@@CarlBugeja seen walking with voice coils (speaker coil). can go much bigger. one thing you miss here is resonance, eg. if 43Hz was the sweet spot you missed it. sweep and watch. i wonder if the membrane might work for a light water vehicle (one sheet, lots of coils). actually i think it was one big coil, one foot, big jumps.
Use just some normal enamel paint and then draw some line under the pad like 🦓. It will make lot more friction. Thin layer of Rubber/hot glue would works same.
i'm impressed! i never bothered with anything but basically hand doodled or laser printer PCBs, but you're tempting me with crazy possibilities here. and this mechanism is some cool stuff. some input from my tinkering experience: laminating very thin sheet iron, say 0,3-0,5mm, onto the back of the printed coils will propably make them much more effective electric magnets. and the magnet sticking to it, could allow it to "snap" off with a LOT more force. once the induced field overcomes the static sticking, it should fling the magnet away with surprising kazamm. if the coil can handle enough current to make this work, the bot will jump into the air. the "stickyness" can be adjusted, by using thicker tape to stick on the iron to increase the distance to the magne.. but the easiest propulsion for a PCB-mobot are vibration motors on a cutout that forms a spring. angling the motor makes it a rather effecive leg. i can recommend waterproof motors, those are nicely enclosed. pancake style vibrators are powerful, (moving mass is further away from the axis of motion) but would have to be solderer to the edge of the pbc. PS: 3M VHB literally keeps the world together.
Hello! What a unique idea and great work. I had some idea after watching. 1) Thinner neodymium magnets 2) Glue PCB coils to thin sheet plastic to hold flat
Did you thought about not using permanent magnet at all ? I'm thinking about a sheet of spring steel cut and bent to shape with electromagnets to get oscillation.
Slightly stiffer magnet arms would transfer more vibration energy to the rest of the body. I know that will make it heavier, too, unless you can find a stiffer light weight plastic. Also, it was probably the heat of soldering that warped the flexible pcb coils. Maybe a lower heat would flatten them again?
have you tried programming it like a hovercraft instead of a walker? high frequency motion like that tends to lack control for steering, but if one actuator provides the HF, the two adjacent actuators could steer left and right in theory
and some capasitors in series with the coils. If you are running them and a higher frequentie. Other wise the coils will just increase there resistance on te imagineary plane and this will cause the current to flow out of fase with the voltage. overal nice progress on your project.
Hey i know LEDs have become far and away more efficient in the past few years but if you want your coils to have a better fighting chance I'd remove that LED to lower the current that's being wasted. Cheers!
You can mount the magnet to stronger body and make the flappy thing free that will make air billed under the flap thing and the robot will flow on the table And you need to billed like a frame under the flaps so it can trapp the air . Keep going this is the first video i sow for you abd i am excited to see more 👍🏻
A suggestion for a way to drive the coils just in case you haven't tried it yet : try to drive multiple coils at the same time with some dephasing in the pwm signal for each leg (for example add a PI/4 phase delay between each leg clockwise to get a rotation in one direction and -PI/4 to get the opposite rotation, or drive only two opposite legs with some dephasing to get a better forward motion)
Driving two coils at the same time would be ideal but I cannot do this with this battery because of its current limitation. But adding a bit of a deal would help I think
Have you ever considered getting a reflow oven? Can you reflow those flex-PCB's? I use a popular toaster-oven to reflow oven conversion kit and it's been working great for years.
I know it would limit the environment that the robot could travel in, but would you get enough current if you used an inductive coil like they used on an inductive charger on a mobile phone. maybe something to have a play with just out of curiosity.
So there is a physical resonance effect? What happens if the coils are operating at different frequencies, is there some harmonic mode that is optimal and others that allow for rotation and direction control?
