Category: Stance Stance Revolution 2

Stance Stance Revolution 2 at Motorama 2024: Yes, It Did Happen!

chorl

Quick note: I re-organized how the individual project threads were categorized on the left sidebar. Yeah, “Done!” and “Not Done!” stopped being useful some time in 2009… But hey, the best time to make a change was 15 years ago and the second best time is now. Everyone’s grouped by type of build, whether it be van or robot or silly scooter. My next insurmountable multi-year sit down for an hour challenge is to update the static pages and make new ones.

Many moons ago, I finished a Stance Stance Revolution. I did in fact go to Motorama right after the fact, even if I’m years late with the event update! In fact, another entire Motorama is happening right this moment, but I’m sitting this one out due to having focused too much on Real Life and the Robot Trap House, not getting around to putting either SSR or Susquehanna Boxcar back together yet.

So, after Vantruck’s “redemption trip” in 2022, and doing a clean in-and-out with Coronavan in 2023 due to surrounding life and work obligations, it was obviously time to debut Operation IDIoc….

Nope, guess who hasn’t been above the Mason Dixon line since 2019? I decided it was Mikuvan’s turn to make the run back up to my weird culturally adopted second home (as my friends joke all the time that I’m a closeted central Pennsylvanian).

At this point in time, I hadn’t yet trusted Vantruck enough to make the long trip to Harrisburg. In retrospect? It would have been just fine. But I decided risking it at the time wasn’t worth the cascading costs – see, at least the Tail of the Dragon was within one AAA tow radius, and with Mikuvan they don’t have to send out a different truck after finding out the first one was too small despite me telling them exactly what it is.

Somewhere north of Roanoke, stopping both for lunch and to peruse the hardware store for stuff I forgot to bring, such as….

… a weapon lock for Stance Stance. I grew fond of using these big D-style hitch pins after doing it for Sadbot for Comicpalooza. I just got one big enough to thread through both discs at once. Bringing it into the store and test fitting pins one by one definitely drew a lot of conversation, and there were even some BB watchers!

The rest of the trip was completely uneventful, which is an abnormality given Mikuvan’s history of enjoying the Shenandoah Valley scenery just a bit too much. I cruised into Harrisburg around dinner time and got my usual hotel room at “whoever still has vacancy and not bugs”.

Here we go! It’s Saturday morning and everything’s passed inspection. I got to drive it in the real arena surface it’ll be on, and I confirmed that the worm drive was a surprisingly good idea. Weapon spinup was rapid, a consequence of me designing more for torque than raw speed because I’ll likely be using these things as reaction wheels often.

The only issue was the roughly 3mm of ground clearance. That was not something I could make better on the fly here, but I figured I’d use the weapons as said reaction wheels if it got hung up briefly.

The first fight was against a two-part multibot, two little TPU bricks working together named BDT. Not to be confused with Jamison’s legendary DDT. The fight is at 1:54:40 in the day 1 beetle stream.

This was an endurance fight, and SSR didn’t endure that well. It spent much of the time (as designed) bouncing around the arena on the blades, but either that or one of the hits caused a weapon motor shaft to break off.

If I had to guess, something bent just enough that the gear’s hub got caught on the disc. This was something I had thought may happen because the gear hub sits outside of the swept circle of the blade’s hub, so it could in principle hit the spokes. A lot of gear-driven weapons make sure the blade hub is a little larger than the center axis of the pinion for this reason, but I didn’t have the weight this time around to make the blade hub larger in diameter. Something else will have to give here.

The other weapon broke down because one of the motor wires got loose and was minced by the gear drive or blade. That’s just first-fight debugging finding all the ways you didn’t zip tie everything like you thought you did.

Knocking down the weapon drive system was quick, allowing me to assess the roots of the damage. I think the two-piece motor and weapon shaft mount (that bracket) might still be just a little too flexible, so maybe I can redesign it with the carbon fiber plate also covering part of the area to increase rigidity and unify things more too.

Because I didn’t want to sacrifice a spare weapon motor yet, I simply retrieved the shaft from one of the spare drive motors (of which I had more) and smashed it together again and sent it.

