Projects‎ > ‎

Combat Robot: "upChukk"

Overview

Design idea:
  • Weight class: 3-pound, aka "Beetleweight"
  • Drive train: standard two-wheel drive
  • Weapon: vertical spinner
The plan is to complete it in time for Dragon*Con 2011 and the Robot Battles event there.

Dragon*Con 2011 Microbattles: RIP upChukk

posted Sep 5, 2011, 5:31 AM by Lewis Baumstark   [ updated Sep 7, 2011, 4:37 AM ]

"...odds are good that your first robot will -- and there is no nice way to say this -- suck..."

So says someone who's been at this longer than I.  Actually I had a pretty decent run for a first-timer.  

First round vs. Melty-Bot

Win.  Opponent was a wooden bot from the GaTech robot club.  My spinner tore him to pieces, literally.  I took a few chunks out of him early on and then he managed to get stuck in a corner.  I turned my spinner on him in the hopes of knocking him loose but ended up dealing a final blow.  (Dude, if you're reading this, I'm really sorry!  I didn't realize it'd do that much damage -- I guess I turned the juice up a little high.)  Swept his shattered remains off stage and that was that.

upChukk vs. Melty-Bot


Second round vs. Misdirected Agression

Lost.  This guy had a horizontal spinner and it delivered a beat-down.  If I read the video correctly, he knocked out my safety disconnect on the first hit, leaving me without power.

Ours is the first match in this video:

Misdirected Aggression



Damage report:
  • Gearbox on one of the Beetle B16 motors separated from the motor, just like in practice (in the pic see the motor dangling to the left of the glare spot and above the red-white-blue wires; the gearbox casing is the circular piece below the wires)
  • One of the hex bolts on my spinner weapon sheared off.  This is probably what put a hole in the roof of the arena.
  • Spinner weapon shaft bent.  I expected this -- even had spare shafts ready to replace it but didn't since the spinner itself was hosed.
  • Mangled emergency disconnect.  My weak(est) spot and he totaled it.  This is what left me dead-in-the-water.

Frantic Patch Job

After the aforementioned beat-down I had my doubts I'd have anything left for the rumble, but I tried anyway.  The biggest issue was the gearbox.  Fortunately, MOST of the pieces were still in my case, so I carefully swept them out and rebuilt the gearbox.  What I learned is that a 3-stage planetary gear will work okay (if loudly) if two stages are each missing a spur gear.  Wow.

While I had the case open I removed the weapon motor and ESC -- no chance getting the weapon ready for the rumble, so may as well save my $70 motor and $50 ESC for a later 'bot.

I had a random piece of angle aluminum in my toolbox that was the perfect size to build a quick-and-dirty fence on my weapon protrusions.  (That sounds dirtier than it really is.)  It offered some protection to the remaining electronics and gave the 'bot some added stability.

Kudos to Brandon of Team Shrewsbeckistan for soldering me a new safety disconnect.  I was now ready for the Royal Rumble!


Royal Rumble

Lost.  Not much to report here, mostly because it was (predictably) chaos.  One 'bot was made entirely of LEGO's, which ended up in pieces all over the arena after about 5 seconds.  So most of us ended up stranded on top of LEGO bricks.  I maybe had 30 seconds of movement (I'm being generous here!) before my rebuilt gearbox once again disintegrated and my drive motor was left dangling and twitching.  In the end, I was still on stage with a couple other immobile bots and one mobile bot who was declared the winner.  

3-lb Royal Rumble



Damage report:
  • Gearbox exploded.  Again.  No pieces found this time.
  • Wedge skirt ripped off, along with a large chunk of the Lexan "armor" to which it was attached.

Lessons Learned

  1. Thread-lock on the bolts that hold B16 gearboxes to the motor.  So says the guy that beat me. (Thanks for the tip, dude!)
  2. Lexan is not armor, at least not in the thickness I used.
  3. There are some brutal spinners out there.  Ponder how to defend against these.
  4. Use grade-8 (I think) bolts on my spinner.  Hardware-store bolts are too soft and shear off at inconvenient times.  (Thanks to the guy I was talking to in the pits whose name I don't recall.)
  5. Bite the bullet and buy hardened weapon shafts.
  6. Have a minion who can pay attention to announcements for me.  More than once I nearly missed my bout because I was too focused on repairing, recharging, etc.
  7. Related to #6: bring a second battery (or, at least, quick-charge in place).
  8. Also related to #6: hack my transmitter such that the "mixing" switch is external (and doesn't require removing a battery and flipping a minuscule DIP switch).  Or, remove mixing from the Scorpion ESC.  In the end, it should take no more than a couple of seconds to re-configure my controller when switching it to a different bot.

