Skuld V3

Version 2 was fine for driving around my makerspace, but the shoving power = 0, the timing belts are annoying to maintain, and I just wasn’t happy with it. SO! 3 weeks until Motorama YOLO let’s do this!

  • Go 4WD so I don’t have to worry about shoving matches

  • Keep using the brushless rage babies but up them to 4x

  • Non-polycarb top plate

I iterated so much on this design that I ended up with a half-done version 4 when I went to Motorama.

New color, new wheels and a stupid amount of hammer weight

New color, new wheels and a stupid amount of hammer weight

The picture above is as after an all-nighter with my pal Ryan (part of Valkyrie this year). He helped me with some NOT best practices wiring (that we both agreed on, and a new issue for V4). We left for competition the next day and had some last minute changes that needed to be made, like, OH HEY! The hammer only works in one direction, not the other. Doublechecking the design:

Consider the forces at play for the positions of a hammer. First, the forces and torque at play;

Borrowed from my friends,    REV Robotics

Borrowed from my friends, REV Robotics

P = power; D = diameter; T = torque

So in the case of a hammer robot, it can be considered as T(sprocket) > T(inertial force of the hammer)

Hammer in the back position

Hammer in the back position

This is the starting position for the hammer. We have a big sprocket and a small sprocket; the small sprocket is attached to a Aerodrive 6374 - 149kV brushless motor (I figured if its strong enough to be retrofitted onto my old intern’s longboard, it was definitely more than needed for my 30lb robot… and i had a few lying around).

CodeCogsEqn.gif

The first hammer I used was a plushie head that Amazon says weighs 57 grams, so, the arm is all that the motor has to drive (assuming negligible weight seeing its attached directly to the sprocket). I tried to go big with an ~8lb aluminum brick.

A hammer for my hammer

A hammer for my hammer

When I tested it, it worked in one direction and not the other. I adjusted my settings to 100% power on my transmitter and the hammer still couldn’t swing back. (I learned later from Xo that the settings for the 12-fet rage I was using doesn’t do the same forward as it does backwards, with the forwards being the stronger of the two.) Luckily, I had brought some spare 1/2” aluminum to quickly smash together a new hammer.

It almost looks like I intentionally had a hammer setup like this

It almost looks like I intentionally had a hammer setup like this

While I scrambled to get everything sorted for competition, I forgot to switch out my batteries, so! First match, I ran out of juice before the three minutes were up. Second fight I won, but not particularly fairly (the team started smoking when I shoved them). Third match I lost to one of the event organizers (and Blacksmith teammate) James. Going into a fight knowing you’re going to lose makes it way more fun. I learned why folks don’t use stainless as armor.

oops. Stainless steel: a pain to work with, crazy easy to bend in combat

oops. Stainless steel: a pain to work with, crazy easy to bend in combat

My top plate got pretty dented and then it also acted like a brake pad on the weapon outrunner so I lost my hammer a minute or two in. I also lost a resistor on one of the drive controllers as well as one of my Markforged corner blocks (I’ve been lazy and not wanting to drill/tap my wall ends.. but no longer). I know the stainless was a bit thin going into it but I didn’t have any other material lying around so I figured it’d be worth the try.


Next version will have a wedge, better armor, a different hammer head, more space for smarter wiring decisions and likely another color change.

Skuld V2

(Written before the 2019 Battlebots season…) So, seeing first gen Skuld was a nice, toasty embarassment fire in Atlanta back in September, I had motivation to not procrastinate version 2. Immediately I went back to the drawing board to rectify things before Motorama 2019 snuck up on me. Goals for this next version:

  • Make a real weapon or else I’m gonna die real hard (Tako-san will likely still tag along as a spare option at Moto)

  • Get rid of those China super cheap drone ESCs

I added a bit of thickness to the arm because I can just see the head snapping off. For the prettyifying, I think I’ll keep the rails cranberry but I haven’t decided about the top plate. All black robots are a bit too common but most actual colors look silly. Maybe I should Millenial Pink the top if Home Depot still has the paint…? Similarly, I could see myself doing a less yellow-y bronze/copper. No one takes their aesthetics seriously (well, almost no one—- Alex also cares).

IMG_3192.jpg

My good friend, Charles, talked me into his (pricey) 6 FET Brushless Rage Bridges for my drivetrain and a 12 FET for the weapon. After some testing, I think this isn’t the version that will be going to Motorama (or anywhere outside of my makerspace for that matter). Stay tuned for a very too-late-scheduled V3 >.>

Skuld V1

With DragonCon just over a month away and Battlebots far enough in the past, its time to get my CAD hustle back on and build a new robot. I decided to go the slightly expensive (but not impossible) 30-lb route. A sportsman 30, to be exact. I figure there are more competitions a year (DragonCon, Franklin, Motorama, silly Mass Destruction demos, etc) to really hone in on a design versus the competition or two max a year the 12-lb class are (so RIP, for now, Tough Love, you unbroken heart, you).

The goals I had with Skuld were the following:

  • Produce a hammerbot so you can understand the challenges and limits (Blacksmith is one of my favorite robots in Battlebots, to date).

  • Design a drivetrain that uses belts. In all of my many years of roboting and automotive design, I have never played with timing belts

  • Make an adorable hammer head from DragonCon seeing its a shoving match. I’m out to win the hearts of the fans, not the competition. Competitive robots tend to look boring.

  • See what all of these crazy kids are so hyped on brushless drive for.

IMG_2536.JPG

Version 1 was finished after a week of all-nighters between my incubator and makerspace in Somerville, MA. I used 1/4” and 1/2” aluminum for the frame, base plate, and hammer module structure. I found I had a few challenges I wasn’t expecting, besides the unfamiliarity of developing a chain-not-for-drive system that had physical limits and seeing how belts want to behave/tension/wrap.

