Ball Magnet

[MrForbes]

We tried our idea for a "ball magnet" basically a way to POSSESS the ball.

http://www.youtube.com/watch?v=KRqj0Mgbq10

Have fun!

[Phoenix Spud]

How does the robot do that??? According to the Youtube post, it has something to do with the wood, but it sounds like there is a vacum behind it.

[Pausert]

I think its that the conveyor belt is running, which pushes the ball into the block of wood.

[Ian Curtis]

How does the robot do that??? According to the Youtube post, it has something to do with the wood, but it sounds like there is a vacum behind it.

It appears to me that something that is not spinning when they do the close up starts spinning when they pull the camera away. :D

Nice Work!

[Ice Berg]

That's quite impressive. I'm assuming you don't break the 3" rule from the front of the roller to the piece of wood. I ask because it looks pretty deep in the video (not that I can tell much from that angle).

[MrForbes]

This is set up to demonstrate the principle, and use the old robot with no modifications other than screwing in a piece of wood. When implementing this on the new robot, we will take all the rules into account.

[Dave McLaughlin]

I guess that Robocup robots come standard with "Ball Magnets" too...

[MrForbes]

We'll get our ideas wherever we can ;)

[Joe Schornak]

This is close to something we are considering implementing.

Your robot still looks intimidating.

[fuzzy1718]

how exactly does this not break the rules, you are controling the direction and position of the bal???

[Dave McLaughlin]

You are POSSESSING the ball, which you are allowed to do. <G43>.

[James Tonthat]

how exactly does this not break the rules, you are controling the direction and position of the bal???

This is how.

POSSESSION: Controlling the position and movement of a BALL. A BALL shall be considered in POSSESSION if, as the ROBOT moves or changes orientation (e.g. backs up or spins in place), the BALL remains in approximately the same position relative to the ROBOT.

<G43> ROBOT BALL POSSESSION - ROBOTS may POSSESS only one BALL at a time. Violation: PENALTY.

Please read your rules.

[Jones571]

Very nice we should have some video up sometime this weekend of our endeavors...

[Elgin Clock]

It's ridiculous how many people I have heard point out the fact that some new design POSESSES the ball. Thank you for pointing out something that is legal people! lol :rolleyes:

[falconmaster]

We tried our idea for a "ball magnet" basically a way to POSSESS the ball.

http://www.youtube.com/watch?v=KRqj0Mgbq10

Have fun!

is it a roller spinning backwards
when we did ours we spun the ball into the robot no matter what direction we traveled

[NOV8R]

Here's what were doing.....http://www.youtube.com/watch?v=K3Nb9NAxdeE.
We are building a full scale prototype with Banebot 20:1 26mm gearmotor and their 3 7/8 30A green wheels. Our students have played with this concept using one 20:1 Banebot gearmotor and a soccer ball. That playing was promissing enough for us to order the parts for a full size prototype that we are going to mount on our prototype holonomic drive.

[MrForbes]

Yes Fredi, it's just the intake roller trying to suck in the ball, but there's a piece of wood blocking the ball so it can't go all the way in. Just a quick and dirty test of the roller idea to POSSESS the ball. We need to play with it more to optimize it, and fit it into the new robot chassis, but it works well enough as is that we are confident we can use the concept.

see, we did get one simple prototype done

[Nawaid Ladak]

Nice job

I actually bought that same exact soccer ball yesterday. The Wilson HEX size 5 soccer ball

Im not sure, but wouldn't it be a good idea to try experimenting on carpet instead of a hard floor

[=Martin=Taylor=]

Cool. We did the same thing with our 2009 robot.

We got most of our inspiration from here => http://www.techunited.nl/index.php?p=11

If you mount some slides on either side of the ball you can even turn with it.

Running it on carpet changes things. Its harder to get the ball to spin on carpet, and if it stops spinning you lose control. Using small wheels instead of a board will allow it to spin more easily.

[NOV8R]

Here's another video showing the concept in action in Robocup juniors....http://www.youtube.com/watch?v=sJAue5eAOkQ

[GaryVoshol]

Too bad U6 soccer players aren't legal mechanisms to use. They play "magnet-ball", so there must be player/ball magnetism.

