Rookie Team 3992 Launcher!

[ksafin]

100% consistent!

We're going to change speed of the wheel and angle, to get into the top hoop.

Stay tuned!

http://youtu.be/ULOqQCh59vA

Feedback and comments appreciated!

[aeniug2]

Looks really good... Although I am curious... Don't you get issues with the ball feed using only the one belt? Even with the 2 pipes?

[ksafin]

Not so far;

We've tested using 3 balls in the belt at once, and everything worked fine.

Feeding is pretty consistent.

[aeniug2]

Originally Posted by ksafin

Not so far;

We've tested using 3 balls in the belt at once, and everything worked fine.

Feeding is pretty consistent.

Sweet... Ill keep that in mind when we assemble our launcher tomorrow... Thanks for posting the video btw... Without videos of other people's launchers I don't think our team would have ever started on ours...
(they don't like how everyone else is ahead of us, lol)

[pfreivald]

That looks really great, guys and gals! The mechanical consistency is impressive. (We also have a belts-and-two-rails system that is very consistent, though we use more belts... Something to consider is when you get hit by other robots, will the balls pop out?)

If you can aim, you'll have a very competitive bot!

(Yay, rookies!)

[ksafin]

Originally Posted by pfreivald

That looks really great, guys and gals! The mechanical consistency is impressive. (We also have a belts-and-two-rails system that is very consistent, though we use more belts... Something to consider is when you get hit by other robots, will the balls pop out?)

If you can aim, you'll have a very competitive bot!

(Yay, rookies!)

Thanks!

We'll definitely look into more belts. As it is right now, I think the biggest problem will be that, you can put a ball in directly centered, or you can put a ball in so that the belt is slightly on the left or right of the "equator" of the ball. So I'm wondering how we can make sure each ball goes in exactly centered.

Similarly we're finding the problem with consistency to be affected by how used the balls have been. We're looking toward the "window" we have where the slightly newer balls rebound off the backboard, and the slightly older balls go straight into the hoop. If we can get it to aim within this window, we think we should be set..

We'll order belts today

[dellagd]

What is the top part of it? Is there a hood that follows a radius from the wheels or just a bar

[ksafin]

What is the top part of it? Is there a hood that follows a radius from the wheels or just a bar

It's copper pipe that we bent around the radius of the wheel.

It's easy to forget in the excitement of your first build that your robot is likely to get the living crapola rammed out of it during games. We ram ours into walls, flip it over onto its sides, tug on cables... I suggest that, as soon as you have it working, you do the same -- and then fix whatever no longer works, and then do it again, and again, and again, until 2-5 times in a row there's nothing to fix.

Will do; we just want it to work first though

[Consider some kind of funneling system from your ball pickup, if you haven't already. The dual rails help a great deal, and one belt might be sufficient -- three-point centering is older than Newton! (The scientist and the field).

I see. I'm thinking the bottom of the conveyor can start open, then we can put two angled sheets of lexan that will gradually compress the ball to be the same each time.

Yup. That's true with everyone. Your 'sweet spot' technique is a good one, to be sure!

Thanks!

We made our own from surgical tubing, 12 gauge wire, and superglue.

Impressive!

[nitneylion452]

I think you're showing me the same shot over and over again. You guys have a really good video editing team! You can't even tell!

Seriously though, great work! That looks fantastic!

[ksafin]

Originally Posted by nitneylion452

I think you're showing me the same shot over and over again. You guys have a really good video editing team! You can't even tell!

Seriously though, great work! That looks fantastic!

Haha thanks so much!

Very encouraging words

[Lil' Lavery]

When you get on the field, the conditions end up being pretty much nothing like your test space. It's the unfortunate reality of FIRST (and 99% of engineering in general). You guys have definitely done some impressive work for a rookie team, but there's a couple issues that could crop up once you set it on the field of competition.

Your feeding belt takes an awfully long time in FIRST scale. Five seconds doesn't seem that long in the real world, but in a 2:00 match having to wait five seconds for balls to traverse the final two foot span of your robot isn't a good idea. Especially considering that there's five other robots on the field complicating things, three of them actively working against you. You may want to attempt to speed that process up.
It looks like you're using a Fisher-Price motor into the stock gearbox, directly feeding a rotating barrel (PVC?). You could increase the diameter of the barrel to get faster belt speed, or you could gear up your current barrel after the gearbox. This is assuming that you are not currently reducing the power to your motor via software.

As mentioned and you guys realize, centering the ball can be a big deal. Fortunately, it may also serve to help protect your balls from being dislodged or repositioned during a match. Try to prevent as much unwanted motion of your payload (the balls) as possible. But at the same time, avoid causing any jams in the system. Make sure your centering and protection devices don't cause balls to get stuck.

Aiming from a distance is difficult. Especially considering your drivers are another 40+ feet from your robot when you're attempting to shoot. You might want to consider features to help your robot aim, and perhaps from a closer distance (such as actually in contact with the fender). The holy grail in this situation is camera tracking and rangefinding of the targets, but very few rookie teams have the time and experience to code, test, and debug those type of systems.

Good luck!

[pfreivald]

Originally Posted by ksafin

We'll definitely look into more belts.

It's easy to forget in the excitement of your first build that your robot is likely to get the living crapola rammed out of it during games. We ram ours into walls, flip it over onto its sides, tug on cables... I suggest that, as soon as you have it working, you do the same -- and then fix whatever no longer works, and then do it again, and again, and again, until 2-5 times in a row there's nothing to fix.

Originally Posted by ksafin

As it is right now, I think the biggest problem will be that, you can put a ball in directly centered, or you can put a ball in so that the belt is slightly on the left or right of the "equator" of the ball. So I'm wondering how we can make sure each ball goes in exactly centered.

Consider some kind of funneling system from your ball pickup, if you haven't already. The dual rails help a great deal, and one belt might be sufficient -- three-point centering is older than Newton! (The scientist and the field).

Originally Posted by ksafin

Similarly we're finding the problem with consistency to be affected by how used the balls have been.

Yup. That's true with everyone. Your 'sweet spot' technique is a good one, to be sure!

Originally Posted by ksafin

We'll order belts today

We made our own from surgical tubing, 12 gauge wire, and superglue.

[Alex.q]

I like the design; we are doing something similar. It appears that there is only a sprocket on the belt at the top and it is just wrapped around a pvc roller at the bottom. Is this accurate, and if so, how well does this work. Does the bet transfer enough torque to the roller, or is that not necessary in your design?

Originally Posted by pfreivald

Consider some kind of funneling system from your ball pickup, if you haven't already. The dual rails help a great deal, and one belt might be sufficient -- three-point centering is older than Newton! (The scientist and the field).

Are you suggesting that 3pt centering would be better than using 5 points of contact (belts + 4 pieces of aluminum in a similar set up to this)? if we had 2 belts, should we put them closer toegether to keep the contact close to one point or spread them out (2-4") apart so that the ball isn't affected as much by being slightly misaligned?

[pfreivald]

Originally Posted by Alex.q

Are you suggesting that 3pt centering would be better than using 5 points of contact (belts + 4 pieces of aluminum in a similar set up to this)? if we had 2 belts, should we put them closer toegether to keep the contact close to one point or spread them out (2-4") apart so that the ball isn't affected as much by being slightly misaligned?

Not to sound glib, but this is something you'd probably have to experiment with yourself. This is what we went with: http://www.youtube.com/watch?v=ZTQ-dU96c2A