This is a project that 1503 has been working on for several months, and just finished this past week (apart from some shifting hardware). It conveniently also serves as the testing platform for our new control system. Some video: http://www.youtube.com/watch?v=itPLTFa1Ivc
Very Very nice work! Time for some questions. How heavy is the entire thing without electronics and with electronics? Are those the banebots wheels? If so how are they holding up and how is traction? Does this drivetrain have the ability to climb ramps and if not would it be easy to modify it? How are the chains tensioned?
Overall I love this design and it looks very similar to many of mine. Good job guys!
Love the double banebot wheels–we had a similar idea.
Nice looking transmissions as well–are those 4-speed?
Could you give us some performance numbers for those of us who cannot access YouTube from our non-home computers?
Nice clean look! A great layout example for other teams with the new control system.
Pretty impressive perfomance, seen it over the weekend…
Clean design as usual. Looks like the team is doing great.
30 Shore A wheels! :eek:
And to think we had too much traction with 70-75 Shore wheels, and they wore through .5" of rubber over the season :eek:
It does look very cool though. The apparent lack of tensioners is especially clever
What are you planning to use those extra drive sprockets fore? Drive-train powered mechanism?
All I can say is WOW. Great job!!:ahh:
Does that thing have a 4 speed? And i love the layout you have no idea, very slick and very compact
I’m lovin the green and red wiring.
That is a very impressive drivetrain. Nice work!
A couple of questions of my own:
Since you’re running traction on all 6 (12) wheels, how much have you dropped the center wheels?
I’ve heard some reports of the banebots wheels blowing out, have you done any durability testing on them yet (getting pushed sideways and such)?
I agree, the layout is amazing!
The mix of 80/20 and custom frame is really well done too.
I wonder how well the wheels hold up while making a sharp turn though…
That’s hot Pat… You deserve another cookie for this.
Interesting that you used 80/20 and #35 chain, looks like a 4 speed shifter. Good work again by Spartronics, see you guys at GTR.
What’s the weight on that?
How easily can wheels be swapped out? These wheels were field tested by a lot of combat robots at robogames, and they found them to wear down quite fast (and that was on steel).
and purely out of curiosity, is the electrical board garolite?
At Robogames the wheels themselves proved to be rather durable, the tread layer is just a lot thinner than they appear and since they are a very soft rubber, wear down rather fast.
Thanks, everyone, for the comments.
We never measured it without electronics; with them, it’s sitting at about 60 lbs (sans compressor). Of course, there are 26 steel gears in there, most of which we’d weight-relieve for a competition robot.
Yes, they’re the 30 Shore A Banebots wheels. We’re really happy with the traction, and may switch to a harder grade if they wear down too easily. We’re prepared to switch the wheels out fairly often since they’re $5.25 each and a couple of minutes on the CNC takes care of the mounting holes.
It should be able to climb most ramps just fine. If not, we’ll just mess around with ground clearance (currently 0.75") and wheel spacing.
As with our 2008 robot, the wheel spacing is such that the chains are an integer number of links. I’ve always been wary of over-tensioning chain, and would rather avoid the friction and added complexity of tensioning systems. If the chains stretch and become too loose, we’ll just replace them.
Yeah, they’re 4-speed transmissions, inspired largely by AndyMark but entirely custom (with a shifting mechanism that addresses several maintenance-related issues we had with our Gen 2’s last year as well as making the parts much easier to machine with our shop). From a drive strategy perspective, I wanted a 3-speed, but decided that the best way to accomplish it would be to design a 4-speed and just not use one of the speeds. The (calculated) speeds are 5, 9, 11 and 19 fps, with the 9 fps speed being ignored.
We mostly picked those wheels for now because they match our team colours. The sprockets are there for now so that the front and back wheels are interchangeable.
The center wheels are dropped 5/32", same as our 2008 robot. Given the flat nature of the wheel tread as compared to the roughtop we used last year, though, I’m thinking we could probably get away with less.
We haven’t done any extensive field testing yet. We cut some wheels apart and were happy with their composition and with the way the treads are bonded to the wheel, but breakage from a side load is still a potential issue. Our attachment point to the wheels is fairly far from the axis, though, so I’m not overly concerned.
Geez, all this talk of cookies, and yet nobody ever follows through…
The extrusion is actually Bosch; we were lucky to obtain it for free. As for the chain, I know that by the numbers, #25 is perfectly acceptable for drivetrains, but it’s less forgiving of slight misalignments and stretches more readily. I’d rather waste the weight on #35 chain and be secure in the knowledge that we’re never going to have issues with it in the heat of competition.
It’s just a matter of sliding the axle out. We’ll have an extra set of wheel assemblies ready and a bunch more pre-machined wheels for swapping with those on the inactive set.
Not positive, but I think so. It was a piece of some sort of fibrous composite that was kicking around our back room. On a competition robot, though, I’m leaning towards using carbon fibre.
Very cool chassis design! The four speed transmissions look really neat, what kind of shifting mechanism are you guys using? I wasn’t aware of any issues with the AM dog shifting design, but I’d like to see what you came up with.
