Hello , Recently I have reallised there is a lack of strong reliable RC gearboxes, so I guess the point of this thread is to share ideas of strong RC gearboxes that have more than 2 speeds ,

Cheers - Dan

I made last year a 4 speed one, it's pretty strong and reliable:

http://www.brickshel...ry.cgi?f=518300

http://www.eurobrick...7000&hl=gearbox

You can see the controller in the last part of the topic. Needs two PF remote controllers, this is the better controller setup:

http://www.brickshel...y.cgi?i=5742902 (better than this: http://www.brickshel...y.cgi?i=5698446 )

You just pull the lever, and the gearbox does it's job. Doesn't matter, which is the current gear in, you can shift also from fourth to second.

Edited December 21, 201311 yr by Mbmc

I created a strong gearbox which you can see here: http://www.eurobrick...4

I still haven't incorporated it into a completed MOC yet, but someone may be able to use it as inspiration.

Edited January 16, 201510 yr by Splat

Thank you both , I was thinking something like this mechanism

,

with piterx's 3+R gearbox , I'm currently experimenting with these two ideas

Edited December 22, 201311 yr by Dans lego

This may not be what Dans lego meant, but in case it's of use to others, I built a compact 3-speed synchronized linear transmission with 20:12 ratios between gears. I have not motorized the gear changing mechanism in the pictures below, but it's easy to do with a linear actuator.

The one unusual part is "Technic, Liftarm 2 x 4 Fork with Pin", used to move a sliding 16t gear along the 10L axle.

More photos, including a parts list, are here. If anyone finds this useful, please let me know - and if you find a way of shrinking it further let me know that too!

Looks pretty cool, that's a neat design! I know Mahj has yet to use that 2x4 piece...

I love that 2x4 piece! I used it in a solid axle design, with 2L pin jointers joining those pieces on either side for a steering kingpin holder. I've never seen them used in a gearbox before, though.

Thank you both , I was thinking something like this mechanism

www.youtube.com/watch?v=TGZjHmQkH9k

with piterx's 3+R gearbox , I'm currently experimenting with these two ideas

This gearbox isn't 100% reliable. Works well with fresh batteries, but it's not for high torque. Why this isn't good for you?:

http://www.brickshel...ry.cgi?f=518300

- handles high torque, easy to use and reliable.

I try to share my gearbox (transmission) building experiences:

There is two part of an RC gearbox:

1. The gearbox itself. Can be linear, synchronized, etc. This is the easier part to make, you can find a lot of gearboxes on brickshelf.

2. A shifting mechanism. This is the essence. If you use a linear actuator for a linear gearbox, then how can you know, the gears's teeth are exactly an precisely overlapped? Every time you watch it? If the teeth ar overlapped 60% instead of 100%, then they are more able to deform, then it can handle less torque. This is the problem with continous control. You don't know exactly, how the gears are connected to each other. You hear that the new gear is in, then it's ok. But sometimes it isn't. That's why I made with lot of experiments the definite controlled gearboxes ( http://www.brickshel...ry.cgi?f=518300 ). You can control them in 100% darkness. If you push one lever up, that always do the same gear change. In a linear control, you have to know which is the current gear, to change it. For example, from 3 to 2 or from 4 to 2. If you want to shift to 2, then you have to know which is the current gear, to know how much shifting you need to do.

Control mechanism is the most important. You can do a definite controlled 4 speed gearbox also with two PF servo motors. I tried out, works well.

Another thing:

When two gears are close to each other in a linear gearbox, then it's able to scratch a lot during shifting under load. The sliding gear is connected to two other gear at the same time.

Edited December 24, 201311 yr by Mbmc

I should have mentioned that my gearbox was very much inspired by Splat's. The fork can be replaced with Technic, Steering Arm with Pins and there are probably other good solutions.

Mbmc makes some good points about the advantages of definite control over continuous control; my gearbox was a continuous one because I was trying to make it small and minimize the number of motors used.

In theory, it seems like a definite control gearbox with N motors could have 2^N gears since each motor has 2 directions. I suppose you could do 3 modes per motor (forward, reverse, off) if you used a spring and were willing to hold down the motor controller to keep it in a particular gear, but that sounds like a pain and a waste of electricity.

