selectable gear engagement

[electrolust]

Imagine a bicycle hub, turned by a motor. The motor itself is the central axis. Think, slot machine wheel.

Now imagine a linear array of 10, or 25 or 50 of them. If each is an independent assembly, that's a lot of motors.
Now imagine 10 or 25 rows of these arrays. The user can stop or start any wheel via a control board.

I'm thinking, rather than build each individual hub as an independent, self-powered assembly, use a single long axis to power the rotation, and engage each hub to the single powered axis via a gear train. I'd need a bigger motor but overall it's likely more efficient and also less maintenance.

But I would want to stop and start each wheel independently of the others. So, I need some kind of actuator to move a sliding "engagement" gear into place, much like a car's manual transmission engages a gear. Or, instead of sliding along the axis of rotation, it could rotate in from the "edge".

The cheapest linear actuator I could find (in 30s) was Lego Technics at $10. The motor is under $5 so this is a tough sell. The first search results shows actuators in the $50-$70 range. I did find pneumatic ones for $11 right away, but still that feels too expensive.

I feel like the lego part will have enough precision, but not enough long term durability.

Lastly, all the ones I find, even "micro" are actually very long, 5" or more. I don't think that's going to fit well as my hubs are 2" width.

At this point it seems easier to go with independently driven hubs, but wanted to see if anyone has input.

[Gyro]

https://www.eevblog.com/forum/chat/what-on-earth-variable-field-motors/

[electrolust]

Thanks Chris, but I've no idea how that is applicable. Can you add some explanation?

I did some more research and think what I would want is a linear solenoid. This little guy seems just about perfect.

https://www.ebay.com/itm/DC-12V-300mA-5N-10mm-Pull-Push-Type-Frame-Linear-Motion-Solenoid-Electromagnet/232855233494?

10mm throw, 5N should be enough force, the overall package is compact enough, price is good. Less current is going to be better but I can make anything work. I'm assuming I can find a non-ebay supplier for about this price.

Does anyone think the overall complexity is better using an actuator, or should I stick with independently driven hub "modules". As an example motor, Google seems to think a 28byj-48 draws 100-200mA so if the linked solenoid actually draws the full advertised 300mA, it is quite a bit more power hungry..

[Gyro]

Sorry, I should have added some text.

It just seemed sort of akin to what you wanted - torque vs speed adjustment, the same function as a gearbox.

When I initially read your op, I didn't realize you were talking real gears / mechanical engagements.

Chris

[Gregg]

From a recent post https://www.eevblog.com/forum/projects/threaded-axle-stepper-motors-as-lock-pins/
This might be what you are looking for

[electrolust]

This guy shows low power solenoid control.

Thanks Gregg, that post does give me ideas.

[ajb]

How much mechanical power do you need to transmit?

Also, the inertia of the load(s) will matter: if you just slam an intermediate gear into place the full torque of the motor plus the inertia of the already-spinning loads immediately collides with the full inertia of the not-yet-spinning load via the first gear tooth to make contact.  That's going to be awfully hard on your gears.  Also note that in many of the obvious configurations, the torque being transmitted via an intermediate gear is translated into a force along the axis of your linear actuator, which increases the amount of force your actuator will need to apply in one direction or the other.  This is partly why most transmissions move things axially rather than radially, because then the translated torque is always orthogonal to the motion of the meshing/unmeshing components (it's also partly why you have a clutch to unload the transmission before you mesh/unmesh anything!). 

A better solution is probably a friction clutch that can be let in softly and allowed to slip a bit as the load comes up to speed, although this makes the clutch a definite wear item.  Electromagnetic clutches are commonly used on automotive AC compressors, but those are probably a bit large and power-hungry for you based on the tiny solenoids

[Gregg]

Just another thought: How about using brushless permanent magnet pancake motors with the coils stationary between each wheel.  Each could provide soft start as well as braking independent of all the other wheels.
A long time ago I remember taking apart an automotive electrical radiator fan that used permanent magnets for the field and coils printed on a circuit board.  Although it was brushed and used ferro magnets, I was impressed by the power it produced.

[electrolust]

How much mechanical power do you need to transmit?

Also, the inertia of the load(s) will matter: if you just slam an intermediate gear into place the full torque of the motor plus the inertia of the already-spinning loads immediately collides with the full inertia of the not-yet-spinning load via the first gear tooth to make contact.

Right now I'm planning to stop the motor before engaging or disengaging any gear. I haven't tested but I'm under the impression that I can stop the motor, engage/disengage, and restart the motor fast enough to be imperceptible.

most transmissions move things axially rather than radially, because then the translated torque is always orthogonal to the motion of the meshing/unmeshing components

yeah. in my case my idea to engage radially is to allow the gears to be cut in 2d. I fear that with axial engagement, I'd need a 3d gear profile. I'm hoping that with radial engagement (my solenoid would push a bit of springy steel so that the solenoid can fully extend even if the gears don't mate) the light axial pressure would engage more readily. I'm imagining that at least one gear would engage and then the other would when it turns. Whereas with axial, if one gear isn't in place then neither gear will ever engage and so i need an aggressive 3d profile.

planning to use plastic gears BTW

[electrolust]

Thanks again Gregg! definitely going to give this a try. Even if it's power hungry I think it's superior given the cost and the mechanical simplicity.

[mc172]

I am picturing something like this:



...But perhaps not with independent steering and suspension. Am I close?

[janoc]

I really can't see how this sort of contraption, with all those gears, mechanical linkages, clutches, solenoids to actuate all that and what not is going to be cheaper to build and simpler to maintain than having a motor per disk and a bit of electronics to control them. Use small brushless motors and apart from a bit of grease here and there the contraption is going to be pretty much wear and maintenance free.