Based on prototypes we’ve seen many people are going for the bottom and side shooters with either 2 or 4 wheels. Not to state the obvious but I just wanted to make the topic clear.
I just want to get everyone’s opinion on these different Designs. If there are other similar designs to the side and top rollers bring them up in this thread too.
What are the clear advantages of having a bottom vs. side roller, and vise-versa?
In our team we’ve been discussing the fact that a side shooter will have less backspin and thus may be less forgiving when either hitting the backboard or basket. We’ve thought about the possibility of having a third wheel on either the top or bottom that was idle, with no power, and would just be the provider of spin on the ball as it goes through the shooter.
On the other hand, a bottom shooter has only one side of the shooter itself being powered, the bottom wheel, and thus instead of having wheels on each side of the ball to propel the ball either a curved piece of metal or lexan or even just a bar of metal must be on top in order to provide friction for the ball to exit the shooter. Thus, the ball is not propelled as far.
There are clear advantages to both options but in terms of accuracy it seems that that depends on the angle of the shooter itself, not the style of it.
What are everyone’s thoughts? Regarding not just both of these types of shooters, but all methods of getting the ball into the basket?
A bottom shooter can have another set of wheels on top and get the same distance. It has the added advantage of not curving your shots like a side-shooter will. After all, it’s pretty hard to get two motors to spin the exact same speed. A top/bottom shooter will simply result in some top spin (or better) some back spin.
Of course, a top/bottom shooter will be taller.
Other than that, there isn’t really a whole lot of difference.
We are working on a single wheel shooter to give the ball backspin, so it will bounce downward and go in the basket when it hits the backboard. We’ll aim for the backboard. We also are hoping for a relatively flat trajectory, since in our experience it’s not too hard to aim side-side, but it is pretty difficult to accurately judge distance, so we’re trying to minimize the effects of how far the robot is from the goal.
We haven’t got there yet…this is all still theory! but it’s based on watching 2006 robots play Aim High
Well… I think it has more to do with the context of the shooting. Firstly, spinning wheels to shoot a ball are always unpredictable based on:
Spinning wheels, as that on a pitching machine, rely on deformation of a ball in order to get the desired friction to shoot the ball. (even baseball pitching machines deform the ball!) And this deformation is different from ball to ball, let alone balls that are quishy in general!
Spinning wheels need to speed up to the desired RPM before a ball can even be placed into it, and the balls then put into it slow it down enough to where you again have to wait to put the next ball in
The efficiency of a wheel powered shooter is probably very low considering the power is distributed by friction (which also means more compression of the ball would be needed to make the ball go further thus affecting points 1 and 2)
Every single ball shooter that operates by wheels has either been right up to the goal hugging the side of the fender, where every robot will try to shoot from as well if they go with a shooter and even there the balls in youtube videos has scuffs on it which means its only a matter of time until something gives.
When shooting a fastball from a pitching machine, there’s a reason why the wheels are different speeds for the slope of the ball. When shooting a crazy basketball that weighs more than twice as much, the slope will be incredibly hard to tune to the wheels speed and compression of the ball and even THEN the distance will play a HUGE role in the speed of the wheels as well which means the wheels would have to be tuned while moving which requires a lot of time, just like a pitching machine warming up
Now for a horizontal v. vertical shooter:
Horizontal shooters basically are side spin magnets meaning if you even have the motors a little off the somehow magically set speeds, there will be a curve left and right. That curve isn’t and will never be enough to get around objects like a boomarang either since the curve on a ball is actually made by a low pressure system on one side.
You also cannot shoot a ball with a slope since the horizontal motors means there is no y component. Then shooting at a basket becomes dodgeball, and the ball’s compression would make a nice bounce right back, and never make it from more than 5 feet away because the speed would need to be too high to make it the right distance in a straight shot.
Vertical are basically the same, however, they, being vertical, means they have little affect on left and right movement. That’s good, but then like in the first 5 points on wheel shooters, there is so much unpredictability, that it’s better to go a more direct mechanism to ball shooter such as a catapult. However, the effect of a tension/spring method weakens with every shot which means its impossible to tune as well without the fear of fatigue affecting shot.
We went with a rotational shooter. Direct, and doesn’t have fatigue.
With our teams prototypes we have found that if you power both the top and bottom and supply 30% power less to the top then the ball gets the perfect backspin. Though these change with different designs and such. This will power enough for distance and backspin though so its what we are going for.
