0
\$\begingroup\$

I am working in a robotics project and I am looking way to make my robot "dash" as soon as possible. Therefore I need to have a quite high acceleration for a brief time.

To be more specific, I am trying to have my robot (~0.5-0.7kg, dimensions are ~ 0.3mx0.2m) to move ~0.3m as fast as possible (best is like 0.1sec). (I am free to choose wheel diameter freely.)

I tried to estimate it roughly as follows: 0.5 * a *0.01 = 0.3 a = 60m/s^2 F = 0.5*60 (if the robot is 0.5kg) = 30N torque = F*x = 30*0.3 = 9*Nm

Therefore I am looking for a light motor which can supply an instant torque of ~9-10 Nm. I might try using 2-4 motors together if I need to. I looked for the stepper motors but as far as I saw, this is unlikely to happen with them.

Is this possible, or is it impossible to find any small motor type that can achieve this?

Edit: I would prefer ~20$ but can go up to ~40$ if it will improve everything significantly.

\$\endgroup\$
6
  • \$\begingroup\$ Approximately, what is the budget? Also, can you make it lighter? If you search the web for Micromouse competitions, you'll find small robots which move quite quickly. They often use FAULHABER or Maxon motors for their high torque, and optional position encoders (so the robot can track its position) \$\endgroup\$
    – gbulmer
    Oct 10, 2015 at 12:42
  • 1
    \$\begingroup\$ Does the robot need to 'continuously' dodge many projectiles, or is it good enough to dodge one or two projectiles? If it is only dodge a few projectile, maybe consider using compressed gas and a few 'rams' to dodge in different directions. \$\endgroup\$
    – gbulmer
    Oct 10, 2015 at 12:46
  • \$\begingroup\$ Just added the budget, I have no idea whether it is realistic or not. Do you know what position encoders do they use? No robot has the time to prepare for another dodge but without manual reload. Can a compressed gas achieve that, wouldnt it be too heavy? \$\endgroup\$
    – ozgeneral
    Oct 10, 2015 at 13:08
  • \$\begingroup\$ The budget will likely help folks. If you search for 'robot wars', you'll find examples of weapons (like pick axes) which use compressed gas. I have never used that stuff, but I knew a guy who did. IIRC he spent a lot more than $40. The position encoders are built into their motors. In some cases they are mult-axis Hall effect with 1000's of pulses/rev. They are well beyond your budget unless you can beg some. I was thinking of moving the cameras off the robot to reduce weight, in which case the robot would need to track its position. \$\endgroup\$
    – gbulmer
    Oct 10, 2015 at 13:44
  • 1
    \$\begingroup\$ I think I have dealt with your questions, and also suggest you think about the whole system, not just the motors, and given your modest budget, especially consider the power supply. \$\endgroup\$
    – gbulmer
    Oct 10, 2015 at 16:52

3 Answers 3

Reset to default
1
\$\begingroup\$

Advice: Try to keep track of all the constraints/goals and your design decisions. Aim for a balance, and try to keep track of the impact of your changes. I often create a mind-map to capture the scope of all the components/subsystems, or some 2D drawing, so that I can see 'everything' and their relationships in one glance.

Making the robot as light as practical should reduce the cost and complexity of the robot's motion system.

Getting the cameras off the robot would also allow yourself to use a lot of cheap, heavy computer power, so that seems like a good idea.

If the cameras can see the robot and the projectile, it probably only needs to know the robots orientation, so that the robot can be driven in a direction to avoid the projectile. Robot orientation might be doable if the camera(s) see different coloured LEDs on the robots (visible) corners.

BLDC's have a lot of torque once under adequate control. However they are more complex to drive than simple DC motors. In my (limited) experience, getting BLDCs to start is the hardest part, which would be a disadvantage in your application.

For a fast start, the motor drive needs some feedback to know their position. I have some off the shelf BLDC controllers, intended for use in models. They take take a noticeable fraction of a second to get the BLDCs moving, and I have seen this to be the case for several other combinations of BLDC and motor control.

With a limited budget, where you may only be able to afford the cost of one set of parts, I strongly recommend you do plenty of investigation before trying BLDCs. Personally, I think they may be a bit too expensive for your budget; my low-cost BLDCs with controllers (supporting reverse, many BLDC controllers don't do reverse, but I suspect your robot will need it) were over $40.

If the robot can be reduced in weight to a few hundred grams, then small geared DC motors are powerful, popular, modest cost and easy to drive. E.g. 12mm gear motors which are available from many suppliers, though the quality of the very cheapest can be different from these (they can be had for about $6 each). You may be able to afford 4 of those. I have seen small 4WD robots (150gm) using them with off the shelf 'wide wheels', and they 'wheeley' because there is some much traction and torque.