[lynca]

We got most of our inspiration from here => http://www.techunited.nl/index.php?p=11

We have also been looking closely at RoboCup robots.

The TechUnited Design has an awesome way to switch between a flat shot and a lob shot.

http://www.techunited.nl/index.php?p=13&film=Qualification_Movie_2009

[Jones571]

What motors have people been using for there magnets in testing? We had sucess with the 550's

[Chris is me]

What motors have people been using for there magnets in testing? We had sucess with the 550's

Seeing as a ball magnet is "just" an intake roller, pretty much any motor in the kit will do. The RS-555 is a motor you're not likely to need anywhere else, and I'd be hesitant to waste a CIM or FP unless you don't need it somewhere else.

[MrForbes]

The one in our video is a Fisher Price, driving the gear from the first stage of the plastic transmission, then a further chain/sprocket reduction (I think it's about 1:3) to the roller.

[=Martin=Taylor=]

What motors have people been using for there magnets in testing? We had sucess with the 550's

I believe the 555 motor this year is much weaker than in years past. The new motor has around 40 watts while the old ones were well over 100.

We were thinking of using the FP motor for our wheels, geared to give a surface velocity on the ball of 17 fps.

[M80]

Number 1 will the ball exceed the distance allowed into the robot? and if it doesn't will the magnet protrude to far from the robot?
Number 2 Does it count as being in contact with more than one ball if a ball is kicked into you while your carrying the ball?

[NOV8R]

Answer #1......The ball is allowed to penetrate 3 inches into the robot. The design of the mechanism can manage this. The wheels you would typically use are only about 4 inches in diameter. That leaves room to mount the mechanism inside of the robot but yet have the mechanism wheels touch the ball.

Answer #2.......A ball that bounces off a ball in possession is not in possession since the robot is not controlling its orientation of the second ball. The rule says you can't be in possession of more than one ball.

[Jeffy]

I don't mean to be asking you to post something that you don't want to reveal yet, but this has been the only application of this concept I have seen on an FRC bot.
Our team was inspired by some of the robocup bots we saw that used a similar mechanism to control the ball. But what we are having trouble with is figureing out how this works.
WE have a setup with a 1.5" tube spining at about 400 rpm, and it just seems to spin the ball a little, and then it kinda bounces off the roller and away.
What is the concept of how this works? (if you don't mind) or if someone else can fill me in.

[Jones571]

I don't mean to be asking you to post something that you don't want to reveal yet, but this has been the only application of this concept I have seen on an FRC bot.
Our team was inspired by some of the robocup bots we saw that used a similar mechanism to control the ball. But what we are having trouble with is figureing out how this works.
WE have a setup with a 1.5" tube spining at about 400 rpm, and it just seems to spin the ball a little, and then it kinda bounces off the roller and away.
What is the concept of how this works? (if you don't mind) or if someone else can fill me in.

one thing to make sure you play with is different contact points between the ball and roller as well as different "sticky" surfaces

[MrForbes]

What is the concept of how this works? (if you don't mind) or if someone else can fill me in.

The roller is mostly above the ball, it spins in the direction that drives the ball "into" the robot. We don't yet have the spec ball nor spec carpet, so we dont' know how well it will work...but the idea is to have the ball actually spinning the whole time, and as you drive the robot around the ball is always spinning on the carpet towards the bot.

As for the dimensions on our prototype...if the roller were at the very front of the FRAME PERIMETER then the ball would ingress just over three inches. We need to experiment with roller size and location to see what works.

I did notice that the ball was marked up from sitting still with the roller spinning on it, so there is more work to be done to see if this is really feasible.

Also the roller should be made only wide enough that one ball can contact it. If two balls can both be POSSESSED by the robot at once, bad penalty things happen.