The chain tensioning (rather, the lack of it) is cool, too. How often did you have to change out chains on your 08 robot, and how loose did they get before you changed them? It seems like a great set up if they don’t loosen up too incredibly fast. Also, are those custom one piece sprockets on the wheels?
I like how organized everything is, even on a congested electronics board.
I also like how the 80/20 ‘standoffs’ are tapped for bolts for easy attachment of the outer side rails. Even better is that these 80/20 pieces could easily serve as chain tensioner mounts for a light-weight chain tensioner, should the design not have an auto-tension system.
Only criticism (and it’s small!): I don’t think it has enough ground clearance to traverse an elevated field component, nor is this particular design versatile enough to easily change a component to allow it to.
depending on the type of field element, I believe this design would do just fine with ramps and the like. Ground clearance isn’t quite the defining factor as the frame will almost always be sitting on one of the 12 wheels present. Also depending on how they accomplished the lead in and trail exposure of the wheel this could climb very well
I’ll preface this post by stating that I don’t mean any criticism at all to AndyMark shifters; we’ve used them unmodified in a few of our robots with great success. Most of the “improvements” we’ve made are because of stuff we want to do which most teams don’t (such as shift at 60 psi and shift two transmissions with one cylinder), or just to make manufacturing easier for our student shop, since, with the exception of the 12 and 40 tooth gears, this design requires gear and shaft sizes that AndyMark doesn’t offer.
That being said, here’s a list of the issues we encountered and the solutions we came up with. Here’s an Inventor screenshot of the shifting interface.
Issue #1: The split pins connecting the dog and shifter block to the shifter shaft (the small shaft that slides inside the output shaft) are hard to add/remove during maintenance, and tend to shear when shifting at 60 psi.
Solution: Our shifting shaft has a #8-32 thread on one end which screws directly into the dog (and is prevented from unscrewing by thread-locking compound), and a #10-32 thread on the other end which goes through the 3/16" bearing which is held on with a nut. This has the added benefit of being much easier to machine, since you don’t have to drill perpendicular holes in a tiny shaft.
Issue #2: The AndyMark dog gear is difficult to machine, as are the surfaces on the gears with which it mates.
Solution: Replace the dog gear with a piece of 1/4" steel square bar, 1" long, with a hole in the middle tapped to #8-32. Since all our gears come with hubs that we have to machine off, leave 1/8" of the hub and cut a cross into it slightly wider than 1/4".
Issue #3: We need longer equivalents to the AndyMark output shaft, which is difficult to machine (for our students who do the work on manual lathes, anyway) since a hex shaft needs to be parallel-turned to round along most of its length and then have a keyway milled into it.
Solution: Use pre-keyed 1/2" hex shaft, and just mill a slot into it where the shifting dog needs to go. To space our driven gears apart like the hex in the original shifter does, we instead have a ring-shaped spacer that fits around what’s left of the hub of the smaller gear, and rubs against the larger gear (right now it’s aluminium, but the plan is to make it out of bronze later).
Issue #4: The tiny screws that hold the shifter block on to the 3/16" bearing are very difficult to access on the robot when the transmission needs to be disassembled for maintenance. Also, the 1/4"-28 thread tends to strip from the non-axial forces when we shift both transmissions with one cylinder.
Solution: Replace the aluminium shifter block with a custom steel one (not pictured). The screws that hold the 3/16" bearing in now come in from the top, and it’s their shank that holds the bearing in instead of the head.
I don’t remember exactly, but it was something like once or twice over three competitions. They weren’t anywhere near loose enough to come off the sprockets, but enough to be occasionally contacting the ground.
This looks great! Kudos on a successful pre-season project that will pay dividends come the build season.
Just a quick question about the banebots wheels, because we were very strongly considering using them (we ordered several in different durometers, etc.)… Did you do any kind of testing with them? Such as, seeing if the hex within the polypropylene hub would strip out? Or if the urethane shears off when being pushed sideways?
We were in the process of designing tests to determine if these wheels would be able to handle the rigors of FIRST season.
Thanks for the great details about the gearbox, it’s a very clever design. The simplicity of the engagement mechanism seems like such an obvious design now, it’s strange that this is the first one like it I’ve seen.
We’ve really only driven it around the hallway a few times, just to see if everything fits and works as planned. There is still no shifting hardware, but everything else is there.
On the middle wheels, we are noticing some notable peeling on the rubber itself, only from maybe 15 minutes of usage, on a smooth concrete-like surface. Also, these wheels like to pickup everything and anything on the concrete that we run it on. We have yet to do any heavy testing such as being pushed sideways or anything. We also have yet to run it on carpeted surfaces.
At any rate, we shall see what these wheels stand up to. I personally love the fact that we have finally built a robot that actually rolls smoothly
Also, the centers of the wheels are machined so the holes are enlarged, and we simply have spacers and bearings riding on an axle on the corners, and a hub for the centres driving the other four wheels. Thus, we are not concerned of stripping the drive configurations that the wheels are shipped with because we dont directly apply power to the center of the wheels.