I decided to go with a brushless setup across the drivetrain and weapon motor for a few reasons:

  • I’ve been living the brushless life since high school

  • I honestly don’t know what controllers to get for a 30-60lb robot

  • Relying on a physical end stop for my hammer module means I want a forgiving motor if slammed to a stop

I’m grateful I have access to some of the waterjets around Cambridge or else this wouldn’t have been done on time. I likely also wouldn’t have been able to finish if I didn’t have so many Markforged resources nearby as well (thanks again for being our Battlebots sponsor…!). The cost of this robot was noticeable, but WAY cheaper than it should have been due to local resources and stubbornly making everything myself.

Night before Dragon Con, everything looked good to go! I burned out one of my Hobbyking Red Brick 100A ESCs, but had one 200A lying around that I swapped it out with. The gears in my gearbox were too tightly packed because i was off by ~0.7mm of clearance between levels and my inrunner. I also used WAY TOO POWERFUL of an inrunner for how I had my SimonK settings on my ESC. They always say the easiest way to learn is the painful way… especially as an engineer.

IMG_2613.JPG

Competition day! And first fight… that remaining Red Brick 100A ESC burst into flames when I was shoving the kiddo in lime green behind me (he took a picture with me after… I yucked it up. Kid’s first fight and I gave him a story!). When I realized the drive was acting pesky, I tried to just continue forward throttle and waggled the hammer around for good measure and then… fire. The First at DragonCon in years! Thank you Randy (one of the many wonderful folk on CHAOS Corps) for hustling Skuld into the street. I still have the toasty bits as a reminder to ask for more help/understanding when flashing/doing anything with controllers that I’m not confident in.

Rainbow Brite V1

This was my first ever <120lb robot I've ever built. I've never done the whole RC Airplane/Drone thing so setting up a receiver and baby motor controllers (without firmware) was a first for me. I asked a friend for help and he ended up just creating a from-scratch chaotic board with an Adafruit Teensy.

I knew my v.1 of whatever I'd make would be stupid (and I had only a week or two to will this thing into being while traveling a fair amount for work) so I set a bunch of goals to achieve:

  • Must have all electronics easily accessible for tweaking and maintenance
  • Must fit the dimensions of this board (roughly 51mm x 81.5mm x ???)
  • 3D printed! because car bills were pricey, time is limited and I just got a Mark Two at work (including an Onyx spinner because I want to see what this carbon fibery material looks like if/when it shatters
  • Underweight so revisions and additions can be made in the future
  • Mad bling LEDs mounted on it so it can have character and flare

So after hours of CADing and sorting out the minimum possible McMaster order I could throw in, I ended up with a good enough CAD.

She's mostly black onyx/Makerbot PLA due to time...

She's mostly black onyx/Makerbot PLA due to time...

Being the A+ procrastinator I can be with personal projects, I barely had time to knock this almost fully printed robot out. I assembled everything with dowel pins, plastite screws and a few M2.5s. The motor belting I made out of o-ring cord stock and super glue (it works pretty well... especially with cure time). I decided my wheels had to be ridiculous and cute because I wanted to try out Markforged's nylon dyeing tutorial after printing the STL in nylon on the Mark Two.

At least fuchsia is one color of the rainbow...

At least fuchsia is one color of the rainbow...

If you like these wheels, you can print them by downloading the STL. I lost track of time so my dyeing endeavors ended up waaaay more concentrated than I was planning.

like hair dye, if you leave it for too long, you get a way, way darker shade than planned... only this won't wash out with cheap shampoo.

like hair dye, if you leave it for too long, you get a way, way darker shade than planned... only this won't wash out with cheap shampoo.

For someone wandering into electronics land, the board my friend built for me couldn't be more confusing to follow or grasp what in the hell is going on.

even now, knowing what you need for motor and battery link-up's, this is still a little hard to follow...

even now, knowing what you need for motor and battery link-up's, this is still a little hard to follow...

Once I attached whips to my motors, I was able to lay out all of my electronic guts before setting them into my 3D printed beast.

my friend's  TeensyBot code  for Rainbow Brite--- loads of Neopixel iterations that attach to one of the dials on my Spektrum transmitter. She can have moods...!

my friend's TeensyBot code for Rainbow Brite--- loads of Neopixel iterations that attach to one of the dials on my Spektrum transmitter. She can have moods...!

Next was setting the parts into the robot...

just need my motor/board mount to set all of the electronics in.

just need my motor/board mount to set all of the electronics in.

Overall, Rainbow Brite is easy to assembly and I was really lucky my two matches were against even more derpy of robots than mine so I took 0 damage.

crazy hair electrical wires/exposed electronics = not the best idea...

crazy hair electrical wires/exposed electronics = not the best idea...

Things I learned from this build:

  • Overestimate your dowel pin lengths for wheels/spinner bars
  • Markforged printers take FOREVER to print (worth it)
  • Superglue is going to work on small components better than set screws in plastic
  • Makerbot PLA does not make good spinner pulleys (the material started to melt)
  • wheels should be mounted onto the chassis base-- I was going to do this but the increased print time for the onyx was going to make me not have a robot in time
  • Always have UHMW on hand kicking around in 1/16" to add a bumper on the front for smooth driving (added day of)
  • Set your motors to be opposite directions when soldering leads onto them and then LABEL them. Otherwise you'll just drive in circles.

So with that, I was able to compete in Mass Destruction 5 at Artisan's Asylum. My first insect weight (weighing in at 1.7lbs for the 3lb class... significantly less than I planned). For now I won't be revising Rainbow Brite but I'm sure I'll get around to it at some point.