Skirting the Issue

posted Aug 21, 2011, 10:38 AM by Lewis Baumstark   [ updated Aug 21, 2011, 10:52 AM ]

I learned something yesterday.  I learned that wedge-bots are @#$%! irritating.  The robot teams I sponsor at UWG had a work-day and made amazing progress, especially given the competition is in two weeks.  Near the end, the three 3-pound bots that had mobility went head-to-head-to-head for some push-each-other-around-tests (i.e., no powered weapons in use).  The wedge gave upChukk and Twistys fits.  We simply had too much ground clearance -- the wedge would get under us and lift our wheels off the ground every time.

So I had a little weight budget left over and added a floating wedge/skirt to the backside of upChukk:

 

Mainly this is to eliminate ground clearance on what will now be the front of my bot (front? wha? see below...).  If it results in a little wedge-like behavior, well, too bad.

Yes, I realize this makes me one of the enemy now.  At least I have a fighting chance.

I also noticed, during out little bot rumble, that I had a heck of a time getting behind the wedge so I had a good place to use my spinner (had it been active, which it wasn't).  It was just too easy for them to keep the wedge pointed straight at me.  I don't think the spinner would have been much use head-on.

So my thinking now is to re-orient my controls and make the wedge-side of upChukk the front and the spinner side the rear.  Maybe I'll have more luck driving past my opponent and then quickly backing up to hit them with the spinner.  I doubt I'm a good enough driver to pull it off, but it's worth a try, at least when fighting those blankety-blank wedges.

Underweight!

posted Jun 16, 2011, 5:19 PM by Lewis Baumstark

2.87 pounds, ready to go.  Can't go over 3 pounds, so I'm in excellent shape.  The balance comes out to just a tad over 2 ounces, which might be just enough to put a piece of plastic in as a stiffener.

Battle-Ready

posted Jun 16, 2011, 5:46 AM by Lewis Baumstark


It's not easy to tell from the picture, but I re-built the weapon (again!) using a chain-sprocket drilled and tapped to accommodate two hex-bolts (one of which also functions as a beefier set-screw for the shaft).  The newer weapon has better balance (and thus less vibration) than my earlier attempt.  I perma-installed roller bearings to hold the shaft and, finally, added a piece of thin plastic on the backside to keep the guts from falling out.

I hate how the kill-switch is hanging out.  It will be an Achilles Heel if I go against a saw weapon.  But I can't come up with a better way to mount it and still have it accessible.

Chucking Achieved!!!

posted Jun 1, 2011, 10:11 AM by Lewis Baumstark

It seems I may have figured out one (possibly both) of the problems in my last update.  Using about 1/3 to 1/2 full speed on the weapon motor, I was able to throw my wood dummy a respectable distance.  The key here is collision speed; in yesterday's tests I "crept up" on the dummy so all that happened was the weapon would scrape at the sides.  Today I pushed it into the dummy at a speed comparable to combat movement which apparently is enough for the blade to "catch" instead of "scrape."

As I said, this was with the weapon running much less than full speed, which cut down on vibration a good bit.  I'm headed to McMaster-Carr in a few to pick up some bearings, which should help, too.

Since I don't need full speed for effectiveness, I need to investigate programming a max RPM into the ESC.  Failing that, I'll put some sort of mechanical block on the transmitter's weapon throttle so I don't redline it.

My Own Worst Enemy

posted May 31, 2011, 1:40 PM by Lewis Baumstark

Equal parts ROFL and GRRR....

Today was the day I finally hooked up and tested the spinner weapon.  The first few tests I got it up to speed and had it "attack" a piece of wood that was roughly 3-lb.  The spinner would scratch at it but would not throw it like I wanted.  That could mean it doesn't have enough energy (unlikely, as we'll see in a moment), that it doesn't have an adequate shape for lifting, or that it's spinning too fast to catch the dummy and throw it.  And as added problem, it was spinning so fast the 'bot was vibrating something fierce.

