Hi from the AM32 team! 😊 Thanks a ton for the shoutout! We’re really excited to see AM32 being used in fighting robots-it's awesome to see our work in action! Keep those bots spinning strong, and reach out anytime if you want to swap tips with other robot builders. 💥🤖 Best of luck in the arena!
This is the second use I've heard of for AM32... My first time was when I went to pick up my desktop CNC mill... He uses it for the spindle control. Using a 500W brushless motor, it gives great results! 😁
Hey chat, if you're doing anything combat robot related, please do it inside a test box. I'm sure that Angus has taken the proper precautions here, but the kinetic energy stored in these types of weapons is equivalent or higher to common pistol bullets, and if something goes wrong drywall or an interior (hollow) door will not save you.
I think it starts getting to that energy in the 3lb weight class, but i'm sure even a typical 150 gram weight class spinner could make short work of human flesh. Yes, always use a test box.
@@pedro.alcatra No, that is absolutely bad advice. The gyroscopic forces of a spinning weapon can make it behave completely unpredictable. In the blink of an eye a stationary clamped bot with the weapon spinning can flip around and launch a piece of debris at any angle which can bounce off any surface with enough energy to do some serious damage. If you test a spinning weapon of any weight class, you either need distance (so that ricochets lose most of their energy) and a shield or have it completely enclosed in bullet proof materials. Trying to predict where shrapnel will fly is a fools game.
With Solidworks, if you select the center of mass and the center of the hub ring, you can lock them to coincide, then the program will automatically adjust the unlocked components of the drawing to keep them on top of each other. Then you can drag them around to whatever shape and position you want, locking the parts you like as you go.
@@DanielCook-h6r it's still a few grand for the most basic package. SolidWorks is a few grand for the most extensive package with generative design, FEA and CFD simulations. It's a good option for hobby use, and basic 2.5D designs, but not for a complete design suite. It has weird constraints and often undefinable (want a plane 45° to another plane? Better hope the default angle offset is what you want). It also doesn't allow for backwards constraint changes. ALSO there's certain angles you just can't rotate to, that one's super annoying
Tip for a perfectly balanced weapon blade design: 1. Create a roughly balanced blade design without your center mounting cutout(s). 2. Display its center-of-mass. 3. Create a side view sketch, zoom all the way in on the center-of-mass, and drop a point right on it. 4. Center your mounting cutout(s) on the sketch point. Much easier than trying to tweak your sketch in tiny increments, and you can get 'perfect' balance this way (not that the manufactured blade will be perfectly balanced, especially after some combat, but at least the design is this way with little fuss).
Angus could get a bigger "bite" with his tooth if he were to reduce the counterweight size by inserting a tungsten slug, like we do with automotive crankshafts.
@@jimurrata6785 That would definiteley work, but then center of mass calculations become very hard and tolerances would mean that it will not be balanced after assembly. Having to then fix the center of mass by grindig away material requires a balancing tool and a lot of patience. Thats all to say that this idea, while great in theory, would make this design a lot more difficult. Maybe when the time comes for V2 and hee needs things to improve this could be among them.
@nicholasweiss4662 I'm really surprised there's not a way to do it in software, given we know the density of tungsten or Mallory Metal. But the rpms are pretty crazy and impact will go up in square to velocity
@@jimurrata6785 In practice, he probably has plenty of potential bite from the blade geometry.. it's usually more a matter of relative closing velocity to the other bot vs weapon RPM / number of cutting 'teeth' the blade has.
@@dogjutsufpv3813 "it's usually more a matter of relative closing velocity to the other bot vs weapon RPM / number of cutting 'teeth' the blade has.". I get what you're saying, but I disagree. You want to hit with the teeth at maximum depth, so your closing speed must be enough to travel the tooth depth in full. The fewer teeth on the blade, the more time you get per revolution to reach maximum depth for the impact. Of course the larger the tooth, the larger the requirement of closing speed.
I think the most valuable bit of Information in this Video was the thing about Bambu Lab now stocking Maker related stuff like Bearings, Motors, Slip Rings, etc... one might be tempted to buy in a pinch when already buying Filaments from them - *_BUT_* - Bruh... Those prices for a *_single_* Bearing / Micromotor... I'm gonna stick with AliE for my Maker Needs cause for maybe a little bit more I'm usually getting on 5-10x the amount of Bearings / Micromotors on AliE.
@@Jehty_ "Jack of all Trades, Master of None" is an old adage that tends to hold true. When a company focuses on what they do best, they can be the best. Comes with the risk of one day being obsolete though. A good example to watch for: A lot AC companies also do plumbing (and vice versa) but they ultimately end up being good at neither. The best AC companies focus exclusively on AC/Heating systems. And the best plumbers focus exclusively on plumbing. The very worst try to do both.
13:37 seems like it slides around a bit between off and max speed. like the rotor is being pushed off of the stator from the intensity of the magnetic field. Because it slides right back once it's dialed down.
@@1992jamoi think in such a small application it would be better to just size the shaft for a press fit with the inner race of those two ball bearings so that they can't slide around on the stationary shaft. those bearings aren't really meant to take axial loads but it's only a small amount of force pushing the rotor around side to side.
This looks like ABSOLUTE FUN. As a 5-year manual lathe machinist... a bit of sand paper goes a LONG way.... hand polishing shafts and bores to +/- .0002" can be done... I had to do it all the time running manual lathes. The next step is understanding proper fit tolerances. Just a few tricks for you to hide up there in your noggin for the future. Maybe if i start to find some extra time in my future the ant weight battle bots sound like something I would love to do. Thank you for the video and the Today I Learned moment for me.
Cheers! Yeah my downfall was trying to drill out the 8mm bearing holes with,,, a drill bit. Obviously it wasn't going to precise but I was hoping epoxy would let me get away with it. A reamer would solve the issue, but costs more than the part from PCBWay haha. There's robot events all the time all over the world! Robotcombatevents . com is a great place to look.
was recommended this rather randomly... figured it was a military combat robot, and figured that the bearing in the thumbnail would be an amazing design I could incorperate into a project... not sure what tbh... have very few at the moment. Most exotic project atm is making a salt water battery that can be made in the event of societal collapse, just cause it would be cool, while the least out there one would be a desk I am rather proud of and nearly finished... did cut my hand open a bit this morning, shaving part of the legs (uneven wooden lengths that jut out due to cutting it in small passes, flipping, and repeating); only had a pocket knife and it slipped on a knot in the wood. Rant over. Hope it boosts engagement enough to recommend to more people lol. Super cool video. Love the little crafts people make.
