2.5 Years of Ringspinner Combat Robot Development
ฝัง
- เผยแพร่เมื่อ 30 ก.ย. 2024
- This video documents my progress in developing a competative Antweight Ringspinner combat robot, discovering new technologies and improving my engineering skills.
Covering the progress from Mk 1, my first ever combat robot, to Mk 3, a unique and powerful walking machine. Ringspinners are a rare archetype due to their complexity but a clever application of Hubmotor technology can help alleviate that. Custom machined parts using titanium, carbon fiber and aircraft alloys, the final product is quite a piece of kit for only weighing 225g.
If you are interested in combat robotics, there may be a club near you! In australia, many are active on facebook. Feel free to reach out.
Music: Keygen Church from Ultrakill
Awesome to see the rapid evolution into something totally unique, featherweight version when? 😂
With Brisbane coming up, it's damn tempting....
This deserves way more views, amazong look at the whole development process.
Agreed, if it can be refined it might be the way for ring driving.
Would love to see the CAD.
Algorithm algorithm algorithm. Algorithm algorithm?
Having almost no idea what actually goes into this stuff, and not even really close to my lane of enjoyment, I stumbled across this channel via MakersMuse and as you were describing the robot, I thought 'wow, that seems dangerous AF, lets see it!' Then you show the robot, and it looks like an angry coaster, and my concerns drop signifigantly; thinking to myself "that little thing cant be dangerous."
But watching it compete with other robots, and seeing the way chunks are flying off, the fully enclosed plexiglas arena, and even seeing these tiny blades gouging chunks out of the arena- it really humbled me at the level of engineering and work that goes into these tiny things. Toys they are not.
You've got a sub from me, my good man, and I cant wait to peruse your other videos.
that bot went from "oh its a normal bot" to "why do I hear boss music?"
Ring spinners aren't a hobby, they are an obsession. I've been working on mine (US 454g 3d printed) since 2018. I have been banging my head against different shuffle designs for 18 months trying to find the right formula, but while working on v10.3, I have already started working on v11, which I intend to be a true walking bot :) I've considered the stator ring concept, but I don't think it would stand up well with just a 3d printed ring and no metal to constrain it... it would be hell on magnets.
Keep up the awesome work.
I looked at your channel, you should upload some of these, it would be really cool.
Love it! I proposed this exact same concept (antweight ring spinner with all the electronics placed inside a custom frameless brushless motor) on the robotwars101 forums way back in 2016, so it's great to finally see someone with the skill to actually build one of these, taking up the challenge!
Surprised you eventually found this, I recall viewing it at some point when researching ring spinners and may have been a subconscious push to make this as it is now. Only just remembered!
Wow, you could almost use the spinning disk itself for motion instead of legs. Controlling the balance and then learning to drive it would be tricky.
TH-cam recommending me a small creator making great quality content about a topic I love?
Did I die?
Is this what a miracle feels like?
ULTRAKILL ost did not go unnoticed... Awesome bot too.
I don't really follow combat robotics personally, so this was super interesting! The idea of turning the ring into the motor itself was very cool.
Also loved seeing Pancake show up and going "hey, I know that one!"
You definitely should! NHRL have some good stuff on their channel. Imo the smaller bots produce much more devastating fights than the heavyweights
Is that... the Ultrakill soundtrack I hear?!??!?! Beautiful choice.
@IHaveYourAttention it was the only fitting choice
Huh, the gyroscopic force put up by the ring is a boon to the bot's stability. Once it's in motion of course, the pit dodges are so fun to watch.
Great video! 👍 Hopefully, once you've cracked that concept in the antweight class, maybe we could see it in heavier categories. I can just picture you entering a beetleweight version of Revolve at something like NHRL.
saw your FIRST shirt, as a current frc team member myself looking into combat robotics during the offseason, what would you say is the biggest thing to learn for combat robotics, and the biggest thing from FIRST to forget about when doing combat robotics?
Glad to see I'm not the only one doing that
I imagined using three pneumatic tires with separate motors for redundancy, in a horizontal equilateral arrangement driving a grooved/channeled ring internally, with the idea some impact would be absorbed before it gets to the chassis.
That bot is great, amazing battle! Keep up superb work mate!
Love your different design approaches.
Did You consider symmetrical placement of coils like three in front and three in the back? It is possible that this configuration can lead to more symmetrical forces in the ring - maybe even easier start of rotation.
Awesome to see the behind the scenes, the evolution of the custom design and the outcomes/assesment/remedies!
What FIRST team were you apart of?
FRC7561, later FRC5985. Also a bunch of FLL and FTC teams
Fascinating! That was fun to watch
Beautiful build, love the combination of faults and fixes. Definitely hope to see more of your builds.
Clearly Mr. Bean is a very good engineer. But seriously, this is an amazing video (and I see you do FIRST stuff which is really cool, I'm doing FRC rn)
Fantastic breakdown! Idk if I'm ever gonna have enough time to pick up another hobby, but combat robotics are honestly tempting, haha!
