I realize this is a couple years old but why wouldn't you use the rudder channel to rotate the Zero point? Then your directional stick could still work as an x/y directional against the zero point which is now adjustable (rotation) via the rudder.
Yes, that is the correct equation for rotational KE, I used the linear version because it is simpler to understand and its something most people have seen before. I was aiming to use simplified ideas to explain a complex control system. Also for those wondering, in the rotational KE equation: I is the Moment of inertia w is the angular velocity
When I first heard 'melty brain', I imagined it was an illegal, totally concealed secret weapon system that trashed your opponent's electronics by lobotomising his control gear's CPU whilst your robot pretended to be a legitimate flipper or drum spinner. Sort of a mini, carefully directed E.M. pulse... Um... Yeah, you're right. Stupid idea. Sorry. (I'll get my coat.)
I mean, EMP or elettroshock would work really well, even if the main body is a faraday cage, motors would still burn their coils, so effective it is banned
Why don't you control it like how rc drones/heli controls their movement? The stick with the throttle will control the 'turning' while the other self-centering stick will control all translational movement. Plus you have the shoulder switches to quickly shut off the power if your bot got stuck and needs to rest etc and another switch to maybe 'idle up' to max power.
He even acknowledges the X axis of the right stick at 17:31, but then ignores it and decides to go with a fully custom transmitter due to said X axis not being available, even though having proportional control over the robot's 'forward' direction would be quite a bit better than having just two buttons.
I have an idea to use a radio beacon to orient the robot. If you have an antenna lying on its side both inside the robot and in the control booth, the signal strength of a carrier wave received in the robot will be proportional to how parallel the two antenna are. As the melty spins you could measure this during the rotation and you'd have a signal peak both when facing the operator and when facing away. Probably would break competition regs about radio control though.
Hey Ben, I'm late, but I hope you see this. I have wanted to build a "Melty Brain" for a while, but I didn't know how to get started. This video is helping a LOT. My problem is, I don't know where to find parts. If you could tell me what the parts you used are, that would be awesome. If you don't mind, a link to the .stl file of the print would be nice too.
Honestly, I'm not the best person to be asking. I have tried a melty but I've never got anything to work. So I dont want to lead you astray by giving you links to the stuff that has failed for me. Good luck with it though!
While all of Nuts spins, only the mass of flails, contributes to the energy of it's weapon. Obviously, that's quite a lot of weight, but I would think less than some of the bar spinners. So Nuts' advantage is more, that the length of the chains increase it's effective diameter and by extension, the tip velocity of the flail. Thanks for a the brain melting info! :D
Yeah very true, the chains increase the length (hence the speed as you mentioned) but also decouple the weapon edge from the main weight of the robot. Looking at it from a KE optimisation standpoint Nuts would be better to be a ring spinner with the chains attached to the ring. They would be able to sping the flails faster and hence hit harder. But Melty Brain provided hardware simplicity and some awe factor
Yep, you're spot on. The really weird thing, though, is just how often Nuts managed some almost unbelievably 'lucky' hits. Despite the lack of mass, the flails unerringly found an opponent's weak spot, smashing an unprotected wheel rim here, dislodging a drive chain sprocket there, or hitting a safety link. I'm sure none of the attacks were actually targeted - it was all just hit-and-hope - but it's that kind of ridiculous good fortune that made Robot Wars worth watching. I liked Nuts for daring to be different. (And I liked the Nutty-Boys in the team, too; they should've used a ♬ Madness song ♪as their theme tune!) ;-)
I bought it that way. When buying a transmitter you are often given the choice of Mode 1 or mode 2 Mode 1 means throttle on the right mode 2 means throttle on the left. Currently mode 2 is the most common, but I personally prefer mode 1 www.rc-airplane-world.com/rc-transmitter-modes.html
11:00 Actual explanation of "how it works" starts here. First 11 mins just saying the same thing over and over in different ways. Also explaining how a robot spins on the spot, which to be honest if you are researching how Melty Brain technology works you already know how wheels work and what melty brain is. Great video but please get to the point or at least link to the individual "chapters" of your video. Maybe reserve "What it is", "How it works" and "Why you'd use it" to their own individual videos. I just want to know how it works and that's what your video says but its a lot of waffle and not enough showing me how it works.