From the test I made the 10Hz frequency was optimum. The leg's mechanical response for the higher frequencies was not matching for the two legs. Given that I have no sensing feedback on these legs it sort of assumes that both legs are matching. Driving the coils with different frequencies should make one leg go faster than the other so in theory it should turn
You could stabilise/straighten parts of the flex pcb like the large circles by adding in solder lines. Try thickening the rigid pcb too, the weight gain looks like it would be worth it.
good start man and well done. Because the movement needs precise vibration, you'll have to be more precise at placing the magnet and rigid pcb, it will help. Also watch out for the heatgun, flex pcb doesn't like too much heat, it tends to warp sometimes, especially on the extremities, did you try to solder it by hand?
Thanks! and thank you for the tip - didn't know about this effect it has to heat and the edges look a little warped.. i will try to resolder the next one by hand
I think you can try to bend flex PCB down to form some kind of spring to hold robot's weight (or do the same with anoter way: add some between pcb and flexpcb or add some additional flexible material). Previous robot version did that thanks to bended flexbcb. With "springs" you'll need much less power to move the robot and you can use larger battery (or lesser, as u wodn't need much power). Also you can try to stack couple of flexpcb in one robot to drive more current/induction.
Haven't considered it no - adding more legs would increase both the weight and the area touching the ground surface increasing the friction.. But I have considered going for 2 legs should post a video about it soon!
The price of Multilayer piezos is too expensive + to actuate piezos one needs to boost the voltage to at least 200V so I would also have to enlarge the middle space of the robot to make more room for a boost convertor. So yes it is possible but this is way more simpler :-)
That would be very interesting.. However it doesn't have any sensors yet for positional feedback but it can still be done with external sensors and transmit feedback via bluetooth
I think you could make a fish robot with these flexible pcb's. That would be awesome. If you could make a remote control fish that we could place in an aquarium and play around with, and see how the real fish would react.
Hi Carl! You are the only one I can imagine to know something like that: I'm planning to build a toy robot that moves by vibration (very very simple). Do you know if there is a critical mass, until which a robot can be moved by vibration? I thought of 3V Cr2032 and a standard vibration motor like they are in cellphones. Thank you if you have time to answer!
Good work so far! One suggestion: in the independent tests around ~8:00, it looks like the arms themselves are vibrating as opposed to the PCB traces moving away from the arm. Something to try is to add a vertical rib along the length of the arm to increase rigidity. Doesn't have to be more than two or three layers, but that should help test whether the movement is due to the mass of the permanent magnets moving or the PCBs flexing.
The bending PCB is definitely something I noticed but I only mentioned it briefly in the video. During the wired test the flexible pcb was actuated but when adding the mass of the battery the actuation reduced and the repulsion force was being absorbed by the Fr4 pcb.. It still worked this way but there's definitely a relationship with the total mass of the robot. I will analyse this further when I get more batteries (hopefully lighter once)
Hello ! beautiful project. Are you modelist RC ? Can you reduce the tickness of pcb ? With abrasif paper Nr 1200 ou 2400 .... To have flexibility on pcb and increase the frequency because the surface seems weak to me. I made a 0.95gr (1g) dragonfly which flies, rubber engine Total wing area x40
Hi! I like your idea! What do you think about using three legs instead of four? All legs have the same distance to the ground and it should still move in a controllable way.
I believe that stiffening your magnetholders might quiet improve your performance as well. It seems to me that the cantilever structures still flex a lot. That might also acount for a lot of the improvment of your system at the higher frequency. But it is a really cool little robot.
Definitely will improve the system! I might redesign its structure to make it more rigid or make it more thicker - both of these will increase the weight so the first step is for sure finding a lighter battery
Carl Bugeja I understand, just do think about what happen when your arms aren’t flexing any more. Your efficiency goes up, depending on the frequency, a lot. You might wanna put your design in a FEM and estimate how much energy is lost. You might not even need a better battery. Well I am looking forward to your next iteration in any case. :)
You're correct 🙂 like I said I will probably redesign the Fr4 pcb because of the flexing and to make it less fragile. But I still think there are other factors that are effecting it, if you look at the 5V wired test their is a noticeable difference even on how the Fr4 pcb is behaving
I've always thought you could make very thin speakers using an array of printed spirals like the ones you have for feet on your robot. In your robot, the spirals create a magnetic fields that are opposed by permanent magnets. But instead of a permanent magnet, you could just use another printed coil wired in reverse so the two printed coils repel each other. That should give you a basic acoustic speaker.