So I tinkered a bit and went in for a another test.  This time, the vibration was so bad it vibrated one of my drive motors off its gearbox.  Literally vibrated the screws -- which were held on with Locktite! -- loose.  Tiny (appr. 3 mm diameter) gears went everywhere.  At this point I think I've recovered about 2/3 of the gearbox.  Naturally I don't have a spare...

Mobility Check #1

posted May 24, 2011, 6:13 PM by Lewis Baumstark

It runs!

YouTube Video



I've yet to test the spinner weapon.  Everything I read says I need to condition the battery with several charge-discharge cycles before running it under heavy load...like a 40-amp spinner.

I also have a weird issue with the drive system.  Straight-back reverse is spinning in place like perhaps one wheel isn't getting the speed it needs to keep up with the other.  I can compensate with steering but it bugs me.  :)  Further investigation needed...

Awake-in-the-middle-of-the-night Thought

posted May 13, 2011, 12:24 AM by Lewis Baumstark

If I use a separate battery for the weapon system, I can use a 3-cell LiPo (11.1 volts), potentially doubling the energy stored in my spinner.  This merits serious consideration...

Ten Steps Forward, Two Steps Back

posted May 12, 2011, 12:47 PM by Lewis Baumstark   [ updated May 13, 2011, 1:12 PM ]

Current Status


Current weight: 2.57 pounds (measured with the scale at work).  Only thing missing from that calculation is the back plate, some bits of wire, and a little bit of mounting hardware for the electronics.

Everything is in place except the battery (more on this in a minute), the weapon ESC, and a rear armor plate to protect the electronics.  You may notice the layout has changed a bit -- I re-designed and re-built it all yesterday.  This design is simpler, likely more combat-durable and, best of all, I think it's a bit lighter!

Drive ESC & Rx Shock Mount

I wanted to mount the drive ESC on PCB stand-offs but I also wanted to shock-mount them.  Here is what I came up with:


The stand-offs are mounted to a piece of plastic I scavenged from the junk bin.  That in turn is layered on top of some mouse pad foam (left over from my Electric Funkatron project).  The whole thing mounts to the 'bot's case with Velcro.  It may not be clear from the pic, but the Rx is tucked immediately underneath the ESC, giving it, too, the benefit of the foam.

Safety Block

I wanted some way to mechanically keep the spinner from spinning up when not in the arena.  So a $0.71 piece of 8-32 threaded rod, two wingnuts, and a couple of holes later: voila!


Battery Power (this is the 2 steps back part...)

I goofed selecting my battery.  There is no way the ones I bought will do the job, so back to the drawing (and ordering board).

The math:

 Device
Continuous 
Current (amps)
Max Burst Current (amps) 
 Beetle B62 gearmotors (2x)*0.4 x2 = 0.8 4.3 x2 = 8.6
 EFL-200 Six-Series 2700 kV brushless motor34 42
 Rx 0.030 0.030
 Total 34.83 50.63

*Using the 12V values, which are WAY too big for the max burst current, since I'm running them at 7.4 V.

I think what happened is I spec'd my batteries for an older weapon design (yeah, that's my story and I'm stickin' to it).  The max current available to the batteries I now have is 25-A.  I probaby need something in the 36+ volt range.

Status

posted May 1, 2011, 12:39 PM by Lewis Baumstark

I weighed everything I have, which is *most* of the bot, and it came in at 2.79 pounds.

Caveats:
  • This is using "back board" (kind of a thin MDF) for the top and bottom armor plates.  When I switch to something more durable that will probably add weight.
  • Using the wrong weapon ESC, since I don't have the correct one yet.  It should be very close, however.
  • One of my support pieces is a bit big, so I'll shave a few grams there.
  • Doesn't include wiring or some screws.
A more sinister problem: I'm running out of wheel clearance.  I've had some 0.220" acrylic (i.e., lexan/plexiglass) that I'd planned to use for the top and bottom armor plates.  Two thicknesses of that, plus the 2" high supports mean I'm at 2.44" thick overall, not counting any protruding screw heads and nuts.  My wheels are 2-7/8" in diameter (2.875"), putting the body much closer to the ground than I'd like. 

I have some thinner acrylic that should work, but it seems too flimsy.  My hope is that with the internal supports it will be ok, but I won't be convinced until I see it.

1-10 of 19