I think farmervilleRob’s prediction of 5 feet, fender only capability for spinning wheel based shooters is a gross underestimate of what the concept is capable of. When looking just at 2009 examples one could get a false impression of what is possible because we were aiming at moving targets. Close shots where a practicality of the game. In 2006 we saw much longer shots hit with consistancy. It is interesting though to consider how chewed up these balls will be allowed to get before they are replaced between matches.
Okay, so the reason we chose a bottom rotational wheels and then just a bar over top was for back spin. Adjusting speed and bar location give more spin vs more distance. Back spin is definately wat u want.
Now some reasoning behind why I think u shouldnt have a wheel unpowered on top, wheels have much more friction over a bar. The bar is used for back spin, thats why it works. The wheels that r underneath it are use for propulsion.
Hmm… an interesting thread. Buried in here we’ve got a rookie clearly and confidently stating what can be done, and a veteran making a conservative estimate based upon observation of the 2006 game play, and carefully limiting his comments to the “subject to further testing” clause.
I’ve posted a link to this video of our 2006 robot before, but I’ll do it here again just to show that a single-wheel shooter is more than capable of hitting from half-court.
But I’ll also point out that just because you can hit the target from half-court in practice, doesn’t mean that you should count on being able to do it during the game. We were resoundingly whupped by a a few teams who made it their goal to get close fast and score with every ball.
Jason
P.S. I should also point out that properly designed shooters will not damage the balls. In fact, I’d expect that to be part of the tech inspection process. Although a week or two into the '06 season I didn’t believe it was possible to build a shooter that wouldn’t damage the ball, it turns out I was wrong about that… our 2006 poof balls have been fired hundreds of times and are still in pretty good shape.
But how heavy were the nerf balls? This year’s are 11.2 ounces. Thats very heavy and will make a HUGE difference in the torque to rpm ratio gearbox. Especially when watching these threads go through and the prototype shooters are barely making the 8 foot shot. Which is also a big change in distance considering that gravity cannot play a part in the full slope like it could in 2006. As well, there is a compression difference pointed out in my previous point that also is a huge affect on the needed torque to compress the ball while taking it in. 2006 was a rpm game. This is a torque game
So it seems there is a huge difference by only 50% weight difference. I’m not against trying a new shooter, but based on these, it’s hard to gear down to the desired torque and keep rpms. I suggest instead of looking to the past for solutions in the 2006 and 2009 games, look into a new design for the new challenge. Who wants to see some 2006 and 2009 remakes? I’de rather watch new prototypes of flying cars fail than watch dependable wagons slowing go in circles.
I’ll echo the previous comments in that the main difference between side vs. top shooters is the potential spin on the ball, and the height of the shooter. We are leaning towards a top shooter because of the ease in imparting backspin.
As for skepticism on the design working at all, mine is gone after testing a prototype of ours today. We built a vertical shooter (2 wheels above, 2 below, all 8") with 2 CIMs through CIMple Box transmissions direct driving the wheels and placed a basketball hoop such that it was the equivalent of firing at the top hoop from the key. I was concerned that there would be too much variation in shot placement with all speed/angle variables being the same, and in our initial tests where we fed the ball in by hand the results weren’t too pleasing, with some shots flying over the backboard while others fell short. We saw ball feeding as a big issue and added a guide chute behind the shooter to consistently feed the balls in the same spot, then fed them in by pushing with a PVC pipe during our second test. This made an enormous difference as we tested about 30 shots and made 50% of them into the basket. All the missed shots landed on either the back or front of the rim as well so even for the shots that missed there was relatively little variation in ball placement. Considering that we were shooting at less than a 45 degree angle and had no spin on the shots I’m sold on the feasibility of a shooter working this year.
Do I read this as I should use my cims for a shooter, or can multiple FPs or RS-775-18s provide the necessary torque? We don’t yet have a finished prototype of this system, but our goal is 20 ft (hopefully). We do not however, wish to lose our cims from the drivetrain unless absolutely necessary. And in regards to copying 2006/09 I don’t mind copying it if* it works.
