m(_ _)m ... and many thanks! I'm starting to find my legs with shooting things like this, and I actually kind of hate being on camera, so I really appreciate the encouragement
@@androofroo keep on doing it! It's really awesome seeing this mechanism! And the evaluation you did and everything! Would love to see you tackle some big project like a rover or robot. And you look good on camera too, don't be shy
Excellent! The improvement from the larger pulleys is obvious. Well... it is now you have so clearly demonstrated it! 👍 😜 How are you going to explain the bruising and lacerations on the hospital A&E admission form? 🥴 Looking forward to future iterations... Like & Subscribed.
Here to see your most up-to-date progress on this. If you wish for a use case, I'd suggest looking into ways to transfer human scale movement into smaller scales. If you have a miniature arm with a gripper, for instance, then a full scale one that you wear, so long as you do not twist your arm too much, you can significantly increase the precision of your movements. This can have benefits when drawing (replace gripper with pen), precision assembly, etc. Just be sure to make the pulley ratio the same (else your 90 degree arm rotation could transfer to multiple full rotations on smaller pulleys at the other end). If you scale things down, consider using a bolt with fishing wire wrapped around it, simply wrap the excess slack around a small bolt and screw it down, that will remove the tension and secure the wire. I have not looked into its creep factor, however, which may be a concern.
@@Gaaaaaame I love this suggestion, it reminds me of something MC Escher writes about in terms of the limitations of our own precision ( he built a stack of magnifying lenses to enhance his precision at tiny scales for woodcuts so now I am picturing Escher with a robotic arm).... also, LOL... everyone else wants me to go the other direction and build a mech
@androofroo mechs are overrated (source: insufficient funding to produce my own at the current time) Jokes aside, string mechanisms are one of the more effective ways to assist body motion on an exoskeleton, lightweight and low cost. People have been itching for it online for a while now, as "true" mechs are immensely difficult to design due to control (how do you follow body movements properly), weight, cost, and sources of control (good luck matching muscle strands on the body in terms of mobility options, this is why mechs are stiff compared to humans). Making a puppeted mech is doable, but unlikely to work if free to move due to the significant wire connections, limited movement options, and lack of feedback for the operator.
Also worth noting that the piezoelectric market is steadily growing, I wrote a paper on it last year and it appears to be something worth looking into if you wish to assist human motion in the coming years. Their primary issue is heat dissipation and low relaxation speed if I recall correctly, perhaps best suited for situations already covered by hydraulics for heavy loads, though less likely to damage the user on failure and able to fit smaller form factors with less piping overhead.
One of your main issues with your assembly is rigidity. Your improved tensioning system has assisted in this, but the material and shape of the arm itself is doing you few favors in that regard. The V3 pulley assembly itself is tall, but immediately connects to a significantly thinner arm section. While you get enhanced leverage, you also end up with an underbuilt arm connection for said leverage. Inherently speaking, the larger the arm, the more weight you sling around AND the larger moment of inertia, both working against you. Solutions: 1. optimize models -> cut weight on the arms and pulleys, lowering moment of inertia and weight to improve responsiveness, ease of control, and stability. 2. improve arm moment of area -> in the same way an I-beam gains efficiency per weight in terms of preventing bending, you can make your arm geometry more efficient for a specific axis, improving its efficacy in certain directions. Your rectangle setup works, but could be improved for additional effectiveness 3. supplement arm leverage -> Your arm segments are acted on by the small connector at the end of the pulley. To improve stability, consider extending that connector's length and rigidity. Additionally, you may consider using a wire assembly outside of your pulley to act as another method of rotation transfer, similar to what you have already, except arm segment to segment instead of assembly to assembly. I'm working on some files for you, but I need the following: 1. diameter of wire sheaths 2. length and diameter of your bolts (I use M4x6x6 threaded inserts and 12mm grub screws, for example) 3. pulley diameter 4. arm segment width and height 5. pulley connector dimensions (the part that slots into the next arm right at the end of your pulley) As of now, I have ideas for improved generic arms (similar in function to yours, with weight savings that hopefully lead to improved responsiveness and lower slop), as well as specialized arm geometry (larger footprint than your current ones, pick based on the arm section's purpose) for lifting type loads (alongside the pully direction, such as for when your arm goes up and down a lot or needs to move a heavy object), torquing loads (forces that do not line up with the pulley but instead try to twist the joint sideways, like if you extended your finger and then pushed against the side of your nail), and both (essentially just a larger arm segment with good geometric stability). Your pulley assembly is mostly fine, I have yet to investigate its internal wire pathing to see if sharp corners exist, but other than some weight savings you can add, the pulley seems alright. The fork is less than ideal for stability, but you lack good options without increasing the footprint significantly, and/or utilizing larger connections at the rotation points.
