No Cables or Batteries, Just Pure Mechanical Engineering - This Swiss Cooling Fan Cools You For Free

Added: 2026-05-12

[00:00:00]You know how nowadays when you open the YouTube  app on your phone, YouTube just instantly gives you a short form video even though you didn't  choose to watch that particular short form video or any short form video for that matter. To be  honest, I I find this feature kind of annoying because it makes me feel like a a human version  of a foie gras duck where my brain is just being forcefully fed useless, contextless content that  I didn't ask for. I think that there should be like maybe we should try like two apps. One  for long form content and the other one for short form content. All that being said, yesterday  YouTube has forcefully shown me a short video that I really liked and I found really interesting.  This video. And as you can see here, it says, let me read it. "Before electricity reached the  tropics, the Swiss solved cooling with clockwork.

[00:00:57]One full crank of this spring motor fan gave you  30 minutes of continuous breeze. No power source needed". Of course, like most short videos, this  video is just a reformatting of somebody else's long video. And I did find the original video  eventually. And as you can see, what we're looking at has a name. It's called a Paillard clockwork fan. It's a Swiss hand cranked fan from the early 1900s. And again in the original  video, the statement that it does 30 minutes of cooling on just one full winding is repeated.  Unfortunately, that statement is never proven. So, we are forced to trust the uploader because he  never shows the fan actually cooling you for 30 minutes with like a stopwatch next to the fan  or something. I don't know. I personally find it highly doubtful that it can do 30 minutes based on  the size of the fan. And I did find another video of a similar device from a similar time that did  like 14 minutes. But even 14 minutes is honestly kind of impressive on just one full winding. And  I thought, hey, these are ancient fans from like the early 1900s. We got we must be able to do  something much better today. So I decided to look around to research to find the modern equivalent  of these manually wound fans. And I found nothing.

[00:02:18]There is no modern equivalent of this. Nobody  makes this anymore. All we have are fans that are plugged into the electricity grid or powered  by batteries. The only manual fans we have are either you have to constantly wind them or  you have to constantly squeeze them. So you have to constantly operate them which is honestly  kind of annoying and useless and your hands get tired after just a few minutes. So, there's  nothing that's the modern equivalent of this, which I find pretty surprising because there's a  market for this. We're all about energy nowadays.

[00:02:56]And what's better than free energy? You always  have your hands with you. You just wind it and you have free cooling. A lot of people want to live  off-grid nowadays and or at least pretend to live off-grid. And electrical motors are pretty power  hungry. So, this would really be useful. There's a lot of people who live and or work in remote  hot areas where there's very little electrical grid coverage. This would also be really nice  during like a power outage. In general, it's something useful to have. So, I decided  to try and build one myself. And if successful, I will share the files for free with you.  Obviously, the first step towards building it is understanding how it works. Fortunately, the  video gives us a lot of very useful codes. So, the first part here is our gear drain, and all  this does is that it increases our output speed.

[00:03:54]The fan is our output, and we want it to spin  fast enough so it can move a substantial amount of air quickly in order to cool us down. If  we observe the mechanism while it's running, we can see that our input is this gear right here.  As you can see, it rotates very slowly. And as we move up in the gear train, the gears rotate  faster and faster. So, we have high torque and a low speed here, which we then convert to high  speed and a low torque here at the fan using our gear train. The next question is what is inside  here in the box underneath the gear train? Well, this is called a clockwork fan. We can see  winding and hear ratcheting. And the short video mentions a spring motor. All that  tells us that inside here there must be a mainspring. And I know just a place where I  can get one easily and for very little money.

[00:05:01]Obviously, this is a tape measure  and inside it there's a main spring because when I pull out the tape, something  magically pulls it back every time. Now, in order to get to the main spring, we  obviously have to disassemble the tape measure.

[00:05:31]This part, the measuring  part is I don't need that.

[00:05:38]Whoa.

[00:05:42]Okay. So, here we can see a  little locking mechanism. Now, this is pretty thin, pretty sharp,  so be kind of careful with it.

[00:05:53]There we go. Finally. So, it unlocks like this.  We open it up and inside is our main spring.

[00:06:13]Yeah, I wanted to show you how it  works, but now I kind of ruined it.

[00:06:25]It took forever, but I finally managed to wind  the spring back. Now, this is the state in which it's in the factory before they install  it into this part, which is the barrel.

[00:06:39]This is the natural state of the spring, this  coiled state. And unfortunately, in this state, it's pretty useless because, as you can see,  I can't wind it anymore. If I try to wind it, it just becomes a tighter coil. I can't use it  to store energy anymore. If I try to unwind it, I can store some energy. But the more I  unwind it, the more space it occupies, and that makes it useless, too. Previously,  when it was inside this barrel, it occupied always the same amount of space, and it was  stuck inside the barrel. The spring never gets pulled out or in of the barrel. That's what  this indentation here does. It gets trapped by this indentation. Let me show you.

