WEBVTT Kind: captions Language: en 00:00:00.080 --> 00:00:04.963 Today, I want to show you a weird light bulb of the not-too-distant past. 00:00:05.560 --> 00:00:06.766 Ta-da! 00:00:07.580 --> 00:00:08.815 If you’re thinking, 00:00:08.815 --> 00:00:13.862 "that’s just a CFL stuffed inside a glass bulb to look more like a regular light bulb!" 00:00:13.862 --> 00:00:15.347 you’re right! 00:00:15.347 --> 00:00:19.444 But that simple visual trickery isn’t what’s special about this bulb. 00:00:19.444 --> 00:00:21.762 Lots of manufacturers were doing that. 00:00:21.762 --> 00:00:22.929 This one? 00:00:22.929 --> 00:00:29.929 Well, for a brief while, GE was marketing a line of compact fluorescent lamps with a trick up their sleeves which they called “bright from the start” - 00:00:29.929 --> 00:00:34.009 and this is one of those bulbs. 00:00:34.009 --> 00:00:38.480 They developed it to overcome one of the central issues of compact fluorescent lamp technology: 00:00:39.023 --> 00:00:42.940 the rather slow warmup time of a cold CFL. 00:00:42.940 --> 00:00:47.289 And they did it by taking this already light bulb-in-a-light bulb lookin’ light bulb 00:00:47.289 --> 00:00:49.331 and going full turducken on it 00:00:49.331 --> 00:00:56.034 by stuffing an incandescent light bulb in the middle of the fluorescent light bulb inside of the faux light bulb. 00:00:56.034 --> 00:00:59.317 With two lighting technologies in the same package, 00:00:59.317 --> 00:01:02.899 this is literally a hybrid light bulb. 00:01:02.899 --> 00:01:04.489 Why does it exist? 00:01:04.760 --> 00:01:09.387 Well, about 15 years ago in the way back when known as 2009, 00:01:09.387 --> 00:01:15.479 we still hadn’t really figured out how to make decent LED drop-in light bulbs. 00:01:15.479 --> 00:01:19.663 If anybody out there remembers this groundbreaking LED bulb from Philips 00:01:19.663 --> 00:01:23.520 with its yellow phosphor coating on the exterior? 00:01:23.520 --> 00:01:30.490 Yeah, this thing came out in 2010, and adjusted for inflation, these were $60 a pop. 00:01:30.490 --> 00:01:34.210 We’ve come a long way in a very short time! 00:01:34.210 --> 00:01:38.474 But in the years prior to the LED bulb becoming feasible and cheap, 00:01:38.474 --> 00:01:43.104 we had gotten very good at making compact fluorescent lamps. 00:01:43.104 --> 00:01:46.546 These energy-saving wonders weren’t perfect by any means - 00:01:46.546 --> 00:01:49.938 the quality of light they produced was a downgrade from incandescent, 00:01:49.938 --> 00:01:55.631 unless of course you like deathly cold, certifiably institutional daylight-balanced lighting 00:01:55.631 --> 00:01:59.262 in which case they unlocked that for the first time in many household applications, 00:01:59.262 --> 00:02:04.391 but either way the mercury content of the discharge tube was an environmental trade-off. 00:02:05.124 --> 00:02:09.359 Still, they used about a quarter of the energy of their incandescent counterparts 00:02:09.359 --> 00:02:16.564 and the not-crap ones anyway lasted a very long time if used correctly so they had their appreciators. 00:02:16.564 --> 00:02:18.263 Including yours truly. 00:02:18.263 --> 00:02:21.524 And a few models remain in production to this day. 00:02:22.040 --> 00:02:27.922 But even if you like or merely tolerate the quality of light they produce and aren’t bothered by the mercury, 00:02:27.922 --> 00:02:30.184 they still have an Achilles heel: 00:02:30.184 --> 00:02:34.496 the bulbs themselves are pretty ugly. 00:02:34.496 --> 00:02:40.742 In some light fixtures this doesn’t matter at all but anything leaning towards decorative where you could see the bulb itself 00:02:40.742 --> 00:02:47.273 would be at least somewhat ruined by the presence of one of them twisty boys instead of a nice round light globe. 00:02:47.273 --> 00:02:50.655 And then there are the specialty bulbs like directional flood lights 00:02:50.655 --> 00:02:54.680 which use reflectors to direct the light they produce in a more directed fashion. 