Thanks for watching! If you would like to support our work and get access to additional videos and articles, please check out our Patreon page: patreon.com/ChronovaEngineering PS Loving all your suggestions - I can't respond to them all but I read every one. ~Mike
Before you drop in the CA glue, put a nylon dummy shaft in the middle. Basically a thread of nylon, that won't glue itself and give you a clear shaft hole. Also... maybe consider making the rotor from 3 pieces. Two disc halves and a shaft from normal metal (or maybe titanium). You could "rebond" it all together using a drop of glue on top, and to make sure it doesn't wick to the sides/other side, bury the assembly in rodico. Just a thought.
I don't know if it would be considered cheating or not, but instead of machining the shaft, you should use a thin wire. Just cut to the required length, magnetise it and pass it through the centre of the stator. You can make a loop at the ends of the wire, so you can attach another wire that would serve as a crank or rod in order to drive something.
This is Amazing! I Know there are ways to Basically Get Free Energy at a very Small Scale. Either from the Atmosphere or by collecting Residual Charges with "Simple" Capacitors. With a Motor This Small, I am really Interested to see what the Minimum Volts/Amps requirement is! Wow, man, Just Amazing!
@@anthonyanglim7147 I thought about this alot while staying the night at a studio (recording sesh) next to highway 610 ⚠️😳 the roar of 70mph+ cars could power the light poles 🤌🤷♂️
I assume that he's just making the filming process more complex for the sake of making videos more interesting to watch,if you get what I'm trying to say.
@@TheGreatAtario I dunno, aren't most lenses coated with a thin film of something or other? ;) (I know that's not what you mean, but I also presume that what you mean is not what the OP meant.)
McLellan gets all of the credit for this motor, however Jim Frassrand from Arcadia, CA was responsible for the winding of the electromagnets and assembly. After the first motor was built and demonstrated, ten more complete motors with controllers were built and sold. The motor was driven by a large, battery operated, motor/generator that generated the multiphase outputs for driving the electromagnets. I viewed the motor operating through a 10x microscope eyepiece. The rotor was painted half white and half black and I estimated that it rotated at roughly a few hundred RPM. I got to witness this in 1974 as Jim's son Michael brought it to our workplace. Good luck on this most difficult challenge!
That's really interesting - I had no idea, I was under the impression it was a 1-person effort, so it's comforting to know he had some help! Thanks for sharing.
Can any of these mothers be viewed today? Any videos showing them operating? And did Feynman ever pay the reward? Thanks for the background information. I’ve lived in and around Arcadia CA my whole life.
@@glennac, I haven't communicated with Jim Frassrand's son Michael in almost 50 years. I can try and contact him as I think that he's still living in Arcadia. Most likely he still has his father's motor. I'll chime in if I learn anything. Jay
Brilliant video! I've always wondered why no one has tried to replicate the McLellan motor. I'm a Ph.D. student who designs and builds microrobots, so I have some experience in this area. A couple of suggestions: 1. As someone else mentioned, UV glue is awesome for assembling things like this. My lab loves Bondic [1], which is readily available online. It is definitely weaker than CA. If I need a strong joint, I'll use it to fix a part in the correct position, and then reinforce with low-viscosity CA (Loctite 495 [2]). 2. For applying glues at a microscale, we use either super-fine tungsten wires [3] held in a pin vise [4], or micro spatulas [5] (my favorite). Residue can be burned off with a lighter. 3. As you said, soldering fine-gauge magnet wire is...challenging. I've never really gotten the hang of it, so I prefer to carefully scratch the insulation off with a razor blade under a microscope, and then adhere it to a PCB pad using conductive epoxy such as MG 8331D [6]. Others have had luck using lots of flux and very low temperature solder. 4. For poking prodding and positioning, we have had great success using dental picks (along with super fine tweezers). Reverse action tweezers (which are bit hard to find with super-fine sizes, but possible) also seem to reduce how often parts go flying by applying a constant pressure. 5. For the record, I think that a jewel bearing is an excellent idea for reducing friction. Centimeter-scale microrobotics has moved away from rotary actuators because friction scales poorly as size decreases. However, I think that another reason is just how difficult it is to build small rotary motors, so thanks so much for your awesome work here :) If you have any particular questions, feel free to shoot me a message. There are some other people in my lab who have ideas that could help you, but I don't want to publicize them here without their permission. Links to products mentioned: [1] www.amazon.com/Bondic-Activated-Waterproof-Resistant-Adhesive/dp/B018IBEHQU?ref_=ast_sto_dp [2] www.amazon.com/Loctite-Bonder-442-49550-Instant-Adhesive/dp/B00065UAH2?ref_=ast_sto_dp [3] www.mcmaster.com/3775K36/ [4] www.mcmaster.com/8455A16/ [5] www.digikey.com/en/products/detail/excelta-corporation/260AA/16691210 [6] www.digikey.com/en/products/detail/mg-chemicals/8331D-14G/13170710
Great write up! Thought I could contribute a few thoughts also: I haven’t used these specifically, but I have had great success with Excelta tweezers (or any Swiss made precision tools, really) in my lab for wafer handling and micro soldering. They make reverse tweezers [1] with the smallest size being “very fine point” (0.2mm x 0.1mm tip). They also have a “Straight Cobaltima®” line [2] that which go down to a ludicrous 0.01 mm x 0.05 mm tip size. On-wafer probes (specifically those for DC) might also be of interest as they have extremely fine tips (~ 5 micron tip radius) and can be placed on an XYZ micro-positioner for micron level control. Probes like the Form Factor DCP 100 [3] are fairly cheap as far as probes go and have replaceable probe needles/tips. You can also get more of a whisker style probe too. Typically made of tungsten wire with beryllium copper tips (though the metallurgy varies wildly so don’t quote me on that). They are also of course conductive ;) [1] www.excelta.com/very-fine-points-31-sa [2] www.excelta.com/straight-tweezers-and-forceps-extreme-precision-cobaltima [3] www.formfactor.com/download/dcp-100-datasheet/?wpdmdl=3382&refresh=6797d2040202a1738002948
love how people make use of the youtube and comment with bright ideas. it is hard to see nowadays. youtube should be a base for people to connect with one another to achieve something great like what you people contribute today. thumbs up guys.
He just made a joke. It's visible in any quality since the video has all the details. It's shot perfectly to avoid such issues in lower quality. But he made a joke about how tiny the motor is.
Yeah, that's perfectly reasonable tbh. I will marry any female electrical engineering type who appears. There can't be many fields more depressingly sausagefestive than EE.
How do you know it wasn't more like the following? Hazel: "What are your skills?" Chronova Engineering: "Machining at extremely small scales, microscopic video making". Hazel: "Marry me". ... or any of myriad other possibilities? 😊😉
AlNiCo magnets are quite ubiquitous in guitar making, and their resistance to machining was also a problem for Fender engineers when they wanted to develop a guitar pickup with adjustable pole pieces. Their solution to the problem was using CuNiFe magnets, which are supposedly a lot easier to machine.
@@Oldtanktapperthe strings and their ratio of thickness to length would be the main issue that I can see. That and it would be impossible to play, lol
@@Oldtanktapperbased on some quick calculations the minimum string length that would produce an audible pitch would be around 1 cm, so it would be very hard to build one any smaller than that. Also at that size you would have to have such drastically different diameters of strings in order to be able to tune all 6 strings to different pitches I'm not sure it would be possible even at that size.
Speaking of such things, I once watched an interview of a man, who would scratch microscopic writing on a grain of rice and such. The level of dexterity required is so high that even one’s own heart beat would cause problems, by shaking the hand. So to remedy this problem, he would swim many hours per day, to improve his cardiovascular system. He had achieved a record breaking heart rate of 30 beats per minute. The slow heart rate gave him time to write in between the heart beats. The writing was so small that a whole paragraph would fit on a pin head.
@@aleksjenner677his name is Graham Short and his work is mind blowing. most of his work is almost invisible to the naked eye, he engraves portraits onto tiny discs and positions them in the eyes of needles
This is one of the most fascinating RUclips videos I’ve seen in a decade. I’m absolutely speechless at what can be achieved with just analog tools and skilled hands.