Great video. Ringspinners, not easy.
Which brand and model is your mill? Is it a CNC mill? Thank you
Yes CNC, is a base model Tormach 770M
Re: walking mechanism... I'm printing a pla tribute of revolve, still trying to get the walking mechanism together. Are you just using steal pins for the 3 holes around the motor shaft/idle pins? How did this hold up through fights? You didn't mention any issues with the walking mechanism, but are there any changes you would make if designing this again? In one of your test shots, before the ring was on it looked like you had a weight on top, was this needed for proper grip, or were you just making sure it would walk at full weight? My pipe dream is to take the weapon from subdivide and slap it on a modified revolve frame. Any suggestions?
The pins were cut paperclips. Worked fine, didn't fail in a fight and we're just superglued in. Weight was to check it worked with a full robot on it. If you have the space/weight, driving the second cam with gears from the drive motor will reduce binding.
@BrokenLinkRobotics thanks as always for the reply. I ended up using some 18ga Brad nails. They didn't want to fit, but some convincing with a soldering iron and a small hammer changed their mind. They don't work great, but good enough to learn from. I have learned more from printing and tweaking your step files than 3 years of TH-cam alone. I cannot thank you enough and look forward to your next video... and playing with your 3 wheeled omni design which I will be stealing fr... errr... learning from next...
With the way the robot is able to skip and roll around just with the outer ring.... what about removing the drive system entirely and just relying on the ring for movement? With some fixed anti-skid pads on the top and bottom of the robot, a tire on the top and bottom surfaces of the ring, and some little solenoids that can intentionally kick the robot up at an angle to put the tire in contact with the floor.
Could do, some similar gyro bots exist, but the main problem would then be loosing the weight bonus gained by the walking mechanism. Just about cant gain enough weight back to make it worthwhile I think.
@@BrokenLinkRobotics Is there any other sort of weight bonus available for any other unconventional propulsion system? I'm not familiar with the ruleset, but I'd assume the whole reason for the walking bonus is as a way to promote more diverse propulsion systems i.e. by offsetting the extra costs that "other than basic wheels" systems would incur, so maybe there's some way of classifying a system like this to get some other bonus?
@@AJMansfield1 US rules sometimes give a weight bonus for 'No wheels', with 50% for shufflers, but in australia we've come to the concensus of it being a spirit of the rule thing. The weight bonus is for the walking legs, here the weight bonus would go to unique locomotion yes but that locomotion is still fudamentally the weapon. Could argue the solenoids are the walking mech but they aren't 2DOF legs or shufflers so dont fit the leg definition either. Tricky rule lawyering one
So do you ever measure the watts your drive can put out? Max voltage? Effective weight of the drive pieces?
About 3W per drive motor, and at 8.4 max volts. Roughly 10g per shuffle pod not including motor
Oh I'm sorry, I meant the hub drive on the ring. But that is interesting too.
Always love your detailed facebook posts, thank you for the detailed youtube post!
I knew I recognized master boot record for the music.
So when are you going to upgrade to a titanium spinner?
Tough to make weight with Ti and a nightmare to machine, but it could solve the creep. Will consider it.
i honestly love the idea of ring spinners and full body spinners in general. like having your biggest defence literally hit back is just a cool concept. they're pretty impractical as far as im aware since they take a while to spin up and have a lot less actual armour, but that doesnt them from being cool
Did you... make a robot Beyblade?
The most powerful of all the toys? Even Moses split the sea with a Beyblade you know!
Omg your hand at 3:00 😬
It's kind of like watching a beyblade with this weapon choice and in this size category.
this looks interesting, but my degenerated mind can only see Mr beans son being exited about fighting bots. XD
Do what they did at tesla (carbon fibre wrap around the ring).
just a thought, i don't know if it is possible. But perhaps you could create down force with the spinner in order to keep you grounded and add more standing weight to the blow when you hit somebody. Perhaps just setting the teeth at an angle would be enough to create the downforce. Interesting video, keep them comming!
many people have used that
Definately worth considering. Adding 'weight' has clear benefits, the problem I could see is ensuring that the added downforce multiplied by the friction with the ground is less than the increased apparent spinup torque generated by the aero device. Need to invest in some fluid simulation software now too...
@@BrokenLinkRobotics how about instead of using fluid dynamics software you could perhaps simply 3d print a couple of test rings with the blades at interval angles. By placing the robot on a weightscale you can see how much downforce the ring is creating. I could never build something as your spinner robot so i'm totally in awe and i hope my suggestion is usefull. Thank you for your time making the videos and sharing your robot journey with us.
I wonder where blendo is right now
Your pain is our gain
Adult bayblade lol
YAYYY TEBEBRE ROSSO SANGUE
🇧🇷
🇦🇺
awesomeness
Enjoyed the video! Fyi, your audio is really quiet.