And send it did. SSR died a valiant warrior’s death against Mulsanne, a P1 beetle with a scorpion’s tail style spinning disc. This video is at 4:36:32 in the stream capture. This was a fight full of knocking chunks off each other until Mulsanne lined up the best possible killshot and cut through the 1mm carbon fiber baseplate and 3 out of 4 cells of the battery!

I honestly think I hit him more off-kilter and upside-down than driving on all 4 wheels. One of the weapons died about midway through, leaving me to actually take myself seriously and use the remaining disc as a reaction wheel to turn and drive for the last 30 something seconds.

Oof. What a way to go through. SSR skating around identifying as an undercutter or gyro walker for much of the time somehow made the whole affair worth it. I have enough spares to put this thing back together as-is, but am going to make some mods to the weapon designs as I mentioned earlier.

Not too sure what as of yet, since the thing is so packed full of gear that there is scant little to just remove weight from. Plus, I’m not sure how much thicker of a baseplate could have stopped that weapon anyway. It might be better to just trim some beef from the weapon hubs and go for reliability changes. Just don’t get hit, duh.

On my way back down, I decided to detour after staying a night near Roanoke again. This time, I went back through a different portion of western North Carolina to explore some roads that I wanted to visit, but because the east-of-Asheville area is a little long for a day trip, haven’t gotten to.

I do need to just spend a week wandering the whole place. I keep saying that, but am very bad at actually putting things down and doing it. Hence my recurring mention of day trips here and there.

It turns out the ass end of Route 197 isn’t paved, but no big deal. With excellent sightlines down the mountain, I just played some Rally Van Simulator.

A friend somewhere at the bottom of the mountain before I found my way back to US 74.

The ship landed some time after midnight Monday into Tuesday, upon which I neglected to unload any of the robots or gear for a solid week. No issues at all out of Mikuvan this time (which I again emphasize how unnatural this is and may indicate a credit line building for a future epic road trip explosion), except maybe the new radiator I installed the year before was too effective. I may have to check the thermostat and see if it’s stuck open some.

These events transpired right before I turned my attention to Vantruck to finish “Patch 1.1” and hit the Tail of the Dragon in April. So I spent a bunch of time after that in post project hangover mode, not touching a robot again until August in preparation for Dragon Con.

Building the Stance Stance Revolution 2

chorl

It’s about two weeks before Motorama 2024! Plenty of time to finish a robot and test it and get to know it…. right!? Theoretically, most of this bot could just appear overnight after a bit of press play and walk away. We’ll see that it wasn’t THAT easy, but it was still a super streamlined build, with machined parts being made while the 3D printables were materializing.

As alluded to in the previous post, my Ender 3 has been the workhorse for making small power transmission parts. I don’t think I’ve bought a gear or pulley in a long time, except things like extruded pinion wire I used on Susquehanna Boxcar. I equipped this Ender 3 V2, a little outdated these days but perfectly functional and smooth, with a 0.2mm nozzle so it can really hit all of the smaller details in gears down to about 32 pitch (0.8 module) or timing belt pulleys down to roughly the 3mm HTD pitch.

I’ve tried things as small as 2mm GT and MXL, but they’re not that clean. So, much of my “all in house” mechanical work focuses on the larger sizes.

This build was an opportunity to bring the Markfrogs back online after sitting decrepitly in a corner forever after the big move. This thing here is basically a classic car – it’s technically a “Mark 1.75”, being in the chassis of a Markforged Mark One but the extruders and controller of a Mark Two.

It was provided to me as the original Overhaul 2 sponsorship package (that’s 2016 if you’re keeping track) and has produced basically all the bracketry for it and Overhaul 3!

So let’s just stuff some carbon fiber in it and GO! The Eiger slicer software has evolved a lot since I last used it a bunch, so the UI took some getting used to. I was able to use the 3D representation of the fiber laying to fiddle some dimensions on this weapon motor mount. Basically I wanted more perimeter loops of carbon fiber, so I had to adjust the wall thickness of the mounting hole area to let it run some lines through.

Here’s a visual test fit of the weapon motor bracket and the drive gears. Somehow, an almost 9 year old machine that’s been sitting mostly dormant the last handful of years and got thrown in a van and then into a corner just wakes up and pounds out a carbon fiber inlaid part.

While making very crunchy bearing sounds, too – those belt tensioners are facing a desperate deferred maintenance condition that I really need to address!