Amazing, as always. When milling PCBs, I like to cover the copper with machine oil. It facilitates the cut, but most importantly, keeps the awful fiber glass dust in the oil and away from the air and peoples lungs
Nice tip, I will remember that. Isopropanol works for machining aluminum, do you know if that would work for copper too? That would leave less residue than cutting oil
So good! Just FYI, regarding the wire "melting", that's not actually what's happening. I've experienced the same thing when connecting to the ~20um windings of surface mount 100:1 transformers for failure analysis, and I determined that what's happening is that the wire is dissolving in the solder, not melting. I had considered trying to pre-alloy the solder with copper by putting it in contact with a large copper wire before doing the work, but in the end, like you, I just figured out a way to make the connection without completely dissolving the wire. I honestly forget exactly what I ended up doing (it was more than a decade ago), I think I just did the work faster. Having abandoned a project to make a ~mm scale coreless motor for a micro RC plane, I'm very excited to see someone persevering and making headway on an even more challenging project! Actually, I would recommend that you consider a coreless design (windings on a nonmagnetic material). The forces will be lower, but removing the magnetic cores would eliminate any tendency for the rotor to stick to the poles. Oh, by the way, I'm guessing this doesn't fulfill your/Feynman's "no new technology" stipulation but have you looked into the squiggle motor and its relative the hollow tube ultrasonic motor? The hollow tube variant is a piezoelectric ultrasonic motor that can be made at the ~mm scale relatively easily. I did successfully make one of these using hand tools (a watchmaker's lathe, and a special grinding fixture to hold the stator against abrasive paper in the right position to grind the flats for the PZT ceramic plates, which I scored/snapped out of a piezo buzzer with an xacto knife). The one I built loosely replicated the motor described in the paper "A 1.6-mm, Metal Tube Ultrasonic Motor". But the squiggle motor is based on modern piezoelectric materials rather than traditional electromagnetism, so in my mind it doesn't win the prize.
This is super interesting! I've noticed the same thing with wire of that size, but assumed the wire broke off due to thermal stresses and became engulfed into the bead of solder. I'm curious, do you remember how you determined it was actually dissolving?
@PuchMaxi Oh no, it was leaded solder. Lead free solder already has copper in the alloy, I actually imagine maybe lead free is better in this respect? Lead free is a ternary eutectic with tin, copper, and silver. But honestly I don't know enough about metallurgy to know offhand whether a eutectic will dissolve more of a component metal when it contacts a lump of the pure metal. I think if it was exactly at the melting temperature it probably would not, but if it were any hotter it seems like maybe it could dissolve more copper? Honestly don't know.
@cameronurban3653 I started looking really closely under the microscope and noticed that in cases where I did successfully strip the enamel and tin it, the wire had thinned out to a substantially smaller diameter, and actually had a shape very similar to pictures I've seen of etched tungsten scanning electron microscope stylus tips (where the metal wire tip is electrochemically dissolved to make it sharp). Then I started poking the wire into the solder blob under the microscope and watched it happen.
I'm only a few minutes into this video when all of a sudden I realize that the inside of my mouth feels cold and dry. Then I realized my mouth was open, and probably had been for most of the video. This is amazing! So tiny, it's smaller than I had ever imagined trying to make anything. Thanks for sharing with us!
In the summer of 1985 I visited a friend who was a student at Caltech. You could just walk on campus into a building where you could see this motor, press a button and see it spin for a few seconds under a microscope. There were other projects on display in the hall, but I only remember the tiny motor. You could also go down to the sub basements and access the steam tunnels to go from building to building. There was such an amazing vibe on the campus. I have never experienced that anywhere else.
If you had three pole-pieces and set them up with a three-phase drive to create a rotating field, it might be possible to use a rotor machined from fully-hard Sterling silver rod, or possibly Tellurium copper. No need for it to be magnetised, it would be driven by eddy currents. The drive waveform could be an analogue pseudo-sine, with eight or so steps, although at the frequency you'll be running, it could probably use a real cosine function. Simpler to go with PWM, but it depends how many DAC outputs your microcontroller has. Each coil would need to be driven from two antiphase outputs. I guess you checked for temperature rise in the coils? That might be the limiting parameter. Supporting the rotor using surface tension seems a good idea if you can find a low-viscosity non-evaporating fluid. Great stuff. Mad, but great!
This remains true of the iron rotor (even demagnetized), but the catch in both cases: an induction motor requires a high enough drive frequency that the eddy currents don't go all the way through it (or in transient terms, don't stop before the phases switch). For a part of this scale, it might take some 100s of Hz, even kHz, to spin -- at which speed you won't see anything! Even a high-speed camera might be challenging due to the intense light needed and in such a small area!
@@T3sl4 Well, that should be fine, there was no constraint as to what speed it had to operate at. If you can see the black half side when stopped, then it turns to a blur, you know it's moving. You can always worry about how fast later.
@@T3sl4what about reluctance motor? No eddy currents, no peramanent magnetisation needed - the "only" problem is how to create rotor (cannot be simple disc).
@@grzegorzdomagala9929 Also thought about reciprocating reluctance (i.e. solenoid + mechanism). A flexture might make one more practical at this scale? An oscillator is still a motor -- but if one wants rotation specifically, converting one into the other will be similarly nontrivial (replace cogged rotor with ratchet rotor? heh).
Your patience levels are off the charts. There's absolutely no way i could do this. Certainly a great first attempt, i look forward to seeing how much further down this rabbit hole you go.
Ooof, I don't hate myself enough to attempt something like this. BTW add photocuring UV adhesive to your arsenal - you can manipulate the part till you're happy with it and a bit of UV light will set it in place in an instant.
@@michaelludvik2173 just did a quick search and it seems they're only available in Europe... unfortunately. They do offer large 50 gram bottles though which could make transatlantic shipping worth it. Note that a lot of UV CA glues are destined to be used for metal/glass combinations. Very few are meant for regular wood/metal/plastic etc. And... make sure to store inside a fridge to maximize their longevity
That is unbelievably impressive. I do take some pride in my ability to do some fairly fine detail work. But that is so far beyond my capabilities it's in an entirely different dimension. The fact you weren't able to get it fully functional makes no nevermind as far as I'm concerned. Just getting those coils spun on the posts is just bewildering and absolutely fascinating to watch. So I do wish you much luck in your next iteration of this project. Just absolutely wild to see.
This is by far the most impressive machining I've seen in a long time. Not to mention your patience 🤯Well done! Edit: the idea of a bunch of precision machined rotors just kicking around your workshop never to be seen again is hilarious to me
Such Hardworking Dedication In Creating magnificent Art of Science ❤And The level Patience You Kept All During Such Journey Of Creating art Is Mind-blowing.....I Watched whole Video With Single Skipping ❤
Please check that your Copper wires are actually pure and annealed ETP grade or better. Chinese suppliers love to swap pure Copper out for Copper-clad Aluminium, as well as to just skip annealing and leave the wire work hardened from the drawing operation. I do have a few tips for a second attempt: Use pure Gold wires (the type commonly used as bond-wires in IC packaging) in stead of Copper. It's not that expensive since the wires have very little mass, while also being more easily wound and handled without degrading its electrical or mechanical properties. You can also easily get it as uncoated wires, which would allow you to use heat to soften and anneal the metal if/when required, as well as giving you the ability to seal and insulate the whole coil (with a solution of whatever insulator you desire) after all other operations have been completed. Calculate the driving waveforms' maximum expected frequency content based upon what RPM you'd like to acheive and pick a suitable magnetic material to use as a core and backing plate (I'd imagine either something like Silicon Steel, also called Transformer Steel, or a soft and low-frequency optimized ferrite would be the best choice; However pure Iron or even just a basic Mild Steel, as long as it doesn't contain any carbon, might also work.). Putting something like brass in the path of the magnetic field not only results in very large induced eddy-losses, but also decreases the produced magnetic field density for a given applied EMF *and* causes the magnetic field that is produced to spread out significantly, instead of being confined to a well-defined magnetic gap. Imagine you're starting with a simple, single electromagnet. Wrapping a coil of wire around a random piece of steel will produce a very poor electromagnet, even though that's how electromagnets are often depicted. A coil of wire will produce a circulating magnetic field, with the total integrated magnetic flux that flows out of the center of the coil being equal in magnitude to the amount of magnetic flux that flows back around the outside of the coil. Therefore you really need to provide a proper magnetic path to conduct and focus the return field as well, which is why most commercial electromagnets produce a magnetic field that lies mostly in-plane with the exposed face, flowing in-between the center magnetic core material and the circular outer sheath material, in stead of extending far into the space in front of the electromagnet (as it is often incorrectly pictured). It's not like you're just missing out on something like half of the theoretical flux (by only using an inner magnetic core) either; since most suitable magnetic materials have permeabilities that are *thousands* of times larger than that of air (or brass), only using a permeable core for half of the magnetic path length actually yields a final magnetic field strength that is on the order of 0.1% of what it would have been with a full-path Core. Similarly, if you have a magnetic path where (for example) only 0.1% of the total length is an air-gap, then doubling that gap length will roughly halve the magnetic flux density in the gap. That's why you want to minimize the gap length (or use something like ferro-fluid to decrease the magnetic reluctance of the gap, although that would obviously not be a viable option in this project) and focus the produced magnetic field to avoid fringing field lines spreading out into the air. Essentially, you want to provide a high-permeability, closed magnetic path and only interrupt it at the contact surface between the stator and the rotor with the minimum possible gap.
