FRC 2012 Ball Launcher by iR3

Check out our 2012 ball launcher / shooter: http://youtu.be/DO0XzkthMUs?hd=1

The purpose of this particular launcher was to try different configurations and parameters such as wheel rpm, wheel spacing and wheel width. We also wanted to be able to adjust these configurations precisely and quickly. Over the next few days we will be trying various wheel configurations and materials to improve launch distance and consistency.

We believe this exact design could serve as a final shooter with the correct control parameters and motor selection but there is obviously a fair amount of structural / mechanical optimization that could take place.

/edit With Tech Specs

Tech Specs:

Wheels: 6" FRC 2008 KOP wheels
Motors: 2x CIM
Gear Used: 14 tooth to 40 tooth 1 stage
Center To Center Wheel Spacing: 12" ( creates 6" opening )
Distance between wheels: 3.625" insides

Cool. Jealous.
Now to compact that design and strap it to a turret.

That looks nice. I assume the point of being able to adjust it so much is to change how far the ball goes, right?

Do you know why the ball didn’t shoot through immediately at :50?

Good job. What is the furthest you guys have launched the ball and at what angle did you shoot it from?

Cool, I’d never thought about the modulox system as a prototyping system. Very clever!

How much is it effected by losses in momentum as you feed balls in one after another, and have you tested it with worn balls yet?

How consistent is your range/accuracy? How do the different factors you tested affect consistency?

I suspect that it wasn’t pushed quite towards the gap, but rather instead against one wheel, which threw it against the other and bounced it back.

I have both a question and an opinion as a response to the question. My question is, do you think it is necessary to use two of the precious four CIM motors to run the launcher? They do have the best balance of speed and torque, but are there any better alternatives out there? I was considering using the AndyMark motors - similar to the size of the FisherPrice motors - because they have very high rpm, but I am not sure about the torque in order to put the ball through with the same velocity as the CIMs would. My team uses mecanum drive, so we are going to use the CIMs for that. I’m just curious if a two wheel drive system will be popular this year because of the necessity for the CIMs on the launcher.

EDIT: I just looked at some motor comparison charts and found that the FisherPrice motors have both higher rpm and torque. I redact my previous statement about the AndyMark motors.

No need to be jealous! We were lucky in that we had the parts around but the only unique parts that you need is a Modulox Channel and a handful of Modulox not in the box parts (we used a lunchbox). Also, this is meant to be a community resource! The step file of the assembly is on our website www.Modulox.com. If you have any specific questions as we post our testing videos feel free to ask.

Oh and turrets? ::cringes::

I assume the point of being able to adjust it so much is to change how far the ball goes, right?

Do you know why the ball didn’t shoot through immediately at :50?

Well, in our experience, when trying to move squishy conformable spherical objects the amount of compression is an important factor. To little, the ball slips through, too much and it may not go through at all. So the wheel spacing plays a major factor, especially when it comes to distance as you mentioned. The last bit of adjustability allows us to change our gear ratios and wheel speed which can have an even bigger effect on launch distance. We didn’t firmly mount it to the ground so we could play with angles and trajectories.

In competition, many teams don’t have the resources to constantly change out for fresh material, simply adjusting the wheel spacing will compensate for wear. So even tho this is a prototype, we’ll probably leave that part in on the final design ( if any of our teams shoot that is :slight_smile: )

Oh, and at the 50 second mark, slijin is correct, he fumbled the ball and it kicked back off wheel which brings me to my next topic, consistency:

How consistent is your range/accuracy? How do the different factors you tested affect consistency?

For the most part, the factors we played with seemed to only effect distance, we’ll let you know more as we go one.

A major observation from our experimenting is the importance of consistency in ball feed to the shooter. We noticed if the ball was fed slightly to one side or the other the angle and distance could change considerably. Teams will need to ensure that the ball is consistently guided into the shooter if they want to be accurate and precise!

What is the furthest you guys have launched the ball and at what angle did you shoot it from?