Well, it’s very difficult to do the math on motors without the stress, without the launch parameters (angle, number of motors and wheels), and rpm relative to a urethane foam basketball lol but I would say definitely more torque than any 2006 or 2009 bot. And based on prototypes, I would say test both. The 775 Banebot probably when geared down to the appropriate torque doesn’t display enough rpms. At 20 feet and 45 degree launch angle you need ~30.008 feet/sec of muzzle velocity. Based on our excel spreadsheet of glorious math!
ps. That takes into account for distance and hoop height from launch (in our spreadsheet its set to 6 feet) and ball weight
It’s pretty easy to get close numbers…find the no load rpm of the motor, find the wheel diameter, decide if you want a wheel on each side of the ball or just one side, and figure out the gear ratio needed to make the wheel(s) spin that fast, or just a bit faster.
If the motor speed drops too much when shooting the ball, add another motor, or add mass to the wheel and wait longer for it to spin up
But isn’t it different with no load rpm? Is no load rpm affected by the load of the wheel? Even if the wheel is 90 degrees vertical, it should still give fight back to the motor thus lowering the no load RPM. Also, the torque would have to pick up the weight of the ball and at that moment in time, the torque is also having to speed the ball up “through” the launch plane and that reduction also slows down the RPM.
If calculating proximities though, like squirrel is suggesting, I would like to add that taking the max efficiency RPM and torque would leave wiggle room for all the complicated math if you don’t want to run a long algorithm for precise numbers even if preciseness is probably the best if you don’t have the financial or physical means of prototyping.
I know it will change based on design, but how will having wheels on the bottom vs wheels on the top and bottom affect the range/accuracy? Will two sets of wheels provide more power/range as one would believe, or would power be entirely dependent on the motors themselves?
Sorry if this is hard to answer based on little actual design.
Well to my knowledge, it would give more range. Think of it this way: If you have a pitching machine on, to throw a fastball with the appropriate slope through the strike zone, the top wheel is set at 40 mph and the bottom wheel at 65 mph. That gives you a 60 mph fastball and that is because of the bottom wheel still having to put spin on the ball otherwise you have a frozen rope (linedrive) at the same level it came out as.
It would, however, make the top motor run at about 30% efficiency compared to the bottom one because it is programmed to run slower. You could throw a frozen rope at the goal, but if that worked, then basketball players would do it… and they don’t.
So, you need the curve. The bottom only motor gives you the curve, but with a plate on the top using friction to create the slope, it gives a lot more rotational translation of power than linear. This slope actually works against you. It’s like programming for a 45 degree shot from a cannon and getting a 60 degree angle anyways with less power and that huge slope. So, you need one for shots further than 5 feet, if you can make it. And they would have to be run by a CIM motor, otherwise there isn’t enough torque. But, you could actually make a super light bot that runs on 2 CIM motors no problem. It’s all up to the rest of the bot, but torque and RPM is your focus for the whole competition. Torque for driving. RPM/torque for shooting. RPM for ball collecting/grabbing. Torque for bridge pushdown. Torque for arm/shooter raising/lowering. Every team needs to balance their motor layouts to correspond with every other part of the bot, otherwise it won’t doing any good if you have an amazing 4 CIM shooter with a slow body and vice verse.
The wheel itself offers little load when it’s up to speed…design it to run at 90% of no load RPM (or max power, which isn’t far from there), and you can always reduce the speed with the control system.
My point is that you shouldn’t worry too much about it, just put something together that’s estimated to maybe work, and see if it does work.
I’m not surprised at the 50% mass increase in the ball… amazing what a little change in radius does!
The shooter video I linked to, however, was driven by only one FP, geared down about 4:1 using a banebots planetary, if I recall. The design was originally set up so that a second gearmotor could be placed on the other end of the axle shaft, but we didn’t need it… in fact it would have broken the shooter speed limit that was in place that year so we left it off.
I wish we had known more about programming and PID loops when we built that robot. We could have added the second motor and used a control loop to maintain a constant speed on the shaft. That way the second motor would have brought us (back) up to shooting speed more rapidly, but then would have dialled back the power to control range. The “new” controller (okay, I guess to most students doing FRC, the cRio isn’t exactly new…) makes closed loop motor control MUCH easier… most experienced teams will have had some experience with closed-loop feedback, but for newer teams looking for a good excuse to learn more about closed-loop speed control, the shooter would be a great place to start.
So while I fortunately/unfortunately don’t have a robot/team to experiment with, I’m reasonably confident that two FP’s, properly arranged and geared, should have good range and repeatability when configured as a vertical shooter.
And I do have to say the backspin was incredible… occasionally the balls would land in the goal, and backspin themselves right out. This year I can only see backspin as being a good thing!