Most of my nylon filament was in abject condition, of course, having picked up a ton of moisture from sitting around. Even a closed container isn’t that weathertight in the long run! So I had to bake my nylon spools (and threw the dessicant packs in for kicks too, to refresh them). I ended remembered too late that gas ovens produce water vapor as a primary emission component and ended up simply steaming my nylon for over 24 hours.

So… uhh, oops. Slight delay in the build as I hunted down a small electric convection oven on Facebook Marketplace to use as my dedicated filament roaster (and future epoxy setter, urethane curer, and so on). Obviously this meant I had to roast each spool for about the same time to get them even remotely printable again…

My supplemental order of Aliexpress worm gears also showed up around this time. The bot will ultimately use 20 tooth gears (these are 30 tooth in the image) but things were cheap enough that I just ordered a small pile. There’s a nice drawer labeled “Gears and Gear Accessories” they can live in!

I designed a few different wheels for the bot as well and tried each design out. They differ in weight as well as philosophical approach. I started with a basic “solid” wheel that could be printed using a sparser infill. This lets the wheel act a bit as a damage sponge as the infill can crush in response to being hit. I also made a basic spoked wheel which weighs a bit less. It also did take less time to print, so I decided to press on with that design for now.

At the end of the week, here’s a pile of completed parts. I decided the weird t-nut pinion gear was not going to stay in the final bot before I ship out. It was definitely too flexible. It’ll do for some basic testing, though.

I merely made a revision to the gear design which used a large hollow bore I could press an aluminum round into and drill and tap like a normal human.

The next parts to arrive were the weapon discs from SendCutSend. I went ahead and ordered a pile of them as well. 10, to be precise. I’m hoping this bot design stays around for a bit or else I’ll have some interesting yard Frisbees.

The long print job was the unibody, which took something like 28 hours on the Mark 1.75. It was the machine I trusted to just blast through the job without poking it or adjusting it!

The completed chassis. It’s not made of Onyx – I can’t afford that stuff any more, so it’s instead just generic Microcenter carbon fiber filled whatever. Definitely smells like nylon, but it makes a substantially different sound when you tap on it and is slightly more flexible. Somehow, despite me putting the most shit-tier mystery filament into it, the print quality was impeccable.

(By the way – that stuff comes pre-wetted. I don’t know if they extrude and reel it up in a tropical rainforest or what, but 10+ hours of 90 celsius or more drying is needed before even TRYING. If you’re not warping the spool, it ain’t hot enough!)

The next op was to make the disk hubs. These are just made of small chunks of 2″ round aluminum. I had remnants of a thick-walled tube that could be quickly turned down and bored out to size.

In lieu of a mill, because over the winter I sold the Benchmaster, Master of Benches and HAVEN’T BOUGHT ANOTHER MILL YET I made several drilling jigs for the parts I’d typically use a mill for. These were dimensionally tuned to tightly grab or press onto the part. I use them only for spot drilling holes, then press them back off and finish the drilling operation without them.

I didn’t want to give up the Benchmaster, Master of Benches, but I cared more about having a quill and having DROs and longer travels than its historical value. So after putting it up for sale at an obnoxious price to deter most people, I got a message from a guy who has a small collection of antique machines who was interested in restoring it as part of his vintage shop. Naturally the sale value was much more reasonable than the obnoxious go-away price. Not trying to turn a profit on it so much as maybe pick up some tooling for the Future Robot Trap Mill.

I went ahead and made three hubs, two for immediate use and one for if somehow an opponent reaches that deep into the bot. By that point I’m guessing the rest of it is hosed already, but…. good to have a backup nonetheless.

The disks interface to the hub using these four dowel pins pressed into the flange.

For the weapon shaft mounting blocks, I did the same get-jiggy-with-it operation on a chunk of 1″ wide, 1/2″ thick aluminum bar. This jig manufactures two blocks. It contains the drilling pattern as well as two marks for where to saw it in half.

I made the driveshafts as planned – just turning a shoulder, snap ring groove, and partial shoulder onto a 3/8″ aluminum hex bar. The small shoulder is for the bottom shaft bearing, and the big bearing sits on the half-10mm-half-hex portion. The gears were drilled out for the 10mm bore, and I added two little dimples to let their set screws get a more positive grip.

