2-speed retro-direct roller drive

[Miles]

I couldn't resist........

Astro 60 brushless motor.

5M synchro belt and 1.5 Mod spur gears (large gear POM).

Two one-way bearings (in opposition) on the motor shaft. Two one-way bearings on the roller shaft.

Reductions: 1.9:1 (belt) & 3:1 (spur gears)

Gear range: 158%

Pivoting mech. still to work out....

 

[spinningmagnets]

Great drawing, Miles. Looks very "do-able" on the mechanical end. Based on your recent posts, it sounded like you and Jeremy Harris were closing in on the motor reversing "on the fly" solution.

Many of the parts that I am using for a friction-drive are also used for a Mud/Bubba left-side-drive (the results of their choices gave me the nudge to also go in that direction). I can see here, that that this motor/trans module of yours would also be easy to configure as a 2-speed FD for medium power, and a 2-sp LSD for a kit with higher power levels.

 

[Miles]

I can see here, that that this motor/trans module of yours would also be easy to configure as a 2-speed FD for medium power, and a 2-sp LSD for a kit with higher power levels.

Indeed. I originally posted this design with a sprocket/pulley as output to a second stage:

http://endless-sphere.com/forums/viewtopic.php?t=8919&p=136337#p136337

 

[Miles]

I chose the two drive ratios that had almost exactly the same centre distance (Specified C.D. for the belt drive is only 130 microns greater than that for the gears! ).

 

[GGoodrum]

This is a great idea. All you need to do is figure out how to mount the drive so that you can have it "auto-engage", like Keplers.

 

[Hillhater]

I chose the two drive ratios that had almost exactly the same centre distance (Specified C.D. for the belt drive is only 130 microns greater than that for the gears! ).

brilliant idea... but you could be more "open" with the ratios, and simply run a tensioner on the belt ?

....it sounded like you and Jeremy Harris were closing in on the motor reversing "on the fly" solution

that would not be necessary for this system on a "kepler" type application, would it ?

 

[Kepler]

Very clever idea. I think it lends itself more to an EVtod, style sliding plate arrangement with the roller constantly in contact but free wheeling when powered down via the one way bearings. This is a good way of solving much of the mechanical complexities associated with shifting to different size rollers. I see this is a mechanical delta / Wye solution. Both have merit. If Hobby City starts selling 6 wire motors, I think the D/W solution would gain momentum. However, using a stock motor, some electronics, to reverse the motor, and a clever bit of mechanics might be something we see first. Either way, I’m excited

When it comes to friction drives, variable contact pressure based on torque is a must have. I am still not convinced 2 speeds are necessary but am more then happy to keep an open mind and even consider building a prototype in the future.

 

[Miles]

All you need to do is figure out how to mount the drive so that you can have it "auto-engage", like Keplers.

That would be ideal but, it's not essential (as it is with the Kepler drive), as the two one-way clutch-bearings on the roller shaft give you freewheeling. Assuming variable pressure, the drag when freewheeling could be quite low. A simple lever dis-engagement, for longer periods of un-assisted use, would be acceptable, I think.

 

[Miles]

I chose the two drive ratios that had almost exactly the same centre distance (Specified C.D. for the belt drive is only 130 microns greater than that for the gears! ).

brilliant idea... but you could be more "open" with the ratios, and simply run a tensioner on the belt ?

Yes, you could certainly use an idler pulley/snubber. I'm just trying to minimise the number of parts, here.

 

[Miles]

....it sounded like you and Jeremy Harris were closing in on the motor reversing "on the fly" solution

that would not be necessary for this system on a "kepler" type application, would it ?

You've lost me....

 

[Miles]

Very clever idea. I think it lends itself more to an EVtod, style sliding plate arrangement with the roller constantly in contact but free wheeling when powered down via the one way bearings. This is a good way of solving much of the mechanical complexities associated with shifting to different size rollers. I see this is a mechanical delta / Wye solution. Both have merit. If Hobby City starts selling 6 wire motors, I think the D/W solution would gain momentum. However, using a stock motor, some electronics, to reverse the motor, and a clever bit of mechanics might be something we see first. Either way, I’m excited

When it comes to friction drives, variable contact pressure based on torque is a must have. I am still not convinced 2 speeds are necessary but am more then happy to keep an open mind and even consider building a prototype in the future.

Thanks Kepler,

I agree, variable pressure with lever engagement would work well with this system. Complete dis-engagement of the roller also avoids one of the drawbacks of retro-direct systems - if you reverse back-drive them (roll the bike backwards) they become a structure and lock up.

Mechanical gears are more useful than Delta/Wye, of course. You can't get more continuous torque with Delta/Wye, it mostly helps the controller out and gives better throttle control.

Whether or not 2 speeds is necessary, depends on the power available. If you were designing for Australian or European power limits, then I'd say it was essential, if you wanted to get up the steeper hills.

 

[Kepler]

Whether or not 2 speeds is necessary, depends on the power available. If you were designing for Australian or European power limits, then I'd say it was essential, if you wanted to get up the steeper hills.

Now that is a very good point. On this forum we tend to presume we have all the power want, when we want it.

 

[Miles]

Re. Delta/Wye - Drewjet's experience :

I really like the delta wye switcher. It makes no difference in acceleration. In fact I can actually accelerate faster to top speed staying in Delta, rather than starting in wyen and switching, because I don't have to let go of throttle hit switch than accelerate again. However, it allows me to have a lower top speed when desired. so that I have better throttle resolution, or if I want to go slower to get better range.

http://endless-sphere.com/forums/viewtopic.php?p=312895#p312895

 

[adrian_sm]

I couldn't resist........

Astro 60 brushless motor.

5M synchro belt and 1.5 Mod spur gears (large gear POM).

Two one-way bearings (in opposition) on the motor shaft. Two one-way bearings on the roller shaft.

Reductions: 1.9:1 (belt) & 3:1 (spur gears)

Gear range: 158%

Pivoting mech. still to work out....

Have you worked out what sort of power can the belt handle? My quick spec sheet search suggests it would be a sub 1kw limit. Especially at low speeds. Source: http://www.bsc.com.au/site/DefaultSite/filesystem/documents/Timing%20Belts.pdf



- Adrian

 

[Miles]

Have you worked out what sort of power can the belt handle?

Yes. I use the Gates and ContiTech software and MITCalc. Typically, for our use, the limit is set by the number of teeth engaged on the drive pulley rather than the capacity of the belt itself.

 

[Hillhater]

....it sounded like you and Jeremy Harris were closing in on the motor reversing "on the fly" solution

that would not be necessary for this system on a "kepler" type application, would it ?

You've lost me....

I was just suggesting that "fast" reversing would not be essential if this was used in a "kepler" style set up where the roller auto disengauges . :wink:

 

[Miles]

As you have freewheeling [edit:inherent in a retro-direct system], you can take as long as you like to change gear.
Or am I still missing the point?

 

[Hillhater]

Exactly .. :wink:

 

[Miles]

Hmmmm I'm obviously missing something.