Just a tip from a hiker who has been using Dyneema for years. If you want your rope to last longer, don't let the rope contact itself. In the video it looked to be touching where it was wrapped around. Dyneema is very strong, but if it is rubbing against itself it will get damaged. I learned this from putting a Dyneema tent into a Dyneema stuff sack. Cool video! Also, the Your Mama joke took me by surprise and cracked me up. :)
I've used dyneema a lot and noticed it degrades less when it rubs against itself than anything else. Hell i have really old whoopie slings that look brand new apart from loop that hooks on but clutch looks fine. I get that rope against rope contact is bad but with dyneema if you don't create huge temperatures i don't really see it being worse than any other rope... hell it would create less of an issue due to very low friction coefficient of UHMWPE. Could you tell me how does your damage look like and is it the case of rubbing when mostly parallel or is it the case of low diameter bend of quite inelastic ropes?
@@Przemo-c Abrasion is the weak point of Dyneema. So much so that every tent manufacturer that makes them with Dyneema recommends folding rather than stuffing your tent in the stuff sacks. Tarp-Tent, Gossamer Gear, Zpacks, Durston , to name a few who recommend it.
As a spearfisherman I have used the same dynemma line for years and year in harsh conditions, big fish in highly abrasive environments such as around coral. Dynemma is actually extremely good at abrasion resistance, more than any other line, you just need to the the correct type. There are ones that are impregnated with wax which significantly reduces failure due to abrasion, also it's important to get the cored dynemma as that extra stiffness reduces wear due to abrasion.
I'm not a climber but I like to watch the channel *"How NOT 2."* I've heard the warning about having ropes rubbing against itself many times (particularly with Dyneema). I highly recommend the channel to anyone interested in ropes and rope work. I also like the channel *"Hard Is Easy."*
What is this, a scientific testing approach for an actuator, a solid testing phase, a good explanation of physics, and freely available files? You have earned yourself an instant subscriber, that was top notch!
well said. science applications and explanations will trump (hah 'Trump...) most channels...SmarterEveryDay is cool. but it was you that got the sub. TY
As a sailor, I saw you messing with bent steel wire and paracord and was like "noooo!" but then you found what I believe is the right answer, and I was "yesss!", not because I agree, but because it says a lot about your process, and that I will learn from your other work. Subbed, and cheers!
I have a master's degree in electrical engineering, and it never ceases to amaze me just how many similarities exist between the mathematics of electromagnetics and the mathematics of mechanics. On one hand, that makes perfect sense (since they both boil down to physics). On the other hand... just when I think I've found a perfect analogy between the two, I think of an edge case that makes my analogy nearly useless. My intention is always to give someone who doesn't have EE training a kind of "horse sense" for how electricity behaves. Consider how water moves through a pipe compared to how electricity moves through a wire. That analogy is fairly intuitive and helpful away from extreme cases. When I started thinking about how a "short circuit" could be represented in the "water through a pipe model", though, that "helpful intuition" began to fall apart very quickly. :(
@@slidingconstant your circuitry is represented by pipes with a huge diameter and your resistive load is represented by pipes of a much smaller diameter. The water throughput is resisted by the small diameter of the resistive load. if you replace the small diameter piping with a the big diameter piping your waterflow increases drasticly and you have your "short circuit water flow"
+1 on this comment, I had amazing electrical, computer and software engineering classes in uni but my statics/dynamics/thermo classes were so poorly taught I just did the minimal amount of work to get by. Awesome work on this video, I'm going to find a project to work on using this and I can't wait to find time to work through the other videos.
I'm a retired EE, with robotic experience going back as a sub-contractor to Disney's Imagineering, in the 1980s. Around the time of the Mike Eisner power grab. Back then merging electrical and mechanical parameters, was part of control systems classes, in school. That said, this channel an its creator's engineering skills are top notch, and he explains it excellently. Back in the old days, with 8-bit microprocessors, and microcontrollers, running only at a few MHz, we'd do the kinematic calculations on a spreadsheet, to generate look up tables, that in turn, would be embedded in the code. Which you can still do, if your application warrants it. But you have the option of using a beefier ARM processor, and get a bit flexibility, running those calculations, in the actual runtime code. I still tinker, so this capstan drive, may show up in one of my future toys !
The drive is brilliant. Your capacities as a designer and engineer are obviously evolving, but most impressive is your capacity to explain things clearly.
Could not agree more, awesome build awesome and clear explanations and story telling! I just found some of the in between shots where you just twiddle the pen a bit distracting as i was trying to figure out what you were trying to show me and I was listening less. Maybe something to keep in mind. Would love to see this channel grow for you to build ever cooler things as I am sure you will ❤
You are going to give Stuff Made Here a run for his money in no time. I like how you share the math instead of completely glossing over it. Amazing work.
Even earlier than settlement-- cave dwellers were rather advanced in terms of materials and craftsmanship, having mastered both stone, dyes and fire. Technologically, pottery isn't far off. The oldest ropes are found fossilized, from the very start of the Stone Age, when the only structures we find were probably semi-permanent work-huts (judging by the materials).
@@ontoverseThat's because people were relatively advanced before they started living in caves. Most early people dwelt in tents, only moving to caves as they expanded out away from Eden.
lol Then proceeds to reinvent the gear replacing the sheer force from interconnected teeth for rope and friction which goes back even farther to the most ancient grandfather technology that started it all. The mastery of friction to first control fire allowing us to illuminate, heat and become the superior troglodyte society finger painting with bat poop no one could have expected would one day dominate the food chain. We make fun of them now, but the cavemen were the sophisticated gentlemen of there time and I respect the nod to our first golden age.
I am impressed. As a retired engineer with a hobby Prusa MK4 printer you have opened my mind to some very neat concepts. I have no desires to make anything this complex but your work shows the possibilities. Thank you for making this informative video.
I laughed out loud during the first jumping part. It just looks so much like a little creature that is inexplicably frustrated with the ground and trying it's hardest to beat it up
It's all in how you look at the shape it did draw. Maybe it wasn't done yet, like how maybe that skateboard rider just took a push ups break before doing two more kick-flips in a row...
I found this very interesting. I am a retired gearbox engineer, consultant, repairer, etc. We designed and built capstans for high-end yachts. These were hypercycloidal gearboxes. They were manufactured to a high degree of accuracy and we achieved minimum backlash, of around 0.002" maximum. Have you determined the maximum loading that rope can withstand ? Probably the most accurate gearbox we made was for satellite tracking, where the backlash was basically 0 ( reality check it was around 0.0005" ). Everything in the gearbox was precision ground and was hand built. For some of the smaller satellite dishes I could see a rope driven gearbox system would be advantageous, as long as the 'stretch' factor was addressed. I hope you do more on the rope gearbox as I personally find it very informative. That's life when you've been an engineer all your life, your mind cannot retire. Thank you.
Hey, I'm looking for a gear ratio of 1:4000 (yeah...). Is that even possible, or would the forces on the 'inner' most gear be so high it's just not feasible? Also, what can you tell me about low friction gear boxes?
I make roller bearings and we can achieve precision and sometimes require precision of 0.0001, how do my old machines do it with what looks like a normal train driving the motions?
Looks like ~1.5mm dyneema, under basically ideal loading conditions, it should handle >1.5 kN - the 3D prints will pretty much always break first in both overload and fatigue
Satellites may complicate this, but to deal with creep couldn't you just have another spring, motor, hydraulic, etc. that is pushing/pulling the drums apart? Then couldn't any creep be cancelled out by just pushing them further apart?
Awesome drive, testing and storytelling! As a sailor, we use dynema and spectra quite a lot as well as concepts of friction and pivot points, and many rope knots are based on the capstan principles of friction. Those principles have been in play for hundreds of years surviving in the harshest conditions on the planet. Given the right rope and your drum made out of stainless, I could almost guarantee your motors will fail before the rope and drum. Looking forward to your next update!
