AndyMark Parts Feedback 2015

[JesseK]

Now that champs has completed, what were the strengths and areas of improvement for the AndyMark line of COTS products? What new products would teams like to see that are in the same core vein as the current AndyMark or their other suppliers (such as 221LLC)? In what useful or unique ways did you use these products that you wish to share? What nuances or undocumented features did you find with them?

I'd like a way to purchase the Churros in bulk. They are the #1 used COTS part on our robot this year, as funny as that sounds.
Our DART actuator didn't get onto the final bot, but it was great for prototyping. It doesn't seem to like axial loads, but that can be dealt with via proper design. It would be nice to swap out different motors for it, however.
I know that 3373 used the Swerve & Steer module, and the only problem they had was zeroing the absolute encoder. The worked through that, but otherwise had no mechanical problems in their swerve this year.
We also have really liked specific gearboxes for their toughness over the years. I'd like to see either mount holes or a kit made available to use a pneumatic cylinder and a locking pin directly into the initial stage's steel gear for some of them. Holding a game piece steady at a specific height seemed to be a barrier for some teams this season since they didn't realize they needed it until after then entire subsystem was assembled.

[marshall]

900 ran into a lot of issues with the GEM gearboxes this year. Granted, we were doing derpy stuff with them so derpy issues were bound to spring up. We shocked the AM crew in St Louis with the multitude of failure modes that we found for the GEMs. We are going to try and write a paper about them once we get some time.

To be clear, 900 <3's AndyMark. They are easily one of the best suppliers out there and amazing to work with. We do derpy things to their gearboxes because we can, not because we are looking to blame them for making bad gearboxes. Seriously AM folks, you rock!

[wilsonmw04]

We used two new AM parts this year.

1. 4" mech wheels: We found them to be robust and very smooth. We only had one roller failure. It was on the practice robot. We wore out one of the brass rollers/rivets. It was easily repaired with a bolt. Not really complaining about the failure because that robot was driven very hard by my drivers. We actually caught air going over the scoring platform in an attempt to tip the robot over. We will use these wheels again.

2. Raw Box: This was used to power our two cim elevator. The worm screw was ALMOST un-backdriveable (is that a word?). With a little programming it easily kept a 6 tote stack at the proper elevation. No failures and would use again.

[marshall]

2. Raw Box: This was used to power our two cim elevator. The worm screw was ALMOST un-backdriveable (is that a word?). With a little programming it easily kept a 6 tote stack at the proper elevation. No failures and would use again.

We broke two of those as well. They are good gearboxes provided you capture the output shaft solidly. Overhung loads on those gearboxes cause problems though.

[IKE]

As an LRI, I came across several teams that were getting hot motors and/or voltage drop outs. A common issue I saw was:
Kit bot modified to 6" wheels with no other mods. The effective free speed for this arrangement is quite high (something on the order of 20 FPS). This lead to a lot of hot motors.
I also helped a few teams that had placed a gear backwards in the gearboxes. There is a small offset that helps keep the gears from dragging on the housing. this drag resulted in hot motors as well. It might be good to add a current check on blocks for a chassis to see if you have un-necessary drag. this can be done with the PDP board and your laptop. Might be good to add a diagnostic tool.

[wilsonmw04]

We broke two of those as well. They are good gearboxes provided you capture the output shaft solidly. Overhung loads on those gearboxes cause problems though.

We didn't support our shaft this year. It was running two different chain loads. I will keep that in mind for next year.

[marshall]

We didn't support our shaft this year. It was running two different chain loads. I will keep that in mind for next year.

Well, we were trying to drive a 7 foot arm with 20 pounds of dynamic load on the end of it... so, yeah... YMMV. ;)

[mrnoble]

We tried both the AM mecanum wheels and the Vex models. We used the 6" standard AM's on a demo bot without incident for 40+ hours before breaking a roller axle. We decided to go with the AM HD wheels for our comp bot and were very happy. Smooth and very strong and durable. We are planning to make hubcaps for them for Cow Town; they look like they deserve something pretty.

[Pratik Kunapuli]

341 used originally used the Banebots 220 series gearbox and they seemed very good to get an initial idea of reductions, especially with the add on 4:1 reduction. Since then, we changed to a SuperShifter for our elevator and have had no complaints since. We locked it into high gear, but the shifting was smooth when we were playing with it. The best part of the SuperShifter was the ability to set it and forget it. We never had to open it up to service it, and once the encoder was set it worked for the rest of the season.