[00:07:36]As you can see, it prevents it from being pulled  in or out. But still you can repeatedly store and release energy while the spring occupies always  the same space. How is that possible? Well, it's possible because the spring wasn't wound  into its natural state. It was wound like this in the opposite direction like this. So, it was  wound into its unnatural state, if you will, and the confines of this barrel prevented it from  ever leaving the barrel while at the same time it always wants to I'm really afraid to release this  now. It always wants to revert back to its natural state. So, when you word it like this, it sounds  like we're kind of cruel towards springs. But no, springs main springs don't have feelings. We  don't need a movement or a hashtag for them.

[00:08:47]Thank you very much. And while main springs don't  have feelings, they definitely do have history.

[00:08:55]You see, main springs can be considered the first  battery. Before we learned how to chemically store electrical energy, we learned how to  mechanically store kinetic energy. Actually, no, main springs are not the first batteries.  The first battery is literally a rock that you take and raise it up. That's what we did first.  Because when you raise a rock, you do some work.

[00:09:21]you create an input and that input results in  potential energy and potential energy can be released. So what we did in the past is that we  raised a rock, we tied a rope around the rock and then we tied a rope to a mechanism and as the rock  was pulled back down towards the earth by gravity, we use that potential energy to operate our  clocks. Of course, this was very inconvenient because it occupied a lot of space and you would  have to raise the rock up again every now and then. Yeah. But then fortunately somewhere  sometimes around the 15th century in Europe, somebody invented the main spring. The main  spring allowed us to store more energy much more conveniently and in a much smaller space.  Mainsprings revolutionized humanity because they allowed everyone to take timekeeping devices  with them on the go. Mainsprings made pocket watches and wrist watches possible. But even main  springs were kind of inconvenient all the way up to about the 1960s because it took us centuries  to figure out the right steel alloys and the right heat treating processes to give main springs the  yield strength and the fatigue life that made it possible for them to be wound and unwound pretty  much an infinite number of times. something that we take for granted today. But up to about the  1960s, mainsprings were actually the main cause, the main reason for watch repair because they  simply didn't last nearly as long as they do today. Okay, so tape measure number one is a  failure. And honestly, that's a good thing because it allowed me to get tape measure number two.  And this one is 10 m whereas the previous one was where are the pieces? Whereas the previous one was  only 3 m. So this one can store a lot more energy.

[00:11:44]It's a bigger battery and I'm going to need that  if I want my fan to run as long as possible.

[00:11:55]Okay, here comes the tricky part.  This is already preloaded. Watch this.

[00:12:06]Okay, now we have to remove  the tape again. Be free.

[00:12:20]So, here we have our barrel again. This time  with a much wider and much longer main spring, which means that we have a lot more potential  energy stored here. And I've also learned my lesson. I'm not going to try to take this main  spring out of this barrel because it's probably going to just unravel into a ball of mess.  Instead, I'm going to make have ready another different main barrel. And then I'm going  to try to somehow safely transfer the main spring directly from this barrel to that barrel  without winding or unwinding or any of that.

[00:13:08]Okay. And here is the new housing for the  main spring. And now I'm going to transfer the main spring from this one to the new  one. And because this is kind of big, very sharp. I'm not sure what it can do.  So, it's better to be safe than sorry, I'm going to get some gloves  and some face protection, too.

[00:13:50]Okay, this is a problem. As you can see, a few  just a few coils are out. The rest is in, but that is enough for it to start escaping. And once it  starts escaping, it's over. I will not catch it.

[00:14:06]I'm so happy I have gloves right now.

[00:14:13]Okay. Okay. I have almost  everything in except this part.

[00:14:23]There's no other way. That's what  I'm doing. I'm going to cut a bit Okay, it is in. It is finally fully  in. I have captured a main spring.

[00:14:58]As you can see, it is totally secure.

[00:15:01]out with this tab here. It cannot go out or  in. But just to be extra sure. Nah. Yeah, I took off the gloves. I took off the gloves. What  an idiot. load-bearingsuper extra sure, I'm going to do this, too. And so I So this part doesn't  stay like this. I want it flat and non dangerous.

[00:15:40]Done. And done. Our power source is  ready. Now let's design the rest of it.

[00:15:55]Okay. So here we have almost all the parts already  out of the printer. The only thing that's missing is the fan itself that's being printed right now  as we speak. I have already uh pressed most of the bearings. As you can see here, there's just one  bearing missing left to be pressed pressed in.

[00:16:23]And done. Now here we have the gears. As you  can see, we have two gears fused into one.