Since the servo has 15 positions, theoretically you could have a 15-speed gearbox with just the servo, but given the slop and flex and backlash inherent in the parts, the maximum number of practical speeds is probably significantly lower. Does anyone know of a single-servo design? Of course, a single-servo design would be sequential so you couldn't shift from, say, 1 to 4 directly. I'd like to see Mbmc's multi-servo design, too.

I believe a strong and reliable RC gear box design is out there. Something with both the necessary torque to go up hills and speed that will exceeded the range of the IR receiver in a second. Reverse will be carried out by the motor just by reversing the direction of rotation. There are several crucial aspect for the transmission you must consider first if you want to achieve great performance out it.

Boxerlego has right, a reverse gear is unnecessary in a gearbox because of the electric drive motor.

Since the servo has 15 positions, theoretically you could have a 15-speed gearbox with just the servo, but given the slop and flex and backlash inherent in the parts, the maximum number of practical speeds is probably significantly lower.

Yes, with reliable control you can do 3 gears with the PF servo. The speed controller is very unstable and isn't reliable for shifting.

Does anyone know of a single-servo design? Of course, a single-servo design would be sequential so you couldn't shift from, say, 1 to 4 directly. I'd like to see Mbmc's multi-servo design, too.

A servo operated gearbox is a very simple and easy to build construction. You have to build a 4 speed gearbox (for example this: http://www.brickshelf.com/cgi-bin/gallery.cgi?i=2513289 ) with 16t clutch gears and synchronizing rings, and put two servos to each side. The servos's 0 position is neutral:

http://www.brickshelf.com/gallery/mbmc137/pictures-videos/servo-gearbox.jpg

The 4 gears on each side aren't necessary, I just had no space for the servos sideways.

(sorry sor the bad picture, but it's built in to a "secret" project )

There are several ways to go at building a shifting mechanism for the transmission. A simple compact shifting mechanism would be awesome. I attempted to build a automatic 6 speed transmission with a direct dive motor that works of the drive motors rpm to make all the shift possible. The size of everything together was very large fit for any large scale design out there and feels difficult to go through all the gears on ground. Theoretically it is the best idea to me because the only thing that is electrically connected to the transmission is the driving motor.

Boxerlego has right, a reverse gear is unnecessary in a gearbox because of the electric drive motor.

that depends if you have a one way clutch or not :)

I built a compact 3-speed synchronized linear transmission with 20:12 ratios between gears. I have not motorized the gear changing mechanism in the pictures below, but it's easy to do with a linear actuator.

I like your design - it is nice and compact.

I have one question though - it looks like your input from the motor is the black 10L axle in the middle, and this has a 16T gear on it which should fit tightly on on the axle, so how does it slide back and forward to change gears?

I should have mentioned that my gearbox was very much inspired by Splat's.

Thanks for the credit It's always good to see someone take one of my ideas and continue to develop it further.

Another part that can be used for the shifting mechanism is Technic, Steering Knuckle Arm with Ball Joint

Some new 16t gears have lower clutch than the older-style ones. I guess this would be considered an 'illegal technique' by Lego, but it doesn't require much force to move. Although it wouldn't work with older 16t gears.

It works also with older 16t gears, you just have to find a loose one. I have a few. I also have a very loose 20t gear, I used it in a moc to build a sliding gear gearbox (not uploaded yet).

Another part that can be used for the shifting mechanism is Technic, Steering Knuckle Arm with Ball Joint

I don't think it's good because of the ball joint. That can only transmit force, so it stresses more the sliding gear and axle.

It works also with older 16t gears, you just have to find a loose one. I have a few. I also have a very loose 20t gear, I used it in a moc to build a sliding gear gearbox (not uploaded yet).

I don't think it's good because of the ball joint. That can only transmit force, so it stresses more the sliding gear and axle.

Plus, the axle hole won't let the axle rotate inside it.

I like your design - it is nice and compact.

I have one question though - it looks like your input from the motor is the black 10L axle in the middle, and this has a 16T gear on it which should fit tightly on on the axle, so how does it slide back and forward to change gears?

Thanks!