Thank you for the comment! Back then i made something with rope for robotics and stumbled into rope tensioning issue. Went to camping knots and how they tighten rope to solve my issue, but seeing ur comment on sailor rope which is the higher level of rope knots just opened my eyes
I appreciate the willingness to test until it breaks and showing that purposefully. Many creators often depict breakage as accidental and not necessary to the process.
Dyneema is a crazy material to work with. It’s absurdly strong, absurdly low stretch, absurdly light weight, and freakishly low friction. Not sure if you ran into any issues with its low friction, but it can be weird in how different to other ropes it is, because in some situations well known knots don’t behave as expected because of its lack of friction, something to keep an eye out for if you use it more. Also it doesn’t like fusing to itself when you melt it, annoying for trying to keep the ends from fraying. Super glue works though if you ever need that knowledge.
You are right about knots being useless, in fact high load applications require rope splicing as it creates a vastly increased surface-interface to leverage friction.
I bought a bunch of drone parachutes at an auction and the shroud lines are made of Dyneema. I thought it was going to be a nightmare to untangle the chutes as most of them are wadded up. Actully it was super easy and fun because the Dyneema slides across itself so easily. Most of the tangles come out with just a bit of shaking. A great feature for a parachute, where tangled lines could mean it does not deploy properly.
As an electrical/control engineer, I always regretted not getting an ME degree or minor. Mechatronics was not a "thing" in the 80's. I recently retired after 40 years of control engineering and have been working on my ME skillset with help from UA-cam folks like yourself. I've been screen saving your math and walking through it. Keep up the great work!
Mechatronics still feels like a burgeoning field today. I'm taking an ME degree with a mechatronics certificate. It's the closest I can get without a double major with EE! Fortunately, I think motor control and electronic drive systems will continue to become more common and thus the electromechanical field should grow to match! Not to mention how important control systems are in our modern world. The topics in this subject span the breadth of the EE and ME worlds--so the crossover is pretty clear
@@Trenz0 My only caveat for an actual Mechatronics degree is that many states do not allow those folks to obtain a professional engineer (PE) license which is required for many industries. Having a (pure) ME degree (or EE in my case) is typically a requirement. To be honest, having a PE these days is like printing money but it takes a few years working under other PE's to obtain one. The PE test is also equivalent to passing the bar exam. I obtained a Control Engineer PE license which was very rare in my state (WA) for the majority of my career.
Your channel is a gold mine but open for the public for free! These videos will inspire a lot of people into robotics (including me lol). Thanks for your work sir!
Yes, but Dyneema DM20 is new to me. (Hmmm... unsurprisingly, it's very pricey rope.) Aaed, I'm very surprised that the reduction ratio wasn't a bit *_under_* 8:1, instead of a bit over 8:1, since it seems like you should add the diameter of the cord to the diameter of the drum, to get the true effective drum diameter. The rope diameter will increase the small drum's effective diameter by a larger percentage than it increases the large drum's effective diameter, which should reduce the ratio of the two. It is a puzzlement.
@@nadirbalena691 he said everything mate. moving, hoisting, reefing, trimming the sails is all done by rope. And since a sail with wind in it can exert several tons of pressure on the main boom or the foresail you need aid to pull them in. For the mainsail mostly block and tackle, or similar pulley systems, and for the foresails capstans to hold them tight. A capstan drive along with bowden cables is used for the steering on sailing yacht. Smaller boats can do with a tiller.
@@ArfurFaulkesHake Everyone forgot the most iconic one. the ANCHOR! At least on older boats like you see pirates had. Would be cool if the navy still had a huge manual anchor.
scientific testing approach for an actuator, a solid testing phase, a good explanation of physics, and freely available files? You have earned yourself an instant subscriber, that was top notch!
I have utmost respect and admiration for an engineer confident enough in his skills and knowledge that he chooses to film himself working while still in his pyjama bottoms. Great video. Subbed!
This is pretty much the same comment that I was going to compose. That is to say, it covers every point and sense that I wanted to (Especially the confidence/pajama conveyance). I do not think that I could quickly come up with something better. So, with a sincere nod, I tip of my hat to you, OxABADCAFE. I do hope you will pardon me as I hop on to your coattails and offer up a hearty, "Hear, hear!". Oh, and, of course I've subscribed.
I think this might be it, although It doesn't make sense why the gap between the drums matters and not the thickness of the rope. I'll eventually do some testing to confirm if this is the correct way. Until then @HighFlights is hereby declared to be my favorite subscriber.
The issue is the cable and will not be at the exact center. The issue is the cable has limited compression strength (in theory zero). Your highest tension location is at the furthest from the bend center but the cable ope will deform to an oval because of this tension changing the effective diameter. It will also change a different degree on a different radius bend. You can test this by taking a known diameter spool and wrapping a known length of cable around it. You will not get the exact turns you calculate. This should be enough for you to visualize the problem. It is what we referr to as a "hard" problem because it will be application specific and there is no uniform formulaic answer. Just think if you had an infinitely stiff straight fiber cable the only part with load would be the fiber furthest from the spool center and the effective diameter of the spool would therefore be at the spool diameter + 2x rope diameter - 2x rope deformation from round.
That's really interesting. Years ago I took apart an old hard drive from the 1980s and it used what I have now learned is a capstan drive to move the head across the platter, back when they used stepper motors instead of magnetic voice coils. Thanks to this video I also learned why they used a thin metal band (where the rope is used here) and that they probably chose the capstan drive due to the minimal backlash. Thanks for explaining everything so well in this video!
If anyone is interested, Dyneema is a popular performance upgrade in the yachting community. by replacing the steel wire mast stays with something lighter, you reduce the weight at the top of the mast. as you want the thing to stay upright, you make the bottom bit (the boat) heavy, and due to the massive lever of the mast, even a small weight reduction up top can improve stability. If you want a small bit to experiment with and dont want a whole roll or to wait for shipping, you may be able to get either offcuts or by the meter from your local marine supply store (chandlery). very cool vid, subbed.
As someone who’s used capstans on yachts for the best part of 10 years never thought I’d see it used like this, yeah a few turns on the capstan will pretty much hold the whole yacht under load. Great video!
I’ve been in robotics and automation product development for 30 years and see huge value in your combination of skills, talent, and communication ability. You represent the future of where we need to go. Can’t wait to see what you can do next after finishing school!
Superficially, yeah, a high quality engineering channel, but when you look a little deeper, you'll see they're totally different. Let me ask you this: Do you see anybody trying to figure out ways to murder wife on this channel? Do you see the difference, now?
I just wish that the American Industry would go back to the high-quality-long-lasting-stuff-made-here-production instead of the cheap-throwaway-consumption-driven-market we have today. The fact that the rope can’t be found (unless ordered from overseas) here in the US is really a downer, especially when the US government wants to impose more tariffs to imported goods. Sorry for this last diatribe, just had to be said.
When I saw "rope" in the title I thought, "Oh, jeeze! He's going to use paracord." So I was not let down. But thank you for demonstrating the weakness of Nylon rope. I had to prove it for myself on my boat back in the 90s. I started with 50-ft of nylon rope. Every time it got wet (which was every time I sailed), it stretched and released tension. By the end of the sailing season I had pulled an additional 150 feet of rope out of the original 50 feet. The next season I switched to Dacron rope and sailed for many seasons with the same rope and without having to retighten it.
@@marleyjanim5033 Amazon price for 90' bundles of 14" nylon = $0.05 per foot. For two 50' bundles of 1/4" polyester = $0.12 per foot. So $5.99 for nylon versus $11.95 for polyester. Polyester rope is getting harder to find at hardware stores. In-store prices are often less than online. If Lowe's doesn't have it then I try Home Depot and then Ace Hardware. It's always available online from those stores and Walmart. I have to spell out polyester every time, because polypropylene rope is also called poly, and it really only has use as a waterski rope if you ask me.
