Shooters: What have you learned?

[LeelandS]

Originally Posted by pfreivald

Interesting... If our shop wasn't getting knocked down next week (and rebuilt hopefully by September) we were going to add a bunch of weight flywheel-style to our shooter to see if that increases accuracy!

Also, this is just an anecdotal observation, but single-axle hood shooters seem to be much more reliable than double-axle shooters.

Yeah. We used a pretty heavy custom wheel for our shooter (modeled off our 2006 shooter). I noticed we really need to be careful of the timing we shoot at, or else the balls don't go nearly far enough. Again, we're working on it for later in the season, but it kind of hurt us at FLR. Especially before we got our shooting to be somewhat accurate.

[JesseK]

1.) The amount of contact arc between a single-wheeled shooter and the ball doesn't appear to matter much with respect to shot consistency, though this also makes the assumption that entry speed/position/angle is identical from shot to shot.

2.) Re-calibration of shooter distances is a non-stop thing to do. Expect to test calibration every few matches.

3.) There were two teams in DC that used back-EMF in unpowered motors to detect how fast the shooter was spinning. This seems like a much more elegant solution than an encoder or tachometer/light sensor since the Voltage across the analog sensor port is directly proportional to the SPEED of the shaft the motor is on. Both teams used KOP-legal motors (one was a window motor without its gearbox, the other was a RS-395).

4.) Flywheel vs no Flywheel, that was the question. I'll post back after I talk to a few teams at championships about their shooter controls.

5.) A single 8"x1.5" IFI wheel with roughtop tread was a very good choice for a shooter wheel this year, and it is a choice we will revisit in the future.

[jblay]

Originally Posted by jspatz1

Ummmmm......maybe not so dumb.

Our wheels have scuffs on them in odd places from years of prototyping. They are beyond broken in.

[daniel_dsouza]

Our shooter was designed in such a way that rotation was handled by the drivetrain, and distance by shooter rpm.

If using a CANjaguar, extremely repeatable results can be obtained by measuring the voltage instead of the rpm.

Shooting at a very high angle can be a lifesaver, because you can avoid shooting for the backboard, which makes shoots somewhat unreliable. Not to mention shots that are all net look pretty cool when you make them every time!

Make sure that you tighten bolts/other fasteners on your shooter whenever possible, because the vibration + excessive movement = inaccurate shots.

Recalibration is a must. During our regional, the balls would keep getting squishier and squishier at a rate which we did not expect. So between matches (when our robot wasn't being repaired ), we would head down to the practice field and change constants in our program.

[techtiger1]

Our team made 2 diffrent shooters this year one was a wide belt design using really heavy pulleys and the other was a single axle shooter. Interia will only help you so much, you need the power to bring your wheel up to speed faster. Compression of the ball or object your shooting was a great variable as well but once again there is a limit on that as well. We ended up going with a custom wheel that was 3 seperate pieces and effectly built our fly wheel into the middle of the wheel powered by 3 rs 550's. I think we determined that for us, at least for this game, a single wheel rather then double wheel is much more effective.

[roystur44]

This is what we learned last night.

When you think the shooter is tuned just right. Hold on to camera very tightly.

http://www.youtube.com/watch?v=OiAo1Oiw5X8

[Ether]

Originally Posted by Grim Tuesday

So what less than (or more than) obvious facts have you learned about shooter design this year?

Control the shooter wheel speed with a bang-bang controller rather than PID.

Tip of the hat to Martin for reminding me of this.

[jblay]

Originally Posted by roystur44

This is what we learned last night.

When you think the shooter is tuned just right. Hold on to camera very tightly.

http://www.youtube.com/watch?v=OiAo1Oiw5X8

You sir just made my day.

[BigJ]

Originally Posted by Ether

Control the shooter wheel speed with a bang-bang controller rather than PID.

Tip of the hat to Martin for reminding me of this.

I wish I saw this earlier! I don't follow the LabVIEW forum on the portal because we don't use LabVIEW

[billbo911]

Originally Posted by Ether

Control the shooter wheel speed with a bang-bang controller rather than PID.

Tip of the hat to Martin for reminding me of this.