How pray tell can one insulate the coil after winding it? There's already turn-on-turn contact! A problem as old as Faraday; back in his day, linen-wrapped iron wire was easiest; which sounds terrible in the modern day of enamels, but, back when these sorts of applications didn't even exist, you had to manage. Forging a new original solution carried similar difficulties; leveraging standard products (like very fine enameled wire) will be most helpful. Eddy currents aren't a problem for a display piece, where the excitation frequency will give a skin depth hundreds of times larger than even the brass base plate / bar used here. (Also why a synchronous (magnetized rotor) design is preferred over induction as another comment suggested. The higher speed required for the latter, does have the possibility to make it run better -- but it also can't be seen..!) A soft steel base might be preferable just to better optimize the magnetic path, but I don't have a problem with brass per se. I certainly wouldn't suggest ferrite here; it's much too brittle to handle. Especially if even the finest diamond saws are simply snapping off parts otherwise (or themselves as the case may be)!
@@T3sl4 The insulation can be done similarly to how some Tesla Coil secondaries are constructed. For each layer, you first wind all of the turns as normal, then you coat it in a very thin resin (something like a low viscosity, UV-cure resin or a fast-cure epoxy). Then just slightly pull on the ends to separate the turns from each other and wait until the resin has cured to a tacky, semi-solid consistency. Finally, you release the tension on the coil and accelerate the curing with intense UV light or by spraying an accelerator onto the coil. Since you have much more control over fine gold wires than ones made of any other metal, it will be much easier to wind and leaving a small gap in-between each turn won't be as detrimental in this application. Brass is a very big problem, because it is placed right in the magnetic return path, where a high-permeability material is needed. The issue with eddy losses has to do with the motor scaling laws. If you want the same amount of output power, however small, then you need to increase the RPM as the motor's physical size shrinks down. Furthermore, as you scale the rotor and bearing surfaces down, you end up significantly increasing the relative strength of friction, while simultaneously decreasing the rotor's moment of inertia. You essentially need to design for higher RPMs if you want such a small motor to look like it's not on the verge of stalling the entire time, but even more so because the size of this project precludes the use of the normal bearings, which would normally have helped to center the rotor shaft and to reduce sticktion at low linear contact speeds. Also, remember that the stator coils are scaled down by the same amount as everything else in the motor, that means that even if the skin depth is greater than the thickness of all the parts, you're still going to be limited by how much current you can force through those tiny wires, while being opposed by eddies that have the advantage of a much thicker and radially larger piece of metal that *ALL* of the magnetic flux is being forced to go through. Even if you don't want to (or can't) make a proper magnetic core that extends the whole length of the magnetic circuit; It should at least be possible to make the same mounting hardware out of a machineable plastic material, then bend a couple of iron wires into a C-shape, and finally press them into the plastic base plate, so that the magnetic field has at least a somewhat low reluctance path, from the bottom of the coils, up to the outer radial edge of the rotor.
I once was told by a man who made coils with 40 micron wire that common super glues do work for a period of time, but pretty soon with such thin wire it destroys wire's isolation. After a lot of testing now he uses shellac for such wire.
@@mr.perfect1067 Shellac is a resin secreted by certain bugs(often used for jelly candies). It easily dissolves in alcohol, so one can modify viscosity and even redo if necessary. I have even seen flux sensors made with 20 micron wire that are 30 years old which were successfully used for cryogenics.
This is one of the most mind-blowing projects I've ever seen on RUclips! It's incredible how all machining theory and mechanics change with scale. Thank you! :)
Your parting tool geometry is what I was taught when running brown and sharp 00G screw machines in the 1970's at Advanced Screw Products 😅. Just recently I tried to share this trick with some young manufacturing engineers.....there are many manufacturing tricks which have never been recorded and are doomed to be lost.
Same. I once played with 0.06mm thick enameled copper wire, and it was such a pain that I never toughed it again. I wanted to make costum guitar pickups, and now with 3D Printing I really would like to visit this topic again
Very cool man. I can only imagine the level of frustration in building and dealing with such small components - especially given the engineering/mechanical nature of the project. I applaud you for being so patient and sticking with it. Well done! :)
Absolutely unbelievable work! I don't know if this can help but they do make micro manipulators for inverted microscopes. You may have seen pictures of glass micro syringes injecting stuff into single cells. This equipment is available on eBay but unfortunately guaranteed working examples are usually expensive.
I can't decide what's the bigger feat...your persistence and expansive problem solving skills to build this device or the unbelievably compelling videography in capturing the scale of the technical difficulties one faces when working at this level. Nevertheless, making nano-scale, functional mechanical devices....both the fabrication of the individual components as well as the assembly of them is a key enabler to more affordable deep, deep space exploration. While advanced lithography methods taken from the semiconductor industry have enabled the nano-scale component challenge, the work you've done here to assemble these tiny, fragile components into a functional device is a real enabler to this application as undoubtedly to many other applications that are outside my own awareness. Amazing, amazing work.
This is an insane show of craftmanship. I wonder if a machine would even be able to create something like this. This feels like a task that only a human is capable of.
Just today, I spent almost an hour trying (over and over again) to install the 'Seconds' hand onto an NH35A watch movement. It's about 20 or 30 times larger than what you're working with here. Crazy !! 🙂
Chronova thank you so much I've been wondering for years what was the origin of the cash prize story that I saw 30 years ago in a documentary about nanotech. Finally now I can remember it's Feynman and the winner's name too
Absolutely amazing.. I love this kind of content. The Sheer dedication one must have towards a project like this must be immense. I enjoyed how you explained (and illustrated) the challenges and how you overcame them. Really good stuff!
Brilliant! I'm not sure if these other methods of motor driving are any easier to miniaturize, but I thought I would note them anyway: Electrostatic (corona discharge from a needle which pushes on the vanes of a tiny Crookes-radiometer), Induction ( Similar stator, but the rotor is not magnetized. Instead, a rotating magnetic field from the stator induces eddy currents in the rotor which produce a complementary magnetic field. ) I think the induction one is promising, since I've heard that such a stator can even rotate a plain copper slug.
You have nerves of steel to perform a task as this one ! Wow, even on my big screen I didn´t immediately discern the thin wire next to the electric resistor. Your dexterity even surpasses that of a neurosurgeon !
This takes me back. I remember seeing the original in Bridge when i was a student at Caltech. Didn't it have a fixed pin in the center that the rotor balanced on?
Yes I would love that! Wouldn't get too hot at that scale. Probably some kind of micro stirling engine using alcohol and a catalyst. Or just the heat of the hand!
I once came across an interesting perspective on a perpetuum mobile. The idea was not to eliminate energy loss but to reach such a level of precision that the losses become irrelevant. I am confident that these skilled individuals have the ability to achieve it.
"Wire itself tended to melt" 10:20 --> It's likely "dissolve" instead of "melt" - the copper dissolves into the solder (presumably until an equilibrium), hence when the solder is already equilibrated on the copper pad, it does not eat away the wire.
This is such impressive work. I'm sure everyone watching this has a crush on Richard Feynman just like I do, and its also just such a classic project. I would love to see more micro projects like this.
Armchair engineer here, with zero expertise in anything. I'm amazed at your ability to drill such fine holes in things. Would it be conceivable to drill a hole in the rotor and use a fine wire as the spindle?
Very nice project, just wanted to leave a small suggestion. The difficulty of working with Alnico may be avoided and the original rotor may work, if you make the motor AC instead of DC, as Induction motors don't need permanent magnets, any metallic rotor would work. You will be trading the mechanical complexity with electronic complexity, i.e., the electronic driver will need DACs to create the phase shifted sinusoids, but I think it's a good trade-off. Note: I am not sure how good the induction would be based on the current design and tolerances, but thought to share a possible avenue to explore.
@@user-rs8zg8ey2bI thought it would be more controllable to use a Microcontroller and DAC setup to generate the phase shifted sinusoids for each coil. So, that the amplitude, frequency and phase shift can be programmed for the experiment. It would be interesting to know your thoughts on this as well...
@@rounakdutta6211 99% of brushless motor do NOT control the amplitude. So imo you don't either, it just complicates the whole electronics drive section. On my channel I spin a magnet to 3.8Mrpm with a hall sensor, fet driver, fet and coil. Yes, you can't use a hall sensor, but why complicate something that does not need that level of control? Good luck.
@@user-rs8zg8ey2bI saw a few of your videos, especially the motor ones. Nice work on those. But when I said induction motor, I didn't mean bldc, I really meant "rotating magnetic field induction motor", so the rotor doesn't have to be a magnet. Without a magnetic rotor, hall-effect sensor based driver would be of little use. In your videos, I understand that the rotors are ball magnets, so you are able to use the hall effect sensor for timing the pulses. Also, since the rotor is very tiny, in this project, I thought a well controlled supply would be helpful in finding out the sweet spot in the power input. E.g., in your high speed motors, the rotor disintegrated very quickly, and the aftermath was ofc fun to watch :-), but that may not be the goal here...
