Monday was a slow day. First of, we found out the robotics room was flooded with fluid from a damaged AC unit. Because of this we had to spend a majority of the meeting cleaning out the room. Luckily no parts were damaged. On the brighter side, we were visted by past alumni, our previous President and Vice President
During the meeting, we branched out to each system and worked on small details.
We tested the drivetrain and bump. We noticed that we forgot to consider the size of the tank tread sprockets when designing. As a result, sprockets hit the bump. Though we rationalized that this would be the equivalent of having a third wheel like other bots, we decided to raise the drivetrain slightly. By raising the drivetrain, we can avoid power inefficiencies from elevated wheels without over complicating chain, reducing output. To fix this, we decided to place keps nuts between the pillowblocks and chassis. This small increase in height should be enough to elevate the sprockets off the bump.
Furthermore, we siphoned a motor from the funnels system and placed it on the arm. With this addition, we can theoretically raise an additional 2 pounds of metal. To further reduce stress on the motors, we added a counterweight made of a linear slide and steel 25 x 1 bars. This weight is equivalent to about .6 pounds. We're planning to further weigh this down filling the bar holes with screws.
Meanwhile, we also worked on the de-scoring system. We took a 12.5 inch channel and stuck an axle through a hole to as a hinge. when, we attached 4 lock plates to act as little forks to de-score buckies. The idea with this system is if we lower it into the column goal, the forks will rotate up and as it passes over a ball, the forks flick back out. By raising the lift, we can de-score around 2 or so bucky balls at a time. When we tested the system though, we noticed that the balls would roll as they rubbed against the side of the goal, causing it to slip past the forks. We decided to elongate the prongs of the fork by using scrap, 6 hole 25 x 1 bars, With this new length, the forklift had issues opening up. To avoid cutting, we rationalized that if we extend the balls slightly deeper, then the prongs will probably extend, however further tests will be done.
We further inspected the lift when raising and lowering. We noticed that there was a noticeable backlash in the sprockets. Because of this, the angle of the lift changes and our motors hit the ground. Currently, we do not know if this will be a problem further we'll see deeper into construction. If problems do continue, we plan on applying padding to cushion the impact of the motors, elastic down the chain to keep things tense, and limiting how low the lift can reach. We also ran into torque issues. Even though we added an additional motor we had difficulty raising and lowering the lift. We believe we're controlling the system incorrectly. We're using mecanum code to program all our tests so controlling the lift is a awkward. However, before we could test code, our controller battery died, ending the day
To make up for lost tempo from monday, we hit Wednesday hard. We worked on 3 things: The lift system, Funnels system, and Protobot. Another pleasant surprise was another one of our alumni visited.
Protobot will be our dummy bot. It's equipped with two, 3 wire tank drive on 4 high traction wheels geared at 1:1. For the lift, we plan on using a single 3 wire motor geared at 5:1 on a linear slide. We then plan on turning the linear slide into a boom lift system by attaching another motor to control the length of the slide. At the end of the slide will be a small claw capable of picking up bucky balls. If the lift doesn't have enough power, we plan on using counterweights to add more torque. We hope this robot will have similar characteristics to our competition bot. It'll have similar driving speed, same lift time, same necessary focus on the arm length as you would funnel position, same alignment necessity as a roller. However, inabilities in continuous grabbing and holonomic drive will force drivers to be as efficient as possible. By having this robot, we can both work on the competition robot and have the drivers practice. Eventually, we hope to have both robots running so drivers can practice against defensive play.
For construction on protobot, we only have parts of the chassis created.
