Literally working on an edit of a video where I failed to build a flipper that works as well as I want while you release a video on how you failed to build a flipper that works as well as you want... 😅
ok, so i would think about this in the context of energy conservation: If the motor has peak power output of 1000w, (= 1000j/sec), and you're flipping in 0.1s, then law of conservation of energy says you cannot get more than 100 joules of energy out of it (less considering you're not at peak power output for most of the time) If instead, you had the same motor spinning up a flywheel, which can take 5-10 seconds to spin up, then you can store 5,000-10,000 joules of energy. You'd need some sort of clutch to be able to engage the flywheel into the liftin arm of course, and the flywheel would take constant upkeep (low energy/torque addition from the motor at top speed).
Your calculations are interesting so I looked at the data The motor has a peak input power of 60A*24.4V so ~1464w its obviously not 100% efficient so 1000w output would be a fair assumption From the data the flip seems to happen in 67ms or 0.067 seconds which gives us even less then 100j I am determined to do this as simple as possible and avoid using at least a clutch (because its more weight, more space, more complexity and a weak link in the chain that will likely break) Adding extra mass to the brushless motor might help here because there is slack in the line, so it spins up a little before pulling any load Spin up time is 100ms (before load) so might allow for some energy storage Again I'm trying to do this as simple as possible which meant no clutches or flywheels but maybe a flywheel (or at least extra spun mass) is unavoidable
this. also, having worked on an equally unimpressive spring flipper (turns out, metal springs don't store that much energy), you need way more than 100j of storage to put 100j of upward motion into the opponent, because you have to transfer the energy into the arm, then the arm into them, and they'll usually do a bunch of rotating too, etc, there's lots of losses. bonus, the harder you want to flip someone, the more energy per millisecond you need to deliver, because the only way to flip with more power is to move the same mass faster.
It’s not as much the top speed but the ACCELERATION! That is starting torque. You can either store the energy in a flywheel or spring and release the energy fast or you will have to wait for the motor to get up to speed with a longer rope and except the delay. The rotating mass of the motor is a flywheel. And it takes time to get it up to speed, and don’t neglect decelerating that moving mass when it gets to the end of its stroke like maybe some shock absorbing material.
Would a dog geared clutch assembly that was actuated by a servos engaging his spool work? That way he can spin the motor up to 40kerpm and then hit the clutch to fire the arm?(That way to refire the arm all he needs to do is have a spring or reverse the servo to disengage and then wind up again for the shot... or maybe a simple friction clutch of a rough sandpaper super glued to a flywheel that is engaged on some tpu or rubber? Just spitballing some simple ideas he could try...
Maybe... a clutch between the spool and the motor so it runs full speed then drop the clutch. Another thought, if the cord were attached at either side of the throwing arm pivot point and wound a couple of times around a central pulley it would have better mechanical engagement and more reliable control. One more thing, your motor could easily be bolted directly to a redesigned chassis instead of the base plate.
You’re also losing some of the energy when the back of your robot lifts up, putting forks out the front (not hinged) would help send more of the energy into the opponent. Good luck, love the ambition of it!
You say it has three times the lifting force needed to lift the opponent in theory? So it can at best accelerate them at 3 G over its range of motion, throwing them 3 times its lift height (assuming no energy is lost spinning them or moving them horizontally). Then you anticipate hitting the ceiling which I assume is is far over that. I expect disappointment. Now having watched the rest, it is as I expected: hopefully adjusting the gear ratio can help you get more energy into the target.
with the esc you are using are you able to change the ramp up speed? I am guessing the esc attempts to meet a certain curve for its ramp up based on the feedback sensors. You may also be limited in just how fast the motor can actually spin up from a physics standpoint, the other issue is the momentum of the system - all that is spinning on 1 end is the bell of the motor, where as the "break" of the system is a 15kg weight on the long end of a lever arm. My guess about ways to change this equation would be to have the pivot to the motor be the longer moment, and then a short moment between pivot and the lifting moment. You may be able to add a mechanical advantage lifting mechanism so that a shorter activation range actually gives a much larger movement of your lifting platform. essentially, you need more mechanical advantage on your very light weight motor bell (and i would assume relatively low torque when trying to spin up. Either, adding more weight to the spinning, making the system being a spining flywall and a clutch that dumps it all to the lifting mechanism, changing the configuration of the pivot, adding a second lever mechanism to convert the existing lever into more movement at the end or some combination of the above.