As always, an amazing assessment and very thorough. It'll take me a minute to digest all of this 😅 but thank you for taking the time give me such good feedback.
Ask for the numbers..... that build uses M3 Hardware throughout. No M4 in the design currently.... but possibly in future designs. The wire sheath is PTFE tubing for 3D printing so it's something like 2 mm ID but it's the stuff they sell on Amazon. The cable is something like 1 mm stainless steel. The internal police are I think 10 mm diameter.
@androofroo M4 was an example. I like multiples of 4 due to geometric relations, the strength of M4 inserts are significantly above M3 for a similar size, and M4 bits are the largest bits still considered small scale for most applications (such as for use with pen screwdrivers). I did not intend to suggest that you utilize it yourself, though i would suggest using a larger rod for your pulley. I am looking into 8mm brass rods that work well with 608RS (Forgive me if I misremember their designation) bearings, perhaps look into similar as well? On an unrelated note since we are now both sharing information, would you perhaps know anything about low cost vacuum chambers and/or arc deposition? I am designing a low cost TiN deposition device and the more information the better. High voltage arc generators are high on my priority list if you know much about them.
@@Gaaaaaame I'm printing up a new version today actually that uses M4 pulleys for the internal wire Channel because I like that profile better.... even though I think it's going to add a little weight. And yes indeed as for the Axel of the police I hate riding on that M3 Bolt. My new version which I have not made any videos of yet has a ring bearing instead of a bolt... it seems quieter and more stable.
@@Gaaaaaame as for the vacuum chamber thing... oof, I do not. I don't currently have any kind of welding setup but I'd love to have oxyacetylene again at some point in the future. Maybe if we had more square footage LOL
I would be curious if bicycle shifting cable housing could work even better? The shifting housing has an inner lined tube for low friction and then the outer wall has small, stiff wires embeded in it running the length. The shifter housings are very stiff as opposed to brake cable housing which are made to squish down a bit to reduce the bite of the brakes. Bought in bulk, I wonder if the cable housing could be cost-effective too? Neat project! And it's fun to see the evolution of it!
Thanks for that suggestion! Yeah I really need to look into bike parts as I'm sure they have much tougher tubing I could leverage. I just happen to do a lot of 3D printing and Fabrication so PTFE was my go-to LOL. Thanks for the note (*^ω^)ﾉ
This is amazing! That seems to work really well. Shame that human arms have so many degrees of freedom, I kinda wanna see (or make if I had the time) one that could copy a human shoulder+elbow+wrist and do some juggling exercises..!
Thanks for that! and yeah I'm with you I would love to do a full body manually powered mech 🤘... I'm sure at some point there's diminishing returns with this system... so maybe not that, but it is going to keep getting more complex.
@@DanielSimu indeed it is. In fact it's part of a larger framework which is entirely modular including the software components... it's grown so quickly it's kind of a hot mess actually so half my work is clean up
Ok cool but I think you can offset that with smaller pullies in more of cherry block system You might also line the rolling joints video Rolling contact joints Rolling joint gives different kind of vid lol
Yeah it would be interesting to have a block with multiple pulleys and yeah of course that would help with mechanical advantage for sure but the extra complexity makes me wince a little so I think I'll stick with the single pulleys for now. Also, yeah I searched rolling joint system... 😂
@ that was my thought as well I was considering trying and was like nope pain in the butt When I was doing this I made module That could use to interface with other things
holy effin shhhh! what an amazing video! keep 'em coming! just discovered your channel and I love it!
m(_ _)m ... and many thanks! I'm starting to find my legs with shooting things like this, and I actually kind of hate being on camera, so I really appreciate the encouragement
@@androofroo keep on doing it! It's really awesome seeing this mechanism! And the evaluation you did and everything!
Would love to see you tackle some big project like a rover or robot.
And you look good on camera too, don't be shy
@@gazorbpazorbian 😁
Super nice project, thumbs up and keep us on par.
