How to build this cubed robot.
Id like to build a cube like robot that has motor movements inside the cube. There are several on the internet like the Cubil (*1) and this one form MIT(*2). Before I finish the circuit board I would like to ask the forum a few questions.
1. Would a stepper motor be a better choice or a dc motor? Stepper motor may allow for precision for ballancing but might lack the power needed to propel the robot.
2. How much mass would be needed (again motor selection could be a factor) M=FA where A = motors sped capabilities, to move a robot of 2 pounds.
Please comment and offer suggestions.
(*1) - https://www.google.com/url?sa=t&rct=...91071109,d.cGU
(*2) - https://youtu.be/6aZbJS6LZbs
1. Would a stepper motor be a better choice or a dc motor? Stepper motor may allow for precision for ballancing but might lack the power needed to propel the robot.
2. How much mass would be needed (again motor selection could be a factor) M=FA where A = motors sped capabilities, to move a robot of 2 pounds.
Please comment and offer suggestions.
(*1) - https://www.google.com/url?sa=t&rct=...91071109,d.cGU
(*2) - https://youtu.be/6aZbJS6LZbs
272 x 185 - 6K
Comments
2) In general, you want a huge friggin' flywheel and minimize the weight and moment of inertia of everything else.
Stepper motors have low torque at high speed (they tend to "skip" steps if under too much load and are running at the top of their rpm), which is a bad combination for a robot design like this.
From one of the sideview shots, it looks like they might be using DC fans for the motors, some of which are brushless, which they specify. Mount a flywheel to it (they mention it), and there's your angular mover in a nice almost-off-the-shelf package. Very clever in any case.
You mention some details that I was unable to discover would you be so kind and provide a link. Thanks.
Brushless DC motors according to this video. Here is another good instructional video about these kinds of motors...
Cool post Zap-o. Looks really complex though!
Agreed. Anybody with a beard needs watchin'.
Wait, that may include nearly half of us. Gordon, Whit, me... do I need to start a poll?
And if we tweak the qualifier to "the guy with the beer", the number of suspects grows.
Sorry for derailing. We now resume our regularly scheduled discussion on jumping cubes.
Most steppers do not come with optical feedback. You almost always have to add that. The majority of stepper designs rely on a known load, which is not the case with these cube robots because of gyroscopic forces. When a stepper is used in an application of uneven load, it may tend to skip steps, and then you need optical feedback.
What's more, most steppers are extremely slow for their weight, being designed to operate primarily in the 0-300 pulses per second range. A 1.8 degree stepper requires 200 to pulses to rotate once. That's a very slow motor. Unless the flywheel is very large and heavy (counter productive for a robot), not much angular momentum can be developed at 60-120 rpm. Try 5,000 to 30,000.
Consider: The cube design in the video relies on a certain angular momentum to break the current position, and move to a new one. The motors are likely always spinning, but change speed (e.g. sudden brake) to make the positional change.
A servomotor (as opposed to an RC servo) is simply any motor with feedback, and it's a common motor used in industry. Can't imagine why you're not finding info on them. There are relatively small pancake motors that have so much torque they'll rip your fingers off -- I know, this almost happened when I was working on a piece of equipment powered by a 4" pancake. Because they use feedback, their controllers apply whatever current is required.
Brushless DC motors work in much the same way as steppers (but always have feedback), But they are designed for higher rpm's, and less so for the stop-and-go common in the majority of stepper applications. Unlike steppers, which often use very large metal frames with weights to match, a brushless DC motor can be constructed to be very lightweight. The size (and weight) of the flywheel, and a little bit of Newtonian physics, takes care of the rest.
J