FRC 95, The Grasshoppers, Build Thread

Super-duper late start this year, but here we are!

We took a more ‘main stream’ approach this year than in many past years, but of course we had to make a few things a bit different.

Chassis is our normal plate-spacer-sheet metal construction with live axles connecting 10 6in wheels (colson for now, pneumatic maybe later… if we ever get them!) 6 CIMs geared for 10.4ft/s (ish) single-speed with WCP gearboxes.

Our main mechanism is a shoulder-mounted shooter/intake combo with two stages of belts/pulleys and rollers. The shooter is powered by 4x 775 pros, and the shoulder is controlled with 2x mini-cims.

Almost everything will have encoder feedback. An ardupilot will assist with navigation, and an Axis camera will assist with vision processing and/or manual aiming.

Firstly, here is a CAD cap of our 2016 robot: Kovaka

http://i.imgur.com/xDHLGwi.png

(I don’t have a lot of great piece-part mfg steps, apologies)

Here is the plasma-cut blank of our belly-pan.

http://i.imgur.com/AbPJcmC.jpg

We then bent that in-house into the crazy belly pan shape for our chassis.

http://i.imgur.com/iCedRZE.jpg

And have started to build the shoulder rail assembly, which is a major structural element in the chassis. Although the chassis is (on purpose) a bit flexible to give us a little pseudo-suspension.

http://i.imgur.com/sSFbC5v.jpg

I will try to get a few shooter pictures and other layout pictures tonight.

I have left our numerous details and design choices, but feel free to ask if anything piques your interest.

Thank you for doing this again, its one of my favorite build season threads. Glad its back.

Are those 775s 1:1? I can’t tell if that middle stage is a planetary or an integrated encoder. If it is an encoder, won’t you be exceeding the maximum rpm for velocity control?

Beautiful CAD (as usual!) I have a few questions:
It looks a little too tall, but is it low-bar capable? I can’t quite tell from the picture.
Are you planning to climb? What defenses can you break?
Are you worried about the whole thing racking under impacts?

Nice metal origami work on that belly pan! Looks like you didn’t put in the holes with the plasma cutter, do you do them afterwards with a drill press / mill while the part is still flat, or after you bend it to match the holes to the other pieces?

The ‘booster’ stage of the shooter is 1:1 on 775s. Our software/controls team advises me that the Talon SRX input can handle something above 100krpm with those encoders. I didn’t do the math myself though.

Thank you!

We intend on being low-bar capable. The shooter folds into the chassis.

No climbing right now, unlikely to add it.

We plan on breaking low bar, ramparts, rough terrain, cheval de fries (sp?), moat, rock wall, sally port, portcullis (sp?), and to hold down the drawbridge for another team to quickly defeat.

All critical motion elements are isolated from chassis twisting, and the top/bottom of the shooter can handle a lot of misalignment with respect to each other. Think more gumby and less he-man :slight_smile:

Thank you! It was an interesting project, but all very doable with the right press-brake. We only played the aluminum drums on a few folds (brought the fold to 70-80deg on the brake and hammered them home).

For ‘critical’ plates like the drive pod plates we blank out the part on a CNC plasma cutter and then do all of the holes/fine features on a Prototrak CNC knee mill. For less critical items like the belly pan we will pierce all of the holes with the cnc plasma cutter, then slice the part out (still on the cnc plasma cutter).

I know the picture of the first belly-pan blank shows no pierces, we made a second pan with hole pierces and match-drilled the holes into that first blank by hand.

Also, without electronics, our chassis is coming in at 65lbs.

Our shooter structure is coming in at 13lbs (no motors/shafts/pulleys/belts).

We anticipate being around 100-110lbs fully dressed.

I think they missed a decimal place. The table in section 7.5.4 of the Talon SRX Software Reference Manual states that the CTRE Magnetic Encoder has a maximum RPM of 10,000.

The 775pros have a free rpm of 18,730. Even if you’re running at 75% free speed, that’s still over 14,000 rpm.

Thanks for the info Ty. We clearly worked the problem from the wrong direction. Hopefully we can come up with a decent way of solving this issue!

Just replace the dipole magnet in the encoder with a monopole magnet. That should increase the functional rpm of the encoder by a factor of 2. :stuck_out_tongue:

/s of course

I’ll do that right after I collect my Nobel prize. :wink:

In seriousness, we’re likely going to pull the magnetic encoder and use a Grayhill 63R instead.

This is looking really good!

Are those semi circle slots in the drivebase for chain tensioning?

Woah, that looks awesome! It also looks quite strong. Good job guys!

Thanks!

Winner winner, chicken dinner. I’m going to post a picture of our chain tension cams tonight.

Thank you very much! Watching the chassis flex +/- 3 or 4 inches per side, then restore to ‘level’ is… unnerving, but it works!

Picture of our chain tension cams. I’m told they look like ninja ghosts…

… kissing ninja ghosts.

A shot down the inside of a drive pod.

Mini-cim shoulder drive setup. We’re gearing 108:1 with the versa planetary (if memory serves) with an additional 16:34 sprocket reduction. Should be able to flip Kovaka back on his wheels if things go sideways… err… upside down.

Wiring is starting to get laid out. We invested heavily in electronics tools and components this year, and it’s paying off. No way we could package 12x speed controllers into this belly pan so neatly if we were using old-school victors/talons/jags. Also, we love velcro.

Starting to populate the shooter with axles, belts, and pulleys.

A close-up of the shooter.

It’s a bit… different… than most shooters we’ve seen before. We’ve setup the first ‘booster’ stage (3/4 of the length) to ramp up in speed as it feeds into the ‘firing’ stage, which spins considerably faster. Prototype testing has been quite promising.

A few pictures from today:

The shooter is nominally populated with belts and rollers and stuck on the robot!

How he’ll be able to fit under the low bar.

Wiring is about 70% there, enough for a power-up test!

We’re currently trying to get code on the Athena (we refuse to call it the “Roborio”), which always seems to have some growing pains. We’ll get there…

Make sure you take into account the load guides for the versaplanetary. This gearing ratio is either just barely scraping by or it is outside of recommended loading.

Are you planning to intake from both sides? Is there any significant advantage, that your team saw, to that strategy (if you are indeed capable of intaking from both sides)?

We did use the guide. We are right on the edge of what is advisable for the versaplanetary with a mini-cim. During normal operation we will be nowhere near the stall limit of the mini-Cim. Flipping the whole robot back over will only be a 22amp draw (from JVN) design call.

Though your point is well-taken. The versaplanetary guide is to be taken quite seriously!