@rounakdutta6211 ok, I see what you mean now. It's easy to change supply voltage to limit rpm. Unless you're a programmer (I am not), adding the dac control to vary the amplitude is not worth the extra complexity imo. Using a mpu to ramp up and hold a steady rpm is a good idea, with some trial and error you could find a nice sweet ramp up speed and running speed. A while ago I wrote different ways to drive brushless motors, but yours don't have magnets. If your interested I will find the link. I just had a KISS idea to drive your motor. If your motor has 4 poles, use a 4 pole bldc motor as a generator to drive your femtomotor. Well done! On another note, why not use a tiny magnet, one coil (no core) and spin it, it could possibly be smaller than your current motor as it has less coils.
the hand shots really set sense of scale, 1 coil down, here it is, a spec of dust held by a single fiber of invisible copper thread, amazing work, great camera work too
IMO, you'd be better off having the four "cores" 1:20 made from a non-magnetic material - say a piece of ceramic or glass (or even a piece of non-ferromagnetic metal, if your speeds are going to be low). The "cogging" from the bits of steel shall always imbalance your rotor, and you'd need damn good bearings to counteract this; whereas an "ironless" stator shall be far easier to implement, albeit less efficient.
This video was magical. Love the sticking to it, the tools, and the narration. If you learned something, the project was successful, compound that with us watchers learning and it was Very successful!
Respect to you for having the courage, skill, and perseverance to press on and reach your goal. Thank you also for showing the process, Marin clear that failure is part of learning and success, and finally for acknowledging the inspirations that provided the foundation that you stood on to be successful… and the partnership of your wife of course!
11:11 wear a mask, then you can safely breathe around these kinds of sizes. I use one for soldering, and while the solder smoke pops out, my breathing doesn't disturb it, it goes in a single, clean line towards the extractor (almost like in those laboratory pictures with wind tunnels for the study of hurricanes).
I interpreted his statement to be less about blowing the part away and more about the additional vibration/movement that breathing causes in your body. I find when micro-soldering that holding my breath briefly steadies my hands significantly if I need the extra stability. A mask does’t hurt to prevent air currents though.
Totally Astounding! I love it! Keep on Chronova bro', your project is so cool and tiny! You are learning - and teaching - so much, it doesn't matter if the motor is not perfect!
Its crazy, I do a lot of fine detailed work also that requires me to old my breath often and I am finding myself doing the same just watching this lol. I broke my brain I think.
This is so bonkers (in the best way) I'm in awe at your dexterity and patience. I'd also like to echo the use of UV cure glue over superglue. With a bit of practice, it's possible to shape it to your liking by judicious exposure to UV to adjust its viscosity. ie reduce capillary action while applying. It might also help to fix the windings while still mounted on the mandrel followed by a final fixation of the stator assembly.
Not sure you understand the scale here lol. 5x is nearly an equal challenge for 99.999% of any machinist. Which would mean it would take nearly as long for the first iteration. This IS his first iteration, it was the test to learn from. Also scaling something this sensitive to magnetism, air currents, debris, etc etc, most likely wouldn't transfer 5x smaller. Most of the "learning" was in the method of machining and assembling, not the proof of concept.
Браво ! Много е вдъхновяващо ! Бях останал с впечатление, че прецизността е останало в миналото, във времето на прецизната механика изработвана на ръка. Това, което Вие правите е УНИКАЛНО ! Много се радвам, че човечеството е все още достойно любопитно и можещо. Пожелавам Ви успех !
This was an absolute wet dream to watch Doing all this by hand... I have all the technology & tooling i need at my fingertips everyday, 5 micron bores and 10 micron true positions are standard for me You still had my full attention from start to finish here 👏
I absolutely loved watching this, your work here is amazing :) I had a few thoughts that might help both in terms of building a rotor in AlNiCo or even potentially in neodymium magnet material directly. There are now open source EDM controllers which would enable you to build a miniaturised EDM sinker if you wanted to, or EDM wire cutter, although, obviously you'd then need a wire feeder etc. The benefit here would be -nearly- zero force machining, which could solve your parting problem. Depending on how committed you get to the design, this would also allow you to make custom carbide tools at this scale as well, which might help with your operations on materials like the AlNiCo etc. where you conduct those on a lathe. As an additional suggestion I've build broaching and form tools in the past for (admittedly much larger) jobs which required substantial variation in job diameter for a lathe. The benefit here is that a broach uses symetric cutting, which results in a far lower risk of part damage during turning. Forming things like the rotor and shaft could be made in a single operation and manufacturing the broach would actually be easier than manufacturing the positive you want to produce from it, because you're drilling holes and milling steps, not cutting long unsupported shafts. I hope this doesn't sound like I'm being negative about how I view your work, I was trained up as a toolmaker as a child and worked on relatively small die and tool jobs before I did my engineering degree, I've been an engineer for about 20 years now in a pretty niche high product value field and I can tell you that the work you're doing is unique in my experience. I'm frankly amazed at what you're achieving and I hope that my suggestions might contribute positively in some way.
By Jove! Kudos to you Sir! Again a breathtaking video. Your videos are the highlight of my day, week, month(s) and are very much appreciated. What a skill… 🙏🏻
A simple circular recess ground into the top side of the coverglass to hold the rotor in place should work. Awesome work dude RESPECT. thankyou for sharing
With a Davinci articulated robot arm this would have been childs play,but that would have taken all the fun out of it. Good job man! Unbelievable patients!
Super impressive! It looks like what you need is a pantograph-tweezer setup that lets you transform larger hand movements with mock-tweezers into 1/10th scale movements with tiny tweezers for doing all of the tiny stuff by hand. That would be a fun one to engineer, CAD, and fabricate to have minimal slop and backlash in the system :]
This is so incredible!! The first shaft model was correct, but it needed to be magnetized by a string electromagnetic current like normal magnets, if you use the spinning magnets technique the spins got briefly magnetized and it won’t maintain its orientation for a long time
Thanks for watching! If you would like to support our work and get access to additional videos and articles, please check out our Patreon page: patreon.com/ChronovaEngineering
PS Loving all your suggestions - I can't respond to them all but I read every one. ~Mike
Nice reference to Mr Feynmans book in title pic !😊
Before you drop in the CA glue, put a nylon dummy shaft in the middle. Basically a thread of nylon, that won't glue itself and give you a clear shaft hole. Also... maybe consider making the rotor from 3 pieces. Two disc halves and a shaft from normal metal (or maybe titanium). You could "rebond" it all together using a drop of glue on top, and to make sure it doesn't wick to the sides/other side, bury the assembly in rodico. Just a thought.
I don't know if it would be considered cheating or not, but instead of machining the shaft, you should use a thin wire. Just cut to the required length, magnetise it and pass it through the centre of the stator. You can make a loop at the ends of the wire, so you can attach another wire that would serve as a crank or rod in order to drive something.
This is Amazing! I Know there are ways to Basically Get Free Energy at a very Small Scale. Either from the Atmosphere or by collecting Residual Charges with "Simple" Capacitors. With a Motor This Small, I am really Interested to see what the Minimum Volts/Amps requirement is! Wow, man, Just Amazing!
@@anthonyanglim7147 I thought about this alot while staying the night at a studio (recording sesh) next to highway 610 ⚠️😳 the roar of 70mph+ cars could power the light poles 🤌🤷♂️
The only thing more mind-blowing than the fact that you built this is the fact that you filmed it so well
I have my doubts any film was involved
I assume that he's just making the filming process more complex for the sake of making videos more interesting to watch,if you get what I'm trying to say.
@@TheGreatAtario I dunno, aren't most lenses coated with a thin film of something or other? ;)
(I know that's not what you mean, but I also presume that what you mean is not what the OP meant.)
@@TheGreatAtario so whats the correct term, mister einstein?
@@PM-wt3ye What's wrong with "recorded" or "shot"?
McLellan gets all of the credit for this motor, however Jim Frassrand from Arcadia, CA was responsible for the winding of the electromagnets and assembly. After the first motor was built and demonstrated, ten more complete motors with controllers were built and sold. The motor was driven by a large, battery operated, motor/generator that generated the multiphase outputs for driving the electromagnets. I viewed the motor operating through a 10x microscope eyepiece. The rotor was painted half white and half black and I estimated that it rotated at roughly a few hundred RPM. I got to witness this in 1974 as Jim's son Michael brought it to our workplace. Good luck on this most difficult challenge!
That's really interesting - I had no idea, I was under the impression it was a 1-person effort, so it's comforting to know he had some help! Thanks for sharing.
Brilliant
Can any of these mothers be viewed today? Any videos showing them operating? And did Feynman ever pay the reward? Thanks for the background information. I’ve lived in and around Arcadia CA my whole life.
@@glennac, I haven't communicated with Jim Frassrand's son Michael in almost 50 years. I can try and contact him as I think that he's still living in Arcadia. Most likely he still has his father's motor. I'll chime in if I learn anything. Jay
@@glennac, yes Feynman did pay for the efforts however he was a bit disappointed. He wanted to see this done with a new type of technology.
This is the kind of content that keeps my faith in the power of the internet alive.
