Designing a "belt + roller" conveyor for balls
 

As an off-season project my team is working on a foam ball shooter and we are now looking at making a conveyor similar to what a lot of other teams have made for ball games in the past. We have plenty of good ideas for how to design something like this, but not much experiences with the parts that comprise one.

The top questions I have are

• What are some common grippy belts that are used, and where can I buy them? Polycord? Like this?
  • What about this flat stuff Miss Daisy used?
  • Knowing names, durometers, and sources will help.
• How do people keep belts from wandering on the roller?
  • Miss Daisy's machine has small black wraps of (electrical tape??) on the roller. Does this work by emulating a Crowned Pulley?
  • What are the spacers in between the cords seen on 254's robot and so many others? A name or link would be great.

Any additional tips or tricks that other teams have picked up over the years would be great. We are building a shooter in order to get experience with something we haven't tried before, so any experiences or lessons learned from you all would be great to have.

Re: Designing a "belt + roller" conveyor for balls
 

Most teams source these belts from McMaster-Carr.

http://www.mcmaster.com/#urethane-belts/=zrook3 will lead you to both the flat and round belts.

Re: Designing a "belt + roller" conveyor for balls
 

It's just flat polycord. McMaster carries both kinds in several different sizes, simply go to belts --> flat/round belts --> set material to urethane.

Yes. 469 did the same in 2013. We put an o-ring under the tape to get more of a crown.

Re: Designing a "belt + roller" conveyor for balls
 

The "spacers" seen on 254's intake are actually grooved rollers which I believe can also be found on McMaster-Carr.

Re: Designing a "belt + roller" conveyor for balls
 

Check out 254's conveyor guide to see how they made their rollers.

Re: Designing a "belt + roller" conveyor for balls
 

Our rookie robot (Rebound Rumble) had a polycord elevator/conveyor to pick the balls up from the floor, raise them about two feet, then laterally another foot or so to drop it into the trigger cup (a PVC end cap mounted on a pneumatic cylinder), which would then lift the balls into the shooter wheels. We did both runs from the same motor, running (IIRC) four strands of polycord vertically and three horizontally. Thes ran on shafts with notches made by gluing short segments of foam tubing (insulation, I believe) with even shorter gaps between them. The polycord ran behind the balls on the lift and under the balls on the lateral run. We had sheet plastic (lexan?) as the opposite wall for the balls to roll up and over.

Re: Designing a "belt + roller" conveyor for balls
 

You can also use surgical tubing instead of polycord, we used it in 2009 and 2012, worked great both times. Spacers in between each band will keep them from going anywhere, just hold them in place using a locking collar.

Re: Designing a "belt + roller" conveyor for balls
 

How did you splice the ends of the surgical tubing together?

Re: Designing a "belt + roller" conveyor for balls
 

That is a great resource! Does anyone know what those particular spacers are? A larger plastic roller on top of the under roller?
I imagine we could just 3D print something that would work well but I would be interested to see what is available.

Re: Designing a "belt + roller" conveyor for balls
 

Depending on the ID of your surgical tubing, you can actually get pretty good results by using a short length of pneumatic tubing as a press-in insert to the ends of the surgical tubing. Works pretty well for these sorts of applications, though there are other ways to do it.

Though I'm definitely sold on flat belting after playing around with it the last few seasons. Just a bit of electrical tape around your roller to make a crown and the belts will self-center.

Re: Designing a "belt + roller" conveyor for balls
 

We've found that a bit of glue and pneumatic tubing works well. Put glue inside each end of the surgical tubing, then slide those ends over a piece of pneumatic tubing (a 1 inch long piece should do just fine). Let dry and it will hold, as long as it's not stretched an outrageous amount.

Re: Designing a "belt + roller" conveyor for balls
 

They are custom machined from one piece of solid round ABS bar.

Re: Designing a "belt + roller" conveyor for balls
 

We fused them up with a heat gun, and then used two pieces of plastic clamping them together for a few minutes.

Also, I wasn't thinking straight. I meant to say urethane tubing.

Re: Designing a "belt + roller" conveyor for balls
 

Ah, that makes more sense.