I think if I were to do this over again, I’d just replace the screws entirely with a single 2-3mm roll pin or something and have less things that can wiggle loose.

It’s starting to look like some kind of weird clockwork mechanism now. The future double-ended motor shaft is made from lengths of 304 stainless steel which can be picked up in 3mm undersize precision ground fit. No need for the case-hardened linear bearing grade stuff here since the forces are relatively low!

The shafts are cut to slightly shorter than the total length between the chassis pockets that house its support bearings. Then I just move the worm gears to where I want them and marked locations to put small flats (using a Dremel) so the set screw on each can grip. This whole assembly was already so smooth that the gears could somehow backdrive if I turned them with some effort!

Motors prepped and being installed with the new shafts. This just involved pressing their stock ones out and maneuvering the new one in place.

The next item to arrive was the side plates from CNC Madness. This operation is hands down the cheapest way to get composite flat parts made quickly right now, and I leveraged their 3mm and 2mm carbon fiber for the weapon uprights. This took just under a week from when I hit the Order button.

All I needed to do was add the countersink holes and the bot was suddenly together! I chose to use large #10 thread-forming screws for plastic (“Plastites”, if you will) instead of heat-set or coarse tapping inserts. Keeps the weight down and gets the point across for the strength of the material I’m fastening into!

The wheels are retained by these wide-face E-clips. No screws here, everything is made of laziness.

After the worm gears were firmly secured by the side plates and lubed up, I found out they in fact could backdrive quite readily. I attribute this to floating the shafts on ball bearings so neither radial nor thrust force caused a bunch of extra friction!

Weapon assemblies undergoing final assembly! The ring gear is attached to the hub using eight countersunk screws, with the aluminum hub flange being threaded to accommodate. Overall, it’s quite a compact package. As I said before, the only thing I’d like to change here is backing down from the gigantic 6200 type bearings I already owned. At least this thing will never fail due to that!

The bot’s entered the “difficult to look at from any angle” stage of assembly here, and was mechanically completed the weekend before I left for Motorama.

The electronics of this thing were done dirty. The 660mAh 4S battery just sits in the cavity in the center designed for its size range and the bottom plate holds it all in. The ESCs are leftover SimonK AfroESC 12 amp units from forever ago, when I bought a few dozen and have slowly since used or given them away in various ways.

The receiver is just a shucked Hobbyking TR6Av2 (or FlySky FSR6B) receiver that I’ve used for… oh, I’m not sure, 13-14 years now? They’re somehow still around and being sold, and my transmitters still work! Odd to put it like that, because it would be equivalent to using a 1992 model radio in 2006 when I first could afford/switched to then-new Spektrum 2.4Ghz tech.

The power switch is a random Fingertech switch I’ve had forever. Everything’s just hot glued in place and point-to-point wired. It’ll basically be impossible to service!

As a last touch and nod to its cultural inspiration, I made some cringy heart rims for Stance Stance. I ended up only bringing these and the 5-spoke wheels to Motorama for weight issues.

Check out the finished product! I’m very proud of how this bot turned out. It looks very sharp and clean, in my opinion. I decided to not paint it in its usual magenta and blue for Motorama since I didn’t want to take everything again, but will do so for future events!

Here’s a video of the worm gear drive! At this point, the left side still had a bit of a sticky spot which I never quite figured out, but it went away with a lot of driving.

The handle is extremely linear and predictable, since the worm gears basically act as an amplifier for motor drag and braking. I was definitely satisfied with how it drove, but only the event can show how durable and reliable it is.

Why was I such a stickler for weight? Because this is what the thing ended up at~ That’s 0.4 ounces under the limit. It’s just so full of gear. Figurative AND literally! I’m actually not sure where it can get any lighter. Perhaps the weapon hub bearings will be a good start, because it’s very much two hard shells and a very creamy, soft center. I’d like to try and transfer some weight into protecting the front, rear, and bottom some more.

The rest of the week before I left for Motorama was spent putting some more of the spare parts together and some test driving dancing in the driveway. Altogether, Stance Stance Revolution 2 was a pretty clean and enjoyable build that needed some design thought to execute, but it wasn’t like I had to devote all four of my remaining brain cells to designing a mechanism or making some obnoxiously complex part. That…. is for a different build coming soon.