I totally agree! 👍🏻
Agreed
Well put! I absolutely agree!
I fear AI will shake faith in real videos.
skibidi
Brilliant video! I've always wondered why no one has tried to replicate the McLellan motor. I'm a Ph.D. student who designs and builds microrobots, so I have some experience in this area. A couple of suggestions:
1. As someone else mentioned, UV glue is awesome for assembling things like this. My lab loves Bondic [1], which is readily available online. It is definitely weaker than CA. If I need a strong joint, I'll use it to fix a part in the correct position, and then reinforce with low-viscosity CA (Loctite 495 [2]).
2. For applying glues at a microscale, we use either super-fine tungsten wires [3] held in a pin vise [4], or micro spatulas [5] (my favorite). Residue can be burned off with a lighter.
3. As you said, soldering fine-gauge magnet wire is...challenging. I've never really gotten the hang of it, so I prefer to carefully scratch the insulation off with a razor blade under a microscope, and then adhere it to a PCB pad using conductive epoxy such as MG 8331D [6]. Others have had luck using lots of flux and very low temperature solder.
4. For poking prodding and positioning, we have had great success using dental picks (along with super fine tweezers). Reverse action tweezers (which are bit hard to find with super-fine sizes, but possible) also seem to reduce how often parts go flying by applying a constant pressure.
5. For the record, I think that a jewel bearing is an excellent idea for reducing friction. Centimeter-scale microrobotics has moved away from rotary actuators because friction scales poorly as size decreases. However, I think that another reason is just how difficult it is to build small rotary motors, so thanks so much for your awesome work here :)
If you have any particular questions, feel free to shoot me a message. There are some other people in my lab who have ideas that could help you, but I don't want to publicize them here without their permission.
Links to products mentioned:
[1] www.amazon.com/Bondic-Activated-Waterproof-Resistant-Adhesive/dp/B018IBEHQU?ref_=ast_sto_dp
[2] www.amazon.com/Loctite-Bonder-442-49550-Instant-Adhesive/dp/B00065UAH2?ref_=ast_sto_dp
[3] www.mcmaster.com/3775K36/
[4] www.mcmaster.com/8455A16/
[5] www.digikey.com/en/products/detail/excelta-corporation/260AA/16691210
[6] www.digikey.com/en/products/detail/mg-chemicals/8331D-14G/13170710
Great write up! Thought I could contribute a few thoughts also:
I haven’t used these specifically, but I have had great success with Excelta tweezers (or any Swiss made precision tools, really) in my lab for wafer handling and micro soldering. They make reverse tweezers [1] with the smallest size being “very fine point” (0.2mm x 0.1mm tip). They also have a “Straight Cobaltima®” line [2] that which go down to a ludicrous 0.01 mm x 0.05 mm tip size.
On-wafer probes (specifically those for DC) might also be of interest as they have extremely fine tips (~ 5 micron tip radius) and can be placed on an XYZ micro-positioner for micron level control. Probes like the Form Factor DCP 100 [3] are fairly cheap as far as probes go and have replaceable probe needles/tips. You can also get more of a whisker style probe too. Typically made of tungsten wire with beryllium copper tips (though the metallurgy varies wildly so don’t quote me on that). They are also of course conductive ;)
[1] www.excelta.com/very-fine-points-31-sa
[2] www.excelta.com/straight-tweezers-and-forceps-extreme-precision-cobaltima
[3] www.formfactor.com/download/dcp-100-datasheet/?wpdmdl=3382&refresh=6797d2040202a1738002948
Props to you guys, and comment for the alogryim
love how people make use of the youtube and comment with bright ideas. it is hard to see nowadays. youtube should be a base for people to connect with one another to achieve something great like what you people contribute today. thumbs up guys.
Great response and great resources. Where are you studying?
Fancy seeing you here 🙂
I'm glad I'm watching this at 1080p, else I would be watching a single pixel for 16 minutes.
It's visible in 720p, even the wire.
Even at 144p
He just made a joke. It's visible in any quality since the video has all the details. It's shot perfectly to avoid such issues in lower quality. But he made a joke about how tiny the motor is.
The Connections (2021) [short documentary] ❤🎉
😂😂😂😂😂😂@@bmxrider8284
Chronova Engineering: "What are your skills"? Hazel: "Electronics". Chronova, "Marry me".
😂
😂😂😂
Yeah, that's perfectly reasonable tbh.
I will marry any female electrical engineering type who appears.
There can't be many fields more depressingly sausagefestive than EE.
How do you know it wasn't more like the following?
Hazel: "What are your skills?" Chronova Engineering: "Machining at extremely small scales, microscopic video making". Hazel: "Marry me". ... or any of myriad other possibilities? 😊😉
Your pfp is awesome
AlNiCo magnets are quite ubiquitous in guitar making, and their resistance to machining was also a problem for Fender engineers when they wanted to develop a guitar pickup with adjustable pole pieces. Their solution to the problem was using CuNiFe magnets, which are supposedly a lot easier to machine.
Thanks for the tip - this looks like the ideal material. I will give it a go!
I’m now wondering just how small a working electric guitar could be? Heck, might as well make a miniature amp and speaker for it too!
@@Oldtanktapper Make it happen! Make it happen!
@@Oldtanktapperthe strings and their ratio of thickness to length would be the main issue that I can see. That and it would be impossible to play, lol
@@Oldtanktapperbased on some quick calculations the minimum string length that would produce an audible pitch would be around 1 cm, so it would be very hard to build one any smaller than that. Also at that size you would have to have such drastically different diameters of strings in order to be able to tune all 6 strings to different pitches I'm not sure it would be possible even at that size.
Speaking of such things, I once watched an interview of a man, who would scratch microscopic writing on a grain of rice and such. The level of dexterity required is so high that even one’s own heart beat would cause problems, by shaking the hand. So to remedy this problem, he would swim many hours per day, to improve his cardiovascular system. He had achieved a record breaking heart rate of 30 beats per minute. The slow heart rate gave him time to write in between the heart beats. The writing was so small that a whole paragraph would fit on a pin head.
Sounds unlikely. Source?
@@aleksjenner677 na I also heard something along these lines once
@@aleksjenner677his name is Graham Short and his work is mind blowing. most of his work is almost invisible to the naked eye, he engraves portraits onto tiny discs and positions them in the eyes of needles
@ The world record holder is Daniel Green. 26 bpm resting rate, sat in 2014.
@ I find no information that Daniel Green wrote an entire paragraph on a pin head. Still looking for a source for your unlikely claims...
This is one of the most fascinating RUclips videos I’ve seen in a decade. I’m absolutely speechless at what can be achieved with just analog tools and skilled hands.
One of the coolest parts of this video was hearing that your wife was helping you how she can with her expertise. Its very sweet!
Where does one get a wife like that? Hell, where does one get a friend like that?
🤣🤣
Amazing, as always. When milling PCBs, I like to cover the copper with machine oil. It facilitates the cut, but most importantly, keeps the awful fiber glass dust in the oil and away from the air and peoples lungs
Nice tip, I will remember that. Isopropanol works for machining aluminum, do you know if that would work for copper too? That would leave less residue than cutting oil
So good!
Just FYI, regarding the wire "melting", that's not actually what's happening. I've experienced the same thing when connecting to the ~20um windings of surface mount 100:1 transformers for failure analysis, and I determined that what's happening is that the wire is dissolving in the solder, not melting.
I had considered trying to pre-alloy the solder with copper by putting it in contact with a large copper wire before doing the work, but in the end, like you, I just figured out a way to make the connection without completely dissolving the wire. I honestly forget exactly what I ended up doing (it was more than a decade ago), I think I just did the work faster.
Having abandoned a project to make a ~mm scale coreless motor for a micro RC plane, I'm very excited to see someone persevering and making headway on an even more challenging project!
Actually, I would recommend that you consider a coreless design (windings on a nonmagnetic material). The forces will be lower, but removing the magnetic cores would eliminate any tendency for the rotor to stick to the poles.
Oh, by the way, I'm guessing this doesn't fulfill your/Feynman's "no new technology" stipulation but have you looked into the squiggle motor and its relative the hollow tube ultrasonic motor? The hollow tube variant is a piezoelectric ultrasonic motor that can be made at the ~mm scale relatively easily. I did successfully make one of these using hand tools (a watchmaker's lathe, and a special grinding fixture to hold the stator against abrasive paper in the right position to grind the flats for the PZT ceramic plates, which I scored/snapped out of a piezo buzzer with an xacto knife). The one I built loosely replicated the motor described in the paper "A 1.6-mm, Metal Tube Ultrasonic Motor". But the squiggle motor is based on modern piezoelectric materials rather than traditional electromagnetism, so in my mind it doesn't win the prize.
This is super interesting! I've noticed the same thing with wire of that size, but assumed the wire broke off due to thermal stresses and became engulfed into the bead of solder. I'm curious, do you remember how you determined it was actually dissolving?