I'm not even in Enginnering, yet I felt I understoood everything. Very good video, thank you so much. I learned alot just from you video! I am also realizing that I need to study more in my area if you are this knowledgeable about yours.
As a complete guess, I would suggest your capstan ratio inaccuracy is due to the affective diameter of the drive capstans - this is not the diameter of the drum but, I think, the *outer* diameter of the wrapped drive cord (possibly with another correction in your case for the helical wrap). There is a good YT video (perhaps by Steve Mould) which explores a similar issue with ball bearings rotating around the circumference of a circle.
I would guess (large drum radius + cable radius) : (small drum radius + cable radius) as the theoretical reduction. If you didn't account for the cable size, that could result in the observed discrepancy.
would that effect not strictly *decrease* the effective gear ratio, not increase it, as it increases the diameter of both sides by the same amount, so the smaller gear grows in proportion?
@@ChucklesTheBeard I was thinking the same thing. We need to know exactly where the drum radius is measured... deepest point of groove, or the flat surface between grooves? The latter would explain the discrepancy, since the groove depth is proportionally larger on the small drum.
There's also a difference introduced because the cable/wire around the spool and the drum is not perpendicular to the spool/drum axis, but angled (coiled), so one turn of the spool moves more than its diameter's length of cable.
I'm not here because i'm super interested in robotics. I'm here for the engineering, the thought process and the spectacular filming. Love your videos and how well you put them together, how you plan and explain everything. Good job!
Old school radio tuners used a similar system to tune all the different band capacitors. Instead of metal cables of ropes, they used a string that was a lot like dental floss. Robotic slide stainers use a very thin metal cable with a plastic coating on it.
Any radio with a linear tuning scale and needle has a capstan on the back of the knob, it's under just enough spring tension that it can turn the tuning capacitor but light enough that it will preferentially slip on the (highly polished and smooth) capstan instead of breaking something. This is how it allows the knob to keep spinning even at the endstops
Amend: larger-format radios such as 70s/80s hi-fi that don't have the knob directly on the tuning capacitor axle. Newer small radios tend to have the knob direct drive to the tuner
There is a variation of this drive where you can get extremely high reductions in a compact space. If the rope goes from the drive drum to the driven drum and loops around throug a pulley to the drive drum again. The drive drum unwind diameter and wind diameter should be different as this is where the reduction comes from. By unwinding on one side and rewinding on the other side but with a slightly different diameter the pulley loop is extended only a little bit and this allows the driven drum to rotare. The closer the two diameters of the drive drum are the bigger the reduction, and it tends towards infinity as they tend towards each other. In order for two way motion you need to have two pulleys, one that extends and one that retracts. Like a differential windlass but the extending side and retracting side act to rotate a drum. Great project, I really enjoyed the video.
old 1980s computer hard drives used a capstan drive for the head mechanism, and they uesd a springsteel or stainless steel flat band instead of rope or cable. they've maintained their accuracy and precision to this day, as many of them are still functional as hard drives.
If youre having trouble sourcing vectran or HMPE/dyneema (DM20), just head down to you local cabelas and grab some raw bow string (or have one made to length). There are even several different blends. Compound bow strings have been made out of vectran and dyneema forever since creep causes loss of tune/accuracy
holy crap as a complete layman when it comes to mechanical engineering your video made it extremely easy to understand and digest, awesome job you would be a great teacher
That moment when you realise this isn't the first time you've watched one of his videos. And that you've watched his last two uploads without knowing it was him in all of them, should be enough to make you subscribe
Good video! When I worked at Texas Instruments we used capstan drive to move the printhead in a printer. The wire rope was plastic coated and the capstan was molded plastic. The controller was an 8080A.
UA-cam did the right thing by putting your video in front of me because I really enjoyed this one and will be subscribing to enjoy future ones. I appreciate your math sketches and your willingness to show that mapping onto the real world is not so easy and what compromises allow you to move your projects forward. Thanks for sharing.
I’m just glad I heard you use Dyneema. I have a kayak with rudders that were originally connected by stainless steel cables. The saltwater killed them and they weren’t flexible enough. I tried paracord but it stretched too much and was really thick. Then I found Dyneema. It’s super thin. Never stretched noticeably. And it was suuuper strong.
Sir i believe you've struck my perfect balance between layman's explanation and shown homework. I'll give it a few videos to confirm but i do believe you've earned a subscriber.
Capstan drive mechanism have been used in computers, old floppy drives and some hard drives used a stepper motor turning a capstan wrapped in thin metal strips. It could push or pull the head assemblies with a much bigger stepper motor (faster and stronger) without the worry of backlash. In fact your first prototype demonstration at 5:06 sounds almost exactly like an old hard drive doing seek tests.
On my desk I have an old miniscribe 20Meg (yes Meg) hard drive with the cover removed. I use it to demonstrate to people how a hard drive works. I use it less and less with everything going to ssd/nvme.
Capstan drives are awesome for robotics! In college I work in a lab that produced various bipedal robots using capstan drives. They coupled the drive to fiberglass springs to be able to store energy as the leg stepped and release it back into the next step. In that way they dramatically reduced the energy required for bipedal motion. They have spun off a lot of the tech into a company called agility robotics and they offer some extremely impressive bipedal robots.
Another good option for filament tendons would be parallel braided Kevlar line. We used to use this for high-power stunt kite control lines, but that has migrated to Dyneema due to the high cost of Kevlar. However, the Kevlar line provided greater control because of its smaller diameter, which contributed to less wind drag/catenary than Dyneema does. Everything is a compromise. I like your nod to This Old Tony in the way you present discussion topics. VERY cool content! Subscribed.
I missed the latter 1/3-1/2 of this video because my brain was busy assimilating its newfound awareness that any living human was as smart as this dude.
Nice. I remember that before digital age we used such things in old radios to connect a frequency selector wheel on the front with a variable capacitor on the PCB. It also moved around a needle on a frequency display. For this they used some sort of wire woven into a fabric.
Yes, these old dial cord systems were great and if the rotational friction was low and the selection knob had a good bit of mass you could give it a twist and it would run quickly across the dial without you having to turn it all the way yourself!
Sailors don't know it, but they love the Capstan equation. Intuitively you put on two wraps for pulling in say a jib sheet (rope), but if you want to ease it out you just flick the sheet to temporarily reduce the wrap length. Paracord sounds perfect for home projects and experiments. You can use Spectra chord as used in sails and ropes for large ships. Good video. Thanks.
im a math major and watching this i realized that the type of math i do and the type of math the engineering majors in some of my math classes do are two entirely separate types of math lmaooo no clue what most of these equations mean/are used for but im glad that there are ppl who do know lol
01:15 I would add: irrational transmission. Gears can only be rational, because the number of teeth in both cogwheels must be a whole number. But there's no limitation on the capstan drive.
BTW a Capstan Drive can also be build with 3 mm GT2 belts. In that case you dont have an inner loop but instead use the teeth of the GT2 itself as a fixation. Needs lots of iteration to get the geometry just right, however GT2 Belts are a lot more rigid and less prown to lenghtening.
tugboats use capstans to tighten lines and make up tow equipment. and we have a rule of thumb with winches that 3 wraps is the minimum number of turns to prevent slippage. cool video!
Great video about a type of drive I didn't know about. Some principles I think would be good to investigate on top of the basic Capstan Drive are : • Coil the cable several time around the big drum to have higher angle freedom; • Adapt the Capstan drive in a planetary gear configuration; • Make the big drum coaxial with the small drum; • Make a version where the cable is not clamp at the end for tension/friction but instead the cable is a closed loop where the tension/friction is provided by a free rotating tension wheel like in a bike.