Looking awesome as ever Angus. In my country, RC toys like drones are really strongly regulated. This sadly also results in RC hobby in general being only a small hobby and most people only drive RC cars and boats, but the amount of hobbyists is steadily decreasing. Most RC shops have closed both online and offline. It's a damn shame. Maybe I'll make my own bot one day but I won't have people to fight in my area! It's still a dream for me, ever since I started watching Robot Wars when I was a kid on the BBC in the 90's. Man I loved Craig Charles's enthusiasm!
The rotor may overheat in this design, consider adding "ram air" holes through the rotor/striker. The top of an out-runner motor is perforated to make a crude cetripital fan, to cool the windings and magnets. The striker will cool the magnets, but without good airflow the motor windings can overheat and the enamel fails. This can cause shorts and fry the ESC.
The fights last for 3min max so it's unlikely to saturate in that time honestly, but tiny angled holes might channel air and also make a pretty cool whistle 🤔
given the RPM and the size of both the motor and holes I don't think we can hear the whistle (at least after winding up). but maybe dogs or bats could ;-)
@@dfischer000 Partially obstructing the air entrainment intake with a rubber flange tapered to the apogee, or something like that...basically, a rubber flap, and as the air cycles around, it'll sound like a comical diarrhea or long fart noise distinct and audible enough to pick out. "Which one is yours?" "The one that sounds like it ate a lot of Taco Bell and clobbering the other one to death to use the toilet." "Ohhh, okay."
Maybe add an internal groove for free rolling ball bearings, so if the blade has a bit of material removed they can compensate for the weight difference
I was trying to thinm of a solution to that. Any damage to the claw will result in unbalancing of the motor. I will have to exore the ball bearing route for a sort of reactive balancing. Thats quite cool.
a motor is just a very inefficient speaker and an ESC/VFD is just a very specialized audio amplifier. they're electrically identical, but with diametrically opposed goals in their mechanical design.
I've just upgraded the motor in a cheap lawn strimmer to a brushless motor. Controller was programmed using an adaptor to connect it to my laptop. Worked well.
Being pedantic here. A pound of Steel and feathers actually weigh the same however when discussing precious metals troy ounces and pounds are used which are in fact less than an imperial pound used to measure steel and feathers. TL/DR: Precious metals are weighed with different units and that means a pound of gold is lighter than a pound of feathers or steel.
@mexicosfinest8380 Steel has a higher density than feathers but 1 pound of steel is equal to 1 pound of feathers, the feathers would just take up a LOT more volume.
Idk why but the concept for this video is funny because off the shelf hubmotors do exist. There are two ant sized hubs from repeat robotics and just cuz robotics and they are pretty indestructible. Nevermind he mean 150g "antweights"
Me and a friend made a hubmotor like this with his lathe for a [150g] Antweight a few days before competition like 15 years ago. It turned out about as well as your manual attempt, but on top of that we accidentally broke one of the tiny copper wires on the motor and couldn't get a replacement for it in time for the event and we lost interest in 150g robots after that. Good job on this one making it actually work :)
i told dave moulds to use a magnetic clutch instead of a belt YEARS AGO, back when we were at school. he politely told me to "f off with my poncey idea" im always ahead of my time :(
Cheers! Yeah i've wanted to make something like this for years... took me a while to wrap my head around the mechanical design honestly but in the end it's quite simple.
This is an insane weapon in basically every metric! Loving the slo-mo shots too (I may be biased toward some 😅) ive been considering developing a hubmotor for a y&d mk3 aswell as another ant idea i have... can't wait for the next one
Sine startup plus no stall protection would give fastest spool up from a stalling hit and its won't stop trying. The stall protect is what got you on the other robot, not the sine start.
I see the blade is moving about 1mm to the left at higher speeds? But there shouldn't be that much play in the bearings, right? Are the bearings moving on the shaft? Will the motor have more torque if it didn't move? Maybe machine a shaft with a sholder to stop the bearings from sliding away from the stator? I'd try to get some old, broken leaf spring for this, as it's probably the cheapest you can get for lots of material for future parts that will be super strong and not shatter. Anyway, I completely agree with your approach here! It's a lot easier than trying to make it yourself! I see one improvement possible in the future, and that would be trying to source magnets that are less brittle. The neodymium magnets you're using are sintered, and so are samarium-cobalt (SmCo) magnets, that may be an alternative if you need higher temperature tolerances. Both are brittle. Less brittle are injection molded magnets, but I haven't bothered to figure out how strong they are compared to neodymium or SmCo. They can be molded into basically any shape, though, and since the hole in the weapon can basically be any size and shape, they could be any size too. Also, since you're not machining this part, having perfect slots for magnets of exact size would make placing them accurately a breeze. Do magnets need to be electrically isolated? Hope not! 😄
Those esc are awesome!! I can't believe how far electronics have come, i want to make a small rc car with those motors in the hub, getting rid of the driveshafts and differential. It's probably going to break the magnets from high speed impacts but i would love to try.
by the look of your design you can use WAY bigger bearings with a smal redesign. just make another support on the side where the bearings is that is threaded over the 4mm shaft then have the shaft for the bearing on that. this will up the support and the bearing size to bigger then skateboard bearings if you need that.
I feel like rounding the tooth support on the leading edge might reduce wind resistance (which is considerable at that RPM) and let you get higher speeds or lower power use. (Leave the trailing edge square; it improves flow separation.) I'm not sure how much effect it'd have at scales this small, but I think it'd be worth a comparison test.
It's a good thing that you used lock-tite as the adhesive for the hub, so many RUclipsrs are in the habit of using Superglue to adhere things. However, you might find that the adhesive is the weak link in the chain, and could end up with a spinning hub if you fall into the "perfect" situation. I would suggest using solder to hold the hub in the spinning blade. a little bit of solder paste and a heat gun would lock that hub into place and keep it there for all, but the most catastrophic of impacts. Even then, there is still a solid chance of the hub breaking loose, but you could be positive with a press fit and solder adhesion that you would have done your very best under reasonable circumstances. Great Job designing that blade, a perfectly balanced blade with only 1 angry/hungry tooth is going to give you the best possible bite, just remember speed is everything. if you ramp that motor up to maximum RPM any small encroachments to the blade are still going to act like a "grinder" so perhaps lowering the speed to a few thousand rpm would give the better chance to take a bigger bite or even less rpm. with the mass of your blade it certainly is going to send an enemy robot flying around the cage even at 1000 rpm! But you are well aware of that at this point since it is old hat for you. Great video, thanks for sharing!