Very interesting observation! Is this due to lead-free solder being used?
@PuchMaxi Oh no, it was leaded solder. Lead free solder already has copper in the alloy, I actually imagine maybe lead free is better in this respect? Lead free is a ternary eutectic with tin, copper, and silver. But honestly I don't know enough about metallurgy to know offhand whether a eutectic will dissolve more of a component metal when it contacts a lump of the pure metal. I think if it was exactly at the melting temperature it probably would not, but if it were any hotter it seems like maybe it could dissolve more copper? Honestly don't know.
@cameronurban3653 I started looking really closely under the microscope and noticed that in cases where I did successfully strip the enamel and tin it, the wire had thinned out to a substantially smaller diameter, and actually had a shape very similar to pictures I've seen of etched tungsten scanning electron microscope stylus tips (where the metal wire tip is electrochemically dissolved to make it sharp). Then I started poking the wire into the solder blob under the microscope and watched it happen.
@ Good to know! I think you are right about the higher temperatures, time/duration is also a factor.
I'm only a few minutes into this video when all of a sudden I realize that the inside of my mouth feels cold and dry. Then I realized my mouth was open, and probably had been for most of the video. This is amazing! So tiny, it's smaller than I had ever imagined trying to make anything. Thanks for sharing with us!
In the summer of 1985 I visited a friend who was a student at Caltech. You could just walk on campus into a building where you could see this motor, press a button and see it spin for a few seconds under a microscope. There were other projects on display in the hall, but I only remember the tiny motor. You could also go down to the sub basements and access the steam tunnels to go from building to building. There was such an amazing vibe on the campus. I have never experienced that anywhere else.
1st) Super attempt
2nd) Totally blown away after reading all the knowledgeable comments.
If you had three pole-pieces and set them up with a three-phase drive to create a rotating field, it might be possible to use a rotor machined from fully-hard Sterling silver rod, or possibly Tellurium copper. No need for it to be magnetised, it would be driven by eddy currents. The drive waveform could be an analogue pseudo-sine, with eight or so steps, although at the frequency you'll be running, it could probably use a real cosine function. Simpler to go with PWM, but it depends how many DAC outputs your microcontroller has. Each coil would need to be driven from two antiphase outputs. I guess you checked for temperature rise in the coils? That might be the limiting parameter. Supporting the rotor using surface tension seems a good idea if you can find a low-viscosity non-evaporating fluid. Great stuff. Mad, but great!
It’s amazing how often those two intersecc
This remains true of the iron rotor (even demagnetized), but the catch in both cases: an induction motor requires a high enough drive frequency that the eddy currents don't go all the way through it (or in transient terms, don't stop before the phases switch). For a part of this scale, it might take some 100s of Hz, even kHz, to spin -- at which speed you won't see anything! Even a high-speed camera might be challenging due to the intense light needed and in such a small area!
@@T3sl4 Well, that should be fine, there was no constraint as to what speed it had to operate at. If you can see the black half side when stopped, then it turns to a blur, you know it's moving. You can always worry about how fast later.
@@T3sl4what about reluctance motor? No eddy currents, no peramanent magnetisation needed - the "only" problem is how to create rotor (cannot be simple disc).
@@grzegorzdomagala9929 Also thought about reciprocating reluctance (i.e. solenoid + mechanism). A flexture might make one more practical at this scale? An oscillator is still a motor -- but if one wants rotation specifically, converting one into the other will be similarly nontrivial (replace cogged rotor with ratchet rotor? heh).
Your patience levels are off the charts. There's absolutely no way i could do this. Certainly a great first attempt, i look forward to seeing how much further down this rabbit hole you go.
1:36 this is such a "couple goals" quote. Awesome that you can share such a passion!
As a machinist who struggles when turning
Ooof, I don't hate myself enough to attempt something like this.
BTW add photocuring UV adhesive to your arsenal - you can manipulate the part till you're happy with it and a bit of UV light will set it in place in an instant.
Yes! Microtec Turbotec 911 UV works well for me...
@@bastian6173 Would a glass pipette (like the ones used to interact with single cells) be helpful for applying tiny amounts?
@@billmiller4800 personally I use thin pieces of wire to apply tiny amounts. If the droplet is small enough there is no need for a pipette
@@bastian6173is this product available in the US?
@@michaelludvik2173 just did a quick search and it seems they're only available in Europe... unfortunately. They do offer large 50 gram bottles though which could make transatlantic shipping worth it. Note that a lot of UV CA glues are destined to be used for metal/glass combinations. Very few are meant for regular wood/metal/plastic etc. And... make sure to store inside a fridge to maximize their longevity
I can confidently say that I will not be going down this rabbit hole. thank you for your sacrifice of time :-)
That is unbelievably impressive. I do take some pride in my ability to do some fairly fine detail work. But that is so far beyond my capabilities it's in an entirely different dimension. The fact you weren't able to get it fully functional makes no nevermind as far as I'm concerned. Just getting those coils spun on the posts is just bewildering and absolutely fascinating to watch. So I do wish you much luck in your next iteration of this project. Just absolutely wild to see.
0:10 Literaly 10sec into the video and I just said to myself "YOU MADE WHAT?!?!!?"
The Connections (2021) [short documentary] ❤🎉
This
This is by far the most impressive machining I've seen in a long time. Not to mention your patience 🤯Well done!
Edit: the idea of a bunch of precision machined rotors just kicking around your workshop never to be seen again is hilarious to me
Every single time you showed the "zoomed out" shot I was blown away by how small this was. Insane work dude you should be very proud
This video is literally unbelievable. I watched you do all of this and still can't believe it.
wow. and you said someone completed this challenge decades ago in the 70's with older tech? amazing 😮
Such Hardworking Dedication In Creating magnificent Art of Science ❤And The level Patience You Kept All During Such Journey Of Creating art Is Mind-blowing.....I Watched whole Video With Single Skipping ❤
I’ve never been so amazed and intrigued within the first 10 seconds of anything I’ve ever watched in my life! 🤯
Please check that your Copper wires are actually pure and annealed ETP grade or better. Chinese suppliers love to swap pure Copper out for Copper-clad Aluminium, as well as to just skip annealing and leave the wire work hardened from the drawing operation.
I do have a few tips for a second attempt:
Use pure Gold wires (the type commonly used as bond-wires in IC packaging) in stead of Copper. It's not that expensive since the wires have very little mass, while also being more easily wound and handled without degrading its electrical or mechanical properties. You can also easily get it as uncoated wires, which would allow you to use heat to soften and anneal the metal if/when required, as well as giving you the ability to seal and insulate the whole coil (with a solution of whatever insulator you desire) after all other operations have been completed.
Calculate the driving waveforms' maximum expected frequency content based upon what RPM you'd like to acheive and pick a suitable magnetic material to use as a core and backing plate (I'd imagine either something like Silicon Steel, also called Transformer Steel, or a soft and low-frequency optimized ferrite would be the best choice; However pure Iron or even just a basic Mild Steel, as long as it doesn't contain any carbon, might also work.).
Putting something like brass in the path of the magnetic field not only results in very large induced eddy-losses, but also decreases the produced magnetic field density for a given applied EMF *and* causes the magnetic field that is produced to spread out significantly, instead of being confined to a well-defined magnetic gap.
Imagine you're starting with a simple, single electromagnet.
Wrapping a coil of wire around a random piece of steel will produce a very poor electromagnet, even though that's how electromagnets are often depicted.
A coil of wire will produce a circulating magnetic field, with the total integrated magnetic flux that flows out of the center of the coil being equal in magnitude to the amount of magnetic flux that flows back around the outside of the coil.
Therefore you really need to provide a proper magnetic path to conduct and focus the return field as well, which is why most commercial electromagnets produce a magnetic field that lies mostly in-plane with the exposed face, flowing in-between the center magnetic core material and the circular outer sheath material, in stead of extending far into the space in front of the electromagnet (as it is often incorrectly pictured).
It's not like you're just missing out on something like half of the theoretical flux (by only using an inner magnetic core) either; since most suitable magnetic materials have permeabilities that are *thousands* of times larger than that of air (or brass), only using a permeable core for half of the magnetic path length actually yields a final magnetic field strength that is on the order of 0.1% of what it would have been with a full-path Core.
Similarly, if you have a magnetic path where (for example) only 0.1% of the total length is an air-gap, then doubling that gap length will roughly halve the magnetic flux density in the gap.
That's why you want to minimize the gap length (or use something like ferro-fluid to decrease the magnetic reluctance of the gap, although that would obviously not be a viable option in this project) and focus the produced magnetic field to avoid fringing field lines spreading out into the air.
Essentially, you want to provide a high-permeability, closed magnetic path and only interrupt it at the contact surface between the stator and the rotor with the minimum possible gap.
How pray tell can one insulate the coil after winding it? There's already turn-on-turn contact!