Just want to point out that while the belts do self-track to the center of the crown, it's definitely not going to stay there under any sort of side load. Be prepared to be realigning belts between matches.

Re: Designing a "belt + roller" conveyor for balls
 

1987 Broncobots had a kick@$$ conveyor belt system this past year (kinda reminded me of what 1986 Titanium did in 2013 with their full court shooter :P) but I believed it was a surgical tubing type conveyor with pulleys at each end to keep it in line and to control the flow of boxes. See it folded up. Roughtop is also REALLY grippy (we used it on my FTC team I help mentor last year and literally wore down the tile flooring during a mis-programmed auton.

Re: Designing a "belt + roller" conveyor for balls
 

so here is another question, what diameter rollers do you all like to use? I know that there are minimum diameters for the different thicknesses of polycord, but I have never fell upon a standard roller diameter that seemed to work better than others.

Re: Designing a "belt + roller" conveyor for balls
 

469 tends to go 2"; most of their rollers are 2" x 0.035" 6061 tube. 2012-2014 were all like this.

Re: Designing a "belt + roller" conveyor for balls
 

What do you use to transfer the power from a gearbox to the 2" tube?

Re: Designing a "belt + roller" conveyor for balls
 

We did something pretty small (about 1") for our 2012 robot. This was our rookie year, limited budget, and we spent most of it on registration and shifters. The diameter was selected so that we got the right speed for one of the motor/gearbox combinations we had (I believe it was a window motor).

Re: Designing a "belt + roller" conveyor for balls
 

IIRC, slightly more than 1.9" (OD of Sc40 1.5" PVC).

Going into further detail, we then wrap the 1.5" PVC with shelf liner. We create our spacers/grooves by using rings of 2" PVC over that assembly.

We used this approach in 2009, 2012 and 2013, and were largely happy with it. We still get some amount of wandering polycord, but it's not really a big deal.

Re: Designing a "belt + roller" conveyor for balls
 

The 254 conveyor slideshow goes over a lot of what you're looking for. McMaster sells polycord (we usually use the 3/16" OD): http://www.mcmaster.com/#round-belts/=zsnxr5

We love polycord, especially for rollers, you can cut and weld it together in a few minutes, and it's super easy to replace. It is also good to use polycord in applications where come slipping is need or expected, mainly so you don't strip a timing belt.

Most rollers we use are polycarbonate tubing with a machined or 3D printed hub that is inserted into a bearing which then fits a into a plate. The hub is then riveted onto the tube.

For pulleys, we usually use custom machined "rope" pulleys with a hex or round keyway bore. We last used poly cord on our 2014 robot roller.

Re: Designing a "belt + roller" conveyor for balls
 

*glances around*

I'll let you in on a little secret I figured out this past year.

Get a polycarbonate roller from mcmaster with 1.125 ID. Put a bearing on one end, and put a versa hub on the other. You see, many common FRC bearings are 1.125 OD, and the versa hub's "hub" is 1.125. Super glue them in and you're done. You can either direct drive the versa hub (if you pick a 1/2 bore hub), leave it to run on the plastic with a 3/8 shaft, or bolt a 25 sprocket to the hub if you want to chain drive it.

That's it. Zero work rollers.

http://www.mcmaster.com/#8585k63/=zsp0kc
http://www.andymark.com/product-p/am-0209.htm
http://content.vexrobotics.com/vexpr...g-20141010.PDF

Not very expensive either, and clear so you can put all sorts of gimmicks inside them :D

Re: Designing a "belt + roller" conveyor for balls
 

One off the constant recurring problem with intakes and conveyor belt lifts is jamming and the resulting fried motor. Especially with the RS series. In 2014 we added a ball clutch to our ball intake system. If we had not developed this we would have had bins of fried motors. We used and AM 9015 with a banesbot planetary. Took a solid .5" shaft collar. Drilled six holes radially in the shaft collar and broached it for the BB shaft. The holes were drilled so that a .1875 ball bearing would sit just below the midpoint. A Martin #35 sprocket that is about .375 thick was then drilled with holes matching the shaft collar. These holes were drilled so that the ball bearings would sit above the midpoint. The 2 are put together on the BB shaft followed by a series of spring bevel washers. The number and pre-loading of the washers determine the clutch release point. The whole thing was done by hand on a drill press and worked amazingly well. We had investigated magnetic and friction clutches. The ball clutch performed better. So my advise is that adding a clutch to a intake or convoy is a really good idea. May be able to dig up some photos if any one is interested.