Quick note on the Dyneema as a climber: Dyneema is great in that it's really strong for it's weight in terms of abrasion resistance and static load, but it does struggle with dynamic/shock loading. Basically, because it barely stretches it takes all the force. I don't think it'll be a major problem in this application since you're ramping up to the peak force relatively slowly (by rope standards) but you might need to regularly change out the chord so that it doesn't snap! Nylon gets around this problem, but it stretches too much and I don't think you want another factor limiting your power
This is exactly the idea I was going to propose. Gives the motor time to speed up before the rope actually "engages" this means when it really starts pulling on the arm the system has a bit more momentum (from the rotating motor) *and* is into the motors power band. Obviously you'll have a slight delay when you order a flip but should be small.
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
A flywheel assembly with interlocking teeth like in an impact driver could work - they'd probably need to lock in place like the dog gears in a sequential gearbox rather than sliding like the hammer in an impact driver.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
Cool idea. Maybe a smaller motor with a flywheel and a clutch. Or the same motor if you can cut weight elsewhere. I'm thinking an electronic clutch like the one used to engage pulleys in automotive applications. The thought being, you can charge up a flywheel over a second or two, then dump the energy in a blink by engaging the clutch.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
If you use your data, then do current times the motor constant, you can calculate how much torque you are making. Then torque times distance is work, which is energy. It tells how much height energy that will at most put into the opponent. Thanks for making nice videos :)
You either need a lot more torque or a lot more momentum or potential energy. I am no engineer do excuse my poor explanation but use a small starter motor or add weight to the spinning mass with a longer rope.
Could simply adding more height to the arm holder help (more triangle shaped than pancake)? This would add length to the arm and engagement time (both for rope pull and engagement with bot) and release height as well... this is a simplish fix but make getting flipped over stink. Seems like you could 3d print a double pully (for testing) but i am not sure if you have room for it... this would allow for more time to spin up and would have more torque. Also you might want to consider a simple one way-ish clasp at the front that does not release until a sufficient for is apply this could give you a bit more burst in the release (similar to the ratchet idea below) but simpler.
I am curious how much lifting force the tip of the arm is exerting. That would definitely help figure out where to tweak next to get better flipping performance.
Wonder if a longer lever arm would help. Right now, your pivot point is mounted close to the motor. You need to move the pivot away from it, probably closer to a 50:50 measurement
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
@@TeamPanicRobotics wonder how much merit a clockspring/tpu spiral would have. "String" that returns to a predictable position. Maybe even a short length of measuring tape
Hi Ben @Team_Panic, been watching for a while. So like you said you need to reach the speed faster so how about adding a servo clutch system. You build speed then engage the clutch for max rpm output in a short burst of energy. Got this idea from a Lego Battlebot called Biscuit head Good luck
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
I tried this for nationals a couple of years ago, but using a chain to a sprocket on the pivot. I had a back and forwards with Craig Danby (Foxtrot) who basically said it all comes down to gearing ratio. I was going for full control, not flip, and the gearbox I it was 415:1, but I don't have the motor any more to check the power in - it could lift things, but was too slow. Would love to see yours work so I can use it as inspiration.
That seems like a really low torque setup for what you're trying to do with it. Maybe you could play around with something like a snail cam so it can ramp up the speed/torque as the arm moves through its range of motion.
Did you try spin the motor the other way for the flipper. It appears as though when the flipper is down the angle at which the rope is attached to the flipper is not 90 degrees and is acting as if the motor side of the lever is shorter (which would give less lifting capacity/power but more speed).
Maybe part of the problem is cos your flipper arm doesn't move over a particularly long arc. If you have a bigger flipper arm you can get more travel to build up energy.
Would a dog geared clutch assembly that was actuated by a servos engaging his spool work? That way he can spin the motor up to 40kerpm and then hit the clutch to fire the arm?(That way to refire the arm all he needs to do is have a spring or reverse the servo to disengage and then wind up again for the shot... or maybe a simple friction clutch of a rough sandpaper super glued to a flywheel that is engaged on some tpu or rubber? Just spitballing some simple ideas you could try... After some thought if you want to get fancy you could have a ratcheting mechanism that builds up tension with a simple self rearming trigger to fire the dog clutch at the freewheeling motor...that way you can have the engagement be very fast and not have any slip...