How many of those electro magnets repelling and contracting would it take to push down a bicycle peddle hmmmm? Think about 40 of those like an accordian. Yes you could just use a selinoid but these pads you get printed seem cooler i guess in appeal? Lol
Kinda wanna see something where you use flex pcb and a string of magnets to make a sorta wave movement. Or heck, make a vibration-based hoverboat.... This test seems to produce some level of lift/mobility, but no obvious control. Use a few of those PCB Motors I keep seeing! ;P
I think you could improve the repulsion by adding an iron/steel disk below the magnets. Right now the magnets might be a bit small for the size of the coils.
i think smaller coils will behave much better - i need to play around with smaller pitches to see how small can it get because these pcbs are only 2-layers
I swear for a second there I thought you had made a new generation drone.
Very impressive progress! Can't wait for the first PCB quadrocopters.
I'm an electrical engineering student from germany and I like this so much! Very awesome!
Ich bin keine Elektronik Ingeneur Student whatever aber auch aus Deutschland😂
magnet pcb is too flexible. so all of its energy lost in there.
i wonder if adding copper layers to the pcb would help
I might do it a bit thicker in the next version 🙂 for the 2 legged version the effect is very minner
@@CarlBugeja just drop some superglue on top layer and to the bottom (for test purpose), also u can reinforce it with some salt or baking soda
I might try that in the next test video.. But the flexing issue was not that of a problem for the 2 legged version
@@CarlBugeja seen walking with voice coils (speaker coil). can go much bigger. one thing you miss here is resonance, eg. if 43Hz was the sweet spot you missed it. sweep and watch. i wonder if the membrane might work for a light water vehicle (one sheet, lots of coils).
actually i think it was one big coil, one foot, big jumps.
10-50hz - would a capacitor help battery draw here? I feel like a few large SMD caps would do well to reduce battery loading
Would it help to glue a rough surface on the bottom side of the flexible coils? Maybe to have some sort of better push¿ Keep up good work!
Not sure if it will help it or increase the friction on the other legs
Why do u use ¿
@@kimi8638 spanish keyboard
Use just some normal enamel paint and then draw some line under the pad like 🦓. It will make lot more friction. Thin layer of Rubber/hot glue would works same.
Fsz.
Kj A@@PriyankaMajumdarOfficial ZAA,mj *
i'm impressed! i never bothered with anything but basically hand doodled or laser printer PCBs, but you're tempting me with crazy possibilities here. and this mechanism is some cool stuff.
some input from my tinkering experience: laminating very thin sheet iron, say 0,3-0,5mm, onto the back of the printed coils will propably make them much more effective electric magnets. and the magnet sticking to it, could allow it to "snap" off with a LOT more force. once the induced field overcomes the static sticking, it should fling the magnet away with surprising kazamm. if the coil can handle enough current to make this work, the bot will jump into the air. the "stickyness" can be adjusted, by using thicker tape to stick on the iron to increase the distance to the magne..
but the easiest propulsion for a PCB-mobot are vibration motors on a cutout that forms a spring. angling the motor makes it a rather effecive leg. i can recommend waterproof motors, those are nicely enclosed. pancake style vibrators are powerful, (moving mass is further away from the axis of motion) but would have to be solderer to the edge of the pbc.
PS: 3M VHB literally keeps the world together.
Magnets! This video gives me a lot of ideas. Thank you!
Hello! What a unique idea and great work. I had some idea after watching.
1) Thinner neodymium magnets
2) Glue PCB coils to thin sheet plastic to hold flat
Did you thought about not using permanent magnet at all ?
I'm thinking about a sheet of spring steel cut and bent to shape with electromagnets to get oscillation.
might be worth doing a slower, smoother frequency ramp. see if you can maybe hit just the right frequency and exploit resonance to increase output
You say legs. I see wings. Motor less wings.