A problem as old as Faraday; back in his day, linen-wrapped iron wire was easiest; which sounds terrible in the modern day of enamels, but, back when these sorts of applications didn't even exist, you had to manage. Forging a new original solution carried similar difficulties; leveraging standard products (like very fine enameled wire) will be most helpful.
Eddy currents aren't a problem for a display piece, where the excitation frequency will give a skin depth hundreds of times larger than even the brass base plate / bar used here. (Also why a synchronous (magnetized rotor) design is preferred over induction as another comment suggested. The higher speed required for the latter, does have the possibility to make it run better -- but it also can't be seen..!) A soft steel base might be preferable just to better optimize the magnetic path, but I don't have a problem with brass per se.
I certainly wouldn't suggest ferrite here; it's much too brittle to handle. Especially if even the finest diamond saws are simply snapping off parts otherwise (or themselves as the case may be)!
I think the issue is not that it is aluminium, else it woudn't melt or dissolve, the issue is the wire is so thin that it dissolves in the solder.
@@T3sl4 The insulation can be done similarly to how some Tesla Coil secondaries are constructed.
For each layer, you first wind all of the turns as normal, then you coat it in a very thin resin (something like a low viscosity, UV-cure resin or a fast-cure epoxy). Then just slightly pull on the ends to separate the turns from each other and wait until the resin has cured to a tacky, semi-solid consistency. Finally, you release the tension on the coil and accelerate the curing with intense UV light or by spraying an accelerator onto the coil.
Since you have much more control over fine gold wires than ones made of any other metal, it will be much easier to wind and leaving a small gap in-between each turn won't be as detrimental in this application.
Brass is a very big problem, because it is placed right in the magnetic return path, where a high-permeability material is needed.
The issue with eddy losses has to do with the motor scaling laws. If you want the same amount of output power, however small, then you need to increase the RPM as the motor's physical size shrinks down.
Furthermore, as you scale the rotor and bearing surfaces down, you end up significantly increasing the relative strength of friction, while simultaneously decreasing the rotor's moment of inertia.
You essentially need to design for higher RPMs if you want such a small motor to look like it's not on the verge of stalling the entire time, but even more so because the size of this project precludes the use of the normal bearings, which would normally have helped to center the rotor shaft and to reduce sticktion at low linear contact speeds.
Also, remember that the stator coils are scaled down by the same amount as everything else in the motor, that means that even if the skin depth is greater than the thickness of all the parts, you're still going to be limited by how much current you can force through those tiny wires, while being opposed by eddies that have the advantage of a much thicker and radially larger piece of metal that *ALL* of the magnetic flux is being forced to go through.
Even if you don't want to (or can't) make a proper magnetic core that extends the whole length of the magnetic circuit; It should at least be possible to make the same mounting hardware out of a machineable plastic material, then bend a couple of iron wires into a C-shape, and finally press them into the plastic base plate, so that the magnetic field has at least a somewhat low reluctance path, from the bottom of the coils, up to the outer radial edge of the rotor.
I once was told by a man who made coils with 40 micron wire that common super glues do work for a period of time, but pretty soon with such thin wire it destroys wire's isolation. After a lot of testing now he uses shellac for such wire.
What do you mean by Shellac sir?
@@mr.perfect1067 Shellac is a resin secreted by certain bugs(often used for jelly candies). It easily dissolves in alcohol, so one can modify viscosity and even redo if necessary. I have even seen flux sensors made with 20 micron wire that are 30 years old which were successfully used for cryogenics.
@@ВладіславЗаморський thanks for information🙂❤
This is one of the most mind-blowing projects I've ever seen on RUclips! It's incredible how all machining theory and mechanics change with scale. Thank you! :)
Bravo to you sir. Amazing. Such detail and precision
Machinist for seemingly forever, I admire your persistence. Small nearly invisible parts are the most difficult by far. Good luck.
Your parting tool geometry is what I was taught when running brown and sharp 00G screw machines in the 1970's at Advanced Screw Products 😅. Just recently I tried to share this trick with some young manufacturing engineers.....there are many manufacturing tricks which have never been recorded and are doomed to be lost.
My brain doesn´t take the scale of this project.
Same. I once played with 0.06mm thick enameled copper wire, and it was such a pain that I never toughed it again. I wanted to make costum guitar pickups, and now with 3D Printing I really would like to visit this topic again
The same as trying to understand astronomy, just the other end of the scale (I nearly said telescope).
@@MS-Patriot2 Definitely! Astronomy is so interesting. I´m planing to build radio astronomy satellite for messing with h2 in space.
Very cool man. I can only imagine the level of frustration in building and dealing with such small components - especially given the engineering/mechanical nature of the project. I applaud you for being so patient and sticking with it. Well done! :)
Absolutely unbelievable work! I don't know if this can help but they do make micro manipulators for inverted microscopes. You may have seen pictures of glass micro syringes injecting stuff into single cells. This equipment is available on eBay but unfortunately guaranteed working examples are usually expensive.
Fantastic project! I especially love your tracing machine.
The transition from your microscope to the electron microscope was immensely satisfying.
I'd have lost my mind trying to do this.
Great work!
I can't decide what's the bigger feat...your persistence and expansive problem solving skills to build this device or the unbelievably compelling videography in capturing the scale of the technical difficulties one faces when working at this level. Nevertheless, making nano-scale, functional mechanical devices....both the fabrication of the individual components as well as the assembly of them is a key enabler to more affordable deep, deep space exploration. While advanced lithography methods taken from the semiconductor industry have enabled the nano-scale component challenge, the work you've done here to assemble these tiny, fragile components into a functional device is a real enabler to this application as undoubtedly to many other applications that are outside my own awareness. Amazing, amazing work.
You could probably make a smaller electrostatic motor, as it avoids the need to wind coils.
This is an insane show of craftmanship. I wonder if a machine would even be able to create something like this. This feels like a task that only a human is capable of.
I started laughing out loud a few times whenever the image on my screen showed how truly tiny this is. Amazing!
Just today, I spent almost an hour trying (over and over again) to install the 'Seconds' hand onto an NH35A watch movement.
It's about 20 or 30 times larger than what you're working with here. Crazy !! 🙂
Chronova thank you so much I've been wondering for years what was the origin of the cash prize story that I saw 30 years ago in a documentary about nanotech. Finally now I can remember it's Feynman and the winner's name too
Absolutely amazing.. I love this kind of content. The Sheer dedication one must have towards a project like this must be immense. I enjoyed how you explained (and illustrated) the challenges and how you overcame them. Really good stuff!
Brilliant! I'm not sure if these other methods of motor driving are any easier to miniaturize, but I thought I would note them anyway: Electrostatic (corona discharge from a needle which pushes on the vanes of a tiny Crookes-radiometer), Induction ( Similar stator, but the rotor is not magnetized. Instead, a rotating magnetic field from the stator induces eddy currents in the rotor which produce a complementary magnetic field. ) I think the induction one is promising, since I've heard that such a stator can even rotate a plain copper slug.
You have nerves of steel to perform a task as this one ! Wow, even on my big screen I didn´t immediately discern the thin wire next to the electric resistor. Your dexterity even surpasses that of a neurosurgeon !
This takes me back. I remember seeing the original in Bridge when i was a student at Caltech. Didn't it have a fixed pin in the center that the rotor balanced on?
Watchmaking always fascinated me, for how small it could become. And this... is insane. Great job, I'm definitely looking forward to V2 !
You should tottally try to make a gasoline powered wristwatch someday. If anyone can do it, you can!
lol, why do you want a burned wrist?
Yes I would love that! Wouldn't get too hot at that scale. Probably some kind of micro stirling engine using alcohol and a catalyst. Or just the heat of the hand!
You could do something similar with a methanol powered fuel cell, if you really needed a liquid power source.
I once came across an interesting perspective on a perpetuum mobile. The idea was not to eliminate energy loss but to reach such a level of precision that the losses become irrelevant. I am confident that these skilled individuals have the ability to achieve it.
"Wire itself tended to melt" 10:20 --> It's likely "dissolve" instead of "melt" - the copper dissolves into the solder (presumably until an equilibrium), hence when the solder is already equilibrated on the copper pad, it does not eat away the wire.
This is such impressive work. I'm sure everyone watching this has a crush on Richard Feynman just like I do, and its also just such a classic project. I would love to see more micro projects like this.
Armchair engineer here, with zero expertise in anything. I'm amazed at your ability to drill such fine holes in things. Would it be conceivable to drill a hole in the rotor and use a fine wire as the spindle?
This is the most insane machining project I have seen undertaken by one man, you clearly have a patience of a saint.
Very nice project, just wanted to leave a small suggestion. The difficulty of working with Alnico may be avoided and the original rotor may work, if you make the motor AC instead of DC, as Induction motors don't need permanent magnets, any metallic rotor would work. You will be trading the mechanical complexity with electronic complexity, i.e., the electronic driver will need DACs to create the phase shifted sinusoids, but I think it's a good trade-off.
Note: I am not sure how good the induction would be based on the current design and tolerances, but thought to share a possible avenue to explore.
