, since I'm even considering the possibility of using one of my powerchair motors with right-angle gearbox, slung under the BB just like on CrazyBike2, with the keyways dremelled into the output shaft as needed for lateral movement (it already has one). Very heavy but also very powerful, with enough torque to break chains and chainrings and pull wheels out of dropouts and stuff.

But ideally something adaptable to an RC motor for wider usage by those few that still want to use all their chainrings but don't want to modify the bike itself, etc.

From a quick look you need to get the rpms down to some reasonable number to drive the chain. A 16T drive sprocket at 300 rpms with 11-32 derailleur will give you 34mph with a 26" wheel in the highest gear. I might try a variation of T1Gabriel's idea: I would try a jackshaft arrangement behind the seat tube with a freewheeling sprocket driving the top chain. I would put an idler sprocket below the drive sprocket, between it and the front chainrings. That would improve chain wrap on both the drive sprocket and chainrings. Hook the front derailleur mechanism to the drive sprocket and idler (or the whole jackshaft assembly) and have it slide sideways to shift the chain on the front chainrings. That way you can put the motor any place - in front of the down tube, on the rear rack or where the kickstand was. Your triangle would be free. It's not as stealthy as the whole assembly under the frame and you end up with another chain with its attendant problems but it is another way to look at it - jd

amberwolf said:

for wider usage by those few that still want to use all their chainrings but don't want to modify the bike itself, etc.

Click to expand...

Would be somebody like me ^_^

Your idea has this advantage at least over a normal BB-drive.

To my suggestion for the "active+geared derailleur":

amberwolf said:

1--the pedal chainrings will have a lot less chainwrap than they used to, so there could be slippage...

Click to expand...

It was just a fast draft - I drew the drive sprocket too far to the right side of the sheet and too high, I think it is possible to lengthen the gearbox more to the lower left side, so that the pedal chainrings will have the same chainwrap as before.

But anyhow, your other points are justified, especially the one with the tattered pants - jap, nothing is perfect 8)

amberwolf said:

But assuming I have the space, it might be a good way to simplify it, as long as a setup can be made that will be easily attachable to a regular unmodified front derailer. If the unit is placed so that it has an idler pressing the chain down below the level of the top of the derailer, so that it can still pass within the derailer, and the motor drive sprocket is forward of that with teh chain passing over it and then down onto the pedal chainrings, it would allow operation without modification to the derailer itself.

Click to expand...

Did you ment it something like this?


(Dont know why it does not display the photo, i tried even saving it in a different size, changing the name...

I think through your input this solution is even better, maybe also safer for your pants

Here's also a close up (here the displaying works, odd!):

Ah, now i just saw that jdcburg was faster with the posting
But I didn't understand all of his words so I'm not sure if my draft hits exactly his explaination... maybe I have to re-read it a few times, that helps in most cases

EDIT: Nah, still don't get what he means with that additional chain an that "jackshaft"...

Miles said:

amberwolf said:

I wonder if the gearbox (or whatever it has) could be adapted for use with an RC motor?

Click to expand...

It's just one of these: http://szunitemotor.en.made-in-china.com/product/cMKQfWyjhnVx/China-Gear-Motor.html dolled up to make it presentable at $1100 kit price

Click to expand...

See my GT and Giant ebike builds in my signature. Total parts cost is under $200 (without battery). Nearly the same exact design.