Slightly stiffer magnet arms would transfer more vibration energy to the rest of the body. I know that will make it heavier, too, unless you can find a stiffer light weight plastic. Also, it was probably the heat of soldering that warped the flexible pcb coils. Maybe a lower heat would flatten them again?
have you tried programming it like a hovercraft instead of a walker? high frequency motion like that tends to lack control for steering, but if one actuator provides the HF, the two adjacent actuators could steer left and right in theory
I am so happy there are such smart people like you in the world!
and some capasitors in series with the coils. If you are running them and a higher frequentie.
Other wise the coils will just increase there resistance on te imagineary plane and this will cause the current to flow out of fase with the voltage.
overal nice progress on your project.
Wow! wonderful creation. What are the probable application of this type of robot?
Thanks! Swarm robotics is one of them 🙂
10/10 best maltese youtuber by far.
Hey i know LEDs have become far and away more efficient in the past few years but if you want your coils to have a better fighting chance I'd remove that LED to lower the current that's being wasted. Cheers!
Yeah it could be a good 20ma
Pcbway is my Favorite pcb manufacturer.
I feel like this has so many potential applications...so cool!
put it in a polythene bag (waterproof ) and let it float on water. Did u checked it ?
Amazing work by the way.
Nice to see that
I thought it would be flying
What if you use the flexibility to your advantage by pushing the rear leg up once it is compressed from the first leg pushing up. Just an idea
You can mount the magnet to stronger body and make the flappy thing free that will make air billed under the flap thing and the robot will flow on the table
And you need to billed like a frame under the flaps so it can trapp the air .
Keep going this is the first video i sow for you abd i am excited to see more 👍🏻
A suggestion for a way to drive the coils just in case you haven't tried it yet : try to drive multiple coils at the same time with some dephasing in the pwm signal for each leg (for example add a PI/4 phase delay between each leg clockwise to get a rotation in one direction and -PI/4 to get the opposite rotation, or drive only two opposite legs with some dephasing to get a better forward motion)
Driving two coils at the same time would be ideal but I cannot do this with this battery because of its current limitation. But adding a bit of a deal would help I think
Have you ever considered getting a reflow oven? Can you reflow those flex-PCB's? I use a popular toaster-oven to reflow oven conversion kit and it's been working great for years.
I know it would limit the environment that the robot could travel in, but would you get enough current if you used an inductive coil like they used on an inductive charger on a mobile phone. maybe something to have a play with just out of curiosity.
isn't this just an over engineered hexbug?
Ikr!
The motion and sound during the higher hertz tests kinda made me think of bugs. Would be cool to try and build some sort of robo dragonfly
So there is a physical resonance effect? What happens if the coils are operating at different frequencies, is there some harmonic mode that is optimal and others that allow for rotation and direction control?
From the test I made the 10Hz frequency was optimum. The leg's mechanical response for the higher frequencies was not matching for the two legs. Given that I have no sensing feedback on these legs it sort of assumes that both legs are matching. Driving the coils with different frequencies should make one leg go faster than the other so in theory it should turn
I am new subscriber because i like your project
How often do you subscribe to projects you don't like.
I love this. Lil slappy quadboi.
Adorable.
You're gettting pretty good at editing, good job!!
Using the same technology, can we do power argumentation?
Like for lifting things etc.
Congrats. An over complicated bristle bot.
Why is this genius so underrated
Very interesting and innovative stuff 😘
You could stabilise/straighten parts of the flex pcb like the large circles by adding in solder lines. Try thickening the rigid pcb too, the weight gain looks like it would be worth it.
good idea :) i think a better one would be to have a flexi-rigid pcb instead - not sure if it will improve it by much though
one step closer to me squashing robots i find under my rug.
good start man and well done. Because the movement needs precise vibration, you'll have to be more precise at placing the magnet and rigid pcb, it will help. Also watch out for the heatgun, flex pcb doesn't like too much heat, it tends to warp sometimes, especially on the extremities, did you try to solder it by hand?
