See what matters when choosing the right motor for your robot.
We’ve come a long way from the previous product Turtle to the current one – Leo Rover 1.8. But the journey of developing and improving the robot wasn’t always pretty. At some point, it was the wheel motors that we encountered difficulties with. In this article, you’re going to find out what solution for a wheel motor works best for intense usage of a mobile robot.
Choosing the right motor for a rover is of paramount importance. We’ve learned first-hand that in such an application of wheeled robots that we’re dealing with, that is, “not-so-delicate”, not only the motor itself and its parameters are important, but also the design of the gearbox plays a crucial role. But we’ll get to that in a sec :)
Also, when selecting motors for the Leo Rover mobile robot, we aim at ones that allow for taking some impacts and backdriving (since it’s quite typical usage of this robotic platform ;)) and with gears that are simply durable.
Let us give you some examples from our development of the rover and you’ll see exactly what to pay attention to with wheel motors.
First, there was Turtle that ran on Buehler Motors 1.61.077.414. Then, there was Leo Rover 1.5 in which we changed the motors we’d been using to other ones – Pololu 25D. And then, we switched back to the Buehler devices when the Pololu ones hadn’t met our expectations. But first things first.
Why did we change the motors in the first place? The Pololu motors had built-in encoders, i.e. rotation sensors, which was something that the Buehler devices lacked. We intended to add rotation sensors as the Leo Rover’s functionality so that the robot could tell how much it drove. For this reason, the Pololu motors seemed to be a good solution.
In addition, the specifications of both the Buehler and Pololu motors looked perfectly fine. All the parameters such as speed, power, wheel torques in either case were similar. So was the operation of the robot itself. That’s why while developing Leo Rover 1.5, we decided to give the Pololu motors a shot. Once we implemented the new motors, everything seemed to be great.
On their own, the Pololu 25D motors were perfectly fine. Everything was indeed great. Until it wasn’t. Our customers started asking about replacement parts for their rovers because the motors in their robots had broken down. It turned out that having been subjected to more intense blows, for example, making the rover go down the stairs or drive into a wall, its motors instantly failed. We learned that the problem lay in the gearbox.
In Pololu 25D that Leo Rover 1.5 was equipped with, all the gears are connected linearly.
In a linear gear system, any impact causes the most affected gear (which is the last gear – directly connected to the wheel hub) to absorb the whole force on itself. As a result, its teeth break.
A linear gearbox is self-locking, meaning, it doesn’t allow the robot for backdriving and inertial driving. So, if a Leo Rover 1.5 happened to drive in the way that made it spin its wheels the other way, the motors’ gearbox started to break down. Thus, more intense usage of the robot caused the gear’s teeth to fall apart making the whole rover nonfunctional.
Leo Rover 1.5 was followed by another iteration of the robot – Leo Rover 1.6. With this version, we returned to Buehler Motors 1.61.077.414 – the very same ones the Turtle rover was equipped with – and we’ve been using them ever since. Although, we customized them by adding magnetic encoders to meet the original need for motors with feedback.
Unlike the Pololu 25D motors, the Buehler ones have a planetary gearbox, also called an epicyclic gearbox. Here, you can see exactly how such a mechanism works:
In a planetary gearbox, there are at least 3 gears that take all the force that acts on them. Thus, every impact is spread over 3 gears instead of one.
This type of gearbox, contrary to the one in the Pololu motors, isn’t self-locking, meaning, it allows for backdriving and inertial driving. For example, if you grab a Leo Rover (any model from iterations 1.6 to 1.8) and try to turn its wheel, it will turn (and actually the rover’s firmware will allow for that 'knowing' the wheel is being turned manually). This means that the entire gearbox will reverse the revs and propel the motor instead of the motor propelling the wheel.
We’ve tested many different motors, but we always end up returning to the Buehler Motors 1.61.077.414. Their quality is so good that they can work for a very long time. They can even outlast all the other parts of Leo Rover!
What makes them so durable then? Interestingly, in Buehler Motors 1.61.077.414 we’re currently using (and have been for a long time), there are plastic gears, whereas people tend to to go for metal ones for robustness reasons. However, being made of plastic doesn’t make the gears in the Leo Rover’s motors less durable – on the contrary, their durability is sufficient, perhaps even more than that of metal gears, as the plastic ones are made of self-lubricating POM material that allows for much more „abuse”. In addition to this, being made of plastic makes the gears work much quieter compared to metal gears (though it’s more of an aesthetic reason ;)).
But the key factor to the durability here is the mechanism of the gearbox. Of the afore-mentioned linear and planetary gearboxes that we used with Leo Rovers, it was the latter that proved to be the best choice for Leo Rover. Such a structure of the gearbox makes the motors robust and reliable as the gears don’t break in more extreme use of the rover as it happened with the linear gearbox in the Pololu devices Leo Rover 1.5 was equipped with.
For typical usage though, the efficiency of Pololu 25D is really good. So, if you’re building your own robot and you decided to use these motors, they’ll work just fine provided that your rover won’t be subjected to harsh usage. Otherwise, remember that if you want your mobile vehicle to be able to take harder impacts, motors with a linear gearbox might fail. Interestingly, JPL’s Open Source Rover was initially equipped with the Pololu 25D motors but as of now, the vehicle runs on ones with a planetary gearbox.
If there’s a takeaway from this article, here’s what matters most to us in wheel motors for Leo Rovers:
These factors play a crucial role in more intense use of a wheeled mobile robot and you might want to keep them in mind when choosing motors for your own robot ;)
Now that you know what to pay attention to while selecting the right motor for a rover, check out our article on selecting an Arduino board that will be suitable for your project :)