Thanks for the feedback. I do need to work on shortening my videos a little, I wanted this video to get anyone up to speed, regardless of if they had seen a metly yet or not. Splitting it into multiple parts would have helped, I'll look at that for future videos
Fun fact team whyachi did something with the same idea, 3 outside wheels to spin and an inside nav bot direct movement. however i think it lost its second qualifilying round. Or was it third? I forget
That would be Y-Pout and Why Not, both of which went 0-1 in their respective weight classes. The bots failed due to the extreme G-forces the electronics were exposed to. Herr Gepoünden had something similar in the later seasons called a "tornado drive". It worked similar to melty brain but it used external laser guides which would be picked up by detectors aboard the robot.
Well, it's still around, but last time I saw it in action I didn't see the laser guidance rigging. It may have been switched to a proper melty brain system.
Laser guidance is still "proper" melty brain, its just a different way to measure angle and robot RPM. It's actually more stable than any accelerometer based system because accelerometers drift instant to instant (but are accurate longer term). the only real issue is getting permission from the organisers to have the external laser set up
Thats a very complicated topic and depend on the accelerometer you buy. No two accelerometer's measure the same way or give the same reading for the same acceleration. The best idea is to buy an accelerometer and google its name and Arduino, you should find some test code or a full Arduino library
I had a little read of openmelt and have ordered an orangutan and going to use pwm on brushless motor setup, how are you getting on with the rest of the project?
I realize this is a couple years old but why wouldn't you use the rudder channel to rotate the Zero point? Then your directional stick could still work as an x/y directional against the zero point which is now adjustable (rotation) via the rudder.
Thats my plan for sure!
It works as an analogy, but the equation for rotational kinetic energy is actually 1/2*I*W^2
Yes, that is the correct equation for rotational KE, I used the linear version because it is simpler to understand and its something most people have seen before. I was aiming to use simplified ideas to explain a complex control system.
Also for those wondering, in the rotational KE equation:
I is the Moment of inertia
w is the angular velocity
When I first heard 'melty brain', I imagined it was an illegal, totally concealed secret weapon system that trashed your opponent's electronics by lobotomising his control gear's CPU whilst your robot pretended to be a legitimate flipper or drum spinner. Sort of a mini, carefully directed E.M. pulse... Um... Yeah, you're right. Stupid idea. Sorry. (I'll get my coat.)
i thought it was 'multi-brain' which makes more sense
I mean, EMP or elettroshock would work really well, even if the main body is a faraday cage, motors would still burn their coils, so effective it is banned
Why don't you control it like how rc drones/heli controls their movement? The stick with the throttle will control the 'turning' while the other self-centering stick will control all translational movement. Plus you have the shoulder switches to quickly shut off the power if your bot got stuck and needs to rest etc and another switch to maybe 'idle up' to max power.
He even acknowledges the X axis of the right stick at 17:31, but then ignores it and decides to go with a fully custom transmitter due to said X axis not being available, even though having proportional control over the robot's 'forward' direction would be quite a bit better than having just two buttons.
I have an idea to use a radio beacon to orient the robot. If you have an antenna lying on its side both inside the robot and in the control booth, the signal strength of a carrier wave received in the robot will be proportional to how parallel the two antenna are. As the melty spins you could measure this during the rotation and you'd have a signal peak both when facing the operator and when facing away. Probably would break competition regs about radio control though.
This is a commonly used tecnique with melty brains, and is known as a beacon
Hey Ben, I'm late, but I hope you see this. I have wanted to build a "Melty Brain" for a while, but I didn't know how to get started. This video is helping a LOT. My problem is, I don't know where to find parts. If you could tell me what the parts you used are, that would be awesome. If you don't mind, a link to the .stl file of the print would be nice too.
Honestly, I'm not the best person to be asking.
I have tried a melty but I've never got anything to work. So I dont want to lead you astray by giving you links to the stuff that has failed for me.
Good luck with it though!
