2016/2017 Winter Break Robot Development

 

Goals for out robot after our hosted scrimmage during winter break:

• Create a cap ball mechanism that holds onto the ball stronger and doesn’t wobble as much
• Convert to the new Actobotics chassis robot
• Create a sweeper system that is fast (most likely won’t have a system to hold more than 1 or 2 balls for more space)

 

Tuesday, Dec. 20, 2016

The Actobotics package came, so the chassis was finished:

Chassis description:

This chassis has the same design as the chassis used throughout the scrimmages. It is a 4 wheel drive 2 motor chassis that uses chains. Switching to Actobotics parts for the chassis has the following benefits:

• Wheels spin more smoothly
• Sturdier
• Larger gaps in the H shape

The two channels that are standing upright on the back end are longer than those of the current main robot. On our main robot, there is a channel between the two upright channels that is used to press the beacons in teleop and to create a sturdy structure with the linear lift. However, from the experiences from the first two scrimmages, this channel between the two uprights doesn’t press the beacons fast enough (even though we are already faster than every other team in NTX). Instead, we will use the autonomous beacon mechanism on the left side of the robot for teleop also, since it should be faster. Because the channel between the two uprights isn’t necessary anymore, the two uprights will be taller so that the sweeper system can go higher faster (before the channel in between the uprights was constantly hit by the sweeper, slowing down the sweeper).

Chassis development summary:

• Switched to actobotics to be sturdier and have smoother wheels
• Raised the two uprights because the channel in between will not be used for beacon pressing in teleop anymore, and to make the sweeping faster

 

Wednesday, Dec. 21, 2016

After much thought, the cap ball mechanism will switch from the current claw design to a new ramp locking design.

Problems with the claw mechanism:

• Takes up a lot of space in every dimension
• Wobbly which weakens the linear lift
• Isn’t able to hold the ball low or strong enough to grasp it
• Claw positioning is difficult (doesn’t start at the same place always because of the 1 to 1 ratio use of a 1000+ degree servo)

The ramp locking design is somewhat similar to what most teams use in that the cap ball is controlled by something underneath, instead of from the sides with our current claw mechanism.

Thought process:

1. Definitely not going to use a claw mechanism
2. Many teams control the cap ball from underneath with success, we will do the same
3. How can we make ours unique?
4. One of the current challenges with obtaining the cap ball is that the linear lift (which is at the very front), prevents our arms from reaching around as far. If the linear lift was farther to the back of our robot, then the cap ball could be closer in reach of our robot, since the front of the robot could be flat and could go underneath some parts of the cap ball.
5. How could I get the cap ball to rest on the front part of the robot (which should be a secure hold/position)? Probably a ramp. A ramp wouldn’t take up much room as it can start straight up and then fall down. It would rotate.
6. How can I make sure that the cap ball stays controlled by the robot even when the linear lift elevates? Make the same ramp raise in an angle.
7. How can I raise the angle of the ramp without using servos or motors (it would require a lot of force)? How about drive the robot into the side vortexes, and use their slope to angle our ramp passively?
8. How do I keep the ramp at this angle once it gets pushed up with the side vortexes? Use a mechanism where an object can only rotate one direction. Our ramp will be raised when it drives up the side vortex and it will stay there because of this one directional mechanism. It should be strong enough.
9. How would our ramp fall down at the start of the end game and before we release the cap ball if we use this one directional mechanism? This one directional mechanism uses a gear and a stick. The stick is positioned in a way against the gear so that the gear can rotate one direction and the teeth will slide under the stick and if the gear rotates the opposite direction its teeth will get jammed against the end of the stick. The stick will stay against the gear with a rubber band. We can use a servo that is strong enough to more the stick away from the gear so that the gear can rotate freely either direction.

 

Ramp locking mechanism:

The end of the Tetrix channel will be attached near the bottom of the linear lift. On top of this Tetrix channel will be a wide platform with slight walls at the ends to prevent the cap ball from falling out left/right. Additionally, there will be beams on the bottom near the ends of the platform that will prevent the platform (and hence the linear lift) from wobbling while traveling. The L brackets on the end of the Tetrix channel will be connected to the ramp. Finally, this mechanism uses Tetrix pieces because it doesn’t require the sturdiness of Actobotics parts and needs to be smaller (Actobotics parts are slightly larger)

 

New cap ball mechanism summary:

• Ramp locking mechanism
• Uses Tetrix pieces to reduce space and is simpler
• Requires one basic servo with little torque and range
• Takes less room than the previous claw mechanism
• Requires that the linear lift be moved back several inches

 

Thursday, Dec. 22, 2016

The Tetrix robot that competed in the first two scrimmages was built in the following order:

1. Chassis
2. Autonomous beacon attachment
3. Cap ball
4. Shooter
5. Harvester (which was never completed)

The Actobotics robot will be built backwards (however obviously starting with the chassis), because of spacing. Since the autonomous beacon attachment and cap ball mechanisms were built first, there is a strong understanding of how much space their future designs will take, unlike the shooter and the harvester systems.