00:02:54.680 --> 00:02:57.099 A CFL can’t do that. 00:02:57.099 --> 00:03:00.474 But those mid-2000’s engineers weren’t just gonna give up - 00:03:00.474 --> 00:03:05.449 they decided to stuff the fluorescent coil inside something else to hide it 00:03:05.449 --> 00:03:08.570 or even alter how it releases its light. 00:03:08.570 --> 00:03:14.041 You could just stuff it into a light bulb-shaped thing to make it look more like a regular light bulb, 00:03:14.041 --> 00:03:19.601 maybe even a sphere to CFLarize the decorative globes for a bathroom vanity. 00:03:19.601 --> 00:03:28.935 Or you could stuff it inside a reflector to... kind of anyway mimic a flood light in form and, to a lesser extent, function. 00:03:28.935 --> 00:03:30.519 And we didn’t stop there! 00:03:30.519 --> 00:03:35.254 The days of CFL dominance were a wild time in weird light bulbs, 00:03:35.254 --> 00:03:42.000 and IKEA was particularly fond of cramming CFL tech into stranger and stranger applications. 00:03:42.000 --> 00:03:47.770 This small spot-light is one of my favorites: just look at that wonky little zig-zaggy tube! 00:03:47.770 --> 00:03:50.110 It’s even two-layers deep! 00:03:50.110 --> 00:03:56.220 But while this technique improved the aesthetic issue and made a few more form factors possible, 00:03:56.220 --> 00:03:58.920 there was a huge catch to doing it. 00:03:58.920 --> 00:04:05.583 CFLs did use quite a bit less energy than incandescent lights which meant they produced a lot less heat, 00:04:05.583 --> 00:04:09.862 but the discharge tube still gets pretty hot. 00:04:09.862 --> 00:04:18.475 The standard exposed coil design can rely on the convection currents from moving air to help dissipate that heat and cool the tube 00:04:18.475 --> 00:04:22.280 which means the operating temperatures remain pretty reasonable. 00:04:22.280 --> 00:04:24.540 But all of those decorative ones? 00:04:24.540 --> 00:04:31.120 They are deliberately sealing up their tubes in some sort of cover which functions like an oven. 00:04:31.120 --> 00:04:36.910 The tubes inside these will run quite a lot hotter than if they were left open to air. 00:04:36.910 --> 00:04:41.210 Which is actually a pretty big problem which would lead to poor performance. 00:04:41.534 --> 00:04:49.977 See, fluorescent lights work because of the ultraviolet light produced by the mercury vapor discharge happening inside the glass tube. 00:04:49.977 --> 00:04:58.267 That largely invisible light is then converted by the phosphors coating the inside surface of the glass into light that we can see. 00:04:58.267 --> 00:05:03.397 Those phosphors fluoresce under UV light which is why they’re called fluorescent lights. 00:05:03.397 --> 00:05:12.267 But the effectiveness of that ultraviolet discharge is greatly influenced by the vapor pressure of the mercury inside the tube. 00:05:12.267 --> 00:05:13.587 What is vapor pressure? 00:05:14.239 --> 00:05:16.559 Well, according to this website I found, 00:05:16.559 --> 00:05:24.595 “equilibrium vapor pressure is the pressure exerted by a vapor in thermodynamic equilibrium with its condensed phases (solid or liquid) 00:05:24.595 --> 00:05:27.926 at a given temperature in a closed system.” 00:05:28.234 --> 00:05:30.000 OK, so what does that mean? 00:05:30.000 --> 00:05:33.509 Well, mercury is a liquid at room temperatures. 00:05:33.509 --> 00:05:38.032 And just as room temperature water slowly evaporates when left open to air, 00:05:38.032 --> 00:05:40.690 room temperature mercury does the same thing. 00:05:40.690 --> 00:05:45.036 Some of it will transition to the gaseous phase even at low temperatures. 00:05:45.036 --> 00:05:47.483 But when sealed up in a tube, 00:05:47.483 --> 00:05:56.150 every gaseous molecule that breaks free from the liquid phase increases the pressure inside that tube just a teeny tiny bit. 00:05:56.150 --> 00:06:01.370 And once the gas pressure caused by those mercury molecules builds to a certain point, 00:06:01.370 --> 00:06:04.160 no more liquid mercury will evaporate. 00:06:04.160 --> 00:06:07.460 That equilibrium point is the vapor pressure. 