[00:16:32]You have probably seen this anatomy  on toys, clocks, you name it, many, many things. It is useful because it allows us to  achieve a relatively high gear ratio while keeping the entire system pretty compact. Now, the whole  gear train is supposed to go together like this.

[00:17:00]Okay. So, we store the energy in the  spring and as the spring releases, it rotates and rotates the entire gear train.  We have a 1-6 gear ratio. Meaning that just one rotation of the spring barrel is going to achieve  16 rotations of this top shaft here where the fan will be attached. But there's more to it than  that because I also tried to make the gear train efficient. And that's because every one of these  fused gears is an opportunity. It's an opportunity because these two gears are just joined together.  They don't mesh. And that means that we can change the gear specs from one to the other gear on the  gear pair. So what I did is that I decreased the gear module going from bottom to top. The gear  module is simply the tooth size of metric gears.

[00:17:58]Larger teeth are obviously stronger and  as such have higher loadbearing capacity, but are also noisier and less smooth, which makes  them better suited to low RPM and high torque.

[00:18:12]On the other hand, smaller gear teeth  obviously have a lower loadbearing capacity, but are smoother, quieter, and more efficient  at high RPM, which aligns perfectly with our application because we are increasing speed or  RPM as we go up in the gear train while at the same time decreasing torque. I have also changed  the pressure angle for the last gear pair at the very top which will operate at highest RPM.  A 14.5° pressure angle is smoother, quieter, and overall better suited and more efficient at  high RPM. It has poor shock resistance and is easier to strip with high torque. But we don't  really have any shocks or high torque by the time the torque is transferred from the bottom to  the top of the gear trim. But all of my gear spec cleverness is going to be an absolutely useless  drop in the ocean compared to the losses created by the imperfections of 3D printing. You see when  you print a gear like this, so two gears fused together and then together with the shaft always  one unit, the top surface is always going to be pretty and smooth and almost perfect. Whereas the  bottom surface is going to be much rougher and uglier and that's because it interfaces with  the supports necessary. These supports are a consequence of the shaft. So the correct approach  which I haven't taken due to laziness would have been to print the gear and the shaft separately  and then just print a keyway uh in the gear create a keyway which I haven't done which means that now  I'm going to have to sand all these unders sides of the gears because I can already hear them and  you will now be able to hear them too. the high spots of all the rough parts are touching each  other, grinding against each other. And of course, that's a massive efficiency loss. And we  don't want that because this whole device is an exercise in preserving energy. The energy  created by the spring, we want to preserve as much of it as possible to transfer as much of as  much of it as possible to the fan. Everything we lose on the way means less rotation of the fan  and we don't want that. So now it's time to sand I have good news and I have bad news. The bad  news is that this doesn't work. It's a complete failure. The good news is that I know exactly  why it's a complete failure and I can fix it.

[00:21:42]The problem is that I took the completely  wrong approach when it comes to making this.

[00:21:49]And that's because I I tried to make this whole  contraption as rigid as possible. As you can see, I made even these brackets to secure it. It has  bolt holes right here to bolt it down to some sort of like a wooden plank or something. And that's  because I thought that every kind of movement, every kind of wiggle, every kind of  imperfection in the rigidity of the device is a loss. The number one problem  that this approach caused is that trying to print giant high infill pieces like these  brackets. Let me take everything apart.

[00:22:38]It leads to imperfections trying to print  something as big as this with 35% infill and four walls. On top of this, the whole  thing is was the whole all of this was on supports and such a heavy part as it prints  it warps. So, it warped and when it warps, the alignment of the gears with each other becomes  incorrect. And so, instead of being perfectly spaced because their individual holes for the  bearings determine the spacing of the gears, but when you warp the whole parts, these holes  move. And now the gears are pressing hard against each other. They don't have the right space. they  can't breathe and so they're now it's impossible to turn them. The only method I could use to  successfully get this to rotate was to actually separate the housings to get rid of these  brackets and to give it some breathing room.

[00:23:57]The other problem I have is that this shaft  on my main barrel for the main spring broke.

[00:24:18]The shaft broke, but I fixed it by printing  another hollow shaft and bolting it uh into this.

[00:24:26]But obviously now this isn't perfectly in the  center because uh the hole that I drilled in it's manually I got it as close as I could. But now  this has a bit of up and down motion. But I really don't want to print a different barrel because  then I would have to transfer this to another barrel and I do not want to do that again. But I  think this is good enough and I can use it if I give the whole system some more breathing room. I  was trying to imitate clockwork. I was trying to imitate metal metal precision with plastic and and  that's because I was too overconfident thinking that hey I have access to technologies from from  the present and so I can beat manufacturing from the 1900s that's not true uh metal is still  metal precision clockwork is still precision clockwork even if 100 years old I cannot just  you know take it politely and you know look at it from some high chair or whatever. Now I have  to get back to basics. This has to be designed completely differently with the limitations  and the possibilities of 3D printing in mind.

