Using two servos for two DoF but the strength of both...

[Fredrik Andersson]

I got this idea about a year ago for a science fiction drawing i was working on, and now when i got into robotics i started to wonder if it works, and if it works well in reality. Not sure if it is already done, but thats why im asking.

The main feature is that you can use the strength of two servos for two different directions. The mechanic also allows an 2 DoF arm to be completely in water, if you can work out to keep the two servos outside the water.

Ok, lets get to the picy picy:







So, if the servos roll in different directions compared to each other, the middle pin will roll and if the servos roll in same direction compared to eachother the whole mechanics in the midde will roll... meh, i hope you can see how it works...

I tested it in LEGO, and it seems to work well, the only concern i have is that the coggs on the gears will have to hold all the weight of whats connected to the middle pin...

[Gertlex]

I like this

You'll probably want to find some good bevel gears somewhere... It might be tricky to find some that can easily be connected to the servos though.

[HDL_CinC_Dragon]

I can easily see that the servos moving in opposite directions will indeed help each other disperse the workload between them thus be able to hand higher loads... but what dont see is "if the servos roll in same direction compared to eachother the whole mechanics in the midde will roll"

What do you mean by that? Based on your CAD design there, it looks like the servos would just battle each other and fry themselves...

And I also agree with Gertlex, you would be best off if you found some Bevel gears. Im sure you would be able to find some in a google search fairly quickly. Maybe even try ebay! And about attaching them to the servos you could probably just super glue them to one of the servo horns that hopefully came with your servos... or maybe just order some from the manufacturer of your servos... would probably only cost you shipping...

But the setup your CAD design shows, if your servos traveled in the same direction relative to eachother, they would probably fry themselves...

[Ro-Bot-X]

I understand this design easily, because I have used it many times with Lego.

The thing that is not obvious at first glance, is that the U shaped piece is NOT fixed to anything. Actualy it should be fixed to the top, big U piece. So, the pieces can rotate around the servo's shaft. Now if both servos rotate in the oposite directions, they will be blocked, so all movement will be transfered to the U shaped pieces and they will rotate around the axis going through both servos shafts.

Try Servocity.com for servo (external) gears and sprockets.

[HDL_CinC_Dragon]

Im just not seeing the room for movement... are there then slots in the |_| shaped component behind the servo horns that allow it to move away from the shaft?

Sorry im not understanding what you mean

[Ro-Bot-X]

This is a differential system, the only difference is that instead of the case powering the wheels axle, here the wheels (the servos) power the case, but limited to 180 degree movement (because the servos can't move more than that).

I have used this system for grab and lift mechanism using only one motor (but the motor goes where the center wheel is, and instead of servos there are claws that will rotate towards each other untill there is a obstacle in the way, then the movement will be transfered to the case and the whole thing will be lifted).

Now the only thing that was not mentioned before, is if only one servo is rotating and the other is blocked, there will be 2 rotations: the wheel will rotate around it's own shaft and the |_| piece will rotate around the servo's shaft!

[HDL_CinC_Dragon]

I SEE WHAT YOU MEAN NOW!!!!

I didnt realize the case holding the servos and the case where the output wheel was were hinged! hahaha sorry about that >_< I totally did not see that at first. Now I do I thought the whole case was just one solid piece

man do I feel stupid now!

[Fredrik Andersson]

Yeah, it may take some tome to realise how it works. It was pretty hard just to come up with this solution from the drawing i made, even when i already had the shape of the connectors on it ^^

Here is it:


I was trying to figure out how the mechanics looked inside of those white connectors between the body segments and i thought that there would fit two motors...

I got another idea too

What if we conneted it like in this image, what movements could we create then:

I haven't thought it through completely, but i guess it colud create a wave of rotations and that each pin out could be an 2 DoF leg. Help me out here!

[HDL_CinC_Dragon]

ehhh gotta be more specific than that. I cant tell what you want to move and what you want to be stationary etc

[paulstreats]

This is a good idea. Im definatlry going to use it in the future since it also reduces the leg weight by having both of them connected at the main joint rather than having one mounted onto the output shaft of the other. The only thing is that this system offers a higher torque for the same 2 degrees of freedom, but you have to sacrifice motive time since both degrees cannot be moving at the same time, ie. while the main shaft is rotating, the central shaft cannot spin.

[Fredrik Andersson]

Just realized something, whitch makes it quite unusable. The servos are restricted to 180 degrees. So, if the servos are positioned at 90 degrees, and we want to spin the central shaft as much as we can, the servos ends up on 0 degrees and 180 degrees. How could we ever rotate the main shaft now? Then we would have to reset the servos to 90 degrees or something.

However, i just realized something else. If we can sacrifice some degrees we could still use it. If we restrict the main shaft to 90 degrees (centered at 90 degrees) and the central shaft to 90 degrees aswell (but we will still end up with 180 degrees for the central shaft if we use a gear at the same size on the centre.

[paulstreats]

or just use unlimited range servos

[HDL_CinC_Dragon]

or just use unlimited range servos

Or simply buy the cheaper 90 degree servos and just modify them for continuous rotation.

[Admin]

if you can work out to keep the two servos outside the water.

http://www.societyofrobots.com/actuators_waterproof_servo.shtml

I tested it in LEGO, and it seems to work well, the only concern i have is that the coggs on the gears will have to hold all the weight of whats connected to the middle pin...

yeap, thats the biggest problem with the design. it greatly limits the lifting torque of your device. as Gertlex said, bevel gears will be best.

I remember seeing this design before somewhere . . . but no idea what its called . . .

[paulstreats]

did you ever get this to work?

is there any vids of it in action

[Admin]

I now remember where I've seen this before!

invented by the chinese thousands of years ago . . .

The South Pointing Chariot

Its a bit more complicated, but still works under the same concept:

[HDL_CinC_Dragon]

wow thats cool as hell...

[airman00]

That machines acts as a compass without using any magnets. A figure on top will always point south.

Read that article its really amazing

Why would the chinese not invent robotics and electronics stuff?

Why is it that for 1500 yrs. no advances were made? What took so long? Is it because of exponential growth which technology grows at? OR maybe its all the whole Middle Ages fault, that set us back quite a lot.

I'm wondering what you guys have to say of the matter

[HDL_CinC_Dragon]

Im heavily disappointed at the middle ages where that sort of stuff was pretty much illegal or whatever... I dunno... I just hate periods of time where no technology has made an advancement... dunno if thats sad or not lol

[airman00]

Im heavily disappointed at the middle ages where that sort of stuff was pretty much illegal or whatever... I dunno... I just hate periods of time where no technology has made an advancement... dunno if thats sad or not lol

I am also disappointed, so much time and manpower wasted !   imagine what could have been!

I sure hope nothing like the middle ages ever happens again in the future

[Fredrik Andersson]

Im heavily disappointed at the middle ages where that sort of stuff was pretty much illegal or whatever... I dunno... I just hate periods of time where no technology has made an advancement... dunno if thats sad or not lol

I am also disappointed, so much time and manpower wasted !   imagine what could have been!

I sure hope nothing like the middle ages ever happens again in the future

Oh, i think we eventually will, when our technology turns against us But i sure don't hope so!

[HDL_CinC_Dragon]

Lol at the rate we're going, technology will now be impossible to even slow down. Technology increases literally exponentially (studies show it). The momentum that it has now can not be stopped. No looking back now O_o