Re: Designing a "belt + roller" conveyor for balls
 

I'm a little puzzled at the need for a specially designed clutch in an intake system (especially a belted one). It seems a lot simply to design your system to slip before it stalls your motor, especially if you're using a fan-cooled motor (such as the RS series). Almost every polycord-based conveyor will do pretty much automatically.

In other words, have your intake roller/conveyor act as your clutch.

Re: Designing a "belt + roller" conveyor for balls
 

I was more referring to the ball intakes that were common in 2014. These usually were belt or chain driven. Yes, poly cord will slip but if it happens frequently it stretches and wears. We had similar problems to a lesser degree in 2010 with the soccer balls.

Re: Designing a "belt + roller" conveyor for balls
 

This is what I love about FRC. Before this thread, I had never heard of crowned pulleys. #SuchLearning

Re: Designing a "belt + roller" conveyor for balls
 

This is fantastic. Nice find.

Re: Designing a "belt + roller" conveyor for balls
 

This is right on the edge of the topic, but I've decided it's worth posting on this topic, because it is a "belt + roller conveyor for balls". For our latest off season project, we've broken our team into three project teams to add some systems to our 2015 robot. Our final 2015 robot had a 4-wheel omni tank drive, and a 6'4" tall lift (including signal light height) on the "front". For this exercise, we gave the students autonomy, with the mentors available as "knowledge resources"; we'll base our build season balance of mentor vs student design decisions on the results of this challenge.

The floor of the classroom shall have a great many tennis balls (courtesy of our head coach from his house, so I personally expect a wide variety of "dead" balls with a fair number of "live" balls).

• Team 1 has to pick up tennis balls in the "right half space" outside of the original robot's frame perimeter and pass it to team 2.
• Team 2 has to accept balls from team 1, pass them behind the lift, and deliver them to team 3.
• Team 3 has to accept balls from team 2, and place them into a standard FRC tote in the "left half space" of the 2015 robot.

The tote begins and must end in a fairly small taped box. (If the team decides to carry it around, that's cool, but there are zero points if the tote does not end up back within the taped box.)

I've been working with all three sub-teams on their mechanisms, but it's team 2 I'm going to address here. It was pretty obvious that "passing the ball around the back" was a much simpler task than the other two, so the teams figured out that team 1 would hand off to team 2 fairly low, and team 2 would hand off to team 3 fairly high. "Passing around the back" is currently planned through a length of 3" diameter aluminum dryer hose, driven by gravity. This means that it is up to team #2 to additionally lift the tennis balls up from where they receive them from team 1 to somewhere quite a few inches above where they will deliver them to team 3. So far, I'm digging the teamwork - team 2 accepted additional tasking because they had been potentially under-tasked by the constraints.

So the question is, how do you lift a tennis ball by four or five feet? I would have probably pursued a polycord-and-plexiglas solution similar to our first year's basketball lift, but given the lack of another robot beating them up, team 2 decided to use a velcro loop. We are using a couple of 3" PVC drain covers glued together and covered in "Cro" (the hooks of velcro) to drive a 3" wide "belt" constructed of dual-sided velcro with the "Vel" (loops) inside and "Cro" outside. The opposite pulley may simply be a fixed axle that "loops" rub around, or may be another spool turned by the velcro loop.

*** - for anyone who does not already know, VelCro is brand name constructed as a contraction of two French words (both of which have English cognates). I do not speak French, but I've grown up in Southeast Louisiana, where French is ubiquitous in place names, food, and other random things, only a bit ahead of Spanish and German. When I went to France a couple of decades ago, I was able to puzzle out most of the road signs. Anyway, Vel is the first syllable of Velour, which means loops, and in English means a rather soft fabric which has a bunch of loops sticking up from a substrate layer, similar to but much softer than carpet. Cro is the first syllable of "crochet", which means hooks. There is a style of yarn work called crochet which uses hooks to pull yarn into intricate shapes. My mother was a serious practitioner of crochet in the late 70's and early '80s, and I did a bit of practice in it.