Somewhere around 27 feet in that last shot and consistently around that distance. We didn’t measure the angle on that shot but it looks to be around 45 degrees.

Our next test will be with at a higher RPM as our gear ratio is currently 2.85:1 (14tooth cim, 40tooth .5"hex) we just didn’t have a bigger cim gear on hand.

How much is it effected by losses in momentum as you feed balls in one after another, and have you tested it with worn balls yet?

In this particular configuration the wheels accelerated to speed very quickly but there certainly is a perceivable momentum loss. I would say it’s not negligible especially as we start to raise the rpm but don’t think it will play a big factor in most configurations.

My question is, do you think it is necessary to use two of the precious four CIM motors to run the launcher?

They are certainly an easy choice but I don’t think they are necessary. We probably would have started with a FP if we had one with a AM cim planentary. The 775 motor series or 500 series motors may suffice as well, but the smaller the motor the longer its going to take to get your wheels up to speed, if they can get there at all.

If we can get access to a few of the smaller motors maybe we’ll try it and post a new video.

@ Dan this reminds me alot of 1251’s 2006 prototype. The modulox seems to be the perfect way to test out a shooter this year, very impressive. @ Ninja The wheels will lose momentum when a ball is shot so you either have to have a controlled feed to wait till the wheels ramp back up or have a diffrent shooting system. Dan, why are you so scared of turrents, they worked fine for 1902 in 2009 hehe.

I thought that the balls would be to soft to be used in a machine like that. But that is REALLY nice.

Store Credit?

thats a cool idea my team mentor and builder like it, funny thing is im not a builder im a programer :smiley:

how do i start my own thread? can anyone help me on that?:ahh:

Yes very similar design indeed!

:: The opinions stated in this post are that of the author and not of his company, management or mentored teams ::

I believe at the upper levels of competition turrets will be very successful. However, there are a few complex systems that have to work perfectly to make a turret design successful i.e. drive train, bridge balance, intake, delivery and shooter. Once a design and prototype is developed and those systems are perfected then and only then would I personally start on a turret… which would probably be in 2013.

But hey, that’s just like… My opinion man…

Store Credit?

:ahh:

After making seeing that shooter the dimensions of the robot are starting to look a lot smaller! Hopefully it can be successfully compacted.

Absolutely, I agree! There is plenty of room for optimization in this prototype using these components. I think a great place to start is the packaging, smaller wheels may help the overall size as well.

Though this setup could theoretically serve as a final design, the point of the shooter was to help answer the practical questions while designing the final solution. Questions like what trajectories work best for long rage shots at the top goal. We tested that in a recent video seen here: http://www.youtube.com/watch?v=l8U9u2iqMEs&sns=em

On the other hand, our final design will definitely include room for modularity and adjustability so that we can continue to iterate and improve on the design.

This is very similar to the shooter we got up and running last night. We designed for all the same adjustability, but not having the Modulox components to make that trivial it took several hours. However it works similarly well. We are unable at the moment to get the speed up to what we would like due to gearing constraints (again where Modulox would be awesome), but are VERY pleased with consistency and should be able to get the speed we need just by gearing down less (the torque seemed more than sufficient). No video yet, but soon.

nice job! good music, too.

would you be willing to do some controlled tests to compare it against my trajectory spreadsheet and post the results?

Mostly, I’d like to see you use your fixture to estimate launch velocity based on range. Mount that sucker to a desk to make it the right height and lock down the angle on it. Feed the balls as nicely as possible and check the distance traveled against numbers from my sheet. It will go a long way to helping us figure out what the drag effect will look like.

I like your idea! We did some trajectory testing in the second video I posted but nothing to the extent you’re talking. I think the hard thing will be calculating launch velocity to compare it with theoretical. If I’m thinking correctly wouldnt you need both for drag?

We could calculate wheel speed rather simply but it may be difficult to determine the balls acceleration through the shooter. Right now my only thought is high speed camera or some sort of photo gate.

Am I over complicating this? Any thoughts?