You are correct, the 10L axle is driven by the motor and the (new design) 16t slides on it. It takes a little bit of force to slide the gear - much more than is the case in your design with the clutch gear - but surprisingly the force isn't too great. A properly braced LA can move it without a problem, and it's very easy to slide by hand.

I've uploaded a video here. Using a linear actuator works, but there are a lot of problems.

1) it's slow (which could be changed by gearing at the expense of making it bigger

2) you have to watch it to know what gear it's in and when to stop it

3) the LA exerts a lot of force on the structure, which thus needs to be reinforced

4) the horrible jerking and noise that occurs at minimum and maximum extension

5) LA + PF-M = lots of space, and the whole point of the gearbox was to make it small

Maybe a captive worm gear would work better, but problems 1 and 2 would still be issues.

Technyk32231 is correct about using the wishbone - the axle needs to be able to rotate. You can't pass the axle through the hole in the wishbone, either, since then the wishbone will protrude above the top of the 16T gear and interfere with the 20T gears.

Another option is to lengthen the gearbox by 1 stud, allowing for the mechanism that moves the sliding gear to be 3 studs wide (instead of 2, as it is now). That would mean that you could use, for example, 2x to move the gear - the same way Splat does it in his gearbox.

Edited December 29, 201311 yr by Hrafn

that depends if you have a one way clutch or not :)

Certainly.

Some new 16t gears have lower clutch than the older-style ones. I guess this would be considered an 'illegal technique' by Lego, but it doesn't require much force to move. Although it wouldn't work with older 16t gears.

Not exactly. There is a 8t gear out that is without friction that would certainly work in this case. The bevel gear is probably suited best for this direct style of gear engagement demonstrated by Hrafn and Splat.

I've uploaded a video here. Using a linear actuator works, but there are a lot of problems.

1) it's slow (which could be changed by gearing at the expense of making it bigger

2) you have to watch it to know what gear it's in and when to stop it

3) the LA exerts a lot of force on the structure, which thus needs to be reinforced

4) the horrible jerking and noise that occurs at minimum and maximum extension

5) LA + PF-M = lots of space, and the whole point of the gearbox was to make it small

Excellent video. Sure the transmission is a risky design to pull off and some problems are around it. Tho it does prove several points. One key point it proves is that call for more gears especially bevel gears. I love that it is a sequential gearbox design because one of the benefits of this design is that it does need to use a typical clutch.

Another part that can be used for the shifting mechanism is Technic, Steering Knuckle Arm with Ball Joint

I don't think it's good because of the ball joint. That can only transmit force, so it stresses more the sliding gear and axle.

Plus, the axle hole won't let the axle rotate inside it.

Technyk32231 is correct about using the wishbone - the axle needs to be able to rotate. You can't pass the axle through the hole in the wishbone, either, since then the wishbone will protrude above the top of the 16T gear and interfere with the 20T gears.

In Hrafn's redesign this piece doesn't work, however in my original design it does as it doesn't interfere with anything above/below the central axle, and therefore the axle goes through the holes of the steering wishbone, and it allows the gearbox to be one stud shorter. The changing of gears can then be controlled via a steering link (http://www.bricklink...tem.asp?P=2739a) which can be controlled via some sort of rotating mechanism, which can be controlled by a servo motor. I'm still trying to work all of this out at the moment to get it nice and compact.

Don't forget 9L links. I think they might be too bendy for something like this, however I haven't much experience with this part (I don't have any).

9L links: they are pretty strong aginst bending, it's no problem.

I still don't understand why to use this ( ) piece in a linear gearbox, there are some better one. If you push it with it's ball joint, then appears what I wrote:

"I don't think it's good because of the ball joint. That can only transmit force, so it stresses more the sliding gear and axle."

I certainly have one or two 20t gear somewhere that actually slide down the axle when held vertical. Probably not the drive gear of choice but that is all I have when it comes to a sliding drive gear.

I agree with Mbmc, I don't think steering knuckle arm with ball joint is good either because the ball joint will rotate inside the socket and will distribute itself unevenly on the gear and axle shaft.

I am going to point this out because I've been using the steering knuckle arm with ball joint for this 2 speed transmission and came up with a way to use the ball joint as a way to determine optimal gear mesh position with the adjacent gear.