Just gonna continue to be that guy who points out the awesome looking 3D printed head base in the back there, always teasing us with something each video >.
Your blade is a simple extrusion: you can draw a dividing line and select the profiles to see the area of either side, without having to pop out of sketch mode and calculate center of mass with every adjustment. It's instant, even works with splines, and will help you in deriving a mathematically driven sketch in place of trial and error! :) (with a lot of algebra, trig, time, and headaches, you can even sketch this up "parametrically" with pen and paper!)
while this does work for simple geometries like a rectangle, unfortunately simply having the same amount of material (or area as the sketch is in 2d) on either side of the desired centre of mass isn't enough for every case for a few reasons. most notably because the distance of each point of mass from the desired centre also plays a role in calculations
I think your cutting head should have slot for machining inserts. Just in case you cannot flip the enemy, you can machine a new slot in their body. Just go with roughing insert with a large nose radius and the insert should last pretty long. And if you can have suitable slot in the cutting head, go with negative turning insert.
Too heavy for the weight class. You saw the difference even from steel to titanium, where he opted for the latter. Tungsten is almost 2.5 times the density of steel; if the one made from steel weighed ~60 g, a tungsten one would _literally_ eat up all of the robot's allowed mass.
@@mnxs Excellent point, I forgot about weight class. Ah well maybe in a heavier class some tungsten inserts to mass up a steel stump grinder wheel then.
I would give you a great increase in strength and alignment if you leave a flange between the two bearings. It only needs to be sufficiently thick to handle a quarter of the shock load from the rotating mass. This will give the bearings a solid surface to align to and keep the press fit from being the only real lateral load carrying member. I have a project that I've been working on for months now that I think would interest you with this. It would take the shock loading off the motor and electronics, increase top weapon RPM while decreasing ramp time. I've been looking for a bot builder to partner with to test it out. To be honest, it was designed for Witch Doctor or Yeti (big, heavy vertical spinners), but it works at this scale as well. I work for a company that designs and manufactures prototype motors for electric vehicles, so I'm surrounded by this all day. Are you interested?
Very cool project! Keep up the great work. But I have to say you would be better off using some other glue than locktight, it is simply not the best for a job like this.
It's a pretty snug fit so should be OK, but if there was any free play an epoxy would have definitely been the better choice. Will see how it performs anyhow!
That weapon could give James, Jordan, and Glen the heebie-jeebies! It looks so chonkii! I wonder if the hardox version could work on an American ant, even though 1-pounders are not a thing in Oz.
5:41 dude that center of mass analysis seems to have proven correct so well here during the test.. I don't see any vibration at full speed at 13:22 amazing!!
This is hella cool. I had a thought though and I'd love to know if you'd be willing to experiment with it! What happens if you half-size the hub blade assembly, and make a mirrored version designed to spin the other direction on a shared shaft? That way you could test the effectiveness of having two, less massive, counter-rotating blades of the same design, vs one, more massive one. I would love to see if one is better than the other (I'd assume that besides gyroscopic effects the single monoblade design here would be superior) For example, would you need to double your motor control circuitry, or would both run easily off the same, single output? What gyroscopic effects would be observed? Would there be an impact on damage output or ability to toss around other antweights? Would having two monoblades lead to significantly less reliability due to complexity vs a single monoblade? How much would a two blade design affect susceptibility to vibration from tiny COM changes e. g. from massive impacts, chipped off blade materiial etc? Most importantly: Could you put cat ears on it? i know an etsy seller that makes fluffy cat ear headbands in Eastern Australia and it'd be pretty funny to see nyanobot 2.0 or 3.0 sporting a pair of loosely attached kitty ears if there's weight to spare, lmao. Though i doubt they'd survive a fight without turning into an environmental hazard that would try to clog up motors. That brings me onto another tangent: hair based antweight warfare. Are thin strands of hair or metal designed to foul motors banned in antweight? I'd assume so, but if not you could be the guy that proves the need for such a rule by making Cousin It, an antweight whose only weapon is a big tuft of hair or wire designed to get yanked out and bind motors, with a propulsion method hardened against ingesting its own fibrous weaponry. Silly idea but it'd be hilarious to watch in action i feel.
That esc is 40a max and has stuff all caps on it, adding a low esr capacitor to it might help filter out voltage spikes especially when stalling the motor however there are a lot better escs out there for not much more $. You would have to get a blheli32 esc tho at the moment and flash am32 to it which is pretty easy. Blheli32 is closed source but is now not supported or issuing any more licences so the quadcopter hobby is moving to am32, manufacturers are slowly switching all there escs to am32 but it's taking them a while as I am guessing they are using up the licences that they have already purchased. Also curious why you max out the timing? As this is a setting that has to do with the timing of when a pole is charged in relation to magnet position and having it set to max could lessen the torque of the motor from what I understand.
Haha I wondered if anyone would recognise it! It's kind of insane that something I made is featured in a makers muse video! That was a very fun event although I've got a few videos to catch up before I get there...
@@engineered_cha0s I saw your video on updating tue design to 4 wheels only 2-3 days, that's how the name is still fresh in my mind. I may try and find Ant weight battles in the UK and try it, probably go with PS3 Bluetooth control and ESP32 for control, set a few programmed sequences etc that just take one button to run.
Amazing project, great build and excellent insights! I’ve been breaking a lot of drone motors with vibration, will have to try switching out the ESC. Any other recommendations for hardening motors, short of building your own? 😅🙏🤗
You should look into grinding hss on a lathe. Thers angles that are more optimal for cutting. Implement them so you have a better chance to "dig in" to the competitor
I assume someone else has said this but if you heat up the blade and cool down the motor you get a small amount of growth and shrinkage. meaning you can make your blades ID slightly smaller than the motor od. then put them together and when the parts return to normal temps they are never coming apart. common industry trick
Yeah! Seth is a great guy with excellent products. Those hub motors are quite new and for the larger "insect weight" classes, but a cool thing to see on the market now
I am making a battle bot is this list good? hubmotor for 1lb the am-32 ESC and FS-i6and the repeat mini mk3 the FS2A receiver and your weka 20A esc and should I get the pcb switch or something different this stuff is from just cause robotics. Thanks I don’t know a lot about electronics
@@MakersMuse Agreed; however, if you consider the instant of impact, the "hit" on the enemy is as hard as the hit on your own bearing. Better to spread it out over 40mm, of hardened steel (i.e. 3mm rollers) no? As long as the bearing is an "outrunner" too - I don't see the problem :)
13:37.. It looks like there is play in the magnet housing (moves left and right slightly when applying/removing power). The Magnets may not be aligned/centered on the rotor... or just too much end play.
