We are attempting to use custom motors (powered by Vex Spike) to run our very heavy robot. There cannibalized from drill motors, which give us a huge amount of power and torch. The issue though is getting them to work at the same speed. From playing with Mindstorms kits and general knowledge, its very hard to get 2 motors to turn at the same speed. And we just had one of the motors fail.
I was thinking of trying to have 1 motor give power to both wheels and have some mechanism for removing power from each wheel (which will give us 360 turning abilities). However we have absolutely no idea how to create one simply.
The design I was thinking of is having 1 motor at a 90 degree angle to the shaft. Then using 2 bevel gears we can make the shaft turn. Thats the easy part.
Starting simply, how can we remove power from just one wheel? The only thing I could think of is having a piston push the wheel off the axle (hey, I’m not a car or super robotics person). But there has to be some other way to do it.
I know this is a very unique situation, but does anyone have any idea’s?
You want a differential drive to send power to both wheels, but then allow one or the other wheel to be stopped via a separate clutch or brake.
The diff that Vex sells is too small for your needs, and a car diff would be absolute overkill. I’m not sure of a source for something in between, but I’d start by looking through Surplus Center’s website.
What Chris has said is very true the two motors should really be run both clockwise or both anti-clockwise because one direction is considerably slower than the other.
You may find this difference so huge that seeking a different motor is a better choice. I would try robot marketplace before any other place. They are pretty reasonable and have a good selection plus it is always nice to support another robotics enthusiast.
I think that could explain why that one motor broke. But making it go the other way will require 3 bevel gears, which would be very hard to acquire and mount correctly inside of the assembly (it fits inside of a 3 inch wide aluminum square tube, weird I know).
We actually got that suggestion, but we’ve already drained our small school issued budget. Our teacher is now starting to invest his own money into the project. While one would be very nice, we just don’t have the money.
Doing some quick research, that actually does seem like a great idea. We would need one though that would only let us choose when the speeds are different, not just when it encounter’s resistance.
As you said though, acquiring one will be hard. We will try and ask around. This is the best idea I’ve seen in a while!
You can’t do this with bevel gears along, anyway. Stopping up one wheel would also lock up the motor, which would not allow you to steer.
As I see it, you’ve got a few choices:
*]Two motor skid-steer. This is the simplest mechanically, but getting the two sides to spin at the same rate can be a challenge. You will probably require rotary encoders that your software uses to keep the wheels as in-sync as possible.
*]One motor differential steer. This uses one motor driving a differential, and a brake on each wheel for steering. Note that a differential will not force both wheels to spin at the same rate when you are trying to go strait, but it will allow it if the terrain cooperates. Finding a diff in the right size/price range will be a challenge.
*]It is possible to use a pair of diffs and a pair of motors to get what you want, but it is mechanically complex to hook up. If you hook it up right, one motor provides the drive for both wheels. The other motor is the spin/steer motor - when it is stopped, both wheels are forced to stay in sync. When it turns, the wheels are forced to spin in opposite directions at half of it’s speed. Regardless, finding 2 diffs is probably not an option if finding 1 diff is an issue EDIT: Here is a pic of something similar - not exactly the same, but using diffs as “mechanical adders”.
So you want to cut power to one wheel? Well then use a shifter…
If you dig through some of the gallery (I mean really dig), you’ll find a picture of some shifters. I’ve seen teams accomplish them through servos and what nots. The key will be to shift out into neutral, and not into another gear.
The reason that the drill motors move with considerably more torque in one direction as opposed to the other is because the brush timing is advanced. One motor is moving with it’s brushes in the retard direction, this causes it to run hot, slow, and with less torque. The way you can change it is by rotating the housing holding the magnets (if any). I’m almost certain that the magnets will just be glued in on the motor can with your drills, but it is possible to carefully rip off the magnets (noting their postion, take a nail or a centerpunch, just something sharp, to trace where the magnets were before you rip them off. Then, you can accurately adjust the magnets to your needs.) However, I don’t know what specific direction you should adjust them towards… I might be able to find more on that later. Sounds labor intensive, I know, but it’s better than making some big gearbox to reverse the wheel direction.
Many hardware problems can be solved in software if you have the correct sensors.
We have used gyros on our FRC robots for a couple years now to help them track straight.
If the gyro senses the robot turning to the left, the robot dials back power to the right hand motor, and vice-versa. It takes a bit of practice to get it tuned.
I’m sure there are mechanical solutions, but you’d be hard pressed to find one simpler or cheaper. The gyro can also keep you going straight when other robots are bumping in to you.
If you don’t have a gyro available… how about encoders? Put an encoder on each shaft, monitor axle speed, and if one wheel is going too fast, then dial back power to that motor.
The #1 thing we’ve found, however, to help a robot go straight under driver control, is to seperate the steering joystick from the forward/reverse joystick. Typically left thumb is front/back, and right thumb is left/right. It might not sound like a big deal… but it works.