Wonderful! Brilliant! I am pausing at 3:02 to subscribe, you had me at 'capstan equation". Please never be afraid to post the math, we need more of this. I will be back for a second comment after watching the rest of the video
seeing the drive sling around the weights on the arm so effortlessly blew my mind. i need a motor and ESC that size so i can make one myself! to... you know... move a weight XD
I hate that I am the one posting a stupid comment here, but everything I could think of to say about this excellent demonstration of the Capstan Drive was already said in the others, so I just wanted to say you are a great presenter and that the yo mama joke was possibly one of the smoothest delivered I've ever heard as well. 😂
I suspect it may have something to do with the path the rope actually wraps on not being the nominal diameter. It's probably something strange in between the centerline diameter of the groove, and the minimum diameter of the groove.
@@Max24871There is no reason to physically test this. You can calculate the total length in the 3D model effortlessly with much better accuracy than trying to measure a string.
A big thankyou for this video. I was converting an old oscillating office fan to a mini "waterwheel" to rotate my tracking solar cooker reflector. But the worm drive in its gearbox was plastic and too weak. Took off the gearbox, put the "waterwheel on the back instead of the front, and there is plenty of room on the front where the fan blades were to wind the capstan string! It's way simpler than the other solution that I was going to use. I would never have finished the project without your video. Thank you!
Dyneema and vectran fiber will creep under heat and load. Maybe the woven ropes help to prevent that but limit your ability to miniaturize. In the late 90s we used stainless steel woven cables (like rope but stainless, eg 7x19) and aluminum capstans to run force feedback robots.
This is not true several different ways. First of all, the magnitude of your forces has to be a substantial percentage of break. Second of all, that's only if you are not using pre heat set dyneema. I run dyneema in world class competition sailboats, there is no creep with the modern iterations.
I worked in semiconductor manufacturing equipment in the nineties/twothousands, that used capstans for linear motion. Instead of rope they used stainless steel shims as the tension medium. Going from memory the the shim was one piece of sheet with two cuts in it to look like a stretched E shape. Two legs went to one end of the linear part and the centre leg to the other the part where the joined was screwed to the capstan. This was done to avoid any offset forces (this is all from memory and hard to describe. They were very accurate and coped with continuous operation in a production environment for a very long time.
Broooo. This is the kind of educational content I NEEEED. Absolutely dope video. I want to see you build a Capstan drive 3D printer to see what kind of pros and cons you end up- mostly so I can see you document your thought process and problem solving.
Mate, you've just solved a problem I've been scrambling to solve for years on a "similar" project....now i need to hold of that rope, fortunately I'm in Europe.
Great video. At 6:41 you explain that the gear ratio is not 8:1 but more like 8.5:1, I did not do the maths. But this problem seems very similar to something I encountered before where the diameter of the wire is not accounted for. Let's say your wire diameter is 5mm and your drum is 20mm, then the circumference you should use for the path calculation is not 20*pi, but 22.5*pi.
This is brilliant! Thank you so much for sharing it so generously. Your skills are amazing, and this mechanism is mind-boggling! You have opened a rabbit hole...
Great video! Always love these little fail compilations and I really like how unhinged the jumping bot looks. 17:07 that's why you should look into combat robotics, why destroy your own bot when it can get destroyed by another bot!
I consider myself extremely well versed in science and mechanical concepts. I'm educated in engineering and I'm a technician by trade. But I have to admit, nearly everything showed in this video was brand new to me. I loved it. Our shop at work is about to transition from maintaining older, conventional material transporters to robots, and we will have to learn how all this stuff works in order to repair them. If anything we get uses a Capstan Drive, I am definitely going to send them all a link to this video.
Love the video, especially the info about the 'why' of limitations and design choices! Another approach to the D/d ratio is to use smaller stainless steel wire. I suspect it's significantly larger than needed.
Thank you so much for filling in a huge missing gap in my knowledge. I never knew about such magical things until your video. You also explain the workings at a level that most technically minded people will understand, without feeling like a fool. Thank you!!
Most surprised by the quick and consistent failure of the stainless cable. Most impressed by the DM20 rope's performance. Most satisfied with the 'square/circle/diamond'. Most entertained by the 'your momma' joke 🤣 Great presentation! 10/10 - would learn again
Just a tip from a hiker who has been using Dyneema for years. If you want your rope to last longer, don't let the rope contact itself. In the video it looked to be touching where it was wrapped around. Dyneema is very strong, but if it is rubbing against itself it will get damaged. I learned this from putting a Dyneema tent into a Dyneema stuff sack. Cool video! Also, the Your Mama joke took me by surprise and cracked me up. :)
I've used dyneema a lot and noticed it degrades less when it rubs against itself than anything else. Hell i have really old whoopie slings that look brand new apart from loop that hooks on but clutch looks fine. I get that rope against rope contact is bad but with dyneema if you don't create huge temperatures i don't really see it being worse than any other rope... hell it would create less of an issue due to very low friction coefficient of UHMWPE.
Could you tell me how does your damage look like and is it the case of rubbing when mostly parallel or is it the case of low diameter bend of quite inelastic ropes?
@@Przemo-c Abrasion is the weak point of Dyneema. So much so that every tent manufacturer that makes them with Dyneema recommends folding rather than stuffing your tent in the stuff sacks. Tarp-Tent, Gossamer Gear, Zpacks, Durston , to name a few who recommend it.
@@tinetannies4637 That's exactly what I was thinking.
As a spearfisherman I have used the same dynemma line for years and year in harsh conditions, big fish in highly abrasive environments such as around coral. Dynemma is actually extremely good at abrasion resistance, more than any other line, you just need to the the correct type. There are ones that are impregnated with wax which significantly reduces failure due to abrasion, also it's important to get the cored dynemma as that extra stiffness reduces wear due to abrasion.
I'm not a climber but I like to watch the channel *"How NOT 2."*
I've heard the warning about having ropes rubbing against itself many times (particularly with Dyneema). I highly recommend the channel to anyone interested in ropes and rope work.
I also like the channel *"Hard Is Easy."*
What is this, a scientific testing approach for an actuator, a solid testing phase, a good explanation of physics, and freely available files? You have earned yourself an instant subscriber, that was top notch!
well said. science applications and explanations will trump (hah 'Trump...) most channels...SmarterEveryDay is cool. but it was you that got the sub. TY
Yep good stuff! Liked, subscribed and did that bell shit everyone's always asking you to do.
Don’t forget the perfectly placed and unexpected Yo Mama joke.
He is insane !
@@MaxCrack33that came out of nowhere. Hilarious 😂
As a sailor, I saw you messing with bent steel wire and paracord and was like "noooo!" but then you found what I believe is the right answer, and I was "yesss!", not because I agree, but because it says a lot about your process, and that I will learn from your other work. Subbed, and cheers!
fellow sailor here. Dyneema rope is sooo good.
I am an Electrical Engineer and seeing a simple exponential, explains things so well, despite knowing nothing of mechanics.
keep up the good content
I have a master's degree in electrical engineering, and it never ceases to amaze me just how many similarities exist between the mathematics of electromagnetics and the mathematics of mechanics. On one hand, that makes perfect sense (since they both boil down to physics). On the other hand... just when I think I've found a perfect analogy between the two, I think of an edge case that makes my analogy nearly useless. My intention is always to give someone who doesn't have EE training a kind of "horse sense" for how electricity behaves. Consider how water moves through a pipe compared to how electricity moves through a wire. That analogy is fairly intuitive and helpful away from extreme cases. When I started thinking about how a "short circuit" could be represented in the "water through a pipe model", though, that "helpful intuition" began to fall apart very quickly. :(
@@slidingconstant your circuitry is represented by pipes with a huge diameter and your resistive load is represented by pipes of a much smaller diameter. The water throughput is resisted by the small diameter of the resistive load. if you replace the small diameter piping with a the big diameter piping your waterflow increases drasticly and you have your "short circuit water flow"
+1 on this comment, I had amazing electrical, computer and software engineering classes in uni but my statics/dynamics/thermo classes were so poorly taught I just did the minimal amount of work to get by. Awesome work on this video, I'm going to find a project to work on using this and I can't wait to find time to work through the other videos.