Yeah I just commented on that myself, you only see it at near max speed, and it reverts when slowing down, I was guessing the motor commutation was getting off (i.e. the blade wasn't actually reaching the speed the controller was trying to push, so the magnets were getting repelled axially rather than radialy).
the rotor ring coming free worries me a bit, maybe you could heat the weapon blade and insert the ring while its hot, giving in a interference fit and making it more unstoppable
You should made a robot with a flail type mower weapon that swings back when it contacts something hard it will still hit extremely hard but it won’t stop the spinning and you can put multiple heads on the weapon at seperate angles say like a triangle with three different flail weapons all that weigh the same for balance then when they strike they swing back and the next weapon comes around for the strike without losing any real strength or velocity it’s how our big flail mowers work
2 of those weighted spinny thing with 2 blade thingies on opposite sides, and the spinnies horizontal or vertical whatever you want, when turning the entire robot the spinnies stop at the same time then turn, then the spinnies are turned on again. if that is not enough, make it 3 spinnies with a total of 6 blades. the top can be cherried with a helicopter type of design but instead of flying you spin it so there is more destruction
i think you'll find you should have designed to used bearings with a bigger OD. Also, a step or sleeve between the outer races and some thread on the end of the axle to apply some pre-load to the inner races to get things adjusted perfectly - right now there's some alarming lateral movement visible on your spin test.
I see some proposing a hole to make a "death whistle" - I wondered if the inside of the tooth could not be flat but angled / sloped to reduce aerodynamic resistance and thus increase performance by a fraction - even though maybe at 22000 RPM we are already cavitating
This is fantastic, i can't wait to see the next video. What effect will the difference in mass between the stainless and titanium have in the force of the hit?
@13:30 or around you can see the hub is being deflected by the stator at max speeds, doesn't seem to be detrimental, but I thought it was interesting (I suspect it implies the commutation is off at those speeds, likely the motor simply isn't reaching the rpm the controller is trying to push, so more force is going in the axial direction). I would cautiously suggest pushing the controller to that virtual speed would result in torque dropping due to the commutation issue.
what if you add a hole perpendicular to the blade so it makes a scary whistle sound, psychological warfare, you know?
Ooh I like that idea !
This is a Genius idea !!!
Stuka dive bomber "Jericho Trumpets" 🎺👍
Dark Mode for the Win 🏆
That's sadistic. I LIKE IT
@@edwardfletcher7790 Yuep. Stuka was the first thing I thought of as well. Those must have been horrifying to hear.
@@MakersMuse Please do this It'd be so amazing
Hi from the AM32 team! 😊
Thanks a ton for the shoutout! We’re really excited to see AM32 being used in fighting robots-it's awesome to see our work in action! Keep those bots spinning strong, and reach out anytime if you want to swap tips with other robot builders. 💥🤖
Best of luck in the arena!
Awesome work on some incredible firmware guys!
This is the second use I've heard of for AM32... My first time was when I went to pick up my desktop CNC mill... He uses it for the spindle control. Using a 500W brushless motor, it gives great results! 😁
Hey chat, if you're doing anything combat robot related, please do it inside a test box. I'm sure that Angus has taken the proper precautions here, but the kinetic energy stored in these types of weapons is equivalent or higher to common pistol bullets, and if something goes wrong drywall or an interior (hollow) door will not save you.
Flywheels are scary and should be respected!
@@mandrakejakeEveryone gangsta until the flywheel hits 22,000rpm 😂
I think it starts getting to that energy in the 3lb weight class, but i'm sure even a typical 150 gram weight class spinner could make short work of human flesh. Yes, always use a test box.
Thankfully spin conservation as at our sides. Just be at it sides.
@@pedro.alcatra No, that is absolutely bad advice. The gyroscopic forces of a spinning weapon can make it behave completely unpredictable. In the blink of an eye a stationary clamped bot with the weapon spinning can flip around and launch a piece of debris at any angle which can bounce off any surface with enough energy to do some serious damage.
If you test a spinning weapon of any weight class, you either need distance (so that ricochets lose most of their energy) and a shield or have it completely enclosed in bullet proof materials.
Trying to predict where shrapnel will fly is a fools game.
With Solidworks, if you select the center of mass and the center of the hub ring, you can lock them to coincide, then the program will automatically adjust the unlocked components of the drawing to keep them on top of each other. Then you can drag them around to whatever shape and position you want, locking the parts you like as you go.
Did not know this; that would be very useful. Thanks!
Solidworks is just miles ahead of Fusion. Ever tried to define planes in Fusion? Nightmare. But Fusion is free and SW is $4k minimum
@@150Gianluca Fair point... I must admit that I never payed for it. XD
@@150Gianluca They have a Maker version they don't advertise, worth looking into
@@DanielCook-h6r it's still a few grand for the most basic package. SolidWorks is a few grand for the most extensive package with generative design, FEA and CFD simulations. It's a good option for hobby use, and basic 2.5D designs, but not for a complete design suite. It has weird constraints and often undefinable (want a plane 45° to another plane? Better hope the default angle offset is what you want). It also doesn't allow for backwards constraint changes. ALSO there's certain angles you just can't rotate to, that one's super annoying
Tip for a perfectly balanced weapon blade design:
1. Create a roughly balanced blade design without your center mounting cutout(s).
2. Display its center-of-mass.
3. Create a side view sketch, zoom all the way in on the center-of-mass, and drop a point right on it.
4. Center your mounting cutout(s) on the sketch point.
Much easier than trying to tweak your sketch in tiny increments, and you can get 'perfect' balance this way (not that the manufactured blade will be perfectly balanced, especially after some combat, but at least the design is this way with little fuss).
Angus could get a bigger "bite" with his tooth if he were to reduce the counterweight size by inserting a tungsten slug, like we do with automotive crankshafts.
@@jimurrata6785 That would definiteley work, but then center of mass calculations become very hard and tolerances would mean that it will not be balanced after assembly. Having to then fix the center of mass by grindig away material requires a balancing tool and a lot of patience.
