WEBVTT Kind: captions Language: en 00:00:00.080 --> 00:00:02.557 You may have noticed that every time I say 00:00:02.557 --> 00:00:04.533 “it’s time for a simpler video” 00:00:04.533 --> 00:00:05.538 it’s a lie. 00:00:06.122 --> 00:00:09.013 For the record, it doesn’t start as a lie 00:00:09.013 --> 00:00:10.000 it’s just… 00:00:10.578 --> 00:00:14.088 well, I start writing the script at the beginning. 00:00:14.088 --> 00:00:16.022 Seems to make sense to do it that way. 00:00:16.022 --> 00:00:17.614 And so, I’ll say 00:00:17.614 --> 00:00:19.133 “ah, this is a simple idea! 00:00:19.133 --> 00:00:20.244 It’ll be simple! 00:00:20.244 --> 00:00:21.249 Simple!” 00:00:21.375 --> 00:00:24.654 and then it just snowballs into a five part series. 00:00:25.040 --> 00:00:26.093 Not today! 00:00:26.093 --> 00:00:28.477 We are just gonna talk about this thing. 00:00:29.120 --> 00:00:32.560 Back when I made the humidifier  video I showed one of these.   00:00:32.560 --> 00:00:34.616 It’s a basic hygrometer. 00:00:34.616 --> 00:00:36.389 It has a dial and a pointy bit 00:00:36.389 --> 00:00:39.064 and what the pointing bit points to is, roughly, 00:00:39.064 --> 00:00:41.568 what the ambient relative humidity is. 00:00:41.568 --> 00:00:42.609 Simple. 00:00:42.609 --> 00:00:44.685 Some of you asked how this works. 00:00:45.280 --> 00:00:48.266 And honestly, yeah that's pretty intriguing, isn’t it? 00:00:48.266 --> 00:00:50.675 It doesn’t take any batteries, and if you look through the back 00:00:50.675 --> 00:00:54.307 you can kind of get a peek at a coil-like thing. 00:00:54.307 --> 00:00:57.035 That looks an awful lot like a bimetallic strip! 00:00:57.280 --> 00:00:59.665 What’s a bimetallic strip, you ask? 00:00:59.665 --> 00:01:02.104 It’s a strip made of two metals. 00:01:02.104 --> 00:01:06.193 That’s useful because different materials  have different thermal expansion properties. 00:01:06.193 --> 00:01:09.127 Virtually everything expands a bit when it gets  warmer, 00:01:09.127 --> 00:01:11.074 and metals are no exception. 00:01:11.074 --> 00:01:15.164 But how much they expand depends on the specific material. 00:01:15.164 --> 00:01:17.744 When you take two metals that are not the same 00:01:17.744 --> 00:01:20.539 and sandwich them together in the form of a strip, 00:01:20.539 --> 00:01:23.576 the mismatch in thermal expansion coefficients between them 00:01:23.576 --> 00:01:27.884 will cause the strip to bend slightly as the temperature changes. 00:01:28.240 --> 00:01:30.739 Bimetallic strips are in all sorts of  places, 00:01:30.739 --> 00:01:32.777 from the world’s greatest toaster, 00:01:32.777 --> 00:01:34.333 to blinking Christmas lights, 00:01:34.333 --> 00:01:36.395 to the thermostat of space heater, 00:01:36.395 --> 00:01:37.462 to, well, 00:01:37.462 --> 00:01:39.203 thermostats in general. 00:01:39.203 --> 00:01:42.435 But a classic and simple application of the bimetallic strip 00:01:42.435 --> 00:01:45.215 is in a thermometer like these. 00:01:45.215 --> 00:01:47.322 And let’s just get it out of our systems, 00:01:47.322 --> 00:01:50.295 yes these numbers are meaningless to many of you out there 00:01:50.295 --> 00:01:54.438 and maybe one day we’ll stop using this temperature scale ‘round these parts 00:01:54.438 --> 00:01:57.888 but I can’t think in Celsius any more than you can think in Fahrenheit 00:01:57.888 --> 00:02:02.850 and besides, a zero to one hundred scale that  correlates well with the range of tolerable ambient temperatures 00:02:02.850 --> 00:02:05.370 is a more pleasing way to do it from my perspective and - 00:02:05.370 --> 00:02:06.973 hold on this is a tangent, now. 00:02:07.455 --> 00:02:08.850 I want to show you something. 00:02:08.850 --> 00:02:09.764 See this? 00:02:09.764 --> 00:02:11.628 This is called a search engine. 