Your jumping robots have already surpassed human levels of upstairs neighbor behavior.
I'm a retired EE, with robotic experience going back as a sub-contractor to Disney's Imagineering, in the 1980s. Around the time of the Mike Eisner power grab. Back then merging electrical and mechanical parameters, was part of control systems classes, in school. That said, this channel an its creator's engineering skills are top notch, and he explains it excellently.
Back in the old days, with 8-bit microprocessors, and microcontrollers, running only at a few MHz, we'd do the kinematic calculations on a spreadsheet, to generate look up tables, that in turn, would be embedded in the code. Which you can still do, if your application warrants it. But you have the option of using a beefier ARM processor, and get a bit flexibility, running those calculations, in the actual runtime code.
I still tinker, so this capstan drive, may show up in one of my future toys !
The drive is brilliant. Your capacities as a designer and engineer are obviously evolving, but most impressive is your capacity to explain things clearly.
yessss !!
Could not agree more, awesome build awesome and clear explanations and story telling!
I just found some of the in between shots where you just twiddle the pen a bit distracting as i was trying to figure out what you were trying to show me and I was listening less. Maybe something to keep in mind.
Would love to see this channel grow for you to build ever cooler things as I am sure you will ❤
To make the video more focused he didn't show his face.
You are going to give Stuff Made Here a run for his money in no time. I like how you share the math instead of completely glossing over it. Amazing work.
@StuffMadeHere, Shane, he is calling you out
Lol @@bmx4free
"Gears are sooo third century BC. Its time for something different"
Starts using ropes which is being used since caveman times
Even earlier than settlement-- cave dwellers were rather advanced in terms of materials and craftsmanship, having mastered both stone, dyes and fire. Technologically, pottery isn't far off. The oldest ropes are found fossilized, from the very start of the Stone Age, when the only structures we find were probably semi-permanent work-huts (judging by the materials).
@@ontoverse Thats even funnier now
@@ontoverseThat's because people were relatively advanced before they started living in caves. Most early people dwelt in tents, only moving to caves as they expanded out away from Eden.
lol Then proceeds to reinvent the gear replacing the sheer force from interconnected teeth for rope and friction which goes back even farther to the most ancient grandfather technology that started it all. The mastery of friction to first control fire allowing us to illuminate, heat and become the superior troglodyte society finger painting with bat poop no one could have expected would one day dominate the food chain. We make fun of them now, but the cavemen were the sophisticated gentlemen of there time and I respect the nod to our first golden age.
we are jsut rope machines
Can’t believe it’s taken the algorithm this long to show me this channel.
Hopefully he gets some traction.
I just got shown his channel too.
Hemp rope
Lol, same
yt will never make a come back, it is what it is.
Dude same
I am impressed. As a retired engineer with a hobby Prusa MK4 printer you have opened my mind to some very neat concepts. I have no desires to make anything this complex but your work shows the possibilities. Thank you for making this informative video.
I laughed out loud during the first jumping part. It just looks so much like a little creature that is inexplicably frustrated with the ground and trying it's hardest to beat it up
This, lol
it also had such a nice, calm and cool intro and then it just went wild trying to jump
pixar lamp ass robot
Robot Riverdance is what came to mind for me!
TIL: wven just 1/4 of a robot dog still gets the zoomies
I love the cool confidence of it drawing shapes wrong.
It's all in how you look at the shape it did draw. Maybe it wasn't done yet, like how maybe that skateboard rider just took a push ups break before doing two more kick-flips in a row...
That was the funniest thing I've seen all week.
@@charlesbruneski9670 robot fails are never NOT funny lol.
@spdcrzy yep I laughed out loud and then watched it a couple more times! 🤣
it's like the robot's signature move.
This video really checked all the boxes. Interesting and educational, great narration and editing. Top-tier engineering content.
I found this very interesting. I am a retired gearbox engineer, consultant, repairer, etc. We designed and built capstans for high-end yachts. These were hypercycloidal gearboxes. They were manufactured to a high degree of accuracy and we achieved minimum backlash, of around 0.002" maximum. Have you determined the maximum loading that rope can withstand ? Probably the most accurate gearbox we made was for satellite tracking, where the backlash was basically 0 ( reality check it was around 0.0005" ). Everything in the gearbox was precision ground and was hand built. For some of the smaller satellite dishes I could see a rope driven gearbox system would be advantageous, as long as the 'stretch' factor was addressed. I hope you do more on the rope gearbox as I personally find it very informative. That's life when you've been an engineer all your life, your mind cannot retire. Thank you.
Hey, I'm looking for a gear ratio of 1:4000 (yeah...). Is that even possible, or would the forces on the 'inner' most gear be so high it's just not feasible?
Also, what can you tell me about low friction gear boxes?
I make roller bearings and we can achieve precision and sometimes require precision of 0.0001, how do my old machines do it with what looks like a normal train driving the motions?
Looks like ~1.5mm dyneema, under basically ideal loading conditions, it should handle >1.5 kN - the 3D prints will pretty much always break first in both overload and fatigue
Satellites may complicate this, but to deal with creep couldn't you just have another spring, motor, hydraulic, etc. that is pushing/pulling the drums apart? Then couldn't any creep be cancelled out by just pushing them further apart?
How am I just now finding this channel? This is gold.
Really uh? Me too I subbed so hard on this
I am very impressed by the quality of what I see on the screen. Professional level engineering with very good cinematography.
Awesome drive, testing and storytelling! As a sailor, we use dynema and spectra quite a lot as well as concepts of friction and pivot points, and many rope knots are based on the capstan principles of friction. Those principles have been in play for hundreds of years surviving in the harshest conditions on the planet. Given the right rope and your drum made out of stainless, I could almost guarantee your motors will fail before the rope and drum. Looking forward to your next update!
Thank you for the comment! Back then i made something with rope for robotics and stumbled into rope tensioning issue. Went to camping knots and how they tighten rope to solve my issue, but seeing ur comment on sailor rope which is the higher level of rope knots just opened my eyes
Honestly , with all the effort humanity has done to reduce friction , it is no surprise that we found a way to use it for once !
Honestly not surprised he ended up using dinema
If that rope is what it claims to be it can prbly hold upwards of 2-3tonnes
@@turesunnegardh6697 oh it most certainly is
I appreciate the willingness to test until it breaks and showing that purposefully. Many creators often depict breakage as accidental and not necessary to the process.
The best part is i'm a dum dum who's native language isn't even english and even i got it. That's how i know the man is great at explaining stuff.
Dyneema is a crazy material to work with. It’s absurdly strong, absurdly low stretch, absurdly light weight, and freakishly low friction. Not sure if you ran into any issues with its low friction, but it can be weird in how different to other ropes it is, because in some situations well known knots don’t behave as expected because of its lack of friction, something to keep an eye out for if you use it more.
Also it doesn’t like fusing to itself when you melt it, annoying for trying to keep the ends from fraying. Super glue works though if you ever need that knowledge.
You are right about knots being useless, in fact high load applications require rope splicing as it creates a vastly increased surface-interface to leverage friction.
I bought a bunch of drone parachutes at an auction and the shroud lines are made of Dyneema. I thought it was going to be a nightmare to untangle the chutes as most of them are wadded up. Actully it was super easy and fun because the Dyneema slides across itself so easily. Most of the tangles come out with just a bit of shaking. A great feature for a parachute, where tangled lines could mean it does not deploy properly.
Did you mean ties?
Where/how can it be purchased and what does it sell for?
They use it for kitegen, skysails power and enerkite too
As an electrical/control engineer, I always regretted not getting an ME degree or minor. Mechatronics was not a "thing" in the 80's. I recently retired after 40 years of control engineering and have been working on my ME skillset with help from UA-cam folks like yourself. I've been screen saving your math and walking through it. Keep up the great work!