What about using an electric clutch, and have the motor already spun up? With that the motor can get up to speed, and by timing the clutch it might even be possible to don't completely stall the motor.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
This project reminds me of another video about something called a "capstan drive". It would be difficult to fit into your current design but something you could look into.
I feel like 1 gear ratio could make a major difference. Flywheel and clutch would work too just probably not worth it at the expense of complexity, space, and weight...
Blip was the inspiration for doing brushless winch powered flippers. I'm just trying to do it as simple as possible, which means no fly wheel and no clutch. At least for now
@@TeamPanicRobotics Have you already tested twisting the rope instead of rolling it up? Might give the motor a bit more time to spin up before hitting peak torque
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
i think a yolo flipper can sometime be even harder to design effectively than a flywheel flipper. your goal is no flywheel and no clutchg. i think for this an high speed motor is not the best, and that's why you are going for the added reduction ratio. a pulley system may be a way to do so, but i think gear system would be the best at fast, responsive actuation: think of it like a 6kg verion of joe brown grab crab more than a little blip, as your principle is not stored energy but instant torque you have the oversized motor like grab crab, you now need the equivalent of the ikea reduction (something like 100:1) and in terms of pulley the closest is 2^7=128 so 7 pulleys to reduce the motor, this number if why i think gears would be better however if you try to make up for a lower reduction ratio the only option is trying to give the motor as much time as possible to spin up and use the motor bell as flywheel, maybe some rope slack that will be stored somewhere, or the motor pulling a mechanically separate lever that comes into contact with the flip arm only after some time, cause you will need all the energy at the start of the flipper and not building up while you flip in a mechanically connected one last option i can give is similar to my flywheel flipper torsimm: cdn.discordapp.com/attachments/1115345894003462186/1200508334021820416/20240126_190314.mp4?ex=66fee303&is=66fd9183&hm=f6b22ac2d643ba052dc7e4eb0a815e65a7af023cb254813e7f00f6ca6d28622b& that just use a geared drum and a system to smash the drum into a geared winch in order to flip, maybe you can put a beefy pinion on the motor and tilt the motor into na geared winch only when you want to flip... that's said the poor pinion tooth has to survive the full flip energy
"that's said the poor pinion tooth has to survive the full flip energy" This is the reason I'm not super keen on any kind of clutch, its more weight, more space and one more thing to break in combat In the first tests, I did have enough rope for the motor to spin twice before taking up load, I'm hesitant to put in more slack because I dont want it to tangle Your flipper is very cool! 150g?
@@TeamPanicRobotics a clutch completely eliminates the pinion issue due to a non instant force transfer the issue was with no clutch a no direct drive like my flipper. if you want to keep the big ratio and rope i think a torsion string might be the best direct drive actuator you can create, but i still think a big planetary would be far easier to manage. the spinner is old BANTS ruleset so 175g, full plastic (drum included) and to be fair swapping the 350mah lipo and esp electronics for a malenki and 180mah should bring it down to 145g. it's also a show off of the torsion string, with no spinup the tiny 8000kv motor (blheli_so even less low rpm torque) can act as a lifter without stalling due to the insane reduction it provides
Why don't you make your flipper arm equal length from the axis so you can flip from either side? You could launch the arm using a spring mechanism. The motor could preload the spring's energy. Imagine a larger rat trap.
This design is inspired by the battlebot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design, hence just bolting a winch drum to a brushless motor. My goal here is not just to build a flipper, it is to build a winch powered flipper as simple as possible. A spring flipper is a different design and not what I am shooting for. Maybe one day I'll try one, but not in this build
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
@@TeamPanicRobotics ahh i see, didn't realise it was based on and already existing design. battle bot is a pain to try and watch here so i never really bothered.
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
@@TeamPanicRobotics I get that, a knot would render your weapon useless. You could switch the robe to a belt and spool it up like in old magnetic tape echo machines? Maybe make some experiments on how many rotations your motor needs to get up to different speed settings? You could also put a magnet on the belt / robe to detect when to stop the motor in time so it doesn't self destruct
Brushless motors are great at producing high speeds and less great at producing low speed torque. You can most likely use a smaller motor if you take advantage of its strengths. Take a look at the video Mark Rober did (ruclips.net/video/P4gNS0Iiu0Q/видео.html) of his flipper (the second smaller robot) that used a relatively small motor to spin a flywheel to store energy to produce enormous torque.