Thanks! and thank you for the tip - didn't know about this effect it has to heat and the edges look a little warped.. i will try to resolder the next one by hand
You need to put a fix spacer between the vibrator arms, near the end. This could help stablize the behaviour of how you have calculated.
I think you can try to bend flex PCB down to form some kind of spring to hold robot's weight (or do the same with anoter way: add some between pcb and flexpcb or add some additional flexible material). Previous robot version did that thanks to bended flexbcb.
With "springs" you'll need much less power to move the robot and you can use larger battery (or lesser, as u wodn't need much power).
Also you can try to stack couple of flexpcb in one robot to drive more current/induction.
Super cool! Have you considered 6 arms to have more deterministic locomotion?
Haven't considered it no - adding more legs would increase both the weight and the area touching the ground surface increasing the friction.. But I have considered going for 2 legs should post a video about it soon!
Where I can order custom flex PCBs?
Couldn't you use piezoelectric plates instead?
The price of Multilayer piezos is too expensive + to actuate piezos one needs to boost the voltage to at least 200V so I would also have to enlarge the middle space of the robot to make more room for a boost convertor. So yes it is possible but this is way more simpler :-)
Remember the "Art of UNIX programming": “Make it run, then make it right, then make it fast”.
@@guatagel2454 very useful point rarely mentioned ,thx
Why not a quadcopter with your pcb motors?
Impressive! ..
I'd imagine deep learning being used to teach the robot move around efficiently, what do u think?
Keep up the great work!
That would be very interesting.. However it doesn't have any sensors yet for positional feedback but it can still be done with external sensors and transmit feedback via bluetooth
Try designing it like a toothbrush robot. But with angled bristles maybe?
I think you could make a fish robot with these flexible pcb's. That would be awesome. If you could make a remote control fish that we could place in an aquarium and play around with, and see how the real fish would react.
I wonder if two flexible pcbs and using electrostatic repulsion would give a better force vs deflection?
🤔 i don't think it would be strong enough
can you try to dip the rigid pcb (the one with the magnets) to super glue and sprinkle it with baking powder to make it more stronger.
Hi Carl! You are the only one I can imagine to know something like that: I'm planning to build a toy robot that moves by vibration (very very simple). Do you know if there is a critical mass, until which a robot can be moved by vibration? I thought of 3V Cr2032 and a standard vibration motor like they are in cellphones. Thank you if you have time to answer!
It would be interesting to make a water robot with this idea ...
Nice Project. Only the one magnet working other 3 are not working why?
In the video I powered two legs because the battery cannot supply suffient current to run more legs.. I will retest it once I get the new battery
what kind of programming language did you used? i really like your work. keep moving.
Thanks! I used c
@@CarlBugeja is there any way i can talk with you again? i trying to build something and i need your help. Any social account or anything?
Good work so far! One suggestion: in the independent tests around ~8:00, it looks like the arms themselves are vibrating as opposed to the PCB traces moving away from the arm. Something to try is to add a vertical rib along the length of the arm to increase rigidity. Doesn't have to be more than two or three layers, but that should help test whether the movement is due to the mass of the permanent magnets moving or the PCBs flexing.
The bending PCB is definitely something I noticed but I only mentioned it briefly in the video. During the wired test the flexible pcb was actuated but when adding the mass of the battery the actuation reduced and the repulsion force was being absorbed by the Fr4 pcb.. It still worked this way but there's definitely a relationship with the total mass of the robot. I will analyse this further when I get more batteries (hopefully lighter once)
do carbon fiber substrate pcb's exist? this could eliminate the flex you see with the "rigid" pcb and should up the efficiency
haha don't think so :P might be worth making it a little thicker or even modify the outline design a little to make it more structurally rigid
If you put one way valves on those it will hover!
Hello ! beautiful project.
Are you modelist RC ? Can you reduce the tickness of pcb ? With abrasif paper Nr 1200 ou 2400 ....