Thanks for the note I'm enjoying the process
Cool description and demonstration ❤️
Excellent! The improvement from the larger pulleys is obvious. Well... it is now you have so clearly demonstrated it! 👍 😜
How are you going to explain the bruising and lacerations on the hospital A&E admission form? 🥴
Looking forward to future iterations... Like & Subscribed.
@@PiefacePete46 Thanks! I'm hoping it never does 🤞, but if the inevitable hospital visit does arrive I will know I brought it on myself 😅
@@androofroo : Or the "Ghost in the Machine"? 😱
@@PiefacePete46 maybe... waiting for me... emergent behavior 😲
Here to see your most up-to-date progress on this. If you wish for a use case, I'd suggest looking into ways to transfer human scale movement into smaller scales. If you have a miniature arm with a gripper, for instance, then a full scale one that you wear, so long as you do not twist your arm too much, you can significantly increase the precision of your movements. This can have benefits when drawing (replace gripper with pen), precision assembly, etc. Just be sure to make the pulley ratio the same (else your 90 degree arm rotation could transfer to multiple full rotations on smaller pulleys at the other end). If you scale things down, consider using a bolt with fishing wire wrapped around it, simply wrap the excess slack around a small bolt and screw it down, that will remove the tension and secure the wire. I have not looked into its creep factor, however, which may be a concern.
@@Gaaaaaame I love this suggestion, it reminds me of something MC Escher writes about in terms of the limitations of our own precision ( he built a stack of magnifying lenses to enhance his precision at tiny scales for woodcuts so now I am picturing Escher with a robotic arm).... also, LOL... everyone else wants me to go the other direction and build a mech
@androofroo mechs are overrated (source: insufficient funding to produce my own at the current time)
Jokes aside, string mechanisms are one of the more effective ways to assist body motion on an exoskeleton, lightweight and low cost. People have been itching for it online for a while now, as "true" mechs are immensely difficult to design due to control (how do you follow body movements properly), weight, cost, and sources of control (good luck matching muscle strands on the body in terms of mobility options, this is why mechs are stiff compared to humans).
Making a puppeted mech is doable, but unlikely to work if free to move due to the significant wire connections, limited movement options, and lack of feedback for the operator.
Also worth noting that the piezoelectric market is steadily growing, I wrote a paper on it last year and it appears to be something worth looking into if you wish to assist human motion in the coming years. Their primary issue is heat dissipation and low relaxation speed if I recall correctly, perhaps best suited for situations already covered by hydraulics for heavy loads, though less likely to damage the user on failure and able to fit smaller form factors with less piping overhead.
@@Gaaaaaame 🙌 my thoughts every time I hear the request
One of your main issues with your assembly is rigidity. Your improved tensioning system has assisted in this, but the material and shape of the arm itself is doing you few favors in that regard. The V3 pulley assembly itself is tall, but immediately connects to a significantly thinner arm section. While you get enhanced leverage, you also end up with an underbuilt arm connection for said leverage. Inherently speaking, the larger the arm, the more weight you sling around AND the larger moment of inertia, both working against you.
Solutions:
1. optimize models -> cut weight on the arms and pulleys, lowering moment of inertia and weight to improve responsiveness, ease of control, and stability.
2. improve arm moment of area -> in the same way an I-beam gains efficiency per weight in terms of preventing bending, you can make your arm geometry more efficient for a specific axis, improving its efficacy in certain directions. Your rectangle setup works, but could be improved for additional effectiveness
3. supplement arm leverage -> Your arm segments are acted on by the small connector at the end of the pulley. To improve stability, consider extending that connector's length and rigidity. Additionally, you may consider using a wire assembly outside of your pulley to act as another method of rotation transfer, similar to what you have already, except arm segment to segment instead of assembly to assembly.
I'm working on some files for you, but I need the following:
1. diameter of wire sheaths
2. length and diameter of your bolts (I use M4x6x6 threaded inserts and 12mm grub screws, for example)
3. pulley diameter
4. arm segment width and height
5. pulley connector dimensions (the part that slots into the next arm right at the end of your pulley)
As of now, I have ideas for improved generic arms (similar in function to yours, with weight savings that hopefully lead to improved responsiveness and lower slop), as well as specialized arm geometry (larger footprint than your current ones, pick based on the arm section's purpose) for lifting type loads (alongside the pully direction, such as for when your arm goes up and down a lot or needs to move a heavy object), torquing loads (forces that do not line up with the pulley but instead try to twist the joint sideways, like if you extended your finger and then pushed against the side of your nail), and both (essentially just a larger arm segment with good geometric stability).