That’s cool! what are you currently working on?
@@gokiburi-chan4255 Something I'm hoping to manufacture.
Mechatronics still feels like a burgeoning field today. I'm taking an ME degree with a mechatronics certificate. It's the closest I can get without a double major with EE! Fortunately, I think motor control and electronic drive systems will continue to become more common and thus the electromechanical field should grow to match! Not to mention how important control systems are in our modern world. The topics in this subject span the breadth of the EE and ME worlds--so the crossover is pretty clear
@@Trenz0 My only caveat for an actual Mechatronics degree is that many states do not allow those folks to obtain a professional engineer (PE) license which is required for many industries. Having a (pure) ME degree (or EE in my case) is typically a requirement. To be honest, having a PE these days is like printing money but it takes a few years working under other PE's to obtain one. The PE test is also equivalent to passing the bar exam. I obtained a Control Engineer PE license which was very rare in my state (WA) for the majority of my career.
Oh! Like minded engineering mind here! Who are some of the other channels you've been watching?
Your channel is a gold mine but open for the public for free!
These videos will inspire a lot of people into robotics (including me lol). Thanks for your work sir!
🤪the inserted "yo mama" joke... priceless!
Also, thanks so much for showing the failures along the journey.
No one was ready for that joke, so good!
Started coughing laughing, it was delivered perfectly!
It was so unexpected... I almost died laughing.
I Instantly subscribed upon hearing it 😂
As a sailor, I can tell you that we use this drive system literally everywhere on a sailboat, even steering
Yes, but Dyneema DM20 is new to me. (Hmmm... unsurprisingly, it's very pricey rope.)
Aaed, I'm very surprised that the reduction ratio wasn't a bit *_under_* 8:1, instead of a bit over 8:1, since it seems like you should add the diameter of the cord to the diameter of the drum, to get the true effective drum diameter. The rope diameter will increase the small drum's effective diameter by a larger percentage than it increases the large drum's effective diameter, which should reduce the ratio of the two. It is a puzzlement.
Terrance howard math
only steering... where else do you use it ?
@@nadirbalena691 he said everything mate.
moving, hoisting, reefing, trimming the sails is all done by rope.
And since a sail with wind in it can exert several tons of pressure on the main boom or the foresail you need aid to pull them in.
For the mainsail mostly block and tackle, or similar pulley systems, and for the foresails capstans to hold them tight.
A capstan drive along with bowden cables is used for the steering on sailing yacht. Smaller boats can do with a tiller.
@@ArfurFaulkesHake Everyone forgot the most iconic one. the ANCHOR! At least on older boats like you see pirates had. Would be cool if the navy still had a huge manual anchor.
scientific testing approach for an actuator, a solid testing phase, a good explanation of physics, and freely available files? You have earned yourself an instant subscriber, that was top notch!
I have utmost respect and admiration for an engineer confident enough in his skills and knowledge that he chooses to film himself working while still in his pyjama bottoms. Great video. Subbed!
This is pretty much the same comment that I was going to compose. That is to say, it covers every point and sense that I wanted to (Especially the confidence/pajama conveyance).
I do not think that I could quickly come up with something better. So, with a sincere nod, I tip of my hat to you, OxABADCAFE.
I do hope you will pardon me as I hop on to your coattails and offer up a hearty, "Hear, hear!".
Oh, and, of course I've subscribed.
@@bob2161 You're welcome. I'm glad I'm not alone in this particular appreciation.
I measured the cad files from GitHub and the ratio is between the radius - the gap between the drums (68-0.625)/(8.5-0.625) = 8.55
I think this might be it, although It doesn't make sense why the gap between the drums matters and not the thickness of the rope. I'll eventually do some testing to confirm if this is the correct way. Until then @HighFlights is hereby declared to be my favorite subscriber.
It is because the effective drum diameter is through the centreline of the rope.
The issue is the cable and will not be at the exact center. The issue is the cable has limited compression strength (in theory zero). Your highest tension location is at the furthest from the bend center but the cable
ope will deform to an oval because of this tension changing the effective diameter. It will also change a different degree on a different radius bend. You can test this by taking a known diameter spool and wrapping a known length of cable around it. You will not get the exact turns you calculate.
This should be enough for you to visualize the problem. It is what we referr to as a "hard" problem because it will be application specific and there is no uniform formulaic answer.
Just think if you had an infinitely stiff straight fiber cable the only part with load would be the fiber furthest from the spool center and the effective diameter of the spool would therefore be at the spool diameter + 2x rope diameter - 2x rope deformation from round.
@@aaedmusa Pretty sure the effective radii would be drum radius + 1/2 rope thickness - a fudge for rope squish.
@@frankslee7909 I was looking for this this comment, this would be my thought as well.
That's really interesting. Years ago I took apart an old hard drive from the 1980s and it used what I have now learned is a capstan drive to move the head across the platter, back when they used stepper motors instead of magnetic voice coils. Thanks to this video I also learned why they used a thin metal band (where the rope is used here) and that they probably chose the capstan drive due to the minimal backlash. Thanks for explaining everything so well in this video!
You are so underrated. Your channel is awesome!
Mom and Dad wake up! Bobby Duke just commented on my video again.
Yo its a surprise bobby duke! always cool to see you in the comments of channels i watch haha
well wewd you look at that. agreed!
I WAWNT BOOBY DUCK TO MAKE A VEEDIO!!!
i wawwnt thayyutttuhh
If anyone is interested, Dyneema is a popular performance upgrade in the yachting community. by replacing the steel wire mast stays with something lighter, you reduce the weight at the top of the mast. as you want the thing to stay upright, you make the bottom bit (the boat) heavy, and due to the massive lever of the mast, even a small weight reduction up top can improve stability.
If you want a small bit to experiment with and dont want a whole roll or to wait for shipping, you may be able to get either offcuts or by the meter from your local marine supply store (chandlery).
very cool vid, subbed.
Dyneema is also used for climbing.
As someone who’s used capstans on yachts for the best part of 10 years never thought I’d see it used like this, yeah a few turns on the capstan will pretty much hold the whole yacht under load. Great video!
I’ve been in robotics and automation product development for 30 years and see huge value in your combination of skills, talent, and communication ability. You represent the future of where we need to go. Can’t wait to see what you can do next after finishing school!
The leg looked so life like, and i love when things are tested to their limit. You have impressive skills, nice work!
I've known about gear backlash for 30 years. This is the first straightforward explanation I've heard for what it is! :D
"If you aren't breaking your robots, are you really pushing them enough?" That's some sage wisdom.
Great work. Your content is 'Stuff Made Here' level quality.
100 pct agree. Love the content!
Superficially, yeah, a high quality engineering channel, but when you look a little deeper, you'll see they're totally different. Let me ask you this: Do you see anybody trying to figure out ways to murder wife on this channel? Do you see the difference, now?
Gives the same vibes as his early videos before showing his face.
I just wish that the American Industry would go back to the high-quality-long-lasting-stuff-made-here-production instead of the cheap-throwaway-consumption-driven-market we have today. The fact that the rope can’t be found (unless ordered from overseas) here in the US is really a downer, especially when the US government wants to impose more tariffs to imported goods. Sorry for this last diatribe, just had to be said.
You really have yourself something special here and you have chosen to share it with all of us. All of your hard work. Have a hug 🤗
Highly underrated channel, the educational content alone is phenomenal, the humor and results are just a bonus! Keep it up!
never understood the term underrated. who's underrating, and who determines the ratings aren't where they should be? I need to see the data.
When I saw "rope" in the title I thought, "Oh, jeeze! He's going to use paracord." So I was not let down. But thank you for demonstrating the weakness of Nylon rope. I had to prove it for myself on my boat back in the 90s. I started with 50-ft of nylon rope. Every time it got wet (which was every time I sailed), it stretched and released tension. By the end of the sailing season I had pulled an additional 150 feet of rope out of the original 50 feet. The next season I switched to Dacron rope and sailed for many seasons with the same rope and without having to retighten it.
