3
$\begingroup$

I am in the concept phase of a driving robot. The two wheels on the front axle will be powered, while the rear will be dragged along. The rear is also responsible for steering but this has noting to do with my question.

Since the robot is required to make relatively sharp turns at high speed. Therefore I have two options to compensate the different speeds on both sides. On the one hand, a differential gear in the front axle could be used. It would be powered by one motor then. On the other hand, I could simply use two motors directly powering each a front wheel. This way I could simulate the differentiation in software.

I'd like to go for the first approach, using the hardware differential. But I have the one concern with it. Would a robot vehicle with differential gear still move straight, without explicit steering applied?

My imagination is that, with those wheels not being solidly connected, the robot would move in random curves which I'd have to compensate with a lot of steering then. I know that for real cars, differential gears are standard and do their work, but now I am talking about a small robot measuring about 6 inches.

$\endgroup$
2
  • $\begingroup$ Is it possible to use 2 motors so that both axles will be powered? You could always say "front is where I'm driving towards" then. You could connect both axles with a hinge so that 90 deg driving is possible. You wouldn't need hinges for a single wheel, too. $\endgroup$
    – ott--
    Jul 31, 2013 at 19:49
  • $\begingroup$ This question is not about the steering concept, but thanks anyway. I already have a solution for the back axle, which has only one point resting on the floor. But I won't go into detail here. The question is whether I'll use two motors in the front, or one motor and a differential. $\endgroup$
    – danijar
    Aug 1, 2013 at 12:05

1 Answer 1

3
$\begingroup$

If I understand your question, you are asking whether a vehicle balancing on two wheels (or two wheels and one caster) will be able to move straight, or at least predictably, if both wheels were driven from the same motor and used a differential.

The answer is yes, but only if you have a way to equalize the forces of friction affecting each wheel. For example, by applying a brake individually to each wheel, you could alter the balance of force between the left and right. It would be crude, but certainly possible to control the steering in this way. In fact, many tractors use independent left and right brake pedals to accomplish this.

Without actively braking (or some other method), your intuition is correct: the amount of rotation of each wheel would depend on an unpredictable set of forces, and the robot would make seemingly random turns.

$\endgroup$
7
  • $\begingroup$ It is not balancing, there is something like a caster in the back. Does your answer still apply to this case? And is it realistically to simulate the differential in software when directly powering the two front wheels by each a motor? $\endgroup$
    – danijar
    Aug 1, 2013 at 12:07
  • 1
    $\begingroup$ @danijar, you contradict yourself. In your question you say that the rear wheel is there for steering, but unpowered, which implies that it is steered with a servo or something. Here, you say that it casters, i.e., it follows whatever steering comes to it, and is not attached to any servo or whatever. Which is the case? $\endgroup$
    – TimWescott
    Aug 1, 2013 at 23:32
  • $\begingroup$ @TimWescott Kind of both. Actually, it is an omni wheel attached transversal to driving direction. It casters and steers at the same time. Only an omni wheel can do that, I guess. $\endgroup$
    – danijar
    Aug 2, 2013 at 0:46
  • $\begingroup$ Can you update your question with a picture or drawing of the wheel configuration? $\endgroup$
    – Ian
    Aug 2, 2013 at 13:31
  • $\begingroup$ @danijar: I sincerely hope that we are suffering a terminology collision, because if you're holding the axis of that omniwheel steady I wouldn't call what it's doing "castering". At any rate, from what you describe, you must be steering the robot by driving the omni-wheel. Thus, if you hold the omni-wheel steady it will enforce straight-line motion on the robot, and if you rotate the omni-wheel it will enforce rotation on the robot, proportional to its drive speed. In this case, then you want a differential on your drive axis, so that the drive wheels aren't trying to enforce steering. $\endgroup$
    – TimWescott
    Aug 2, 2013 at 18:19

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.