Nice, quick and easy. I like it!
Mind if I attempt to port it over to LabView?
It can be placed in a periodic task with the interval time of your choice. That way it is one less thing for Teleop to process and it can be used during Autonomous without having to build it in two places.

[Ether]

Originally Posted by billbo911

Mind if I attempt to port it over to LabView?

By all means, please do.

This post explains in a bit more detail why it works for the shooter wheel speed application:

[CalTran]

Originally Posted by Hawiian Cadder

Consistency can be gained through LOTS of compression. We ran our competition robot with 4-5 inches of it in order to ensure precision.

Really? We did about the same thing with our original shooter, and we found DRASTIC variation between shots because of its explosive exit as it uncompressed...

[Jeffy]

Originally Posted by CalTran

Really? We did about the same thing with our original shooter, and we found DRASTIC variation between shots because of its explosive exit as it uncompressed...

Conveniently knowing both systems well allow me to answer.

Our shooter does not shoot the ball with a very high velocity (compared to most). The ball is compressed between a roller and a flat piece of poly carbonate. This allows the ball to be decompressed over a fairly large distance (~5 inches). So, it's not exactly an "explosive" exit like it is from a shooter with dual wheels spinning much faster than ours.

Morale of the story:
Allow the ball to compress and decompress over a distance rather than "instantly".

[CalTran]

Originally Posted by Jeffy

So, it's not exactly an "explosive" exit like it is from a shooter with dual wheels spinning much faster than ours.

Er, yeah? I don't recall comparing the exits from the two shooters...

[Kellen Hill]

In designing a shooter, you must design around the ball you will be shooting and what you will be shooting it into. In 2006, I think the Poof balls were much more consistent in their squishy-ness or it was just a lot easier to shoot at a 30inch diameter near vertical goal. Either way, shooters that year did not require as much work due to a more consistent game piece and/or a large goal to shoot at.

This year however, we are playing with a game piece that changes quite a bit throughout the regional and is not consistent in squishy-ness to start with, along with a much more difficult goal to shoot at (18 inches and horizontal). With these factors, creating a shooter that would make every shot without adjustment took a little more time and effort.

From what I have found, the top three options for having a consistent shot are:

1. Catapult (16 is the best that I've seen: http://www.chiefdelphi.com/forums/sh...d.php?t=103876)
This approach removes ball squishy-ness from the picture. While it's harder to get significant backspin, you can land the ball in the same spot every time if you do it right. The trick is the getting the thing lined up and the right distance away from the hoop.

2. Extended acceleration period (2949 using belts: http://www.chiefdelphi.com/forums/sh...highlight=2949)
With lower compression and extended acceleration, each ball will exit the shooter at a consistent speed regardless of squish. Backspin is tweakable based on your setup. 2949 had a good setup with their shooter on a turret and their laser targeting system.

3. Shoot it hard with a lot of backspin (341 and I believe many others: http://www.chiefdelphi.com/forums/sh...d.php?t=103579)
For this setup, you need a decent amount of compression for this years ball and I think it helps a significant amount to be a tall robot. I will definitely be inspecting 341's shooter at champs to see just how they did it. As long as you can hit the backboard in this case, the backspin should put the ball right into the hoop as long as you are lined up correctly.


As to what we (525) did this season, we had to find another way to do it. We didn't wind up having enough height (we are about 3 ft tall) or enough compression on our shooter to easily do what 341 did. We thought about speeding up the last stage of our feeding mechanism to do more of what 2949 did. And obviously a catapult would have been quite a drastic change from what we had to work with. Our shooter used a 6inch Colson with one FP 0801-0673 with just short of 2 inches of compression on the ball and an adjustable hood. We decided we had to adjust our shooter speed/hood angle on a per ball basis in order to have a consistent shot. To do this we added a compression load cell (http://search.digikey.com/us/en/prod...6948-ND/809394) to measure each balls squish factor. You can find more information about our setup for the load cell here: http://www.chiefdelphi.com/forums/sh...d.php?t=105394. This setup got us to the point where we hit 9/10 balls in tele-op for one of our finals matches at 10,000 Lakes, hopefully something we can carry on to champs.

I'm sure there are other options for having a consistent shot, bot these are the three designs that have stuck out to me.

I hope this is helpful and I hope I remember all of this for the next game involving shooting a ball.

Kellen Hill