I'm new to Arduino. I've been working with the Elegoo "THE MOST COMPLETE STARTER KIT TUTORIAL FOR MEGA 2560" tutorials.
I'm trying to build a toy robot that walks. I'm thinking it would be about 12" to 24" tall and made of plastic or thin wood. I may add some metal for strength. I need motors for the joints (hip, knee, ankle, shoulder, elbow, neck). I'm hoping for something cheap. I've looked all over YouTube and I see a lot of cars but only a couple robots that I don't like.
I'm looking at micro servos but even the micros are pretty big (1"+). Also, they don't seem very strong. I'm looking into stepper motors but I'm worried about high rpm. Can I control them enough to reliably move 45° or 90°? Also, again, I'm looking for a cheap solution.
I'm looking at the motor below. They come with single or dual axles which is great. I don't know if I can control them well enough to just move 45° or 90° accurately and then stop.
What is my best option or how can I make a choice that will work (research)?
I held a journeyman electrical contractors license in Central Florida for 8 years.
I hold a diploma in electronics from Cleveland Institute of Electronics (for what it's worth).
I've been a software development consultant using Web based, database technology for 25 years. VB, Java, SQL Server and lately PHP and MySql.
Yea, I'm old now - lol.
I get the jist of Arduino - WAY COOL! but I can't decide what to do about motors for joints.
I'm thinking a metal (not aluminum) geared micro servo but any good advise would help.
As for torque? What would it take to lift a 6" plastic leg?
RPM? Slow, as fast as an arm or leg would move.
Torque? I don't know how to figure that out. It shouldn't take much.
It needs to lift a 3" plastic tube or maybe 1/8" wood 90° up.
I'm looking at 3D printers so maybe that material, more or less hollow.
"Buying cheap motors as a start is wasting money."
Agreed. That's why I'm here asking questions :o)
Low RPM comes from geared motors but they still need to produce torque on the outgoing axle.
Study basic physics from preliminary school and up. There's no way without it, except buying and scrapping motors until there's one that is powerful enough.
That tells helpers nothing. Start using a scale noticing the weights of the parts and measure the length of bar moving the weight. Torque = weight * bar length.
That reveals a fundamental misconception about how steppers work. In any case, steppers are far too inefficient to be useful for walking robots.
Which incidentally, is about the most difficult sort of robot to make, as a first project. Walking is extremely complex behavior, but fortunately, there is plenty of material on the web that will explain to you just how complex it is.
You have to work out the mechanics of your application first.
One cannot start with the motor and work backwards unless you intend to end up with masses of unusable gear.
Firstly, study how any "walking" robot operates, do some drawings with pencil and paper (preferably to scale....best approach) and then work out the requirements.
Electronics will be a challange but there are a few successful systems out there BUT you will need to adapt to your application.
So a lot of footwork required by you before you even start on this.
As mentioned above, do some groundwork on robots with wheels first.
As just a stab in the dark given my past experience with most things mechanical, your 1" or smaller expectation is way off. Doesn't really compute in the world of requirements and theory anyhow.
Thanks for all your replies.
I realize that a 'walking' robot is a challenge and still haven't decided whether or not to back off of it. That's why I came here first. I think if I can iron out some things I can do it. I've done animations of walking characters. I need 6 joints and big flat feet. I may need something like a gyro or tilt sensor. I have one but it has to lay all the way down to trigger 'tilt'.
"1'' or smaller ...." yea I'm seeing that.
I was watching James Bruton on YouTube who was describing designs for cycloidal motors; interesting but they're big and heavy. I wonder if I can make tiny ones that don't require bearings. I hoped I could just buy something like that cheap but I don't see anything.
"Low RPM comes from geared motors..." The yellow one above is geared 1:48. It has maximum torque: 800gf cm and max 200 rpm. It's powerful enough to easily lift a plastic or wood 6 inch leg. I can get them @ 1:120.
I was hoping someone here had enough experience to tell me whether or not I can control it well enough to make it turn between 15° and 120° and stop reliably. I can use PWM to slow them down & smooth movement. Yes???
If nobody here knows I suppose I can run some tests to find out. I have 4 of them.
With the dual axle at one end I can slip it inside the torso or leg with the axles at the bottom. Can I do something like that and control it?
If I can't control that motor can I make something like that design rather than cycloidal with a small motor that I can control?
"Torque = weight * bar length..." That helps. Thanks. Don't degrees & axis (x y) play into that too? To move it sideways won't take as much as to move it up right?
To move it 45° won't take as much as 90° yes? 2 motors working together will lessen the load on each correct?
Again, I'm convinced that yellow motor can handle anything the leg can throw at it - but can I control it?
I realize I need to study physics and mechanical structures (cycloidal geared motors?) but I came here hoping to learn from you.
PWM driving dc motor should be ok.
Given the gear reduction required they would stop instantly when power was removed.
Worm gear output would be best as beyond around 40 degrees of the gear cut, they will not drive backwards.
Naturally under these conditions , thrust bearings would be essential otherwise the gears will lock up.
With that considered, it would in most cases require ones own design and build gearbox.
Not difficult, done it myself many times but then that depends on your skilset.
Given the query, possibly not great in that respect.
Again a plus for a wheeled robot to cut your teeth on.