Folks who do 'Ant Weight' and 'Mini Sumo' robot wars use those DC geared motors. They come in a range of gear ratios so you could aim for a balance between speed and torque. They need a motor controller, but there are plenty of ready made controllers from internet resellers for a few GBP capable of controlling more than enough power.

Those robot wars folks also use model servos. Again they come in a range of power and speed, can be modified for greater than 180 degree rotation, and have their own motor control electronics, and so might be quick to get working.

As well as the motors, you should think about the cost of powering them. LiPo batteries have the greatest power/weight. However, you may end up spending 30GBP+ for a battery charger on top of the cost of batteries. NiMh are not as good. However, AA or AAA are easy to buy, and the charger is cheap.

\$\endgroup\$
3
  • \$\begingroup\$ I looked up for the small geared motors, but they appeared to have less torque capacity. If the start is a problem, I may have the robot to circle around before the shot and then it might stop/accelerate faster depending on the case. Its okay for me to experiment on different motors, I can afford that myself. I only have a limited budget about what to use in the final version of the project. They want students to work their way around with limited resources. About the power source, thanks for pointing that out, I will remember that when it comes to supplying power. \$\endgroup\$
    – ozgeneral
    Oct 10, 2015 at 18:04
  • \$\begingroup\$ I actually have a "map" as you suggested, my plan is to get started with little performance and improve from thereon, keeping track of the performance improvements, which is similar to what you've suggested. Thanks for all your suggestions, I just ordered a raspberry pi 2 and couple cheap stepper motors for now. I will first have it move in a controlled manner next week, then I will start coding for image processing. Then I will have a long time to improve response times. Again, thanks a lot for sharing your knowledge. Its really nice to see people are sharing similar interests and enthusiasm. \$\endgroup\$
    – ozgeneral
    Oct 10, 2015 at 18:10
  • \$\begingroup\$ Its lovely to learn about peoples fun, exciting and intriguing projects, and a delight to try to help. So thanks to you too. \$\endgroup\$
    – gbulmer
    Oct 10, 2015 at 23:13
3
\$\begingroup\$

Permanent magnet synchronous motor (PMSM) would be a perfect choice if you have budget, otherwise a BLDC motor is also a good candidate. Stepper motors don't have a such high dynamics and short overload capability. Normaly PMSM is rated with 250% overload.
If you want a really good performance, then you hav to match moments of inertia, ideal is J_mot = J_load, if this is not possible then a reducer is used J_mot = J_reducer + J_load/P^2 , p is the reduction ratio.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks a lot, brushless DC motor seems like my best option with my budget. I might run two brushless motors in overload for a short time, which might give at least something closer to what I would need. About matching the moments of inertia, when I type reducer into google I end up with large gear reducers. Is there a term for smaller applications like robotics? Could you send me an example so that I will be find similar options using that lead? \$\endgroup\$
    – ozgeneral
    Oct 10, 2015 at 16:03
  • \$\begingroup\$ F=ma * 0.5*60 =30N; Power at end of acceleration needed P= Fv= 30*6 = 180W; T=Fr=30*0.3=9Nm, are you sure to use 0.6m wheel diameter? J = mR^2 = 0.5 * 0.09 = 0.045 kgm^2, pretty big inertia, you would need large high torqe motor at very low speed, better soultion is to lower the wheel diameter. \$\endgroup\$
    – Marko Buršič
    Oct 10, 2015 at 17:41
3
\$\begingroup\$

I'd write this as a comment, but I'm afraid I don't have enough points.

You have stated that you'd like your robot to be able to accelerate at 60 m/s^2. You also mention that your robot has wheels. I'm afraid it's not feasible to accelerate at this rate with a wheel driven vehicle unless you have some method of producing extra down-force (e.g. magnets, suction fans) otherwise it simply will not have enough grip.

Normal model robot tyres will have a mu value (mu is the ratio of the down-force on the robot to its available lateral force, or coefficient of friction) of between 0.7 (e.g. o-rings) and 1.5-ish (e.g. mini-z racer tyres), depending on the softness of the tyre used. I've seen some mention of custom mini-sumo robots that have tyres with mu value of up to around 3, which is really high. To accelerate your robot at 60 m/s^2 you'd need a mu of around 6.

Cheers, Tim.

\$\endgroup\$
2
  • \$\begingroup\$ that's a useful observation: hope you get more points soon. \$\endgroup\$
    – Jasen Слава Україні
    Nov 2, 2015 at 20:28
  • \$\begingroup\$ I think you should have enough points to comment by now :) thanks for the insight it actually seems not feasible as you pointed. I am able to design my surface though, so magnets might be a very good idea for providing the grip without increasing moment of inertia! \$\endgroup\$
    – ozgeneral
    Nov 4, 2015 at 17:07

Your Answer

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

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