Funny you mention Mark Rober That robot isn't his, it's a repainted version of blip, the BattleBot. Blip was the inspiration for doing brushless winch powered flippers. I'm just trying to do it as simple as possible, which means no fly wheel and no clutch. At least for now
If your goal is to keep things simple, I think having the pulley be driven by a set of Nautilus gears would add a lot of torque and snappiness in terms of the motion of the flipper
The motor cant spool up quickly enough, so over a certain throttle input you wont have any difference in power Sad to say this concept really doesnt seem feasible. It needs a clutch to allow it to be spun up as otherwise you are losing a huge amount of energy moving the arm at a low speed as the motor accelerates.
Seen you are hitting similar RPM you are probably reliant on the motors mass more than the motor power. Why not add more rope and a flywheel so it can store up more energy
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
for weight reasons its an aluminium bracket, I cant weld to aluminium. I know its possible (and tricky) I just dont have access to any of the tools needed
You’re also losing some of the energy when the back of your robot lifts up, putting forks out the front (not hinged) would help send more of the energy into the opponent. Good luck, love the ambition of it!
Literally working on an edit of a video where I failed to build a flipper that works as well as I want while you release a video on how you failed to build a flipper that works as well as you want... 😅
hahaha must be something in the air... or maybe a lack of things in the air 😅
ok, so i would think about this in the context of energy conservation:
If the motor has peak power output of 1000w, (= 1000j/sec), and you're flipping in 0.1s, then law of conservation of energy says you cannot get more than 100 joules of energy out of it (less considering you're not at peak power output for most of the time)
If instead, you had the same motor spinning up a flywheel, which can take 5-10 seconds to spin up, then you can store 5,000-10,000 joules of energy.
You'd need some sort of clutch to be able to engage the flywheel into the liftin arm of course, and the flywheel would take constant upkeep (low energy/torque addition from the motor at top speed).
Your calculations are interesting so I looked at the data
The motor has a peak input power of 60A*24.4V so ~1464w its obviously not 100% efficient so 1000w output would be a fair assumption
From the data the flip seems to happen in 67ms or 0.067 seconds which gives us even less then 100j
I am determined to do this as simple as possible and avoid using at least a clutch (because its more weight, more space, more complexity and a weak link in the chain that will likely break)
Adding extra mass to the brushless motor might help here because there is slack in the line, so it spins up a little before pulling any load
Spin up time is 100ms (before load) so might allow for some energy storage
Again I'm trying to do this as simple as possible which meant no clutches or flywheels but maybe a flywheel (or at least extra spun mass) is unavoidable
this. also, having worked on an equally unimpressive spring flipper (turns out, metal springs don't store that much energy), you need way more than 100j of storage to put 100j of upward motion into the opponent, because you have to transfer the energy into the arm, then the arm into them, and they'll usually do a bunch of rotating too, etc, there's lots of losses. bonus, the harder you want to flip someone, the more energy per millisecond you need to deliver, because the only way to flip with more power is to move the same mass faster.
It’s not as much the top speed but the ACCELERATION! That is starting torque. You can either store the energy in a flywheel or spring and release the energy fast or you will have to wait for the motor to get up to speed with a longer rope and except the delay. The rotating mass of the motor is a flywheel. And it takes time to get it up to speed, and don’t neglect decelerating that moving mass when it gets to the end of its stroke like maybe some shock absorbing material.
Would a dog geared clutch assembly that was actuated by a servos engaging his spool work? That way he can spin the motor up to 40kerpm and then hit the clutch to fire the arm?(That way to refire the arm all he needs to do is have a spring or reverse the servo to disengage and then wind up again for the shot... or maybe a simple friction clutch of a rough sandpaper super glued to a flywheel that is engaged on some tpu or rubber? Just spitballing some simple ideas he could try...
Maybe... a clutch between the spool and the motor so it runs full speed then drop the clutch. Another thought, if the cord were attached at either side of the throwing arm pivot point and wound a couple of times around a central pulley it would have better mechanical engagement and more reliable control. One more thing, your motor could easily be bolted directly to a redesigned chassis instead of the base plate.
You’re also losing some of the energy when the back of your robot lifts up, putting forks out the front (not hinged) would help send more of the energy into the opponent. Good luck, love the ambition of it!