To have flexibility on pcb and increase the frequency because the surface seems weak to me. I made a 0.95gr (1g) dragonfly which flies, rubber engine Total wing area x40
I love your work
Cool project, fresh idea, I'm impressed. Subscribed
Hi! I like your idea! What do you think about using three legs instead of four? All legs have the same distance to the ground and it should still move in a controllable way.
Amazing inventin brother
Amazing man great idea and work
Cool project awesome. May i know from where you. Ordered flexible pcb?
Thanks! I got the PCBs from pcbway. The link is in the description 🙂
good job man! you should try add some spheric wheel in the center of robot to increase movement speed
Cool Project. Great Job!
What is the name of the small PCB you use for the brain of this desgin ?
There’s also the fact that your force is almost perpendicular to direction of desired motion.
What controler are you using
What microcontrollers are you using?
CYBLE-022001-00
Great channel! Subbed!
How much did the flex pcb cost how many layers?
The flexpcbs are 2 layers and were sponsored
Love it! How many mah consume?
Thanks! Currently around 200mA with only powering 1 leg
What is that MCU SoC you use?
Keep doing it. You are doing great.
I believe that stiffening your magnetholders might quiet improve your performance as well. It seems to me that the cantilever structures still flex a lot. That might also acount for a lot of the improvment of your system at the higher frequency. But it is a really cool little robot.
Definitely will improve the system! I might redesign its structure to make it more rigid or make it more thicker - both of these will increase the weight so the first step is for sure finding a lighter battery
Carl Bugeja I understand, just do think about what happen when your arms aren’t flexing any more. Your efficiency goes up, depending on the frequency, a lot. You might wanna put your design in a FEM and estimate how much energy is lost. You might not even need a better battery. Well I am looking forward to your next iteration in any case. :)
You're correct 🙂 like I said I will probably redesign the Fr4 pcb because of the flexing and to make it less fragile. But I still think there are other factors that are effecting it, if you look at the 5V wired test their is a noticeable difference even on how the Fr4 pcb is behaving
What is the white stuff on the actuators?
I've always thought you could make very thin speakers using an array of printed spirals like the ones you have for feet on your robot. In your robot, the spirals create a magnetic fields that are opposed by permanent magnets. But instead of a permanent magnet, you could just use another printed coil wired in reverse so the two printed coils repel each other. That should give you a basic acoustic speaker.
Already done it 😉 ua-cam.com/video/36KwlUBadrg/v-deo.html
How do u code in chip will u make a video on it please
Hi, great project Loved it!!!!
Can you tell me which MCU are you using?
Thanks! Its the CYBLE-022001-00 bluetooth mode
@@CarlBugeja Thank you friend :)
Wow. Your life was my dream. Why didn’t I found this guy earlier? 😱😱😱
Lol thx
Where and how I can learn this stuff
Hello Carl
I want to design antenna array. Please give your guidance
You are number one my friend
Fun things come in small packages!
Good work
How many of those electro magnets repelling and contracting would it take to push down a bicycle peddle hmmmm? Think about 40 of those like an accordian. Yes you could just use a selinoid but these pads you get printed seem cooler i guess in appeal? Lol
Brother your my new role model
Cong freaking ratulations bro! That little creepy robot suprisingly works! :D
Awesome.👍I want to make one robot will you please help me for and circuit diagram.
Nice project
Is that battery puffed out? Failed?
Great scott
very interesting project
Kinda wanna see something where you use flex pcb and a string of magnets to make a sorta wave movement. Or heck, make a vibration-based hoverboat.... This test seems to produce some level of lift/mobility, but no obvious control. Use a few of those PCB Motors I keep seeing! ;P
Nice ! Can't wait to see the next update.
The crazy thing is people are doing this kinda stuff in the nano scale
I think you could improve the repulsion by adding an iron/steel disk below the magnets. Right now the magnets might be a bit small for the size of the coils.
i think smaller coils will behave much better - i need to play around with smaller pitches to see how small can it get because these pcbs are only 2-layers