May I ask the price difference?
@@marleyjanim5033 Amazon price for 90' bundles of 14" nylon = $0.05 per foot. For two 50' bundles of 1/4" polyester = $0.12 per foot. So $5.99 for nylon versus $11.95 for polyester. Polyester rope is getting harder to find at hardware stores. In-store prices are often less than online. If Lowe's doesn't have it then I try Home Depot and then Ace Hardware. It's always available online from those stores and Walmart. I have to spell out polyester every time, because polypropylene rope is also called poly, and it really only has use as a waterski rope if you ask me.
I'm not even in Enginnering, yet I felt I understoood everything. Very good video, thank you so much. I learned alot just from you video!
I am also realizing that I need to study more in my area if you are this knowledgeable about yours.
As a complete guess, I would suggest your capstan ratio inaccuracy is due to the affective diameter of the drive capstans - this is not the diameter of the drum but, I think, the *outer* diameter of the wrapped drive cord (possibly with another correction in your case for the helical wrap). There is a good YT video (perhaps by Steve Mould) which explores a similar issue with ball bearings rotating around the circumference of a circle.
I would guess (large drum radius + cable radius) : (small drum radius + cable radius) as the theoretical reduction. If you didn't account for the cable size, that could result in the observed discrepancy.
would that effect not strictly *decrease* the effective gear ratio, not increase it, as it increases the diameter of both sides by the same amount, so the smaller gear grows in proportion?
@@ChucklesTheBeard I was thinking the same thing. We need to know exactly where the drum radius is measured... deepest point of groove, or the flat surface between grooves? The latter would explain the discrepancy, since the groove depth is proportionally larger on the small drum.
This is why you should always use the double stop method and calculate the ratio to find the center. As the Rope wears out it may change.
There's also a difference introduced because the cable/wire around the spool and the drum is not perpendicular to the spool/drum axis, but angled (coiled), so one turn of the spool moves more than its diameter's length of cable.
I'm not here because i'm super interested in robotics. I'm here for the engineering, the thought process and the spectacular filming.
Love your videos and how well you put them together, how you plan and explain everything. Good job!
Old school radio tuners used a similar system to tune all the different band capacitors. Instead of metal cables of ropes, they used a string that was a lot like dental floss.
Robotic slide stainers use a very thin metal cable with a plastic coating on it.
Any radio with a linear tuning scale and needle has a capstan on the back of the knob, it's under just enough spring tension that it can turn the tuning capacitor but light enough that it will preferentially slip on the (highly polished and smooth) capstan instead of breaking something. This is how it allows the knob to keep spinning even at the endstops
Amend: larger-format radios such as 70s/80s hi-fi that don't have the knob directly on the tuning capacitor axle. Newer small radios tend to have the knob direct drive to the tuner
There is a variation of this drive where you can get extremely high reductions in a compact space. If the rope goes from the drive drum to the driven drum and loops around throug a pulley to the drive drum again. The drive drum unwind diameter and wind diameter should be different as this is where the reduction comes from. By unwinding on one side and rewinding on the other side but with a slightly different diameter the pulley loop is extended only a little bit and this allows the driven drum to rotare. The closer the two diameters of the drive drum are the bigger the reduction, and it tends towards infinity as they tend towards each other. In order for two way motion you need to have two pulleys, one that extends and one that retracts.
Like a differential windlass but the extending side and retracting side act to rotate a drum.
Great project, I really enjoyed the video.
differential windlass?
Do you know any recourses for this one? Like a paper or a scematic?
@@non7sens
The Wikipedia article on windlasses explains it well.
through*
old 1980s computer hard drives used a capstan drive for the head mechanism, and they uesd a springsteel or stainless steel flat band instead of rope or cable. they've maintained their accuracy and precision to this day, as many of them are still functional as hard drives.
Oh wow, cool info.
I know right?!?!/ Kinda makes you feel sorry for the guy... ;-)
That's what I was thinking could work, using a metal that is more susceptible to bending. Like the wound up spring steel ribbon inside a clock.
Exactly what I was thinking
Requiring an optical position feedback I believe.
If youre having trouble sourcing vectran or HMPE/dyneema (DM20), just head down to you local cabelas and grab some raw bow string (or have one made to length). There are even several different blends. Compound bow strings have been made out of vectran and dyneema forever since creep causes loss of tune/accuracy
holy crap as a complete layman when it comes to mechanical engineering your video made it extremely easy to understand and digest, awesome job you would be a great teacher
That moment when you realise this isn't the first time you've watched one of his videos. And that you've watched his last two uploads without knowing it was him in all of them, should be enough to make you subscribe
Good video! When I worked at Texas Instruments we used capstan drive to move the printhead in a printer. The wire rope was plastic coated and the capstan was molded plastic. The controller was an 8080A.
Dude, when that robot is jumping, I bet your neighbors are thinking that you're some kind of stud.
Or just being annoyed at the constant *DAMN BANGING*
Nah, they think he is a stud!
... Because of the constant banging
UA-cam did the right thing by putting your video in front of me because I really enjoyed this one and will be subscribing to enjoy future ones. I appreciate your math sketches and your willingness to show that mapping onto the real world is not so easy and what compromises allow you to move your projects forward. Thanks for sharing.
I’m just glad I heard you use Dyneema. I have a kayak with rudders that were originally connected by stainless steel cables.
The saltwater killed them and they weren’t flexible enough. I tried paracord but it stretched too much and was really thick.
Then I found Dyneema. It’s super thin. Never stretched noticeably. And it was suuuper strong.
Sir i believe you've struck my perfect balance between layman's explanation and shown homework.
I'll give it a few videos to confirm but i do believe you've earned a subscriber.
Capstan drive mechanism have been used in computers, old floppy drives and some hard drives used a stepper motor turning a capstan wrapped in thin metal strips. It could push or pull the head assemblies with a much bigger stepper motor (faster and stronger) without the worry of backlash. In fact your first prototype demonstration at 5:06 sounds almost exactly like an old hard drive doing seek tests.
On my desk I have an old miniscribe 20Meg (yes Meg) hard drive with the cover removed. I use it to demonstrate to people how a hard drive works. I use it less and less with everything going to ssd/nvme.
Wow, I know that sound now that you explain it. I remember the needle going back and fort
Wow, awesome engineering. Hello from Ñew - Brunswick, Canada. I am restoring a 1960 harvester thresher, lots of backlash with 43 pulleys.
Capstan drives are awesome for robotics! In college I work in a lab that produced various bipedal robots using capstan drives. They coupled the drive to fiberglass springs to be able to store energy as the leg stepped and release it back into the next step. In that way they dramatically reduced the energy required for bipedal motion. They have spun off a lot of the tech into a company called agility robotics and they offer some extremely impressive bipedal robots.
Why capstans over timing belts?
Do you have any links to that sort of stuff, please, as it would fit in with a project that I'm working on.
Another good option for filament tendons would be parallel braided Kevlar line. We used to use this for high-power stunt kite control lines, but that has migrated to Dyneema due to the high cost of Kevlar. However, the Kevlar line provided greater control because of its smaller diameter, which contributed to less wind drag/catenary than Dyneema does. Everything is a compromise. I like your nod to This Old Tony in the way you present discussion topics. VERY cool content! Subscribed.
I missed the latter 1/3-1/2 of this video because my brain was busy assimilating its newfound awareness that any living human was as smart as this dude.
Very cool project! I love that it allows a mostly 3D printed actuator to have such great performance!
Nice. I remember that before digital age we used such things in old radios to connect a frequency selector wheel on the front with a variable capacitor on the PCB. It also moved around a needle on a frequency display. For this they used some sort of wire woven into a fabric.
