Let's Make Robots!

Idea Explained



Right, i'll try and explain the idea 







when the other leg is rotated in the opposite direction the same happens.


I found that the gyro was not of sufficient mass to lift the legs.


I also discovered that the feet need to grip well or they slip.


If anyone has a gyro handy, do try just holding it between your fingers and alternately twist it like shown.


what is going on is that the horizontal force is transfered into a vertical force somehow :) 


Like i said i didn't get it walking but i think the idea has merit, if i held down the foot it did lift the other leg.

also my gyro developed wobbles which kept on washing off all my speed.














Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.
I studied gyroscopic precession as part of an engineering paper at uni a few years ago, if you go back to the equations you can calculate the lift force from the flywheel mass and rotational velocity, although it can get a bit complicated for real problems. It doesn't look like you have a motor or anything to keep the flywheel spinning though, so as the bot started to take each step the extra torque applied to the gyro axle would slow down the flywheel, making the step uneven. If you spin a gyro up and then tip it back and forth it looses it's speed a lot faster than if it's just sitting there.

The motor is on the original picture at the top running a rubber O ring to the gyro shaft.

it's a little 5v or so motor and it took about a minute to get to max rev's

when it got there it seemed to maintain the speed alright.

 by the time i got it hooked up to the rest of the bot it had a fair bit of wear on the pivot points and would develop said vibrations.

it really needed propper bearings. i'll revisit the idea when i can make a bigger bearinged gyro... with a lot more speed ;) 

I think that speed is not necessary - and in fact, planning to speed it up sounds difficult to implement, even if you address the  current vibration problems you're already having.

I think it's probably enough to increase the angular moment for the flywheel - i.e. use a bigger one or add more weight to the rim of the flywheel (or conversely, use lighter foot motors etc), I really think that just getting better bearings and spinning it up won't solve the problem - furthermore, simply spinning up the flywheel will make the steps very fast (and probabaly short..?). (having said that, I suppose that would be an easy way to control the step size...)

What does interest me is how you arrange for the torque to be applied to, say, the right foot, BEFORE you actually rotate the left foot - since you need to pick up the right foot before you move it.

at the risk of patronizing for a moment (just to better illustrate my question)

To walk, youdo something like;

1. shift weight to the left,

2. pick up right foot,

3. move right foot foward,

4. put down right foot

5. shift weight to the right, etc. etc.

But since you're using the foot rotation to generate the torque you need to lift the OTHER foot, you'll be combining steps 1, 2 and three into the same step - would that result in some shuffling really?

Thanks for the input, I really appreciate it. My experiance of gyros is just from observation, I'll outline the idea development:


I first started playing with a bike wheel & having it spinning between my hands in the same orientation as in the robot photo.

I noticed if I dropped one hand, the wheel supported itself and rotated on the other hand.

Based on this I intended to make the robot as the one above but with retracting pistons for legs.

ie: lift leg, bot rotates say 80 degrees or so forward, lower leg and lift the other & the same happens in opposite rotational direction.

 I then made up the frame work and discovered that the weight of the extras (legs and pivots) meant that when the leg was retracted the gyro dropped very rapidly and had very minimal rotation - micro steps :(   

I then saw that quickly twisting one leg threw the gyro and other up in the air, even with the added weight.

Then I encountered the problem of the foot slipping and the other failings. 


I think you are correct in surmising that the present arrangement will result in a shuffle.

Further consideration of the physics makes me think that a combination of lifting and twisting is needed if it is to work. Or maybe better still the foot that twists may extend forcing it to grip the ground more.

 I agree with the ideas of making the mass around the gyro wheel greater and a bigger diameter wheel whould improve its capabilities.(and lighter motors)

Although I have seen that the speed of the gyro certainly appears to be proportionally related to the force required to tilt off its spinning axis. This makes sense to me then, that the faster it spins (to a degree) the more weight I can hang off the pivot and have the gyro carry the weight forward.





I think there's a few bits and pieces that can help you decide what is the best direction to go;

angular momentum is the tendency for a spinning thing to keep spinning - more accurately, it's the resistance a thing has to CHANGING the status quo - so this also goes for trying to start spinning something up. Anguar momentum is proportional to  the moment of inertia  (the way mass is distributed about the centre of rotation, and is related to the radius of the spinning thing) and L is also proportional to the angular speed - so the faster the thing is spinning, the higher the momentum.

You mentioned that the gyro dropped quickly when you added weight, this is because the momentum of your gyro is too small for the weight you're applying. I think I mentioned before that you can bump up the angular momentum by increasing I or w - I is probably easier to increase, simply by changing the shape of your gyroscope flywheel, or if that is optimal, just adding more weight to it.

I think it should be possible to continue with your first arrangement - i.e. that one where you were using retracting pistons. Since you'd be trying to generate rotation only in one plane (and not trying to lift the legs by precessing the gyro), then you could even STACK your gyros - have more than one in a kind of ladder, one above the other even three should be possible.

Did you try using compressed air for your pistons? Air is nice and light!


So stacking the gyros compounds the angular momentum? <--- correct terminology? of the system as whole.

or is the effect the same a 1 single gyro of a larger size instead of 3 smaller ones.

Also compressed air sounds like an interesting prospect, my dodgey pistons were to be raised and lowered with the motors I used in the photo, don't ask how, but i was tying to use them like crude servo's (as i have none at present).

 After coming up with the idea I was dreaming (  :P  ) of making it quite large, 2+ meters. I wanted it to walk around the yard.

Compressed air would be eminently suitable for that size, i'd use one of those 12v compressors for vehicle tires & run it of few gel cells. It would also be much easier to make a gyro like that, even beef up the bike wheel.

Yes I think that they add linearly - so having two gyros is the same has simply doubling their size, or their rotation speed.If you stack them though, you have to be sure that the individual gyros have the same angular momentu, and that during the stepping, each gyro has the same torque applied to it (i.e. the axel length is the same).

I think that from a construction point of view, it's easier to stack two gyros than to try to make a new gyro that is twice the size, or even just to double the speed.

Why don't you just do some basic measurements with your bicycle wheel? see how fast you need to spin it so that it can support something like 250grams without sagging too far, and then you can also calculate the total torque on the end of the axel, and estimate the "stepping speed" - or just do it empirically.

weight is probably crucial - so big heavy batteries are probably out, Would you use a separate reserviour for the compressed air? something like big 2l soda bottles might be ideal to hold quite a bit of compressed air -you can calculate how much you need.

I think a 2+ meter is indeed ambitious, but I think large is probably easier than tiny.

It appears as an original form of movement, based on gyroscopic precession, which would almost seem to work.

 In Physics class, I remember a demo of sitting on a stool on a turntable or lazy susan, and holding a bicycle wheel vertically by the shaft that was spinning at a high rpm. Tilting the wheel one way or another would produce a force the spun you around on the turntable

It might be possible to experiment, having the "leg turning" motors move slowly, not sure what gear ratio is used there, and have the gyro speed way up, maybe more mass on the gyro wheel too. And I'd reduce the mass of the legs, keeping the bottoms sticky. 

Interesting experiment.

That makes sense. Now you have me wondering if something is fundamentally wrong with the design, or if the execution leaves something to be desired. Maybe more mass in the feet, or in the flywheel. Maybe better grip. Or maybe the feet are placed too far apart. Or not close enough.

I really cannot tell and it is bugging me!

mikv, PLEASE explain, we're all dying to know what this could possibly be!  i see two motors facing donw, and some gryo (???) in the center... i dont see how this could do anything...yet