[00:25:55]Okay. So here is version 2.0. As you can see,  it's dramatically more simple, more sleek, more elegant, a lot less filament. Now, it  even has a bit of give in it to accommodate all the imperfections. The gears are  no longer together with the shaft. Now, they are using this Lego knockoff axle system.  So, and they also can play a bit on it. So, now we are actually loading the fibers. We are not  trying to rip the shaft off. We are not loading them where they adhere to each other. We're  not attacking them where they're weak. Instead, we are trying to do this to the fibers. And  that's impossible because this is where they're strong. We're going we're going in favor  of their strength. However, once again, I have good news and bad news. The bad news is  that this broken shaft on the original barrel uh is simply too inaccurate. No matter how much  I try to accurately drill it in the center and screw it into the center, human eyes and  hands are not precision instruments and there is always a bit of out of round, a bit of  play. And so when I try to install it into the new housings and I turn it around, I don't  know if you're going to see it immediately.

[00:27:29]As you can see, observe there's  simply far too much play.

[00:27:38]The whole thing is wobbling around more than I  can uh afford. The good news is that I'm having fun now. I'm solving a completely different  problem. And the problem that I want to solve now is how do we save this? How do I save this  barrel? How do I attach a shaft to it once again?

[00:28:04]How do I give it a shaft while keeping accuracy?  How do I put it in the dead center while still keeping this part that I need so much? And I  think I have an idea. So what I have printed is this gear which is exactly the same as this gear.  And then I have printed this alignment sleeve.

[00:28:33]So, the alignment sleeve goes onto this gear. As  you can see, it's a pretty tight fit. In this gear here, we have this hole for the axle. We're  going to put this in. Again, there's going to be a bit of play. And again, we are loading the  fibers in this direction. This is printed. The layers are like this. And we are holding  like this which means that this is virtually unbreakable under normal conditions. And now  we put it all together in the alignment sleeve.

[00:29:14]And this sleeves guarantees that  these two gears are now perfectly aligned. And as you can see I  have these four holes prepared.

[00:29:38]And voila, we have restored  a shaft with a bit of give, too. And as you can see,  the gears align perfectly.

[00:30:07]Okay, here it is in all of its absolute ugliness,  but I'm finally ready to try it. And I've already tested it a bit. It It kind of works. So  obviously the most important thing here is not just whether it works and whether it  you know creates a breeze but the ratio. The most important thing is the ratio between  how much time you spend winding and how much breeze you get out of your fan. So here I  have my stopwatch ready. I'm going to start it.

[00:30:41]I'm going to wind it and I'm going to release  it. And then we're going to see what happens.

[00:31:04]Okay, that's enough. No overkill  this time. That took 39 seconds.

[00:31:09]I can wind it a bit faster. Now, when  it gets to 45, I'm going to let it go.

[00:31:41]2 minutes and 30 seconds. So for  40 40 seconds of winding we got 15 1 minute. So for 40 seconds of winding  we got 1 minute and 45 seconds of breeze.

[00:32:03]Extremely extremely light breeze. I have  this here. Let's see what do we get.

[00:32:26]We're getting like 03 m per second, which is  this was like two days worth of research and development. Uh, and I think it has potential  like with more effort, with more precision, with just better everything. I think I've also figured  out why nobody's making this because it would be expensive. It will be really expensive and you  would you would need to use it for a lifetime, maybe even more to make it pay for itself in  the tiny little energy savings that it would create. It would be expensive because this is  genuinely precision machinery. The only way I could get it to run is to keep it all loosey  goosey and let let it just, you know, work itself into whatever shape it wants because every  little imperfection, every little inaccuracy, doesn't matter how small, it all stacks up and  everything has to be dead accurate. It really is, it's a good name for it. It It really is a  clockwork fan. This also makes you realize just how incredibly impressive the craftsmanship of the  early 1900s was. Think about it. No CAD software, no CNC machines, no nothing none of that. And they  still got it running 15 minutes, maybe even 30 minutes. That requires, I'm sure of it, extreme  accuracy, extreme precision of manufacturing, fully understanding the system, making no  mistakes anywhere. Achieving that back then, wow, maybe we haven't come such a long way.  Uh, I'm pretty sure those things were really expensive back in the day. Probably maybe even  the cutting edge of technology. probably only the very rich colonialists or whoever could  afford it back then. Who knows? But still the people who made them. Yeah. Impressive. And they  would still be expensive today. They would still really be expensive today. That's why that's why  they're not everywhere online. Yeah. Thanks a lot for watching and I'll be seeing you soon with  more fun and useful stuff on the D4A channel.