Yes, these old dial cord systems were great and if the rotational friction was low and the selection knob had a good bit of mass you could give it a twist and it would run quickly across the dial without you having to turn it all the way yourself!
Sailors don't know it, but they love the Capstan equation. Intuitively you put on two wraps for pulling in say a jib sheet (rope), but if you want to ease it out you just flick the sheet to temporarily reduce the wrap length. Paracord sounds perfect for home projects and experiments. You can use Spectra chord as used in sails and ropes for large ships. Good video. Thanks.
im a math major and watching this i realized that the type of math i do and the type of math the engineering majors in some of my math classes do are two entirely separate types of math lmaooo
no clue what most of these equations mean/are used for but im glad that there are ppl who do know lol
01:15 I would add: irrational transmission. Gears can only be rational, because the number of teeth in both cogwheels must be a whole number. But there's no limitation on the capstan drive.
You and Joshua both pushing the 3DP community forward. Love to see it.
BTW a Capstan Drive can also be build with 3 mm GT2 belts. In that case you dont have an inner loop but instead use the teeth of the GT2 itself as a fixation. Needs lots of iteration to get the geometry just right, however GT2 Belts are a lot more rigid and less prown to lenghtening.
tugboats use capstans to tighten lines and make up tow equipment. and we have a rule of thumb with winches that 3 wraps is the minimum number of turns to prevent slippage. cool video!
Love your dry humor and really approachable explanation! This is so cool and I am so impressed with your work.
16:50,18:50 Robots jumping manically is hilarious! Great job with the design, explanation, and filmaking!
Great video about a type of drive I didn't know about. Some principles I think would be good to investigate on top of the basic Capstan Drive are :
• Coil the cable several time around the big drum to have higher angle freedom;
• Adapt the Capstan drive in a planetary gear configuration;
• Make the big drum coaxial with the small drum;
• Make a version where the cable is not clamp at the end for tension/friction but instead the cable is a closed loop where the tension/friction is provided by a free rotating tension wheel like in a bike.
I can't wait to see an update on this project. It looks so smooth and fast. Amazing choice for gear substitution
Wonderful! Brilliant! I am pausing at 3:02 to subscribe, you had me at 'capstan equation". Please never be afraid to post the math, we need more of this. I will be back for a second comment after watching the rest of the video
seeing the drive sling around the weights on the arm so effortlessly blew my mind. i need a motor and ESC that size so i can make one myself! to... you know... move a weight XD
I hate that I am the one posting a stupid comment here, but everything I could think of to say about this excellent demonstration of the Capstan Drive was already said in the others, so I just wanted to say you are a great presenter and that the yo mama joke was possibly one of the smoothest delivered I've ever heard as well. 😂
This channel is underrated.
The thing about 8.5 to 1 might be like that "SAT Problem That Everybody Got Wrong" ?
its because a helical path is longer than a circle of the same diameter. It is NOT surprising.
I suspect it may have something to do with the path the rope actually wraps on not being the nominal diameter. It's probably something strange in between the centerline diameter of the groove, and the minimum diameter of the groove.
@@Brainstormer_Industiresthat would be very easily to test with very thin string
I think the radius of the rope is also important.
@@Max24871There is no reason to physically test this. You can calculate the total length in the 3D model effortlessly with much better accuracy than trying to measure a string.
A big thankyou for this video. I was converting an old oscillating office fan to a mini "waterwheel" to rotate my tracking solar cooker reflector. But the worm drive in its gearbox was plastic and too weak. Took off the gearbox, put the "waterwheel on the back instead of the front, and there is plenty of room on the front where the fan blades were to wind the capstan string! It's way simpler than the other solution that I was going to use. I would never have finished the project without your video. Thank you!
@@Brians-Easy-Low-Tech-Solutions That’s super awesome dude!
You put so much effort into your videos! It’s greatly appreciated. You will go places 😇
Dyneema and vectran fiber will creep under heat and load. Maybe the woven ropes help to prevent that but limit your ability to miniaturize. In the late 90s we used stainless steel woven cables (like rope but stainless, eg 7x19) and aluminum capstans to run force feedback robots.
At an appropriate load DM20 will have essentially 0% creep. You just have to design within specifications.
Dynema stretches under constant tension. Vectran and Aramida no stretches. Vectran need to control heat below 350 degrees Celsius
This is not true several different ways. First of all, the magnitude of your forces has to be a substantial percentage of break. Second of all, that's only if you are not using pre heat set dyneema.
I run dyneema in world class competition sailboats, there is no creep with the modern iterations.
@@Vikingll There are various forms of dyneema, its not all the same. Google DM20 vs SK78 for example.
@@jaro6985 tks
I worked in semiconductor manufacturing equipment in the nineties/twothousands, that used capstans for linear motion. Instead of rope they used stainless steel shims as the tension medium. Going from memory the the shim was one piece of sheet with two cuts in it to look like a stretched E shape. Two legs went to one end of the linear part and the centre leg to the other the part where the joined was screwed to the capstan. This was done to avoid any offset forces (this is all from memory and hard to describe. They were very accurate and coped with continuous operation in a production environment for a very long time.
17:00 "Pass the butter" moment
What an incredibly well-produced video and fascinating concept! Loved when that robot started jumping!😆
Broooo. This is the kind of educational content I NEEEED. Absolutely dope video. I want to see you build a Capstan drive 3D printer to see what kind of pros and cons you end up- mostly so I can see you document your thought process and problem solving.
0:33 “Do you want to play a game?”
Bro just slidin the "Yo mama" joke 😂😂
slidin in yo mama
Mate, you've just solved a problem I've been scrambling to solve for years on a "similar" project....now i need to hold of that rope, fortunately I'm in Europe.
Great video. At 6:41 you explain that the gear ratio is not 8:1 but more like 8.5:1, I did not do the maths. But this problem seems very similar to something I encountered before where the diameter of the wire is not accounted for. Let's say your wire diameter is 5mm and your drum is 20mm, then the circumference you should use for the path calculation is not 20*pi, but 22.5*pi.
I would like to know from the author if that accounts for his 8.5:1 ratio.
You add half the wire diameter... twice. Once on each side of the drum. So 25*pi.
19:48 Sooo.... Aaed Musa.... probably my favorite new UA-camr at the moment
This is brilliant! Thank you so much for sharing it so generously. Your skills are amazing, and this mechanism is mind-boggling! You have opened a rabbit hole...
Great video! Always love these little fail compilations and I really like how unhinged the jumping bot looks.
17:07 that's why you should look into combat robotics, why destroy your own bot when it can get destroyed by another bot!
16:40 it looks kinda cute as if its frustrated xD
You are pretty much onto something in the world of biomechanics. Keep going.
Holy! Awesome work man!
regarding 6:40 I suspect the thickness of the rope adding up in some weird way
I consider myself extremely well versed in science and mechanical concepts. I'm educated in engineering and I'm a technician by trade. But I have to admit, nearly everything showed in this video was brand new to me. I loved it. Our shop at work is about to transition from maintaining older, conventional material transporters to robots, and we will have to learn how all this stuff works in order to repair them. If anything we get uses a Capstan Drive, I am definitely going to send them all a link to this video.
He’s alive!😂
What is the ratio for spider silk?
I get that reference
Love the video, especially the info about the 'why' of limitations and design choices!
Another approach to the D/d ratio is to use smaller stainless steel wire. I suspect it's significantly larger than needed.
How do you know "yo mama"? 🤣🤣
Thank you so much for filling in a huge missing gap in my knowledge. I never knew about such magical things until your video. You also explain the workings at a level that most technically minded people will understand, without feeling like a fool. Thank you!!
Really happy to see that the algorithm has given you a deserved increase in exposure, your work is exceptional.
Most surprised by the quick and consistent failure of the stainless cable. Most impressed by the DM20 rope's performance. Most satisfied with the 'square/circle/diamond'. Most entertained by the 'your momma' joke 🤣
Great presentation!
10/10 - would learn again