00:02:11.628 --> 00:02:13.572 There’s all sorts of them out there these days. 00:02:13.572 --> 00:02:15.405 And most of them let you do this: 00:02:15.760 --> 00:02:17.598 70 f to c. 00:02:17.598 --> 00:02:18.396 Blam. 00:02:18.396 --> 00:02:19.267 There it is. 00:02:19.267 --> 00:02:20.746 You have a smartphone, right? 00:02:20.746 --> 00:02:24.094 It probably has a search bar in a very convenient place. 00:02:24.094 --> 00:02:26.223 Oh and maybe you even have a voice assistant 00:02:26.223 --> 00:02:31.034 you can literally just ask what whatever temperature is in whatever other scale you like better. 00:02:31.034 --> 00:02:33.003 It really is that easy. 00:02:33.600 --> 00:02:34.945 Now, when it's appropriate, 00:02:34.945 --> 00:02:37.895 I’ll continue to provide conversions here on this YouTube channel 00:02:37.895 --> 00:02:39.944 because it would be careless of me not to, 00:02:40.118 --> 00:02:43.893 but when I dare to offhandedly tweet something 00:02:43.893 --> 00:02:46.154 and use one of our silly units 00:02:46.154 --> 00:02:47.385 I don’t wanna hear it. 00:02:47.385 --> 00:02:51.006 You’ll spend way more time composing your complaint to me 00:02:51.006 --> 00:02:54.108 than you will just googling the conversion yourself. 00:02:54.108 --> 00:02:58.424 We are all very aware of how silly the aversion  to the metric system is 00:02:58.424 --> 00:02:59.820 but we can’t help it. 00:03:00.240 --> 00:03:01.691 Oh, and yes I am on Twitter. 00:03:01.691 --> 00:03:03.454 It’s mostly puns and other dad jokes. 00:03:03.454 --> 00:03:05.294 And snark like this! 00:03:05.920 --> 00:03:07.715 OK, let’s get back on track. 00:03:07.715 --> 00:03:10.560 In a thermometer like this, the bimetallic strip 00:03:10.560 --> 00:03:14.498 is formed into a coil with the outside end secured in place 00:03:14.498 --> 00:03:17.771 and the inner part of the coil attached to a pointer. 00:03:17.771 --> 00:03:21.945 Although the amount the strip bends  when the temperature changes is slight, 00:03:21.945 --> 00:03:25.453 forming a coil compounds that effect with each rotation. 00:03:25.453 --> 00:03:28.427 You end up with a coil that contracts as it  gets colder, 00:03:28.427 --> 00:03:30.238 and expands as it gets warmer. 00:03:30.238 --> 00:03:32.021 Or potentially vise versa. 00:03:32.021 --> 00:03:36.720 And so, the center of that coil rotates somewhat as the temperature changes. 00:03:37.360 --> 00:03:39.804 Attach a pointer to the center of that coil, 00:03:39.804 --> 00:03:42.028 and with a little experimentation and calibration 00:03:42.028 --> 00:03:46.184 you’ll have created a simple device  which indicates the current temperature. 00:03:46.184 --> 00:03:48.341 The effect is surprisingly linear. 00:03:48.341 --> 00:03:51.835 The spacing between these numbers is consistent through most of the travel, 00:03:51.835 --> 00:03:54.428 although you can see it vary somewhat at the ends. 00:03:54.800 --> 00:03:58.081 So, if we take this thing apart through The Magic of Buying Two of Them 00:03:58.081 --> 00:04:01.880 we see what appears to be two bimetallic strips. 00:04:01.880 --> 00:04:06.333 One for the tiny, mostly useless  thermometer they give as a token gesture, 00:04:06.333 --> 00:04:08.357 and the other for the hygrometer. 00:04:08.357 --> 00:04:12.305 But this one is actually a bimetallic strip tease. 00:04:12.305 --> 00:04:14.575 Sure, it’s essentially the exact same idea, 00:04:14.575 --> 00:04:17.840 but how can two dissimilar metals give you a humidity reading? 00:04:18.640 --> 00:04:19.905 Well, they can’t. 00:04:19.905 --> 00:04:22.695 In fact not bimetallic, 00:04:22.695 --> 00:04:24.578 but it is bimaterial. 00:04:24.578 --> 00:04:27.630 Take a closer look at it and you’ll see that one side of the coil 00:04:27.630 --> 00:04:30.000 appears to be covered in a sort of paper. 