Thats all to say that this idea, while great in theory, would make this design a lot more difficult. Maybe when the time comes for V2 and hee needs things to improve this could be among them.
@nicholasweiss4662 I'm really surprised there's not a way to do it in software, given we know the density of tungsten or Mallory Metal.
But the rpms are pretty crazy and impact will go up in square to velocity
@@jimurrata6785 In practice, he probably has plenty of potential bite from the blade geometry.. it's usually more a matter of relative closing velocity to the other bot vs weapon RPM / number of cutting 'teeth' the blade has.
@@dogjutsufpv3813 "it's usually more a matter of relative closing velocity to the other bot vs weapon RPM / number of cutting 'teeth' the blade has.".
I get what you're saying, but I disagree. You want to hit with the teeth at maximum depth, so your closing speed must be enough to travel the tooth depth in full. The fewer teeth on the blade, the more time you get per revolution to reach maximum depth for the impact. Of course the larger the tooth, the larger the requirement of closing speed.
I think the most valuable bit of Information in this Video was the thing about Bambu Lab now stocking Maker related stuff like Bearings, Motors, Slip Rings, etc... one might be tempted to buy in a pinch when already buying Filaments from them - *_BUT_* - Bruh... Those prices for a *_single_* Bearing / Micromotor... I'm gonna stick with AliE for my Maker Needs cause for maybe a little bit more I'm usually getting on 5-10x the amount of Bearings / Micromotors on AliE.
I was surprised, but, also, why? I just don't get companies that becomes Jack of All Trades, it's a dangerous game.
@@emuboy85 Why is it dangerous?
@@Jehty_ "Jack of all Trades, Master of None" is an old adage that tends to hold true. When a company focuses on what they do best, they can be the best. Comes with the risk of one day being obsolete though. A good example to watch for: A lot AC companies also do plumbing (and vice versa) but they ultimately end up being good at neither. The best AC companies focus exclusively on AC/Heating systems. And the best plumbers focus exclusively on plumbing. The very worst try to do both.
@Jehty_ the quality of their main venture decrease. Like Google and Microsoft
@@privacyvalued4134 we are talking about selling a few more items.
I don't get how any of what you wrote would apply to that.
13:37 seems like it slides around a bit between off and max speed. like the rotor is being pushed off of the stator from the intensity of the magnetic field. Because it slides right back once it's dialed down.
Just needs something to retain it, maybe a tiny thrust bearing either side?
leet
Incorrect
@@1992jamoi think in such a small application it would be better to just size the shaft for a press fit with the inner race of those two ball bearings so that they can't slide around on the stationary shaft. those bearings aren't really meant to take axial loads but it's only a small amount of force pushing the rotor around side to side.
@@1992jamo you could also just use red loctite and call it a day. it's an antweight after all.
this has real "Be the reason the rules are written" vibes
This looks like ABSOLUTE FUN. As a 5-year manual lathe machinist... a bit of sand paper goes a LONG way.... hand polishing shafts and bores to +/- .0002" can be done... I had to do it all the time running manual lathes. The next step is understanding proper fit tolerances. Just a few tricks for you to hide up there in your noggin for the future. Maybe if i start to find some extra time in my future the ant weight battle bots sound like something I would love to do. Thank you for the video and the Today I Learned moment for me.
Cheers! Yeah my downfall was trying to drill out the 8mm bearing holes with,,, a drill bit. Obviously it wasn't going to precise but I was hoping epoxy would let me get away with it. A reamer would solve the issue, but costs more than the part from PCBWay haha. There's robot events all the time all over the world! Robotcombatevents . com is a great place to look.
That cat at 0:30 is judging every decision I've made in my life
Me too buddy
@ I think a video on the “judges” is in order. Maybe some info on pets and workshop/machines combo too?
@@MakersMuse stop reading peoples thoughts, you demon
@@jsal92 Haha did two videos previous about them! I made a puzzle and remote controlled toy for them to chase.
was recommended this rather randomly... figured it was a military combat robot, and figured that the bearing in the thumbnail would be an amazing design I could incorperate into a project... not sure what tbh... have very few at the moment. Most exotic project atm is making a salt water battery that can be made in the event of societal collapse, just cause it would be cool, while the least out there one would be a desk I am rather proud of and nearly finished... did cut my hand open a bit this morning, shaving part of the legs (uneven wooden lengths that jut out due to cutting it in small passes, flipping, and repeating); only had a pocket knife and it slipped on a knot in the wood.
Rant over. Hope it boosts engagement enough to recommend to more people lol. Super cool video. Love the little crafts people make.
Looking awesome as ever Angus. In my country, RC toys like drones are really strongly regulated. This sadly also results in RC hobby in general being only a small hobby and most people only drive RC cars and boats, but the amount of hobbyists is steadily decreasing. Most RC shops have closed both online and offline. It's a damn shame. Maybe I'll make my own bot one day but I won't have people to fight in my area! It's still a dream for me, ever since I started watching Robot Wars when I was a kid on the BBC in the 90's. Man I loved Craig Charles's enthusiasm!
Yeah, for me Germany is RC Hell. Too many regulations, and Karen's that have PHD's on aeronautics.
Is the regs just on drones/planes or is it on everything RC related?
The rotor may overheat in this design, consider adding "ram air" holes through the rotor/striker. The top of an out-runner motor is perforated to make a crude cetripital fan, to cool the windings and magnets. The striker will cool the magnets, but without good airflow the motor windings can overheat and the enamel fails. This can cause shorts and fry the ESC.
The fights last for 3min max so it's unlikely to saturate in that time honestly, but tiny angled holes might channel air and also make a pretty cool whistle 🤔
@@MakersMuse I want to hear that WHISTLE. dang
given the RPM and the size of both the motor and holes I don't think we can hear the whistle (at least after winding up).
but maybe dogs or bats could ;-)
@@dfischer000 Partially obstructing the air entrainment intake with a rubber flange tapered to the apogee, or something like that...basically, a rubber flap, and as the air cycles around, it'll sound like a comical diarrhea or long fart noise distinct and audible enough to pick out.
"Which one is yours?"
"The one that sounds like it ate a lot of Taco Bell and clobbering the other one to death to use the toilet."
"Ohhh, okay."
Maybe add an internal groove for free rolling ball bearings, so if the blade has a bit of material removed they can compensate for the weight difference
I was trying to thinm of a solution to that. Any damage to the claw will result in unbalancing of the motor. I will have to exore the ball bearing route for a sort of reactive balancing. Thats quite cool.