If you increase the length of the rope it will allow the motor to spin up before the rope pulls tight
You say it has three times the lifting force needed to lift the opponent in theory? So it can at best accelerate them at 3 G over its range of motion, throwing them 3 times its lift height (assuming no energy is lost spinning them or moving them horizontally). Then you anticipate hitting the ceiling which I assume is is far over that. I expect disappointment. Now having watched the rest, it is as I expected: hopefully adjusting the gear ratio can help you get more energy into the target.
with the esc you are using are you able to change the ramp up speed? I am guessing the esc attempts to meet a certain curve for its ramp up based on the feedback sensors. You may also be limited in just how fast the motor can actually spin up from a physics standpoint, the other issue is the momentum of the system - all that is spinning on 1 end is the bell of the motor, where as the "break" of the system is a 15kg weight on the long end of a lever arm.
My guess about ways to change this equation would be to have the pivot to the motor be the longer moment, and then a short moment between pivot and the lifting moment. You may be able to add a mechanical advantage lifting mechanism so that a shorter activation range actually gives a much larger movement of your lifting platform.
essentially, you need more mechanical advantage on your very light weight motor bell (and i would assume relatively low torque when trying to spin up. Either, adding more weight to the spinning, making the system being a spining flywall and a clutch that dumps it all to the lifting mechanism, changing the configuration of the pivot, adding a second lever mechanism to convert the existing lever into more movement at the end or some combination of the above.
Aww yeah the yoloflipper is back!
Quick note on the Dyneema as a climber: Dyneema is great in that it's really strong for it's weight in terms of abrasion resistance and static load, but it does struggle with dynamic/shock loading. Basically, because it barely stretches it takes all the force. I don't think it'll be a major problem in this application since you're ramping up to the peak force relatively slowly (by rope standards) but you might need to regularly change out the chord so that it doesn't snap!
Nylon gets around this problem, but it stretches too much and I don't think you want another factor limiting your power
What if the string was longer so it had more time to spin before a load was placed on it?
This is exactly the idea I was going to propose. Gives the motor time to speed up before the rope actually "engages" this means when it really starts pulling on the arm the system has a bit more momentum (from the rotating motor) *and* is into the motors power band. Obviously you'll have a slight delay when you order a flip but should be small.
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load
In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load
Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
I'm looking at a choo choo flipper mechanism but it's on a beetle scale. It's been a fun side quest
is there a simple mechanism (like a ratchet) that will let you store kinetic energy from the motor and can engage the pulley at speed?
A flywheel assembly with interlocking teeth like in an impact driver could work - they'd probably need to lock in place like the dog gears in a sequential gearbox rather than sliding like the hammer in an impact driver.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
Try capstan drive for high precision and torque works great with rope as reducer
Cool idea. Maybe a smaller motor with a flywheel and a clutch. Or the same motor if you can cut weight elsewhere. I'm thinking an electronic clutch like the one used to engage pulleys in automotive applications.
The thought being, you can charge up a flywheel over a second or two, then dump the energy in a blink by engaging the clutch.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
If you use your data, then do current times the motor constant, you can calculate how much torque you are making. Then torque times distance is work, which is energy. It tells how much height energy that will at most put into the opponent.
Thanks for making nice videos :)
You either need a lot more torque or a lot more momentum or potential energy. I am no engineer do excuse my poor explanation but use a small starter motor or add weight to the spinning mass with a longer rope.
Could simply adding more height to the arm holder help (more triangle shaped than pancake)? This would add length to the arm and engagement time (both for rope pull and engagement with bot) and release height as well... this is a simplish fix but make getting flipped over stink.
Seems like you could 3d print a double pully (for testing) but i am not sure if you have room for it... this would allow for more time to spin up and would have more torque. Also you might want to consider a simple one way-ish clasp at the front that does not release until a sufficient for is apply this could give you a bit more burst in the release (similar to the ratchet idea below) but simpler.
I am curious how much lifting force the tip of the arm is exerting. That would definitely help figure out where to tweak next to get better flipping performance.
