Micro Metal Gearmotor #1093 question

Support Other Pololu products
1

I am new to this forum so if something is wrong with my question please let me know.

I recently bought two 30:1 motors (#1093) to build a line follower robot. I want to accelerate the robot during straight parts of the track. This also means I need to brake before a curve. My question is about a good, proven strategy for breaking with this gearmotor. I use a dual motor driver TB6612FNG (#713). I do not mind to implement an algorithm for optimal braking like a PID controller. But first I need to understand how to use the gearmotor / driver for braking. The driver spec mentions a feature called short-brake but so far I could not figure out if this is something useful for a solution to my problem.

Do I risk damaging the motors by using motor-brake?

I googled for documentation without any luck so the link to the RTFM would be much appreciated.

2

Hello.

It should be fine to use the brake low feature of the TB6612FNG with that motor. Depending on your setup and the course, you might be able to slow down enough for the corners by just using the “coast” feature of that driver.

All the information we have on the TB6612FNG motor driver carrier is on the product page and the driver’s datasheet. However, you might find the “Motor Driver Truth Tables” and “Simple Code to Drive the Motors” sections of this application note on using the TB6612FNG on the Orangutan robot controllers helpful.

- Jeremy

3

thanks for your answer. I understand that you do not have more information but this does not answer my question regarding a proven strategy for breaking with this driver/ gearmotor. I kind of hoped that one of the forum users would share some insights on the topic.

4

I’ve used the slightly higher-geared version of this motor.
In my experience, it’s likely that your wheel friction is going to be a bigger impediment to proper braking than the motor itself.
With a suitable motor driver, you can attempt to brake by forcing current backwards, even. That’s going to switch pretty quickly; faster than your wheels and chassis can keep up due to inertia.
Another challenge will be making the bot go straight at high speeds, unless you use encoders.

I suggest you put your bot together, and then make it go forward on a white piece of paper straight to a black line/area. When it hits the area, make it clamp down on the brake (the “shorted leads” braking.) Measure how far into the black the detector actually goes.
If you have encoders, it will be more possible to “force” the wheels to stand-still, but also more likely you’ll slip against the surface.

5

I think measuring like you explained (black area) is a good idea since as we all know one can not improve what one can not measure.

Before I start measuring I need at last a rough idea how the breaking works with the gearmotor / motor driver.

Three points I do not understand from your answer, could you please elaborate a little:

  • With a suitable motor driver (what makes a motor driver suitable, does that mean that mine is NOT suitable?)
  • That’s going to switch pretty quickly
  • If you have encoders, it will be more possible to “force” the wheels to stand-still (how are encoders related to breaking)

If I understand correctly you advice “blocking wheels” for optimal breaking. Why is that? From my car driving lessons I remember that “blocking wheels” is the worst breaking.

6

When the bot is moving, the motor generates torque, which is transmitted through the gearbox with some loss, into the wheels, where torque is converted to force that accelerates the bot forward, as long as the friction between wheels and floor is high enough to avoid slipping.

When you want to brake, this is inverted. The friction of the wheels against the ground and the inertia of the bot drives the wheels, which back-drives the gearbox, which makes the motor spin, which generates current (like a generator!) There are losses in this transmission, that make a lightweight bot with low inertia stop “pretty quickly” (which you’d have to measure to know exactly how quickly.) The braking comes from both gearbox losses, and motor losses.

The “shorted braking” feature means that the output of the motor, when it is working as a generator, is shorted together. This will present a large amount of load on the motor, which will significantly increase motor losses – the motor-turned-generator will now dissipate more energy, and thus bleed more inertia from the moving bot. Any inertia that is still present will move the bot a little bit, which you can sense as moving wheels as long as there is no slip.

If you want to force the wheels to not move, using motor drive to apply force, then you need to be able to detect movement of the wheels, which can be done by detecting movement of the motor shaft, which is what encoders is for. You could thus implement a control loop that detects movement on the wheels, and applies reverse force from the motor to move the wheel “back” to where it was. Do this fast enough and the wheel will be, effectively, stationary. (See PID control loops for more details.)

When it comes to “blocked wheels braking,” the reason you don’t lock your wheels in a car is twofold:

  1. Locked wheels will just skid forwards; you lose steering. It’s often more important to be able to steer than to stop as quickly as possible.
  2. On some kinds of surfaces (rain/snow? I forget) the rolling friction may actually be bigger than the sliding friction, and thus non-locked wheels will stop faster by bleeding the kinetic energy out the brake pads instead of out the road/tire contact patch.

For rubber tires on dry asphalt, if the shortest stopping distance is the only important thing, locking the wheel is typically maximally efficient.