@@nationofcraig thats really big brain thinking
This is the same principle as "balancing beads" some vehicles use in their tires.
Please explain how that works. It sounds really cool, but I don't understand how it works.
@@paulroberto2286 centrifugal foces position the ball bearings at the right point to get balanced again!
0:36 WHAT?!! WHERE IS YOUR NORMAL INTRO!? I WANNA EXPLORE MICROSCOPIC WORLDS!!
The idea that a speed controller can pulse the motor to produce tones blew my mind.
you can even program in new melodies! it's super neat
a motor is just a very inefficient speaker and an ESC/VFD is just a very specialized audio amplifier. they're electrically identical, but with diametrically opposed goals in their mechanical design.
"Fucking magnets, how do they work?"
Thanks for the AM32 mention, i didnt know it before and i could use rhis in other rc projects ! ❤
I've just upgraded the motor in a cheap lawn strimmer to a brushless motor. Controller was programmed using an adaptor to connect it to my laptop. Worked well.
We are truly honored to be a part of your remarkable campaign!😊
7:30 "The 4140 blades are crazy heavy compared to the titanium blades. This is because steel is heavier than feathers."
"but steel is heavier than feathers...." "yes, but they're both a kilogram" 😟
Being pedantic here.
A pound of Steel and feathers actually weigh the same however when discussing precious metals troy ounces and pounds are used which are in fact less than an imperial pound used to measure steel and feathers.
TL/DR:
Precious metals are weighed with different units and that means a pound of gold is lighter than a pound of feathers or steel.
@@Ixidora But steel is heavier than feathers.
@mexicosfinest8380
Steel has a higher density than feathers but 1 pound of steel is equal to 1 pound of feathers, the feathers would just take up a LOT more volume.
@@Ixidora I don't get it.
Name your custom blade "The Hellephant"
Good one!
Jesus that little weapon is terrifying :D
Love your supervisor laying in front of the window 😁👍🇦🇺
13:40 lol I don't even want to be in the same room as that thing, much less the same arena! Nice design.
Haha cheers! It's got a nice hum to it for sure
Idk why but the concept for this video is funny because off the shelf hubmotors do exist. There are two ant sized hubs from repeat robotics and just cuz robotics and they are pretty indestructible.
Nevermind he mean 150g "antweights"
It's also just more fun to custom engineer stuff yourself (: (aforementioned fun may or may not Include much trial and error and money)
Antweight means 150g about everywhere except the US.
fairyweights you'd call them, yeh, not 1lb
You should use one of the Loctite bearing retainers , not whatever that blue stuff was. The bearing retainers are designed for this purpose.
That's why I only use red Loctite on the lug nuts on my car.
Me and a friend made a hubmotor like this with his lathe for a [150g] Antweight a few days before competition like 15 years ago. It turned out about as well as your manual attempt, but on top of that we accidentally broke one of the tiny copper wires on the motor and couldn't get a replacement for it in time for the event and we lost interest in 150g robots after that. Good job on this one making it actually work :)
i told dave moulds to use a magnetic clutch instead of a belt YEARS AGO, back when we were at school. he politely told me to "f off with my poncey idea"
im always ahead of my time :(
Magnetic clutches work but tuning is harder than belts, and your clutch surface is custom.
This looks impressive ! Congratulations. There's no better feeling than imagining something in your head and seeing it realized.
Cheers! Yeah i've wanted to make something like this for years... took me a while to wrap my head around the mechanical design honestly but in the end it's quite simple.
One of your greatest videos imo, great pacing.
Thank you for the teaser... That's GENUINELY terrifying! I'm now awaiting the robot that can beat this!
This is an insane weapon in basically every metric! Loving the slo-mo shots too (I may be biased toward some 😅) ive been considering developing a hubmotor for a y&d mk3 aswell as another ant idea i have... can't wait for the next one
Can’t wait for part 2! This process is fascinating and inspiring
Sine startup plus no stall protection would give fastest spool up from a stalling hit and its won't stop trying. The stall protect is what got you on the other robot, not the sine start.
Looks sick Angus! Excited to see where you go from here.
Designing a thing around it's weapon is what happened to the A10 and that thing is pretty cool as well. So good luck!
Best PCBway commercial hidden as a cool video. Love those ants!
I try to only work with sponsors that are actually meaningful, PCBWay make some legit good parts!
DUDE! What a ridiculous weapon. Can't wait to see more!
I see the blade is moving about 1mm to the left at higher speeds? But there shouldn't be that much play in the bearings, right? Are the bearings moving on the shaft? Will the motor have more torque if it didn't move? Maybe machine a shaft with a sholder to stop the bearings from sliding away from the stator? I'd try to get some old, broken leaf spring for this, as it's probably the cheapest you can get for lots of material for future parts that will be super strong and not shatter.
Anyway, I completely agree with your approach here! It's a lot easier than trying to make it yourself!
I see one improvement possible in the future, and that would be trying to source magnets that are less brittle. The neodymium magnets you're using are sintered, and so are samarium-cobalt (SmCo) magnets, that may be an alternative if you need higher temperature tolerances. Both are brittle.
Less brittle are injection molded magnets, but I haven't bothered to figure out how strong they are compared to neodymium or SmCo. They can be molded into basically any shape, though, and since the hole in the weapon can basically be any size and shape, they could be any size too. Also, since you're not machining this part, having perfect slots for magnets of exact size would make placing them accurately a breeze. Do magnets need to be electrically isolated? Hope not! 😄
1:48 why is bro's cat staring at the camera tha way?💀
Those esc are awesome!! I can't believe how far electronics have come, i want to make a small rc car with those motors in the hub, getting rid of the driveshafts and differential. It's probably going to break the magnets from high speed impacts but i would love to try.
I'm calling it. Those magnets will be shattered after one super hard impact.
by the look of your design you can use WAY bigger bearings with a smal redesign. just make another support on the side where the bearings is that is threaded over the 4mm shaft then have the shaft for the bearing on that. this will up the support and the bearing size to bigger then skateboard bearings if you need that.
I feel like rounding the tooth support on the leading edge might reduce wind resistance (which is considerable at that RPM) and let you get higher speeds or lower power use.
(Leave the trailing edge square; it improves flow separation.)
I'm not sure how much effect it'd have at scales this small, but I think it'd be worth a comparison test.