The harvester will be 2 sweepers vertically aligned that spin at the same speed, just like on the old robot. However, one of the main changes from the old sweeper system is that the new one will be wider, to increase the speed of particle intake.

There will be neat walls surrounding the inside of the harvester to guide the particles up top, instead of bent aluminum on the old robot.

The picture above is the first set of sweepers that were made that were used on the Tetrix robot. The problem with old sweeper system was that the sweepers weren’t very wide, which means that it is just a little harder sweeping in particles. There will be changes on the positioning of the sprockets in relation to the sweeper and the channels that support them to increase the width.

Before, we used PVC with holes throughout to hold the zip ties and rubber. Now, the zip ties and rubber will be held together with an aluminum beam, with diamond shaped aluminum plates to cover each individual zip ties and rubber. This will make replacing each zip tie much easier than if we were to just sandwich all the zip ties between two aluminum beams (but much harder to make).

Additionally, one set of sprockets from the motor to the first (bottom) sweeper will be on the left side of the system, and the other set from the bottom to top sweeper on the right. These sprockets will be as close to the channel as possible in order to increase the width of both sweepers.

This is a close up of the piece that will hold one end of the aluminum beam. It is made up of two Aluminum L beams that are held together with one screw. This screw will eventually be attached to one end of an aluminum beam that will hold all the zip ties.

 

Sunday, Dec. 25, 2016

The sweepers were finished, and the sprockets were all chained together. Also, the walls were added to guide the walls. On the top of the sweeper system are many narrow beams shaped in quarter archs that are in between each zip tie of the top sweeper. These archs help guide the particles to the front of the robot, where they will roll into the shooter.

Originally, the rubber parts were longer and they hit the ground, the top, and the aluminum beams of the other sweeper. Once they were cut around 1 cm shorter, they spun faster and the sweeper system was much quieter. The archs decreased the time it took to get a particle to the very top by a lot, by pushing the particles into the front.

 

Monday, Dec. 26, 2016

The shooter was mounted onto the robot nearly the same way as the old one. Before, the main supportive beam from the chassis to the shooter was tilted, and this tilt was what angled the shooter. This took up extra room, so the new shooter is mounted onto the chassis with a straight vertical beam. The tilt is created by sawing a piece of wood at the desired angle, and then attaching this piece of wood between the shooter and the main beam.

The shooter was attached as close to the right side of the robot as possible to leave room on the left side for the autonomous mechanism and other things.

After the shooter was attached, a ramp was made at the top of the sweeper system to guide the particles into the shooter one at a time. This ramp is tilted at around 5 degrees. Additionally, walls of tape were placed at the front side of this ramp to prevent the particles from shooting into the middle of the robot. Tape was used since it is very easy to use and it isn’t very hard, so the particles won’t bounce back.

At first the tape only prevented most of the particles from escaping the system, since the sweepers were pretty powerful and shot the particles over the tape walls. Then the aluminum bars that hold the tape up were replaced with longer bars, so that the tape walls could be higher. After this change no particles escaped and every particle that entered the sweeper system would end up in the shooter pretty fast.

Tuesday, Dec. 26 2016

Now that the sweeper and the shooter systems were finished, the cap ball mechanism was next. Because of the space taken up by the sweeper and shooter systems, the linear lift would have to be closer to the front of the robot than expected. Hence, the cap ball locking mechanism from earlier would not work sine it would take too much room front to back (it was too long). And rotating this mechanism to where it would stand vertically would not work because that defeats the purpose of moving the linear lift back, to prevent anything from touching the middle of the cap ball which means the bottom of the cap ball is further out (since it is in the shape of a sphere). The solution is to make this locking mechanism shorter.

The linear lift was attached onto the robot using wood instead of a lot of aluminum L beams to make it easier. It is still mounted on with the same structure as the old robot, a beam across its bottom and another channel connecting the top of the linear lift to the top of the sweeper system.

The locking mechanism was reduced from being around 8 inches long to less than 3. This was done by cleverly using the same axle to hold the passive locking mechanism and the active unlocking mechanism.

Wednesday, Dec. 28, 2016