next topic »

[on:]

Hello guys,

I am looking for a motor controller that can be used in differential drive robots (mixed mode). So far I had seen Sabretooth and Roboteq's AX series. The problem that I face now is that the motors that I would like to get (for example this http://www.robotmarketplace.com/products/0-PDX26.html ) has stall current wayyyyyyyyyyyy above the motor controller limits. And I don't even know where else can I find a controller that can do differential drive and also be able to support that kind of stall current.

And another question, how do I know how much torque I will need for my tracked robot which weighs around 20KG or less? Is the motor located in the link I provided a bit of an overkill for such robot?

[Reply #1 on:]

There isn't going to be any cheap and easy options when dealing with motors with such high stall currents.  The site you're getting the motor from does provide a number of speed controllers that operate at those currents and higher, but they start getting pricey.  Building a large robot is a pricey endeavor though, no way around that really unless you can hand build virtually every component of the robot, and then you're really just trading financial cost for time cost.

As for the appropriate size motor for your robot, that depends on a variety of factors.  How fast you want the robot to go.  What kind of tire setup you are using. Treaded setups require a lot of torque for differential steering as you are doing a lot of skidding and need to be able to overcome the static friction between the treads and the ground, and even this friction will vary from surface to surface.  For the same reason 4 wheels differential steer system will require more torque than 2 wheel setups.  6, 8, and more wheels will all require more torque as well.

[Reply #2 on:]

And another question, how do I know how much torque I will need for my tracked robot which weighs around 20KG or less? Is the motor located in the link I provided a bit of an overkill for such robot?

Two questions you need to answer:
1) What is the radius of your drive wheels?
2) If you tied a rope to your robot and pulled on it, how much force would you need to apply before the robot lost traction and started sliding?

Using a hypothetical answer of 10kg of force to slide the robot and wheels of 5 cm radius, then you'd need a motor with 10kg * 5cm = 50kg/cm of torque.

[Reply #3 on:]

I think there are a number of options on the site that sells the motor  - have you looked through these? 
Maybe:
http://www.robotmarketplace.com/products/RS-RS160D.html

[Reply #4 on:]

At 900+ rpm, you are going to have one quick tracked bot. Below are two h-bridges that might fit your needs.

http://www.parallax.com/tabid/768/ProductID/64/Default.aspx
http://secure.oatleyelectronics.com//product_info.php?products_id=206&osCsid=30093449b694ae247d94c2bcce5d0560

[Reply #5 on:]

hey everyone. I really appreciate your inputs.

At 900+ rpm, you are going to have one quick tracked bot. Below are two h-bridges that might fit your needs.

http://www.parallax.com/tabid/768/ProductID/64/Default.aspx
http://secure.oatleyelectronics.com//product_info.php?products_id=206&osCsid=30093449b694ae247d94c2bcce5d0560

I think there are a number of options on the site that sells the motor  - have you looked through these? 
Maybe:
http://www.robotmarketplace.com/products/RS-RS160D.html

for the two quotes above, they don't seem to be able to do differential steering. so how do i achieve differential steering? Or am i missing something here?

There isn't going to be any cheap and easy options when dealing with motors with such high stall currents.  The site you're getting the motor from does provide a number of speed controllers that operate at those currents and higher, but they start getting pricey.  Building a large robot is a pricey endeavor though, no way around that really unless you can hand build virtually every component of the robot, and then you're really just trading financial cost for time cost.

As for the appropriate size motor for your robot, that depends on a variety of factors.  How fast you want the robot to go.  What kind of tire setup you are using. Treaded setups require a lot of torque for differential steering as you are doing a lot of skidding and need to be able to overcome the static friction between the treads and the ground, and even this friction will vary from surface to surface.  For the same reason 4 wheels differential steer system will require more torque than 2 wheel setups.  6, 8, and more wheels will all require more torque as well.

And another question, how do I know how much torque I will need for my tracked robot which weighs around 20KG or less? Is the motor located in the link I provided a bit of an overkill for such robot?

Two questions you need to answer:
1) What is the radius of your drive wheels?
2) If you tied a rope to your robot and pulled on it, how much force would you need to apply before the robot lost traction and started sliding?

Using a hypothetical answer of 10kg of force to slide the robot and wheels of 5 cm radius, then you'd need a motor with 10kg * 5cm = 50kg/cm of torque.

To jraskell - Thanks for your advice. Actually I don't need my robot to be fast, I just need it to be around 0.5m/s. For the acceleration part, I don't really know what I need. What if I do want my robot to go on an incline, like maybe 20 degrees? Thanks in advance for your advice.

To Chagrin - I always read about the tire radius and stuff, but how do I determine my radius when I'm using tracks? I have these servos which I am going to modify to continuous rotation. The model is Seiko PS050. Do you think it is strong enough? And do you have any idea how I can check the stall current of the servo?


THANKS GUYS! (:

[Reply #6 on:]

Differential steering is a software issue - as long as you can make a single motor go forward and backward with some sort of command from the microcontroller, you can do differential drive.

Track radius is equal to the radius of the driving sprocket - but turning torques are going to be MUCH higher due to the fact that the track is situated perpendicular to the desired direction of movement (and thus has a huge amount of friction). 

No modified servo will be able to drive anything above ~2kg @ 20cm/s or so, much less 20kg @ 50cm/s.  You're limited by physics:  A servo at 6v draws around 1A, making a power input of 6W.  With a drive radius of 5cm and a mass of 20kg, even at an acceleration of say 5cm/s² at a speed of ~900rpm, is an easy 27W.  And that's disregarding inefficiencies, which are unlikely to be above 80% or so.

[Reply #7 on:]

so what do you guys propose that I do?

[Reply #8 on:]

I agree with everything that Aeturnalus said. Furthermore, that stall current value is very questionable. In other forums, people are doubtful of that number. I think that it would be a good idea to measure the stall torque yourself.  The method is to simply connect the motor to a power supply with a current meter in the line, and grab the shaft and stall (or attempt to stall) the motor. Of course you would only do this momentarily (for a second or two) to see what the maximum current is.

Something to note: For certain size and torque motor - you DO NOT want to grab the shaft with your fingers (friction burn, if you're lucky). Instead, build a pony brake (or clamp on to the shaft with a piece of leather and some vice grips).

[Reply #9 on:]

so what do you guys propose that I do?

You probably need to spend some time researching similar projects on the net. Controlling the motors for differential operation is the easy part. Building a 20kg tracked bot may be more difficult than you think.

[Reply #10 on:]

Also - you might want to look into this hobby - and use/abuse the ideas there for large-scale tracked vehicles:

http://www.rctankcombat.com/