@@TeamPanicRobotics Thanks!
While all of Nuts spins, only the mass of flails, contributes to the energy of it's weapon. Obviously, that's quite a lot of weight, but I would think less than some of the bar spinners.
So Nuts' advantage is more, that the length of the chains increase it's effective diameter and by extension, the tip velocity of the flail.
Thanks for a the brain melting info! :D
Yeah very true, the chains increase the length (hence the speed as you mentioned) but also decouple the weapon edge from the main weight of the robot.
Looking at it from a KE optimisation standpoint Nuts would be better to be a ring spinner with the chains attached to the ring. They would be able to sping the flails faster and hence hit harder. But Melty Brain provided hardware simplicity and some awe factor
Yep, you're spot on. The really weird thing, though, is just how often Nuts managed some almost unbelievably 'lucky' hits. Despite the lack of mass, the flails unerringly found an opponent's weak spot, smashing an unprotected wheel rim here, dislodging a drive chain sprocket there, or hitting a safety link. I'm sure none of the attacks were actually targeted - it was all just hit-and-hope - but it's that kind of ridiculous good fortune that made Robot Wars worth watching. I liked Nuts for daring to be different. (And I liked the Nutty-Boys in the team, too; they should've used a ♬ Madness song ♪as their theme tune!) ;-)
Wait, how did you get your throttle to be on the right of the controller?
I bought it that way.
When buying a transmitter you are often given the choice of Mode 1 or mode 2
Mode 1 means throttle on the right
mode 2 means throttle on the left.
Currently mode 2 is the most common, but I personally prefer mode 1
www.rc-airplane-world.com/rc-transmitter-modes.html
11:00 Actual explanation of "how it works" starts here. First 11 mins just saying the same thing over and over in different ways. Also explaining how a robot spins on the spot, which to be honest if you are researching how Melty Brain technology works you already know how wheels work and what melty brain is. Great video but please get to the point or at least link to the individual "chapters" of your video.
Maybe reserve "What it is", "How it works" and "Why you'd use it" to their own individual videos. I just want to know how it works and that's what your video says but its a lot of waffle and not enough showing me how it works.
Thanks for the feedback.
I do need to work on shortening my videos a little, I wanted this video to get anyone up to speed, regardless of if they had seen a metly yet or not.
Splitting it into multiple parts would have helped, I'll look at that for future videos
Fun fact team whyachi did something with the same idea, 3 outside wheels to spin and an inside nav bot direct movement. however i think it lost its second qualifilying round. Or was it third? I forget
That would be Y-Pout and Why Not, both of which went 0-1 in their respective weight classes. The bots failed due to the extreme G-forces the electronics were exposed to.
Herr Gepoünden had something similar in the later seasons called a "tornado drive". It worked similar to melty brain but it used external laser guides which would be picked up by detectors aboard the robot.
VestedUTuber weight did they not test that and it ripped itself apart or did they not shockmount their stuff
VestedUTuber actually uses, its still around
Well, it's still around, but last time I saw it in action I didn't see the laser guidance rigging. It may have been switched to a proper melty brain system.
Laser guidance is still "proper" melty brain, its just a different way to measure angle and robot RPM. It's actually more stable than any accelerometer based system because accelerometers drift instant to instant (but are accurate longer term). the only real issue is getting permission from the organisers to have the external laser set up
After looking at it, I want to place some flashing LEDs indicating virtual "forward"/zero direction
Nice!
Any chance in an explanation on how to code the accelerometer? Would I be able to use an arduino?
Thats a very complicated topic and depend on the accelerometer you buy.
No two accelerometer's measure the same way or give the same reading for the same acceleration.
The best idea is to buy an accelerometer and google its name and Arduino, you should find some test code or a full Arduino library
I had a little read of openmelt and have ordered an orangutan and going to use pwm on brushless motor setup, how are you getting on with the rest of the project?
Because of Christmas and a few other things I've had to put my melti on hold for the moment.
I'm hoping to pick it back up in the next month or so
Did you end up finishing off this bot? Would love to see a full build video
3:17 mass isn't weight it is how much "stuff" there is in an object