[Akash Rastogi]

Wow how did I not see this thread before. We're also prototyping something similar (using our 09 and 08 rollers) but I'm adding tiny tiny ball casters

[Nurnburger]

I don't mean to be asking you to post something that you don't want to reveal yet, but this has been the only application of this concept I have seen on an FRC bot.
Our team was inspired by some of the robocup bots we saw that used a similar mechanism to control the ball. But what we are having trouble with is figureing out how this works.
WE have a setup with a 1.5" tube spining at about 400 rpm, and it just seems to spin the ball a little, and then it kinda bounces off the roller and away.
What is the concept of how this works? (if you don't mind) or if someone else can fill me in.

It's very important that you have a surface with enough friction that it can actually apply force to the ball. Otherwise the roller/wheels don't do anything. And as Jones571 said, the contact point is also important.

[MrForbes]

We've been working on this some more, tonight's video:

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

(wow, the first video has over 7500 views)

[dtengineering]

We stumbled upon the weirdly amazing effectiveness of backspinning the ball entirely by accident at our second team meeting. We'd just finished building a goal and some bumps and were discussing how to kick the ball. We've had good success with spinning "shooters" in the past, so the idea was to try a roller. (Zeki, one of our grade 12 students who has been on the team for four years is coming up with all sorts of really cool ideas this year. One, for a non-functional decoration... we're hoping to be able to fit into the weight and size limits.)

We took our old "Aim High" bot, laid it on it's back, spun up the IFI traction wheel and touched the ball to it. The wheel was spinning the wrong way to "kick" the ball... and the ball started to backspin and was held in place.

It didn't work particularly well, but I think its the first time we've ever figured out an essential component of the game that early. I think 1726 must have posted the ball magnet video about the same time, because we saw their video very shortly afterwards... and account for a few of those thousands of views.

So, of course, we figured if backspin is good... more is better. So we rigged up a test chassis using 6" AM wheels spinning at a couple thousand RPM and disovered that we could actually melt the outer lining of a soccer ball at those speeds. So we tried it with last year's intake mechanism, as 1726 had, and it worked much better... but we lost a couple days of build doggedly researching the "more is better" idea, and wondering why it didn't work.

Now we've got a 2" diameter aluminum roller directly driven by a 25:1 banebots running off last year's FP motor (about 600 rpm-ish) as our "final mock-up" and it's running okay. We've got a couple 26:1 P60 gearboxes from banebots on order for use with this year's FP's for competition. We used "magic tape" the black plastic wrap that sticks beautifully to itself, as a traction surface, but will probably go back to the rough top tread when it arrives, riveting it to the roller in a helical pattern in the hopes that it might help control lateral ball movement. We could probaly do with a little bit more speed, but can order a "faster" P60 and change it out in Seattle if we need to.

I can't believe how well this seems to work... mind you, EVERYTHING always seems to work well on the practice field.

Jason

P.S. We are trying this with an HS500 DTI ball, which apparently has a very similar finish to the HS300 ball on a carpet that, while not official spec, seems to be really close to spec. The HS500's shiny surface works, if anything, better, on this surface than the matte (and now partly melted in places) finish on the old soccer ball we borrowed from the PE department. They'll get a nice shiny HS500 in return after the season ends.

[sanddrag]

We used "magic tape" the black plastic wrap that sticks beautifully to itself, as a traction surface. Do you have any more details on this "magic tape"? It sounds interesting. Have a link?

[R.C.]

Do you have any more details on this "magic tape"? It sounds interesting. Have a link?

Is it also known as "grip tape" that is found on baseball bats?

-RC

[vamfun]

We've been working on this some more, tonight's video:

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



Looks like you are getting closer. We have found mixed results.. its very sensitive to the friction on both the rug and soccer ball. Some balls stick like crazy to both the rug and roller and stall the roller. Others hardly stick. We have fooled with different roller surfaces and geometries but I think our rug must be too sticky compared to what I see in your video. So we have a design that will prevent roller stalling and is tolerant of friction variations...maybe.

In your video you mentioned shelf paper. How would you legally bind that to a roller?