[Munchskull]

We used the gears form the kit of parts drive train, my only comment would be, can they please be made of aluminum next year?

[Oblarg]

The worm in the WormBox consists of a plastic worm on a metal bushing. The connection between these two is not as strong as it should be (it appears to be a rather weak interference fit), and so under load the plastic can slide on the bushing and get pushed into the CIM mounting hole, causing lots of friction and eventual failure.

We discovered this before it became unmanageable and, with the help of the NASA machine shop, were able to push the bushings back into place and secure them there with pins. Such pins should be standard to prevent this from happening in the first place. Other than that, they were surprisingly durable for plastic gearboxes and the encoder mounts were very handy. Would purchase again, provided this problem is fixed.

We also received a hex hub that wasn't bored anywhere close to straight, and was unusable. We seem to encounter little QC problems like this every year, and while they're not crippling they are a constant annoyance.

[Michael Hill]

We ran into a couple of issues with AM parts this year. First off was shearing off the output shaft of the PG-188 when back driving it. That was not expected at all. We ended up taking off the PG-188 in the end anyway. Second was having an issue with the AM Planetary gearbox. The shaft was undersized by a few thousands (we mic'd it out). It looks as if it was made to 5/16 rather than 8mm, so the press fit gear we were using was slipping on the shaft. We tried knurling the shaft, and that seemed to work, but we eventually replaced it with a MiniCIM after finding some weight because we were unsure about how long the knurl would last.

The tough boxes worked great. We CNCed a new face plate out of .050 2024-T3 to lighten them up a bit and that worked a treat. We also used a couple of 32dp gears from AndyMark, and we didn't have any issues with them, so that's a plus.

[Oblarg]

We ran into a couple of issues with AM parts this year. First off was shearing off the output shaft of the PG-188 when back driving it. That was not expected at all.

Had this problem last year when we used two of them, direct-driven, to actuate our roller arm (after both failed as described, we switched to pneumatics). If you look at the spec sheet, it actually does specify that this is the standard failure mode if you exceed the maximum rated torque - I was surprised, too, until I looked it up.

[Michael Hill]

Had this problem last year when we used two of them, direct-driven, to actuate our roller arm (after both failed as described, we switched to pneumatics). If you look at the spec sheet, it actually does specify that this is the standard failure mode if you exceed the maximum rated torque - I was surprised, too, until I looked it up.

If you open up the PG-188 gearbox, the shaft has a couple of triangular notches in them. That's where ours broke. I'm not sure if they're put there to create a known failure mode or not. If so, I'm guessing the thinking is that it's relatively simple and cheap to replace the shaft rather than the gear assembly.

[MrForbes]

As an LRI, I came across several teams that were getting hot motors and/or voltage drop outs. A common issue I saw was:
Kit bot modified to 6" wheels with no other mods. The effective free speed for this arrangement is quite high (something on the order of 20 FPS). This lead to a lot of hot motors.

We did the 6" wheel mod to our kit chassis, and we also changed the gears (we used some gears from older AM transmissions). It worked great! we had some 6" white wheels left over from something else, and we notched the ends of the chassis so they'd fit.

I also helped a few teams that had placed a gear backwards in the gearboxes. There is a small offset that helps keep the gears from dragging on the housing. this drag resulted in hot motors as well. It might be good to add a current check on blocks for a chassis to see if you have un-necessary drag. this can be done with the PDP board and your laptop. Might be good to add a diagnostic tool.

The instructions sort of show the offset on the gears, but don't explain it nearly well enough for most students to notice.

[Oblarg]

If you open up the PG-188 gearbox, the shaft has a couple of triangular notches in them. That's where ours broke. I'm not sure if they're put there to create a known failure mode or not. If so, I'm guessing the thinking is that it's relatively simple and cheap to replace the shaft rather than the gear assembly.

That was my guess, too, though if this is the case the fact that AndyMark doesn't seem to sell the output shaft separately on their website is a bit of a headscratcher.

[Pretzel]

Our lift mechanism was powered by a Raw Box in order to prevent backdriving. With the Talons for the two motors (Rs-775's with CIM-ilies) set to brake mode, no additional current was required to prevent a six tote stack with bin and noodle from backdriving the motors.

We did have one issue, however, with the holes for mounting the motors. Both of our Raw Boxes required us to drill out the mounting holes to get the screw to actually line up with the threaded holes on all of our CIMs. Once we drilled them out to a larger diameter it worked fine, but perhaps this is an issue due to the holes being drilled before bending the sheet metal?