Wonder if a longer lever arm would help. Right now, your pivot point is mounted close to the motor. You need to move the pivot away from it, probably closer to a 50:50 measurement
Looks like you need a lot longer length of dyneema for more inertia
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load
In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load
Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
@@TeamPanicRobotics wonder how much merit a clockspring/tpu spiral would have. "String" that returns to a predictable position. Maybe even a short length of measuring tape
Hi Ben @Team_Panic, been watching for a while. So like you said you need to reach the speed faster so how about adding a servo clutch system. You build speed then engage the clutch for max rpm output in a short burst of energy. Got this idea from a Lego Battlebot called Biscuit head Good luck
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
cool looks like a lot of work
I tried this for nationals a couple of years ago, but using a chain to a sprocket on the pivot. I had a back and forwards with Craig Danby (Foxtrot) who basically said it all comes down to gearing ratio. I was going for full control, not flip, and the gearbox I it was 415:1, but I don't have the motor any more to check the power in - it could lift things, but was too slow. Would love to see yours work so I can use it as inspiration.
That seems like a really low torque setup for what you're trying to do with it. Maybe you could play around with something like a snail cam so it can ramp up the speed/torque as the arm moves through its range of motion.
Did you try spin the motor the other way for the flipper. It appears as though when the flipper is down the angle at which the rope is attached to the flipper is not 90 degrees and is acting as if the motor side of the lever is shorter (which would give less lifting capacity/power but more speed).
Thats a good point, I wasnt paying much attention to which way I was spinning the motor, but I should
Maybe part of the problem is cos your flipper arm doesn't move over a particularly long arc. If you have a bigger flipper arm you can get more travel to build up energy.
Would a dog geared clutch assembly that was actuated by a servos engaging his spool work? That way he can spin the motor up to 40kerpm and then hit the clutch to fire the arm?(That way to refire the arm all he needs to do is have a spring or reverse the servo to disengage and then wind up again for the shot... or maybe a simple friction clutch of a rough sandpaper super glued to a flywheel that is engaged on some tpu or rubber? Just spitballing some simple ideas you could try... After some thought if you want to get fancy you could have a ratcheting mechanism that builds up tension with a simple self rearming trigger to fire the dog clutch at the freewheeling motor...that way you can have the engagement be very fast and not have any slip...
What about using an electric clutch, and have the motor already spun up? With that the motor can get up to speed, and by timing the clutch it might even be possible to don't completely stall the motor.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
This project reminds me of another video about something called a "capstan drive". It would be difficult to fit into your current design but something you could look into.
I watched that same video!
It was actually in that video that I learnt about the rope I am using
Use a longer rope so the motor gets time to spin up before the load is applied.
would it be potentially better to switch to a gear system instead of a rope system?
for testing maybe, but I would like to have this work with rope eventually
I feel like 1 gear ratio could make a major difference. Flywheel and clutch would work too just probably not worth it at the expense of complexity, space, and weight...
Take a look at how Blip from battlebots does it, they also use an electric motor for their flipping mechanism
Blip was the inspiration for doing brushless winch powered flippers. I'm just trying to do it as simple as possible, which means no fly wheel and no clutch. At least for now
@@TeamPanicRobotics Have you already tested twisting the rope instead of rolling it up?
Might give the motor a bit more time to spin up before hitting peak torque
Gearbox?
You could add more slack to the rope which would let the motor reach max speed
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load
In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load
Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
i think a yolo flipper can sometime be even harder to design effectively than a flywheel flipper.
your goal is no flywheel and no clutchg. i think for this an high speed motor is not the best, and that's why you are going for the added reduction ratio. a pulley system may be a way to do so, but i think gear system would be the best at fast, responsive actuation: think of it like a 6kg verion of joe brown grab crab more than a little blip, as your principle is not stored energy but instant torque
you have the oversized motor like grab crab, you now need the equivalent of the ikea reduction (something like 100:1) and in terms of pulley the closest is 2^7=128 so 7 pulleys to reduce the motor, this number if why i think gears would be better
however if you try to make up for a lower reduction ratio the only option is trying to give the motor as much time as possible to spin up and use the motor bell as flywheel, maybe some rope slack that will be stored somewhere, or the motor pulling a mechanically separate lever that comes into contact with the flip arm only after some time, cause you will need all the energy at the start of the flipper and not building up while you flip in a mechanically connected one
last option i can give is similar to my flywheel flipper torsimm: cdn.discordapp.com/attachments/1115345894003462186/1200508334021820416/20240126_190314.mp4?ex=66fee303&is=66fd9183&hm=f6b22ac2d643ba052dc7e4eb0a815e65a7af023cb254813e7f00f6ca6d28622b& that just use a geared drum and a system to smash the drum into a geared winch in order to flip, maybe you can put a beefy pinion on the motor and tilt the motor into na geared winch only when you want to flip... that's said the poor pinion tooth has to survive the full flip energy
"that's said the poor pinion tooth has to survive the full flip energy"
This is the reason I'm not super keen on any kind of clutch, its more weight, more space and one more thing to break in combat
In the first tests, I did have enough rope for the motor to spin twice before taking up load, I'm hesitant to put in more slack because I dont want it to tangle
Your flipper is very cool! 150g?