As expected, the funnels were a pain to build. We had issues with axles hitting motors, pillowblocks colliding, space for chain, gears, sprockets, and the heat inside the garage was plain tedious, We eventually just added a piece of plate metal to mount the motor near the back, slightly right side of the drivetrain. We used a gear ratio of 6:1 for control and for power to control bucky and large balls. We applied 8" rails to eliminate cantilevering and reduce foreseen tipping issues. We used standoffs to bridge the gap between the drivetrain and brace. We ran a test near the end of the meeting and the funnels worked well. However, they could be better. We needed to apply more standoffs to the brace to keep it sturdy. We noticed the brace was mirrored incorrectly. Axles didn't feed through holes completely and thus, chain slipped. The system is structurally complete, so next meeting we plan on taking things slow and hammer small details
We were slogging through lift tests. we ran into issues downloading code. However after about 20 minutes of researching, we identified the problem and wrote some code. We ran a test on the lift with weight and the new intake. With this setup, we could only lift .6 pounds of weight. We tried rooting it to programming. Since we knew this was a low force system, we swap our programming from joystick values to buttons. With this, we hoped acceleration would be more abrupt, giving us more momentum to lift. Similar results were obtained. Next, we look for mechanical flaws. The only flaw we noticed was a missing pillowblock on one side, possibly resulting in more friction. However, we knew this wasn't big enough to cause a drop in 1-2 pounds of force. We feared that we broke a motor. We then grabbed an unused motor and tested out every motor port of the cortex. All good. We then tested out every wire. Thankfully, the second wire made the motor unresponsive. What's even better, that wire led to a 393 motor. We ran the test again with a new motor controller. The chain bar now lifted 2.4 steel channels and 2 linear slides. About 2 pounds of weight. Unfortunately, we didn't have the time to see how much weight it could take without stalling
Satisfied with the work we've had, mix ups with other clubs, and the strain of exams, IB requirements, and officer applications for robotics, we decided to take today slowly to relieve the strain on members and focused on small details on the robot.
We started off by properly spacing the funnels as much as possible. This as a result reduced wiggle within the bot. We also tensioned up the chain by removing a link, reducing even more wiggle. We replaced the hooks on the funnel with scraps of 1 x 25 bar, which should help the robot gather balls and prevent standoffs from angling as they have in the past.
We moved the front wheels of the drivetrain forward, fixing the problem of sprockets hitting the bump. Additionally, by bring the drive forward, our polygon of support increases and our center of balance goes forward slightly. Finally, we also added a bumper which not only looks appealing, but it'll act as a sturdy mount for the lift system
We also added a 17.5 inch long channel on the lift to act as more counterweight and provide as a base for even more counterweight. We also created a hinge for our large ball manipulator. This manipulator needs a hinge to stay in dimensions at the start of the match and expand when the lift is raised for the first time. Also the hinge allows the robot to travel under the barrier without lowering the lift.
Meanwhile, we focused on the protobot's arm and drivetrain. We ran into problems with the drivetrain. We were short on sprockets, so we were forced to swap sprockets for smaller sprockets and give us a slight gear ratio for speed. Despite changing the ratio, we are still close to the current robot's drivetrain of 1:1.25 speed. In addition to the drivetrain, we started working on the arm of the protolift. We planned to make this robot have a 2 motor boom lift, 1 three wire motor to raise and 1 three wire motor to lift. The lift is going to be geared at 5:1 torque, which should be enough, combined with counterweight and elastic, to raise a claw and bucky ball.
In addition to the protobot, we also we started developing another de-scoring system in parallel to our own. We took one of our gateways pieces and two 1 x 25 bars. We curved our bars and bolted them together. By attaching elastic across the ring in the forming a "#", we can lower the ring over balls and the elastic will stretch. With this, the balls will get caught in the net and by raising the lift, we can possibly de-score 3 balls without requiring any expansions.
Lift test: two 393s + 1 269
Shout out to NAR for reviewing our design
Lift test Bloopers
During the weekend, I decided to think more about shooting. The hard part was figuring out how to shoot without using motors. As a result, i decided to reflect on my elastic powered lift systems. the original purpose of the elastic powered systems was to minimize the amount of motor power necessary to raise a heavy weight without jeopardizing axles and still maintain a fast lift. Studying degrees of freedom from mechanisms like the steam engine, i came up with something like this. During the time i recorded this video, i didn't completely grasp the concept. As a result a few things are different. Rather than elastic attached to the mount at the top, it shold be string. This string causes tension in the elastic of an arm system. Another addition is that the center inhibiter is completely unecessary (which is why i was confused near the end of the video). Because the mount attached to the string just hovers there, the hook can overshoot the mount and still relatch without turning. Also, maintaining position as i mentioned in the video is false. The hook will slightly turn clockwise, but thats more good than bad. Hopefully this concept will contribute to shooting and lifting in the future