It's a good thing that you used lock-tite as the adhesive for the hub, so many RUclipsrs are in the habit of using Superglue to adhere things. However, you might find that the adhesive is the weak link in the chain, and could end up with a spinning hub if you fall into the "perfect" situation. I would suggest using solder to hold the hub in the spinning blade. a little bit of solder paste and a heat gun would lock that hub into place and keep it there for all, but the most catastrophic of impacts. Even then, there is still a solid chance of the hub breaking loose, but you could be positive with a press fit and solder adhesion that you would have done your very best under reasonable circumstances. Great Job designing that blade, a perfectly balanced blade with only 1 angry/hungry tooth is going to give you the best possible bite, just remember speed is everything. if you ramp that motor up to maximum RPM any small encroachments to the blade are still going to act like a "grinder" so perhaps lowering the speed to a few thousand rpm would give the better chance to take a bigger bite or even less rpm. with the mass of your blade it certainly is going to send an enemy robot flying around the cage even at 1000 rpm! But you are well aware of that at this point since it is old hat for you. Great video, thanks for sharing!
Just gonna continue to be that guy who points out the awesome looking 3D printed head base in the back there, always teasing us with something each video >.
This is SO cool!
What an awesome blade, that SPEED was killer too!
Looking forward to part 2!
Green locktite 609 is designed for press fit applications. But the reds work too. Red 271 and 277, 277 being the thicker one. Great video as usual 🤘
Your blade looks scalable. Oil your bearings, no grease. Your arena looks cool. 😎
Your blade is a simple extrusion: you can draw a dividing line and select the profiles to see the area of either side, without having to pop out of sketch mode and calculate center of mass with every adjustment.
It's instant, even works with splines, and will help you in deriving a mathematically driven sketch in place of trial and error! :)
(with a lot of algebra, trig, time, and headaches, you can even sketch this up "parametrically" with pen and paper!)
while this does work for simple geometries like a rectangle, unfortunately simply having the same amount of material (or area as the sketch is in 2d) on either side of the desired centre of mass isn't enough for every case for a few reasons. most notably because the distance of each point of mass from the desired centre also plays a role in calculations
13:33 I can FEEL the pucker in that room when it started to resonate
woah us combot fans be eating good today
I think your cutting head should have slot for machining inserts. Just in case you cannot flip the enemy, you can machine a new slot in their body. Just go with roughing insert with a large nose radius and the insert should last pretty long. And if you can have suitable slot in the cutting head, go with negative turning insert.
Impressive work, at one rally I talked to a barrel spinner maker who moved to pin roller bearings after many ball race failures.
Amazing work! Love the design. Can't wait to see it in action.
Cheers! Either way it'll be pretty destructive ;)
Blade of Tungsten next perhaps? High density, very hard, relatively tough, should build a lot of kinetic energy.
Too heavy for the weight class. You saw the difference even from steel to titanium, where he opted for the latter. Tungsten is almost 2.5 times the density of steel; if the one made from steel weighed ~60 g, a tungsten one would _literally_ eat up all of the robot's allowed mass.
@@mnxs Excellent point, I forgot about weight class. Ah well maybe in a heavier class some tungsten inserts to mass up a steel stump grinder wheel then.
The wheel of destruction never stops spinning. That thing is going to do some nasty damage, I can't wait to see part 2!
That beautiful piece of machined steel blade ripping at 22k was SUPER satisfying to watch and listen to
13:48 "so stoked" I was SO NERVOUS that thing was gonna blow apart XD
I would give you a great increase in strength and alignment if you leave a flange between the two bearings. It only needs to be sufficiently thick to handle a quarter of the shock load from the rotating mass. This will give the bearings a solid surface to align to and keep the press fit from being the only real lateral load carrying member.
I have a project that I've been working on for months now that I think would interest you with this. It would take the shock loading off the motor and electronics, increase top weapon RPM while decreasing ramp time. I've been looking for a bot builder to partner with to test it out. To be honest, it was designed for Witch Doctor or Yeti (big, heavy vertical spinners), but it works at this scale as well. I work for a company that designs and manufactures prototype motors for electric vehicles, so I'm surrounded by this all day. Are you interested?
That is the cleanest sound and btw I maxed my headphones to hear ALL of it!
Very cool project! Keep up the great work. But I have to say you would be better off using some other glue than locktight, it is simply not the best for a job like this.
It's a pretty snug fit so should be OK, but if there was any free play an epoxy would have definitely been the better choice. Will see how it performs anyhow!
That weapon could give James, Jordan, and Glen the heebie-jeebies! It looks so chonkii!
I wonder if the hardox version could work on an American ant, even though 1-pounders are not a thing in Oz.
Definitely! If there's ever a vert to bend my blade it'd be this!
Very cool build! Look forward to seeing what's next.
8:30 "huh those looks just like the bearings I put into my A1 extruder assembly..."
It'd make sense for them to offer their own parts for sure
5:41 dude that center of mass analysis seems to have proven correct so well here during the test.. I don't see any vibration at full speed at 13:22 amazing!!
This is hella cool. I had a thought though and I'd love to know if you'd be willing to experiment with it!
What happens if you half-size the hub blade assembly, and make a mirrored version designed to spin the other direction on a shared shaft? That way you could test the effectiveness of having two, less massive, counter-rotating blades of the same design, vs one, more massive one.
I would love to see if one is better than the other (I'd assume that besides gyroscopic effects the single monoblade design here would be superior)
For example, would you need to double your motor control circuitry, or would both run easily off the same, single output? What gyroscopic effects would be observed?
Would there be an impact on damage output or ability to toss around other antweights? Would having two monoblades lead to significantly less reliability due to complexity vs a single monoblade?
How much would a two blade design affect susceptibility to vibration from tiny COM changes e. g. from massive impacts, chipped off blade materiial etc?
Most importantly: Could you put cat ears on it? i know an etsy seller that makes fluffy cat ear headbands in Eastern Australia and it'd be pretty funny to see nyanobot 2.0 or 3.0 sporting a pair of loosely attached kitty ears if there's weight to spare, lmao. Though i doubt they'd survive a fight without turning into an environmental hazard that would try to clog up motors.
That brings me onto another tangent: hair based antweight warfare. Are thin strands of hair or metal designed to foul motors banned in antweight? I'd assume so, but if not you could be the guy that proves the need for such a rule by making Cousin It, an antweight whose only weapon is a big tuft of hair or wire designed to get yanked out and bind motors, with a propulsion method hardened against ingesting its own fibrous weaponry. Silly idea but it'd be hilarious to watch in action i feel.