@@TeamPanicRobotics a clutch completely eliminates the pinion issue due to a non instant force transfer the issue was with no clutch a no direct drive like my flipper. if you want to keep the big ratio and rope i think a torsion string might be the best direct drive actuator you can create, but i still think a big planetary would be far easier to manage.
the spinner is old BANTS ruleset so 175g, full plastic (drum included) and to be fair swapping the 350mah lipo and esp electronics for a malenki and 180mah should bring it down to 145g. it's also a show off of the torsion string, with no spinup the tiny 8000kv motor (blheli_so even less low rpm torque) can act as a lifter without stalling due to the insane reduction it provides
What did muck do for you to beat him up so much?
I think this is Scoop, isnt Muck a truck of some description?
I cant really remember, the only ones I do are Lofty and Dizzy
Why don't you make your flipper arm equal length from the axis so you can flip from either side?
You could launch the arm using a spring mechanism. The motor could preload the spring's energy. Imagine a larger rat trap.
This design is inspired by the battlebot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design, hence just bolting a winch drum to a brushless motor.
My goal here is not just to build a flipper, it is to build a winch powered flipper as simple as possible.
A spring flipper is a different design and not what I am shooting for. Maybe one day I'll try one, but not in this build
A smaller motor a flywheel and a electro/magnetic clutch. Good way to get a lot of power out of a compact system.
This design is inspired by the battle bot Blip, which uses a flywheel and a wound string to power the flipper
I wanted to see how simple you could make that design which means no flywheel and no clutch. At least for now
@@TeamPanicRobotics ahh i see, didn't realise it was based on and already existing design. battle bot is a pain to try and watch here so i never really bothered.
Make the robe longer, give the motor time to speed up and get some momentum going
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load
In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load
Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
@@TeamPanicRobotics I get that, a knot would render your weapon useless. You could switch the robe to a belt and spool it up like in old magnetic tape echo machines? Maybe make some experiments on how many rotations your motor needs to get up to different speed settings? You could also put a magnet on the belt / robe to detect when to stop the motor in time so it doesn't self destruct
Brushless motors are great at producing high speeds and less great at producing low speed torque. You can most likely use a smaller motor if you take advantage of its strengths. Take a look at the video Mark Rober did (ruclips.net/video/P4gNS0Iiu0Q/видео.html) of his flipper (the second smaller robot) that used a relatively small motor to spin a flywheel to store energy to produce enormous torque.
Funny you mention Mark Rober
That robot isn't his, it's a repainted version of blip, the BattleBot.
Blip was the inspiration for doing brushless winch powered flippers. I'm just trying to do it as simple as possible, which means no fly wheel and no clutch. At least for now
If your goal is to keep things simple, I think having the pulley be driven by a set of Nautilus gears would add a lot of torque and snappiness in terms of the motion of the flipper
The motor cant spool up quickly enough, so over a certain throttle input you wont have any difference in power
Sad to say this concept really doesnt seem feasible. It needs a clutch to allow it to be spun up as otherwise you are losing a huge amount of energy moving the arm at a low speed as the motor accelerates.
Seen you are hitting similar RPM you are probably reliant on the motors mass more than the motor power. Why not add more rope and a flywheel so it can store up more energy
In the initial tests I had enough rope slack for the motor to spin twice before taking up any load
In the final test I got the geometry and rope length a little wrong so it only spun once before taking up the load
Adding more slack is a delicate balance, as I agree, more spool up time would help but I dont want to add so much slack that the rope tangles after the first flip
I’d love to make a control bot someday
6:40 why not just weld the nuts to the bracket?
for weight reasons its an aluminium bracket,
I cant weld to aluminium. I know its possible (and tricky) I just dont have access to any of the tools needed
You’re also losing some of the energy when the back of your robot lifts up, putting forks out the front (not hinged) would help send more of the energy into the opponent. Good luck, love the ambition of it!