Ok, i just thought i would throw it out there: quadruple flywheel anyone? I mean it might sound like overkill, but wouldn’t that give you accuracy on both the x-axis and the y-axis? Any thoughts?
Not only accuracy on the x and y axis, but with DIVINE code, you could face the flywheel upwards and using inertial navigation where the robot knows where it is, the robot would be able to shoot from anywhere to anywhere on the field with near perfect accuracy and while the robot is facing any direction. Unfortunately it would take 8 motors to power the flywheel so only 4 are left for a feeding mechanism and wheels… Only The divine code for aiming the ball and an extra motor so you can have a 4 motor drive and motorized feeding mechanism are the only things stopping the creation of arguably the most powerful fielding robot only cons I can think of is fire rate, movement speed, and weight (and the extra motor and the complexity)
A quad flywheel doesn’t sound too bad but also seems like it has a few inherent problems.
It seems like it would take up a lot of space on the robot and take away from other subsystems. From what we’ve seen so far, most flywheel systems(especially horizontal double flywheels) take up a lot of space. A quad flywheel sounds like it would have to be insanely space efficient if you wanted to fit in an intake or any method of feeding into the flywheel itself. I suppose you could build the flywheel at an angle and tilt it back into the robot to make room for an intake of some sort.
While control over both axes of a ball might sound good in theory, variability doesn’t make for a very consistent robot. Double flywheels already face a major issue in maintaining consistent speed across both flywheels if they aren’t chained together. Adding a second set only introduces more variability into the system and possibly more inefficiency if you decide to chain the flywheels together.
Tuning the velocity control seems like it would be a pain too. At the very least, you’d have to write two different control loops with a different set of constants for each that would need to be changed depending on tournament conditions. Doesn’t sound like a lot of fun.
But really, I’m all for a quad flywheel. Just imagine the sick curved trickshots you could do with that level of control. The real DANK MLG NOSCOPES.
Fine tuning that would be very difficult and annoying I think…
I was thinking maybe 4-6 motors and bevel the wheels together. You could easily add encoders and, with a little programming, make it pretty accurate.
It looks like double flywheels have to deal with horizontal as well as vertical accuracy compared to a single flywheel which only deals with vertical accuracy. While possible to tune, I feel like a quad flywheel would face issues with accuracy in all directions.
Correct me if I’m wrong, but wouldn’t you only need 3 flywheels to get the full spectrum of control? I would think that you could then change which flywheel operated at a higher speed to change the path the ball would take, while requiring less motors and eliminating the need to chain across flywheels.
While arguably harder to construct and mount, a triple flywheel with 120 degrees between the flywheels would, I think, be a slightly more feasible alternative to 4 flywheels.
In any case, can’t wait to see what comes out of this
I don’t think a quadruple flywheel is necessary. I can’t find the video, but their was a team with a triple flywheel that was able to shoot from anywhere on the field, and was still quite versatile.
I’m guessing your talking about Wingus and Dingus. He’s got like 9 “flywheels” all lines up and shoots like 3 at a time.
Also, as far we know so far, based on their teaser, 8059 could be a quadruple flywheel…
I think that idea would be kind of a hard-to-do thing. Because since it’s contact with all 4 flywheels, and at the least of 2 motors per wheel, you would need so much programming skill to make it work. Here is pros and cons of the 4 flywheels:
- Allows you to shoot from anywhere from the field, and at least be 45 degrees aimed towards net. No problem with accuracy, just make the robot increase speed of left or right flywheel.
(Since this was brought up too, you should be able to shoot at least 45 degrees aimed towards net even with 2 horizontal flywheels. Huh :/)
(Sadly could only find one pro)
1.After a shot, since it’s only 2 motors per wheel, the time between shots will take longer. So by doing so, you must have an encoder on each of the flywheels and make a PID loop for each one of them.
2.Very inefficient on space and motor usage. If you were to make one, it would be a struggle making an accumulator for it, and affording to have a drivetrain.
So… I think that 4 wheel flywheels is an OK idea, but why not just try 2 wheels because it should have the same pros, but not as many cons.
This is actually an idea that we prototyped at the beginning of the year, it was one of our first launcher prototypes. We never actually got it working, as it got REAL complicated real fast. We encountered so many issues. The complexity of the structure was insane. Think about it. We used 4 inch wheels, so that was 12 inches wide and 12 inches tall. Now, you either have to make 4 gear boxes or connect them. Connecting also makes it insanely complicated, as you have to build it around your wheels and path for the ball. A mechanism required for this would have SO much friction, not to mention the fact that it would just make it bigger. It just wasn’t worth the effort.
No, there was a team that posted an innovate award video that had 3 wheels spinning in a triangular fashion to shoot balls. They were able to shoot from anywhere on the field.
I found the video.
Theoretically shouldn’t it work to just have a normal horizontal two wheel shooter with a third vertical wheel?