00:04:30.400 --> 00:04:32.456 And that’s because it is. 00:04:32.456 --> 00:04:34.152 This style of hygrometer, 00:04:34.152 --> 00:04:37.381 fittingly named the Metal-Paper Coil hygrometer, 00:04:37.381 --> 00:04:41.416 functions identically to that of  a bimetallic thermometer. 00:04:41.416 --> 00:04:46.193 The key difference is that the salt-impregnated paper changes based on humidity. 00:04:46.193 --> 00:04:50.882 This is one of those things that seems puzzling at first but then absolutely obvious once you know it. 00:04:50.882 --> 00:04:54.111 Like, you know what happens to paper on really humid days. 00:04:54.111 --> 00:04:56.049 It gets all floppy and gross. 00:04:56.049 --> 00:04:57.646 Same sort of thing goes on here. 00:04:58.080 --> 00:05:02.560 The paper side of the coil will  deform slightly as humidity increases, 00:05:02.560 --> 00:05:05.341 changing the overall tension on the coil. 00:05:05.341 --> 00:05:07.527 That makes the needle move. 00:05:07.527 --> 00:05:10.979 Watch what happens when I put a bit of water directly on the paper. 00:05:10.979 --> 00:05:15.299 The coil quickly deforms and the needle shoots to “very damp.” 00:05:15.299 --> 00:05:19.482 And here you’ll notice that the scale is very much not linear. 00:05:19.482 --> 00:05:21.359 Low humidities are far apart, 00:05:21.359 --> 00:05:26.080 indicating that the effect of humidity on the paper diminishes as humidity increases. 00:05:26.640 --> 00:05:30.116 And this clock is why I decided to make this  little video. 00:05:30.116 --> 00:05:32.768 I bought it when I bought the other clocks for the last video 00:05:32.768 --> 00:05:34.538 because it would actually be useful to me 00:05:34.538 --> 00:05:36.684 and I had a perfect spot for it outside. 00:05:36.684 --> 00:05:40.448 But as soon as I realized the humidity scale was linear, 00:05:40.448 --> 00:05:42.998 I got suspicious of this thing. 00:05:42.998 --> 00:05:43.903 Sure enough, 00:05:43.903 --> 00:05:45.327 it’s crap. 00:05:45.327 --> 00:05:49.052 It agrees with this one around 50 percent or so, 00:05:49.052 --> 00:05:53.390 but this won’t ever read higher than maybe 80 percent? 00:05:53.390 --> 00:05:56.715 This one read 90 and this only read 75. 00:05:57.505 --> 00:05:58.236 No good. 00:05:58.720 --> 00:06:03.173 In fairness, metal-paper coil hygrometers aren’t super accurate to begin with. 00:06:03.173 --> 00:06:06.820 This website I found says they’re only good to about 10%. 00:06:06.820 --> 00:06:10.936 And it’s not like I could expect the hygrometer in a $10 Walmart clock 00:06:10.936 --> 00:06:12.726 to be any decent to begin with. 00:06:12.726 --> 00:06:16.397 But if you ever see a linear scale on a hygrometer like this, 00:06:16.880 --> 00:06:19.472 it’s definitely not one of the better ones. 00:06:19.930 --> 00:06:22.295 OK, well probably not - 00:06:22.295 --> 00:06:25.506 I haven’t yet tested this one at the time of filming. 00:06:25.506 --> 00:06:28.077 I’ll pin a comment with the final verdict. 00:06:28.077 --> 00:06:31.943 This melding of paper and metal isn’t the  only way to go about it, of course. 00:06:31.943 --> 00:06:35.043 In fact, it seems to be a rather new-fangled way to do it, 00:06:35.043 --> 00:06:38.400 presumably it’s just very easy to manufacture inexpensively. 00:06:39.040 --> 00:06:42.152 All you really need to make a hygrometer like this 00:06:42.152 --> 00:06:47.163 is some sort of spring that’s pulling on something else which will deform based on humidity. 00:06:47.163 --> 00:06:52.268 The first hygrometers that worked on this principle used actual human hairs. 00:06:52.268 --> 00:06:54.718 The hair would lengthen with increased humidity, 00:06:54.718 --> 00:06:57.130 reducing tension on a spring or pulley, 00:06:57.130 --> 00:07:00.240 and thus allowing an indicator of some kind to move. 00:07:00.880 --> 00:07:05.120 As it dried out, the opposite occurred and  the indicator moved in the other direction.   