Re: Designing a "belt + roller" conveyor for balls
 

Thanks. Kudos to the team who integrates a cheap slip ring and puts LED's in one.....

Re: Designing a "belt + roller" conveyor for balls
 

When 341 made extensive use of polyurethane flat belting in 2012 (and 2013), we used exclusively 1.25" OD, 1.12" ID aluminum round tubing from Online Metals for our rollers. 10 seconds on the lathe (or a minute with a deburring tool) and it is a great fit for a hub or 1.125" OD bearing.

Re: Designing a "belt + roller" conveyor for balls
 

In 2014 we used the polycarb tube for our rollers and it definitely suffered from a lack of rigidity. The aluminum upgrade is definitely something I would recommend. Our tubes were around 2ft+ long.

Re: Designing a "belt + roller" conveyor for balls
 

Thanks for all the great feedback everyone, the team working on the beltway for our shooter got a lot of use out of the resources you guys provided.

We currently have the plates that make up the shooter assembly CADed and will sending them out for manufacturing on Monday. If people wouldn't mind giving the design and especially the drawing a look over we would really appreciate it.

Shooter Plate
Shooter Plate Drawing

Feedback I would particularly like include:

• We're going with a 1/4" thick plate, is that appropriate for this application?
• Is the pocketing too agresive? ie. are 1/4" aluminum struts strong enough?
• Is the Drawing too hectic? It looks like I have a lot of dimentions crammed in there...
• Also any overal design comments you have are welcome :)

Re: Designing a "belt + roller" conveyor for balls
 

Has anyone else used hollow core round belt with the barbed connector? Example: http://www.mcmaster.com/#6567k57/=109btha (belt) http://www.mcmaster.com/#6567k22/=109bt5d (connector)

Re: Designing a "belt + roller" conveyor for balls
 

We used this type of belt for our rookie 2012 Rebound Rumble robot to pick the basketballs off the floor and raise them to the cup which would lift them to the shooter. It was probably the most reliable system on the robot. I was not as involved that year, so I don't know why, but the belt was taped together rather than just held with the barbed connector. I suspect that it was because we were running it around about 1" diameter rollers (larger is recommended).

Re: Designing a "belt + roller" conveyor for balls
 

That's my exact situation. Last time we used them was on our 2012 rookie bot and they were also on < 2" diameter rollers and taped. I was wondering if anyone else had because like you I wasn't very involved with the lift and want to make sure we didn't just get lucky with it.

Re: Designing a "belt + roller" conveyor for balls
 

Back in high school (seven years ago!!! :eek: ) my team used these in 2009 for a ball conveyor, just using the barbed connectors to join them, around ~2" rollers. Worked well, I don't think they ever pulled themselves apart. The belts may have stretched a bit over time since they are hollow and relatively thin, so have some extra barbs so you can shorten longer runs over time. Welded poly-cord is probably more robust but this was good enough for 2009.

I personally prefer the orange flat stuff since it's a bit tackier and has nice self-centering properties.

Re: Designing a "belt + roller" conveyor for balls
 

My team also used them for our intake in 2012 and we did use the barbs. When we first cut the urethrane, the barbed connectors constantly popped out. After a few days of sitting and stretching they would be fine until the urethrane would stretch too much, we would have to re-cut it, and the process would start over again. Also, the barbed connectors are a pain to take out. That being said, they worked pretty well for our intake, and I would use them again if the need arises.

Re: Designing a "belt + roller" conveyor for balls
 

Other than the added complexity of machining the bore of the tube ends, would it be worth using the 1.25" OD polycarb tubing with the 1" ID and then machining the bore of the tube ends to accept the 1.125" bearing and hub? Besides being significantly stiffer, this would also allow you to intentionally undersize the bore to get the amount of "press fit" you might want for the inserts.