You don't need DACs for this function at all. Why would you think this?
@@user-rs8zg8ey2bI thought it would be more controllable to use a Microcontroller and DAC setup to generate the phase shifted sinusoids for each coil. So, that the amplitude, frequency and phase shift can be programmed for the experiment. It would be interesting to know your thoughts on this as well...
@@rounakdutta6211 99% of brushless motor do NOT control the amplitude. So imo you don't either, it just complicates the whole electronics drive section.
On my channel I spin a magnet to 3.8Mrpm with a hall sensor, fet driver, fet and coil. Yes, you can't use a hall sensor, but why complicate something that does not need that level of control? Good luck.
@@user-rs8zg8ey2bI saw a few of your videos, especially the motor ones. Nice work on those. But when I said induction motor, I didn't mean bldc, I really meant "rotating magnetic field induction motor", so the rotor doesn't have to be a magnet. Without a magnetic rotor, hall-effect sensor based driver would be of little use. In your videos, I understand that the rotors are ball magnets, so you are able to use the hall effect sensor for timing the pulses.
Also, since the rotor is very tiny, in this project, I thought a well controlled supply would be helpful in finding out the sweet spot in the power input. E.g., in your high speed motors, the rotor disintegrated very quickly, and the aftermath was ofc fun to watch :-), but that may not be the goal here...
@rounakdutta6211 ok, I see what you mean now.
It's easy to change supply voltage to limit rpm.
Unless you're a programmer (I am not), adding the dac control to vary the amplitude is not worth the extra complexity imo.
Using a mpu to ramp up and hold a steady rpm is a good idea, with some trial and error you could find a nice sweet ramp up speed and running speed.
A while ago I wrote different ways to drive brushless motors, but yours don't have magnets. If your interested I will find the link.
I just had a KISS idea to drive your motor. If your motor has 4 poles, use a 4 pole bldc motor as a generator to drive your femtomotor.
Well done!
On another note, why not use a tiny magnet, one coil (no core) and spin it, it could possibly be smaller than your current motor as it has less coils.
the hand shots really set sense of scale, 1 coil down, here it is, a spec of dust held by a single fiber of invisible copper thread, amazing work, great camera work too
@5:35 The background finger adds a lot of perspective into the scale of this ridiculous task.
This is one of the coolest videos I've seen in forever! I love engineering and building, but the patience for something like this... WOW.
Excellent
IMO, you'd be better off having the four "cores" 1:20 made from a non-magnetic material - say a piece of ceramic or glass (or even a piece of non-ferromagnetic metal, if your speeds are going to be low). The "cogging" from the bits of steel shall always imbalance your rotor, and you'd need damn good bearings to counteract this; whereas an "ironless" stator shall be far easier to implement, albeit less efficient.
You have such a pleasant voice. Combined with these awesome videos, this is incredibly relaxing to watch. Got yourself a sub. :)
Wow really amazing!
This video was magical. Love the sticking to it, the tools, and the narration. If you learned something, the project was successful, compound that with us watchers learning and it was Very successful!
Thats mind blowing
Respect to you for having the courage, skill, and perseverance to press on and reach your goal. Thank you also for showing the process, Marin clear that failure is part of learning and success, and finally for acknowledging the inspirations that provided the foundation that you stood on to be successful… and the partnership of your wife of course!
11:11 wear a mask, then you can safely breathe around these kinds of sizes. I use one for soldering, and while the solder smoke pops out, my breathing doesn't disturb it, it goes in a single, clean line towards the extractor (almost like in those laboratory pictures with wind tunnels for the study of hurricanes).
couldnt he just turn his head for a moment and inhale
I interpreted his statement to be less about blowing the part away and more about the additional vibration/movement that breathing causes in your body. I find when micro-soldering that holding my breath briefly steadies my hands significantly if I need the extra stability. A mask does’t hurt to prevent air currents though.
Totally Astounding! I love it!
Keep on Chronova bro', your project is so cool and tiny! You are learning - and teaching - so much, it doesn't matter if the motor is not perfect!
Its crazy, I do a lot of fine detailed work also that requires me to old my breath often and I am finding myself doing the same just watching this lol. I broke my brain I think.
I never get tired of watching you work.
This seems like it should be impossible.
Incredible patience and skill, well done!
0:19 seconds in and pressed the subscribe button!
This is so bonkers (in the best way) I'm in awe at your dexterity and patience. I'd also like to echo the use of UV cure glue over superglue. With a bit of practice, it's possible to shape it to your liking by judicious exposure to UV to adjust its viscosity. ie reduce capillary action while applying. It might also help to fix the windings while still mounted on the mandrel followed by a final fixation of the stator assembly.
could have started with a motor 5x times the final size and learn/refine/improve the processes
Not sure you understand the scale here lol.
5x is nearly an equal challenge for 99.999% of any machinist. Which would mean it would take nearly as long for the first iteration. This IS his first iteration, it was the test to learn from.
Also scaling something this sensitive to magnetism, air currents, debris, etc etc, most likely wouldn't transfer 5x smaller. Most of the "learning" was in the method of machining and assembling, not the proof of concept.
Браво ! Много е вдъхновяващо ! Бях останал с впечатление, че прецизността е останало в миналото, във времето на прецизната механика изработвана на ръка. Това, което Вие правите е УНИКАЛНО ! Много се радвам, че човечеството е все още достойно любопитно и можещо. Пожелавам Ви успех !
What's this? A motor for Ants?
Really appreciate the effort put into making the small motor and this video! Now I truly understand the importance of engineering and the internet.
Why not use a magnet instead of a bunch of water
This is such a cool little project. It's really awesome, and you did great. Thanks for the upload.
WHAT IS THIS?!? A MOTOR FOR ANTS??!??!
So it’s basically a “MoToR fOr AnTs” 😂
Astonishing precision, great video!
Thank you for taking us on this journey with you!!!
Just watching this video made me hold my breath because it felt like my breath was going to spoil it! It's so awesome!
This was an absolute wet dream to watch
Doing all this by hand...
I have all the technology & tooling i need at my fingertips everyday, 5 micron bores and 10 micron true positions are standard for me
You still had my full attention from start to finish here 👏
Now that is dedication for you. Awesome work Mike.😊
I love all the extra tools and parts you make to make your job easier
I absolutely loved watching this, your work here is amazing :)
I had a few thoughts that might help both in terms of building a rotor in AlNiCo or even potentially in neodymium magnet material directly. There are now open source EDM controllers which would enable you to build a miniaturised EDM sinker if you wanted to, or EDM wire cutter, although, obviously you'd then need a wire feeder etc. The benefit here would be -nearly- zero force machining, which could solve your parting problem.
Depending on how committed you get to the design, this would also allow you to make custom carbide tools at this scale as well, which might help with your operations on materials like the AlNiCo etc. where you conduct those on a lathe.
As an additional suggestion I've build broaching and form tools in the past for (admittedly much larger) jobs which required substantial variation in job diameter for a lathe. The benefit here is that a broach uses symetric cutting, which results in a far lower risk of part damage during turning. Forming things like the rotor and shaft could be made in a single operation and manufacturing the broach would actually be easier than manufacturing the positive you want to produce from it, because you're drilling holes and milling steps, not cutting long unsupported shafts.
I hope this doesn't sound like I'm being negative about how I view your work, I was trained up as a toolmaker as a child and worked on relatively small die and tool jobs before I did my engineering degree, I've been an engineer for about 20 years now in a pretty niche high product value field and I can tell you that the work you're doing is unique in my experience. I'm frankly amazed at what you're achieving and I hope that my suggestions might contribute positively in some way.
Wow, that's incredible & not to mention, patience & determination!
By Jove! Kudos to you Sir! Again a breathtaking video. Your videos are the highlight of my day, week, month(s) and are very much appreciated. What a skill… 🙏🏻
That pantograph was really clever! Fantastic work!
A simple circular recess ground into the top side of the coverglass to hold the rotor in place should work.
Awesome work dude RESPECT.
thankyou for sharing
Having all these shots well centered is great because I can watch the video @ 2.5× magnification and not just take your word for it 😂 excellent work!
With a Davinci articulated robot arm this would have been childs play,but that would have taken all the fun out of it. Good job man! Unbelievable patients!
Unbelievable how much patience you have! I get frustrated just waiting for my first coffee in the morning. 😅
Super impressive! It looks like what you need is a pantograph-tweezer setup that lets you transform larger hand movements with mock-tweezers into 1/10th scale movements with tiny tweezers for doing all of the tiny stuff by hand. That would be a fun one to engineer, CAD, and fabricate to have minimal slop and backlash in the system :]
What a joy ! Totally mad, utterly amazing and wonderful ! 👏
This is so incredible!! The first shaft model was correct, but it needed to be magnetized by a string electromagnetic current like normal magnets, if you use the spinning magnets technique the spins got briefly magnetized and it won’t maintain its orientation for a long time