00:07:05.760 --> 00:07:10.394 This style of hygrometer is still made, but  it’s a rather specialist sort of thing. 00:07:10.394 --> 00:07:13.339 However, they can be quite accurate when calibrated, 00:07:13.339 --> 00:07:18.610 with one manufacturer claiming accuracy to within one percent at room temperatures! 00:07:18.610 --> 00:07:22.160 However, I found other sources which aren’t quite so optimistic. 00:07:22.880 --> 00:07:26.320 The worst thing about these cheap  hygrometers is their accuracy.   00:07:26.880 --> 00:07:30.160 While the hair-tension method  might be a little bit more accurate,   00:07:30.160 --> 00:07:33.580 it may still not be as accurate as you can be  with other means. 00:07:33.580 --> 00:07:34.661 Other means? 00:07:34.661 --> 00:07:36.092 What other means? 00:07:36.480 --> 00:07:41.093 Well, the humidistats in the humidifiers I showed  you are electronic. 00:07:41.093 --> 00:07:43.929 The actual component that becomes affected by humidity 00:07:43.929 --> 00:07:47.236 is simpler than the circuitry that interacts with it. 00:07:47.236 --> 00:07:52.201 Typically there will be a polymer or salt of some sort that's exposed to air, 00:07:52.201 --> 00:07:54.693 and the electrical resistance of this material 00:07:54.693 --> 00:07:59.031 changes with the moisture absorbed in it and thus the ambient humidity. 00:07:59.031 --> 00:08:02.137 Measure its resistance, compare that to a lookup table, 00:08:02.137 --> 00:08:05.386 and now you know the relative humidity within a few percentage points. 00:08:05.520 --> 00:08:09.600 These days this rather basic task  can be done at the component level,   00:08:09.600 --> 00:08:14.000 allowing a simple interface between a sensor  and whatever device you’re putting it in. 00:08:14.560 --> 00:08:19.280 There are other techniques involving materials  which exhibit changes in their capacitance,   00:08:19.280 --> 00:08:22.400 but the essence of all of them  is in the material which changes   00:08:22.400 --> 00:08:25.227 in some way depending on ambient humidity. 00:08:25.227 --> 00:08:27.667 It can be something we measure electrically, 00:08:27.667 --> 00:08:32.419 or it can be something that moves a pointer like  a piece of paper or a human hair. 00:08:32.419 --> 00:08:34.365 What we’ve done with these devices, 00:08:34.365 --> 00:08:35.838 including the thermometers, 00:08:35.838 --> 00:08:38.897 is quite simple but also quite profound. 00:08:38.897 --> 00:08:43.214 Figuring out how environmental conditions affect different  materials 00:08:43.214 --> 00:08:49.884 allowed us to work backwards and define those environmental conditions using the material as a reference. 00:08:49.884 --> 00:08:51.840 Those humans, so clever! 00:08:52.560 --> 00:08:56.326 While it may be more technically impressive to  do that using electronicals 00:08:56.326 --> 00:08:58.932 and hook it up to a microcontroller which can beam 00:08:58.932 --> 00:09:01.440 god-awful blue numbers straight into your retinas, 00:09:02.400 --> 00:09:06.563 that’s just not as satisfying as these simple things. 00:09:06.563 --> 00:09:09.633 Turning the physical properties of something into a, 00:09:09.838 --> 00:09:13.040 well, turning motion is never gonna get old to me. 00:09:13.920 --> 00:09:15.803 I told you it’d be simple! 00:09:15.803 --> 00:09:20.825 Now I’ll have fun looking at the YouTube studio and seeing my watch time metric tank. 00:09:20.825 --> 00:09:21.840 It’ll be great. 00:09:23.760 --> 00:09:28.128 Oh yeah, sorry. No music or  credits today. Just… 00:09:28.128 --> 00:09:28.628 this. 00:09:30.534 --> 00:09:33.330 Take a closer look, and you'll see that one of zah 00:09:33.330 --> 00:09:34.300 elbledebuh be deh. 00:09:34.300 --> 00:09:35.041 plerh