Fixed in recent videos
Bro, where do you find the documents for the motor winding?
Just googled about 'BLDC Motor Winding' with key search terms if Wye, Delta and DLRK.
Awesome vid, I was certainly curious about how you pulled it off.
Revolve vs william wagtail was definitely one of my favourites from the event.
It sure is an engineering challenge! It's much easier to spin the entire robot, with two side-by-side motors going opposite directions. But then ... how do you actually move rather than spinning in place?
I think you can do it with a just a single tilted omni-wheel near the center.
First off, imagine a wheel that has the same radius as the distance from the center of mass. If you drive this wheel, the robot will spin in place, with one wheel rotation equal to one robot rotation. So far so good, but you're just spinning in place.
Now make it an omni-wheel by placing four rollers around it. This doesn't have perfect 360 coverage, but that's okay. Anyway, it's still just spinning in place, but at least the rollers now allow the robot to slide around some.
Now the secret sauce - tilt the omni-wheel so it's not properly perpendicular to the motor axle. This means that one roller is closer to the robot's center of mass; the opposite roller is further from the robot's center of mass. Thus, one roller will be moving too fast compared to the robot's spin; the opposite roller will be moving too slow compared to the robot's spin.
This produces motion while spinning, rather than just spinning in place. You can show which direction the robot is "pointed" in by coloring one of the side rollers white. As the robot spins around, the white will flash visible to show the direction of motion.
This "tilted omniwheel spinning robot" only requires one motor and one tilted omniwheel. As a bonus, the design is flippable.
I think two motors can make the robot a lot more nimble, as well as offering redundancy. Four won't fit. Three does fit, but it's a somewhat tight fit so it's probably not worth the effort.
Okay, this spinning robot can be simplified.
So, imagine a motor driving a spinning cylinder, like a smooth F1 tire. You attach two rubber feet to the outer surface, one closer to the motor and one further from the motor. It doesn't need to be an omniwheel. Just a cylindrical wheel with two rubber feet.
The point is, the diameter of the wheel is roughly the distance from the center of mass. One rubber foot is closer to the center of mass. The other rubber foot is further from the center of mass.
It's just one motor and one wheel, but it lets your robot move while spinning.
Okay, the spinning feet robot can be controlled with simple single channel R/C.
Again, the construction is just a circular robot with a single motor and wheel. The wheel is near the center, offset from the center of mass by the wheel radius. It has two rubber feet opposite each other, with one foot offset closer to the motor, and the other foot offset further from the motor.
You first spin up the robot with full forward (or full reverse). Then, to move you reduce throttle until the wheel rotation matches the robot spin. This will make the robot move (while spinning). At this point you increase throttle to steer anti-spinward and decrease throttle to steer spinward.
Quintessentially, you're adjusting from a steady state where the wheel rotation matches the robot spin. If you make the wheel rotate a bit too fast, the next footstep will arrive a bit too early - thus it "turns" a bit opposite the spin. If you make the wheel rotate a bit too slow, the next footstep will arrive a bit too late. Thus it "turns" a bit spinward.
To help see which direction the robot is "pointed" in, paint a white spot 90 degrees from the feet. This white spot will flash visible as the robot spins. This will be particularly important after hitting the enemy or a wall, because after that the robot could be "pointed" any direction.
This scheme keeps the construction and control simple and robust, although there is no redundancy.
Oh well, I'm ignorant. Looking around for spinning robots, I saw "meltybrain" robots, which use two side-by-side wheels and nevertheless manage to control drift rather than just spinning in place.
I still think my spinning feet idea could do better, and possibly be more robust against battle damage. It doesn't use finicky sensors and software to pulse the motor cyclically. It just uses smooth motor rotation and achieves the cyclic foot pushes with geometry.
Also, unlike a meltybrain, the spinning feet don't need to fight wheel friction to drift sideways to the wheel motion. The rubber feet simply do not touch the floor except when their motion is aligned with the robot's desired motion.
Cool design now when are you going to scale of up to the level of a heavyweight? 😬😂
Super cool! Never thought making your own motor was doable.
I love this design, I built a Feather Weight ring spinner in Robot Rumble 2.0 inspired by it.
Thanks for the video really interesting
Really enjoyed hearing about the development journey!
Amazing! Been looking at your designs for a while now
Saw the FRC shirt and knew this was gonna be good
Wonderful presentation!
Keep Going Bro !!!!!!
Great presentation!
Great engineering
more
yes
Could you design the inside of your robot to be like a version of Ripperoni where you can have a counter gyro system which would then make your robot more controllable at the expense of increased electrical demand and internal complexity?
I doubt that would work for a ring spinner.
@@Lumakid100 ah OK just a thought.
awesome insight into some amazing engineering
Seeing it spin and lift itself could you print an air capture cone set below the bot? You might have some torque troubles with the shufflers because of it but it could be a neat thing to check