[Gregor]

We didn't use many AM parts this year, but we watched many teams running the stock kitbot beach themselves on the bump.

In prior years (2010 & 2012), there was been some kind of modification to allow for traversing, or at least not getting stuck on obstacles.

I believe this should be an important requirement for designing future kitbots.

[Michael Hill]

That was my guess, too, though if this is the case the fact that AndyMark doesn't seem to sell the output shaft separately on their website is a bit of a headscratcher.

You might be able to buy it if you call them up.

[InFlight]

On the Nanobox there is inadequate clearance between the upper mounting hole location and the installed CIM motor. The 10-24 cap head bolts heads pushed the CIM down making for a very loud or binding gearbox. We had to machine down the OD of the bolt heads on a lathe to make these gearboxes usable. We used four of these per robot for our Mecanum drive.

[ehochstein]

We had huge successes with AndyMark this year.

We ended up using 5, 6" Aluminum Dualie Omni wheels (http://www.andymark.com/product-p/am-0432.htm). They worked perfectly all season except for one point during practice where the bolts keeping the wheel together fell out. It was easily fixed with some thread locker.

We used the kit bot chassis this year and modified it to be an H-Drive. We found it incredibly simple to make this modification, we just ordered a third inner side plate and an additional Toughbox Mini. We created four gusset plates to hold the center plate on the robot and had the "H" of our drive installed by the end of week 2. Worked great all season too.

Our least favorite AndyMark product of the season was the Stackerbox Double (http://www.andymark.com/product-p/am-0561.htm). It worked really well once we fully assembled it for use but it was a little difficult to modify the gearbox to work with a CIM motor. Overall, we found the documentation for the gearbox to be somewhat lacking. Would be super helpful if there was a document outlining the method to drive this directly with a CIM motor!

I know the team loves AndyMark, we bought a few non-AndyMark products this year and only ran into issues. The students are really happy with the quality that AndyMark brings to the table.

[Ben Martin]

We ran a raw box this year with a custom output shaft--no complaints, worked through every competition. 2 CIMs through VP's were input into the transmission.

We stripped the worm and worm gear on our practice robot's gearbox at one point, but we believe we assembled that one wrong and we were lifting 8 totes with a >5 ft lever arm with a 6:1 reduction against the gearbox (stress testing). So, the load rating seems pretty accurate.

We would certainly use it again.

[Qbot2640]

We used tons of AndyMark stuff this year. My only complaint was the shaft adapter to take the gear-motor shaft (10mm?) to half inch to drive a chain sprocket. We broke two...usually in match, rendering our lift inoperative. It would be nice to have a steel version...or better yet - make all the gear-motors have 1/2 hex shafts.

EDIT: Thought of another thing...The recesses for screw heads on the plastic belt pulley halves are too tight to get any kind of socket tool around the head - and the screws get too tight to drive with a slotted screwdriver.

[Nuttyman54]

1983 used the LJ Bevel Boxes in our drivetrain this year. There were a few quality issues with the bearing bores not being straight/lined up across the box and with some flashing left in a few of the bevel gear teeth that we had to file out. We bought 9 (2 robots + spare), and only 8 were really usable. We put the 4 best on the comp bot and the others on the practice robot. There was a noticeable difference in robot speed and battery draw, but ultimately they ran fine all season on both robots.

I would love to see these with a 1/2" hex output, since 3/8" hex COTS parts don't have the same available selection as 1/2" hex does.

From what I recall, that was the only AndyMark product we used on our robot this year. We had plans to run modified supershifters and all kinds of fun things, but then Recycle Rush happened.

We didn't use many AM parts this year, but we watched many teams running the stock kitbot beach themselves on the bump.

In prior years (2010 & 2012), there was been some kind of modification to allow for traversing, or at least not getting stuck on obstacles.

I believe this should be an important requirement for designing future kitbots.

I don't have the timecode reference handy, but Kate Pilotte mentioned on the Kit of Parts episode of Behind the Lines (https://www.youtube.com/watch?v=E33Gxjdi0PY) that the scoring platforms weren't added until something like October 2014, with the kitbot design already finalized and in manufacturing, so not much AndyMark could do at that point. The GDC did their best to implement a platform design which was capable of being traversed by the kitbot, but ultimately some team modifications were needed to avoid all potential problems.