00:06:07.460 --> 00:06:10.970 Now there are some wonky factors here which I’m skipping over. 00:06:10.970 --> 00:06:17.402 For instance, the vapor pressure of a fluorescent tube is actually determined by the coldest spot of the glass 00:06:17.402 --> 00:06:23.462 because that cold spot will cause condensation which returns some of the gaseous mercury back to liquid - 00:06:23.462 --> 00:06:29.697 which of course then lowers the internal pressure of the tube a bit, so some liquid mercury somewhere else will just evaporate again 00:06:29.697 --> 00:06:32.236 to replenish what just condensed and keep it in equilibrium. 00:06:32.236 --> 00:06:34.423 But anyway, the reason this matters at all 00:06:34.423 --> 00:06:40.451 is that fluorescent lights need a pretty specific mercury vapor pressure to work well. 00:06:40.451 --> 00:06:42.966 If the vapor pressure is too low, 00:06:42.966 --> 00:06:48.613 the ultraviolet discharge that we’re looking for just doesn’t work at all - so we can’t have that. 00:06:48.613 --> 00:06:50.989 But if it gets too high, 00:06:50.989 --> 00:06:57.541 well a higher vapor pressure means more mercury molecules are floating around per given volume, 00:06:57.541 --> 00:07:04.513 and in those close quarters, the mercury molecules start to absorb some of their neighbors’ discharge energy, 00:07:04.513 --> 00:07:11.470 so ultimately less ultraviolet light is produced by the discharge when vapor pressure gets too high. 00:07:11.470 --> 00:07:19.668 For best results, then, we need the lamp to be operating within a fairly narrow goldilocks zone of acceptable vapor pressure. 00:07:19.668 --> 00:07:24.520 Getting to that goldilocks zone isn’t impossible but it is tricky. 00:07:24.520 --> 00:07:29.147 Because even if you get the pressures just right when making the tube, 00:07:29.147 --> 00:07:34.631 you aren’t in complete control of the ambient temperatures the tube will experience. 00:07:34.631 --> 00:07:39.210 And as temperature increases, so does the vapor pressure. 00:07:39.210 --> 00:07:42.085 High temperatures mean there’s more energy in the system 00:07:42.085 --> 00:07:45.698 which means that more liquid mercury will transition to the gaseous phase 00:07:45.698 --> 00:07:53.317 and then, because those molecules are trapped in a tube with a fixed volume, we run into that too many molecules problem. 00:07:53.480 --> 00:08:01.100 In practice this means that if a fluorescent tube gets too hot in operation it will actually start to lose brightness. 00:08:01.100 --> 00:08:07.000 Which for a thing which's whole purpose is to make light is the opposite of desirable. 00:08:07.000 --> 00:08:13.840 In early fluorescent lighting designs, this was just a reality that we worked around and a limitation that we accepted. 00:08:13.840 --> 00:08:18.304 So long as they were being used in an environment somewhat close to room temperature, 00:08:18.304 --> 00:08:25.559 they’d work fine - and their internal pressure was calibrated during manufacturing to meet that usage expectation. 00:08:25.559 --> 00:08:28.575 But we humans are never satisfied with limitations 00:08:28.575 --> 00:08:35.000 so we worked on figuring out how to make the technology functional in a wider range of temperatures. 00:08:35.000 --> 00:08:41.789 In addition to tackling that functional issue, though, we also wanted to make the technology more energy-efficient. 00:08:41.789 --> 00:08:48.501 And it turned out that efforts to improve efficiency would force us to solve the temperature problem. 00:08:48.501 --> 00:08:50.751 As we developed fluorescent technology, 00:08:50.751 --> 00:08:56.571 we found that increasing the intensity of the discharge by running more current through a narrower tube 00:08:56.571 --> 00:09:00.163 could produce more light with less electrical energy. 00:09:00.163 --> 00:09:05.059 That's in part how a CFL can make so much light in this relatively small space 00:09:05.330 --> 00:09:10.628 and why fluorescent tubes kept getting skinnier as time went on. 00:09:10.628 --> 00:09:18.500 But those skinnier tubes dissipated similar power levels as their predecessors through less gas and glass, 00:09:18.500 --> 00:09:21.574 so they got much hotter when operating 00:09:21.574 --> 00:09:29.030 which in turn elevated the internal vapor pressure well beyond the point where the mercury discharge would produce much light. 