[lenny8]

We've been working on this some more, tonight's video:

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

(wow, the first video has over 7500 views)

*gasp* that looks wildly familiar :yikes:

[hipsterjr]

Members on my team are talking about using a brush roller instead of the solid type shown here, but I doubt that it will put the required back spin on the ball. Has anyone tried brushs yet?

[Ryan Dognaux]

Very interesting idea, thanks for sharing the video with everyone.

A few quick questions -

1) What materials on the rollers have people found works best when using this kind of setup?

2) How high from the ground do you have your roller mounted? During your trials, have you found a roller position that works best?

[MrForbes]

In your video you mentioned shelf paper. How would you legally bind that to a roller?

We used clear packaging tape to attach the inner end, and a bit on the outer corners after wrapping the shelf liner around the roller a bit more than one turn.

I have not found any rules prohibiting the use of packaging tape. If you know of one, please let us all know! (rule number)

Has anyone tried brushs yet?

We sort of tried a brush, it was flaps of the shelf liner, like we used on last year's robot to lift the balls up the helix. It did not work at all. But we didn't use "real" brushes. I think it needs more friction than you can get with brushes.



1) What materials on the rollers have people found works best when using this kind of setup?

2) How high from the ground do you have your roller mounted? During your trials, have you found a roller position that works best?

1- We tried wheel tread and it works well to grip the ball, but it seems to need to have a full surface, not individual strips...the ball falls off when it encounters a significant gap. Shelf liner seems to work well, but we are concerned about it's life. It might be just fine, and it is cheap and can be replaced easily if it's secured with tape like we did on the latest video.

2- seems to need the roller pretty far up the ball to get enough traction on the ball to be able to spin it.

[Chris is me]

Your first video convinced me ball posession would be easy, so I stopped worrying about it...

*grumble grumble grumble*

What motor / RPM are you driving the motor at? What speed variations have you tried? You might see something very similar from New York...

[R.C.]

Your first video convinced me ball posession would be easy, so I stopped worrying about it...

*grumble grumble grumble*

What motor / RPM are you driving the motor at? What speed variations have you tried? You might see something very similar from New York...

We were messing around with our roller yesterday and we using a drill that outputs about 1000-1500 rpms. We found out that the faster the roller the more backspin it puts on the ball. Therefore causing the ball to come back to the roller causing it to "stick"

We will be using 2 mabuchi's to power our roller.

-RC

[JHSmentor]

We were messing around with our roller yesterday and we using a drill that outputs about 1000-1500 rpms. We found out that the faster the roller the more backspin it puts on the ball. Therefore causing the ball to come back to the roller causing it to "stick"

We will be using 2 mabuchi's to power our roller.

-RC

Do you have a mock up or pic on how you are going to be using the mabuchi's? like a gearbox, etc.?

we are going to be working on this design type this weekend but we don't have a ton of resources and zero "extra" parts (no stock pile from previous years like veteran teams tend to have).

[R.C.]

Do you have a mock up or pic on how you are going to be using the mabuchi's? like a gearbox, etc.?

we are going to be working on this design type this weekend but we don't have a ton of resources and zero "extra" parts (no stock pile from previous years like veteran teams tend to have).

Bryan,

In the offseason of 2009 we had our banebot gearboxes fail on us, not the new p60 ones. So we ripped the first two stages of the plastic kop gearboxes and made some new gearboxes.

http://www.chiefdelphi.com/media/photos/34184

For this year we are taking two mabuchi's and gearing it to a 3-4:1. To work on your design, I would recommend getting some sort of roller setup and powering it with a drill.

-RC

[NOV8R]

1583 is also using a ball magnet (or as we call it 'dribbler'). Our system consist of two Banebot 20:1 gearboxes powering two of their 3 7/8 inch wheels. We currently are using 385 motors on our prototype but will use the Fisher Price motors on our competition robot. The wheels make contact on the ball about four inches apart and about 2/3 of the way up the ball. We have a 'proof of concept' device up and operating. The device back spins the ball fairly fast and because of this it really 'sticks' the ball to the robot. We are using a holonomic octagon shaped chassis and are confident the ball will turn with the robot. We took some video today of the device 'sticking' the ball and then shooting. I'll load the video on youtube and provide a link.