That esc is 40a max and has stuff all caps on it, adding a low esr capacitor to it might help filter out voltage spikes especially when stalling the motor however there are a lot better escs out there for not much more $. You would have to get a blheli32 esc tho at the moment and flash am32 to it which is pretty easy. Blheli32 is closed source but is now not supported or issuing any more licences so the quadcopter hobby is moving to am32, manufacturers are slowly switching all there escs to am32 but it's taking them a while as I am guessing they are using up the licences that they have already purchased. Also curious why you max out the timing? As this is a setting that has to do with the timing of when a pole is charged in relation to magnet position and having it set to max could lessen the torque of the motor from what I understand.
Angus, have you considered covering the magnets in the rotor with JB weld? I feel like it would increase its resiliancy quite a lot
11:35 To make your programming a bit easier, buy a few pogo-pins and 3d print a holder so you don't have to solder the wires.
I had no idea there were Battlebots of this small scale. - Cool!
This looks AWESOME nice work 👍
I did notice the motor has some horizontal movement though....
gnarly - I also noticed it hitting a resonance frequency on the ramp up so avoid that as much as possible ;)
Yellow and dangerous MK2 at the 0:12 mark
Absolute beast that robot!
Haha I wondered if anyone would recognise it! It's kind of insane that something I made is featured in a makers muse video! That was a very fun event although I've got a few videos to catch up before I get there...
Really interesting design!
@@engineered_cha0s I saw your video on updating tue design to 4 wheels only 2-3 days, that's how the name is still fresh in my mind.
I may try and find Ant weight battles in the UK and try it, probably go with PS3 Bluetooth control and ESP32 for control, set a few programmed sequences etc that just take one button to run.
@@welshdave5263 that checks out, I only uploaded it on Monday 😆 bots with automated functions are rare... I'd love to see how it turns out!
Seeing that blade spinning by itself is terrifying and gives me ididathing vibes, I wonder what you two could make together
Amazing project, great build and excellent insights! I’ve been breaking a lot of drone motors with vibration, will have to try switching out the ESC. Any other recommendations for hardening motors, short of building your own? 😅🙏🤗
You should look into grinding hss on a lathe. Thers angles that are more optimal for cutting. Implement them so you have a better chance to "dig in" to the competitor
I assume someone else has said this but if you heat up the blade and cool down the motor you get a small amount of growth and shrinkage. meaning you can make your blades ID slightly smaller than the motor od. then put them together and when the parts return to normal temps they are never coming apart. common industry trick
Not trying to advertise but there is one place I know that does sell hub motors the place is called just cause robotics
Yeah! Seth is a great guy with excellent products. Those hub motors are quite new and for the larger "insect weight" classes, but a cool thing to see on the market now
I am making a battle bot is this list good? hubmotor for 1lb the am-32 ESC and FS-i6and the repeat mini mk3 the FS2A receiver and your weka 20A esc and should I get the pcb switch or something different this stuff is from just cause robotics. Thanks I don’t know a lot about electronics
I propose a name : "Figonski" from "finger gonski" :) Awesome Project , super impressive, this will be brutal :)
a 40mm bearing instead of a piddly 4mm bearing, and you'd be golden.
haha the entire thing is only 60mm dia swing! that'd be one chunky bearing
@@MakersMuse Agreed; however, if you consider the instant of impact, the "hit" on the enemy is as hard as the hit on your own bearing. Better to spread it out over 40mm, of hardened steel (i.e. 3mm rollers) no? As long as the bearing is an "outrunner" too - I don't see the problem :)
13:37.. It looks like there is play in the magnet housing (moves left and right slightly when applying/removing power).
The Magnets may not be aligned/centered on the rotor... or just too much end play.
Yeah I just commented on that myself, you only see it at near max speed, and it reverts when slowing down, I was guessing the motor commutation was getting off (i.e. the blade wasn't actually reaching the speed the controller was trying to push, so the magnets were getting repelled axially rather than radialy).
That motor test reminded me of those audio tests to see the highest frequency you can hear
Thumbs up for kitty in background!
Always watching!
Beautiful work, Angus!
As someone who recently got in to antweights this is amazing
the rotor ring coming free worries me a bit, maybe you could heat the weapon blade and insert the ring while its hot, giving in a interference fit and making it more unstoppable
Yet another example of why kinetic energy is terrifying.
More robot vids please!😊
You should made a robot with a flail type mower weapon that swings back when it contacts something hard it will still hit extremely hard but it won’t stop the spinning and you can put multiple heads on the weapon at seperate angles say like a triangle with three different flail weapons all that weigh the same for balance then when they strike they swing back and the next weapon comes around for the strike without losing any real strength or velocity it’s how our big flail mowers work
2 of those weighted spinny thing with 2 blade thingies on opposite sides, and the spinnies horizontal or vertical whatever you want, when turning the entire robot the spinnies stop at the same time then turn, then the spinnies are turned on again. if that is not enough, make it 3 spinnies with a total of 6 blades. the top can be cherried with a helicopter type of design but instead of flying you spin it so there is more destruction
Holy cow, dude! 22 krpm? That thing is going to absolutely decimate the competition.
There was some good info here. Thank you
i think you'll find you should have designed to used bearings with a bigger OD. Also, a step or sleeve between the outer races and some thread on the end of the axle to apply some pre-load to the inner races to get things adjusted perfectly - right now there's some alarming lateral movement visible on your spin test.
Interesting video content.
Looking forward to seeing future instalments!👌👍
I see some proposing a hole to make a "death whistle" - I wondered if the inside of the tooth could not be flat but angled / sloped to reduce aerodynamic resistance and thus increase performance by a fraction - even though maybe at 22000 RPM we are already cavitating
seeing this thing turn up to full power is terrifying
If you have space for a larger spinner, distributing the weight far from the shaft gives more KE punch for less weight, but I'm sure you know that.
This is fantastic, i can't wait to see the next video. What effect will the difference in mass between the stainless and titanium have in the force of the hit?
@13:30 or around you can see the hub is being deflected by the stator at max speeds, doesn't seem to be detrimental, but I thought it was interesting (I suspect it implies the commutation is off at those speeds, likely the motor simply isn't reaching the rpm the controller is trying to push, so more force is going in the axial direction). I would cautiously suggest pushing the controller to that virtual speed would result in torque dropping due to the commutation issue.
Awesome concepts and instructions.