00:09:29.681 --> 00:09:32.050 But clearly we figured it out. 00:09:32.050 --> 00:09:34.340 This light bulb is a lightin’. 00:09:34.340 --> 00:09:36.771 So what did we do to make this possible? 00:09:37.341 --> 00:09:38.860 Isn’t the answer obvious? 00:09:38.860 --> 00:09:41.506 Just amalgamate the mercury, Silly Billy! 00:09:41.901 --> 00:09:44.570 For the record, I don’t like this word. 00:09:44.570 --> 00:09:45.660 a-MALL-gum? 00:09:45.796 --> 00:09:46.895 AA-mal-gam? 00:09:46.895 --> 00:09:48.067 a-MAL-gum? 00:09:48.257 --> 00:09:49.271 [synthesized voices saying "Amalgam"] 00:09:49.271 --> 00:09:50.493 We’ll go with that one. 00:09:50.493 --> 00:09:59.318 If instead of just mercury you throw in an amalgam of mercury (which in simple terms means a mixture of mercury and some other metal or metals), 00:09:59.318 --> 00:10:03.438 you can bend the mercury’s temperature vs. vapor-pressure curve 00:10:03.438 --> 00:10:05.309 to better fit your needs. 00:10:05.309 --> 00:10:07.499 Mix in a bit of indium and bismuth, 00:10:07.499 --> 00:10:10.920 or if you’re feelin' fancy some bismuth, lead, and tin, 00:10:10.920 --> 00:10:16.992 and you can run a fluorescent tube at much higher temperatures without diminishing the discharge 00:10:16.992 --> 00:10:23.666 because those extra elements will sort of hold on to the mercury and keep it from vaporizing quite so easily, 00:10:23.666 --> 00:10:26.931 which lowers the vapor pressure at high temperatures. 00:10:26.931 --> 00:10:28.056 Which is great! 00:10:28.056 --> 00:10:33.945 Oh - and remember that thing about the coldest spot of the glass influencing the vapor pressure? 00:10:33.945 --> 00:10:42.000 Well, some CFLs have little bumps formed in the discharge tube specifically to be a cold spot. 00:10:42.617 --> 00:10:44.021 This little pocket of glass 00:10:44.021 --> 00:10:48.117 sticks out beyond the confines of the discharge going through the tube, 00:10:48.117 --> 00:10:50.420 so it stays cooler than the rest of the glass 00:10:50.420 --> 00:10:54.440 and helps to regulate the internal vapor pressure when operating. 00:10:54.440 --> 00:10:56.543 But back to the amalgam. 00:10:56.543 --> 00:11:02.695 These new mercury amalgams for high-temperature tubes are in fact solids at room temperature. 00:11:02.695 --> 00:11:09.990 Note that despite being a solid, some mercury will still evaporate (or I guess sublimate) into the gaseous phase. 00:11:09.990 --> 00:11:13.205 The amalgam came in the form of little pellets. 00:11:13.205 --> 00:11:16.255 Which, you might have noticed these before in certain lamps - 00:11:16.255 --> 00:11:19.945 sometimes it seems like there’s a little ball rattling around inside somewhere 00:11:19.945 --> 00:11:21.143 [rattling] 00:11:21.143 --> 00:11:26.175 and there is, in fact - though usually it’s held captive in a special pocket 00:11:26.175 --> 00:11:31.612 to keep it in one place so it doesn’t roll around and cause damage to the phosphor coating. 00:11:31.612 --> 00:11:38.178 Some designs like that weird IKEA spotlight actually leave that pocket and the pellet visible, 00:11:38.178 --> 00:11:44.161 though most of the time it’s hiding near the electrodes as is the case for this ordinary CFL. 00:11:44.161 --> 00:11:51.385 These solid pellets were much easier to deal with compared to liquid mercury which was tremendously helpful for manufacturing these 00:11:51.385 --> 00:11:57.798 but of course the main reason for their use is to lower the vapor pressure of the mercury when the lamp is running hot, 00:11:57.798 --> 00:12:01.325 as it will with narrow, high-efficiency tubes. 00:12:01.325 --> 00:12:04.034 Except, there was a trade-off. 00:12:04.034 --> 00:12:05.765 There always is! 00:12:05.765 --> 00:12:13.050 With these amalgam pellets, there's less available mercury to form the discharge when the lamp is cold. 00:12:13.050 --> 00:12:19.970 So when these high-efficiency lamps first start, they only operate at partial brightness. 