[JHSmentor]

Bryan,

In the offseason of 2009 we had our banebot gearboxes fail on us, not the new p60 ones. So we ripped the first two stages of the plastic kop gearboxes and made some new gearboxes.

http://www.chiefdelphi.com/media/photos/34184

For this year we are taking two mabuchi's and gearing it to a 3-4:1. To work on your design, I would recommend getting some sort of roller setup and powering it with a drill.

-RC

thanks RC - I appreciate the feedback. I'll use your advice and use a drill to do the mock up - but we will need to get creative I think with mounting something for the competition bot. should be fun!

[NOV8R]

Here's the video I promissed of 1583's chassis/magnet/shooter protoypes in action http://www.youtube.com/watch?v=VC52tUnVAzk

[johnr]

It seems that the roller method is going to be used by alot of teams. Will refs assume that since you have this roller device you are possessing the ball. What if your roller is two or three balls wide and you turn it off to herd two balls? Will a ref be able to tell the difference when you are moving forward? What if your roller isn't that good? If you back-up and the ball rolls toward you are you in some form of possesion? Maybe this method isn't possesion but herding 2.0.

[Chris is me]

What if your roller is two or three balls wide and you turn it off to herd two balls? Will a ref be able to tell the difference when you are moving forward?

Refs can only rule multiple POSESSION if the balls stick with the robot as it turns or moves backward.

If you back-up and the ball rolls toward you are you in some form of possesion?

The rules define possesion as the ball staying in contact with the robot as it turns or moves backward. So by definition, the balls are POSESSED.

[ttldomination]

Here's the video I promissed of 1583's chassis/magnet/shooter protoypes in action http://www.youtube.com/watch?v=VC52tUnVAzk

That's a very good mechanism. May I inquire as to what you are using to $@#$@#$@#$@# back the mechanism and launch it?

[NOV8R]

ttldomination....when the video was made we were using the KOP window motor/gearbox as the power source to $@#$@#$@#$@# the surgical tubing. In 2008 the team next to us at the Denver regional was 1625 Winnovation. We are using the same concept they used on their 2008 robot to launch the ball over the overpass. It's a really simple and effective design.

The only limitation to our prototype mechanism was range. As you can see from the video it would only launch the ball about 8 feet. We've since designed an improved cocker/trigger that uses a CIM hooked to an AM toughbox. Since the toughbox has straight cut gears we also had to design a device to keep the stretched tubing from back driving the gearbox. The improved devices are now installed and our students will test the mechaism Monday. The ump of the improved cocking device should enable us to launch a ball full court. One of our design goals is to shoot all three balls from the far zone in the 15 second autonomous period. I think we have the right stuff to do that now.

Let me know if you need more information. We could take and post a picture of the cocking/trigger mechanism if that would help. You can't see any of that in the video.

[MrForbes]

What if your roller is two or three balls wide

We're going to make ours just less than two balls wide, so we don't have to worry about it.

[ttldomination]

ttldomination....when the video was made we were using the KOP window motor/gearbox as the power source to $@#$@#$@#$@# the surgical tubing. In 2008 the team next to us at the Denver regional was 1625 Winnovation. We are using the same concept they used on their 2008 robot to launch the ball over the overpass. It's a really simple and effective design.

The only limitation to our prototype mechanism was range. As you can see from the video it would only launch the ball about 8 feet. We've since designed an improved cocker/trigger that uses a CIM hooked to an AM toughbox. Since the toughbox has straight cut gears we also had to design a device to keep the stretched tubing from back driving the gearbox. The improved devices are now installed and our students will test the mechaism Monday. The ump of the improved cocking device should enable us to launch a ball full court. One of our design goals is to shoot all three balls from the far zone in the 15 second autonomous period. I think we have the right stuff to do that now.

Let me know if you need more information. We could take and post a picture of the cocking/trigger mechanism if that would help. You can't see any of that in the video.


I would love to see the mechanism you guys are using.

We're having trouble with our release and we'd love to see a mechanism that actually works.

Thanks.