Your pulley assembly is mostly fine, I have yet to investigate its internal wire pathing to see if sharp corners exist, but other than some weight savings you can add, the pulley seems alright. The fork is less than ideal for stability, but you lack good options without increasing the footprint significantly, and/or utilizing larger connections at the rotation points.
As always, an amazing assessment and very thorough. It'll take me a minute to digest all of this 😅 but thank you for taking the time give me such good feedback.
Ask for the numbers..... that build uses M3 Hardware throughout. No M4 in the design currently.... but possibly in future designs. The wire sheath is PTFE tubing for 3D printing so it's something like 2 mm ID but it's the stuff they sell on Amazon. The cable is something like 1 mm stainless steel. The internal police are I think 10 mm diameter.
@androofroo M4 was an example. I like multiples of 4 due to geometric relations, the strength of M4 inserts are significantly above M3 for a similar size, and M4 bits are the largest bits still considered small scale for most applications (such as for use with pen screwdrivers). I did not intend to suggest that you utilize it yourself, though i would suggest using a larger rod for your pulley. I am looking into 8mm brass rods that work well with 608RS (Forgive me if I misremember their designation) bearings, perhaps look into similar as well?
On an unrelated note since we are now both sharing information, would you perhaps know anything about low cost vacuum chambers and/or arc deposition? I am designing a low cost TiN deposition device and the more information the better. High voltage arc generators are high on my priority list if you know much about them.
@@Gaaaaaame I'm printing up a new version today actually that uses M4 pulleys for the internal wire Channel because I like that profile better.... even though I think it's going to add a little weight.
And yes indeed as for the Axel of the police I hate riding on that M3 Bolt. My new version which I have not made any videos of yet has a ring bearing instead of a bolt... it seems quieter and more stable.
@@Gaaaaaame as for the vacuum chamber thing... oof, I do not. I don't currently have any kind of welding setup but I'd love to have oxyacetylene again at some point in the future. Maybe if we had more square footage LOL
I would be curious if bicycle shifting cable housing could work even better? The shifting housing has an inner lined tube for low friction and then the outer wall has small, stiff wires embeded in it running the length. The shifter housings are very stiff as opposed to brake cable housing which are made to squish down a bit to reduce the bite of the brakes.
Bought in bulk, I wonder if the cable housing could be cost-effective too?
Neat project! And it's fun to see the evolution of it!
Thanks for that suggestion! Yeah I really need to look into bike parts as I'm sure they have much tougher tubing I could leverage. I just happen to do a lot of 3D printing and Fabrication so PTFE was my go-to LOL.
Thanks for the note (*^ω^)ﾉ
@androofroo of course!
This is amazing! That seems to work really well. Shame that human arms have so many degrees of freedom, I kinda wanna see (or make if I had the time) one that could copy a human shoulder+elbow+wrist and do some juggling exercises..!
Thanks for that! and yeah I'm with you I would love to do a full body manually powered mech 🤘... I'm sure at some point there's diminishing returns with this system... so maybe not that, but it is going to keep getting more complex.
@@DanielSimu plus I checked out your channel... of course you would suggest juggling LOL
@@androofroo Yeah, every joint adds a bit of slop and play.. but 6 axis may be achievable, if your system is modular?
@@DanielSimu indeed it is. In fact it's part of a larger framework which is entirely modular including the software components... it's grown so quickly it's kind of a hot mess actually so half my work is clean up
this could be nested inside a CV joint to add axial rotation maybe?
Ok cool but I think you can offset that with smaller pullies in more of cherry block system
You might also line the rolling joints video
Rolling contact joints
Rolling joint gives different kind of vid lol
Yeah it would be interesting to have a block with multiple pulleys and yeah of course that would help with mechanical advantage for sure but the extra complexity makes me wince a little so I think I'll stick with the single pulleys for now.
Also, yeah I searched rolling joint system... 😂
@ that was my thought as well
I was considering trying and was like nope pain in the butt
When I was doing this I made module
That could use to interface with other things
@@preppertechnicianee6013 indeed modular component architecture is the best