00:12:19.970 --> 00:12:23.332 It won’t reach full brightness until it’s at operating temperature 00:12:23.332 --> 00:12:27.702 and enough of the mercury in the amalgam has actually managed to vaporize 00:12:27.702 --> 00:12:31.527 and produce the correct vapor pressure inside the tube. 00:12:31.527 --> 00:12:38.530 This is why CFLs and high-efficiency fluorescent tubes take time to reach full-brightness. 00:12:38.530 --> 00:12:48.519 The amalgam of mercury inside them is deliberately suppressing the vapor pressure when the lamp is cold so that it becomes optimal once the lamp is hot. 00:12:48.834 --> 00:12:55.870 Now, with most CFLs and high-efficiency tubes, this effect is noticeable but not that extreme. 00:12:55.870 --> 00:13:01.660 The lamp produces a good deal of light right away but will roughly double in brightness over the next minute or so. 00:13:02.230 --> 00:13:04.554 But if the tube gets very cold, 00:13:04.554 --> 00:13:10.890 say it’s being used outside or you stuck it in a freezer for a few hours for the purposes of demonstration, 00:13:10.890 --> 00:13:17.594 then the vapor pressure of mercury in the tube is far too low for it to do anything. 00:13:17.594 --> 00:13:25.709 To allow the lamp to start in this condition, a starter gas (usually a mixture of argon and neon) also fills the tube. 00:13:25.709 --> 00:13:30.550 That gas doesn’t produce much light - just a dull, pinky-purple glow. 00:13:30.550 --> 00:13:35.860 But it does allow for a discharge through the tube to happen without the mercury contributing. 00:13:35.860 --> 00:13:38.056 And that will produce some heat. 00:13:38.056 --> 00:13:45.004 Therefore, the temperature inside the tube does increase with time, most quickly at the ends of the tube near the electrodes. 00:13:45.004 --> 00:13:48.605 This heat helps more of the mercury in the amalgam to vaporize, 00:13:48.605 --> 00:13:51.028 increasing the vapor pressure in the tube, 00:13:51.028 --> 00:13:59.290 and once it’s correct that mercury contributes a lot of ultraviolet light to the discharge and the lamp is operating as intended. 00:13:59.290 --> 00:14:09.889 And the key thing to this video is that how aggressive you need that amalgam to be depends on how hot you expect the tube to get. 00:14:09.889 --> 00:14:16.209 An ordinary CFL with its tube open to air doesn't need a terribly aggressive amalgam. 00:14:16.209 --> 00:14:20.588 The tube doesn’t get that hot in operation with airflow to cool it, 00:14:20.588 --> 00:14:27.258 so the amalgam is fairly tame and at room temperatures at least some mercury vapor is present throughout the tube. 00:14:27.258 --> 00:14:32.524 Therefore it produces a good deal of light right away unless it’s very cold. 00:14:32.524 --> 00:14:36.445 But if you expect the tube to get hotter in operation, 00:14:36.445 --> 00:14:42.337 you’ll need a more aggressive amalgam which will lower the mercury vapor pressure even more. 00:14:42.337 --> 00:14:46.290 Which finally brings us back to bulbs like this. 00:14:46.290 --> 00:14:51.190 How aggressive do you suppose we need the amalgam to be in these bulbs? 00:14:51.190 --> 00:14:54.002 The answer is very aggressive. 00:14:54.002 --> 00:14:59.737 The discharge tube sealed inside these is going to get MUCH hotter than usual once warmed up, 00:14:59.737 --> 00:15:05.157 so we need an amalgam formulation which will drastically lower the vapor pressure of the mercury. 00:15:05.836 --> 00:15:08.916 That’s no problem - we know how to do that just fine. 00:15:08.916 --> 00:15:09.916 But the trade-off? 00:15:10.160 --> 00:15:11.160 There’s always a trade-off. 00:15:11.730 --> 00:15:18.878 Very, very little mercury vapor is free inside the tube of these bulbs even at room temperature. 00:15:18.878 --> 00:15:22.668 So those decorative or specialty CFLs? 00:15:22.668 --> 00:15:28.067 They behave like the CFL kept in the freezer every time they start. 00:15:28.067 --> 00:15:32.529 This CFL flood spot thing is currently at room-temperature. 00:15:32.529 --> 00:15:34.959 Despite its relative warmth, 00:15:34.959 --> 00:15:41.727 it barely glows when switched on and takes a solid minute or more to reach full-brightness. 00:15:41.727 --> 00:15:44.209 That IKEA spot from earlier? 00:15:44.480 --> 00:15:46.410 Yeah, it does the same thing. 00:15:46.410 --> 00:15:49.886 It’s pretty useless when you first switch it on. 00:15:49.886 --> 00:15:54.450 Most if not all decorative CFLs exhibited this behavior. 00:15:54.450 --> 00:16:02.707 They had to suppress the mercury vapor pressure a LOT because the tube inside would get very hot once warmed up, 00:16:02.707 --> 00:16:08.253 which meant they had terrible light output whenever they were first switched on. 00:16:08.253 --> 00:16:11.450 Which means they’re kind of annoying to live with. 00:16:11.450 --> 00:16:16.599 Although, shoutout to the decorative globe CFLs in the bathroom vanity when I was a kid. 00:16:16.599 --> 00:16:21.813 It was actually pretty nice to have a very dim-at-first light when making a bathroom visit in the middle of the night. 00:16:21.813 --> 00:16:23.492 Helps the eyes adjust. 00:16:23.734 --> 00:16:27.610 In most cases though, this behavior is annoying. 00:16:27.610 --> 00:16:32.220 Now, of course, the easiest solution is to just use open-air CFLs. 00:16:32.220 --> 00:16:38.160 They usually didn’t have this problem unless you were using them outdoors in very cold weather. 00:16:38.160 --> 00:16:44.730 But if you just had to cover up the ugly factor, or wanted to use a CFL in a specific application, 00:16:44.730 --> 00:16:50.649 you had no choice but to deal with the poopy cold starts of an enclosed tube. 00:16:51.165 --> 00:16:55.440 Unless, of course, you do what GE did with these bulbs. 00:16:55.955 --> 00:16:57.610 What did they do? 00:16:57.610 --> 00:17:03.560 They just stuck a halogen capsule inside the fluorescent spiral, added a timer, and called it a day. 00:17:03.560 --> 00:17:07.679 A halogen lamp will attain its full brightness in a fraction of a second 00:17:07.679 --> 00:17:10.807 (at the expense of using quite a lot of power). 00:17:10.807 --> 00:17:16.806 But if you’re just using it to fill in that gap that occurs when a CFL warms up, 00:17:16.806 --> 00:17:23.113 then you could just power it up for about a minute and then switch it off once the tube had gotten bright. 00:17:23.113 --> 00:17:26.760 Which is precisely what this light bulb does. 00:17:26.760 --> 00:17:33.534 When power is first applied, the internal circuitry switches on both the halogen lamp and the fluorescent tube. 00:17:33.534 --> 00:17:37.750 That gives it plenty of usable brightness the moment you need it. 00:17:37.750 --> 00:17:40.365 And then, after about a minute of operation, 00:17:40.365 --> 00:17:42.622 the fluorescent tube has gotten plenty warm 00:17:42.622 --> 00:17:45.621 (with a little help from the halogen capsule, it should be noted) 00:17:45.621 --> 00:17:51.755 so it’s producing adequate light on its own, and the power-hungry halogen capsule is switched off. 00:17:51.755 --> 00:17:54.780 Take a look at this power meter while I switch on the lamp. 00:17:54.780 --> 00:17:57.717 At first, this is a pretty power-hungry bulb, 00:17:57.717 --> 00:18:03.453 pulling about the same power level as the 100W incandescent light bulb it’s meant to replace. 00:18:03.453 --> 00:18:11.860 But before long, a large majority of its power consumption drops off and it operates close to the 25W the package claims. 00:18:11.860 --> 00:18:15.929 It’s a very simple idea, but it’s very effective. 00:18:15.929 --> 00:18:22.461 This bulb gets all of the pros of a CFL with the major con engineered out of the picture. 00:18:22.461 --> 00:18:30.479 And this particular GE Reveal bulb is among the best CFLs when it comes to color rendering and incandescent-mimicry I’ve ever encountered. 00:18:30.479 --> 00:18:32.940 It’s a genuinely impressive thing. 00:18:32.940 --> 00:18:38.136 GE put this hybrid drivetrain(?) in several different bulb types, 00:18:38.136 --> 00:18:41.224 including this flood which I showed you earlier. 00:18:41.224 --> 00:18:44.569 And here, because the front is clear and not frosted, 00:18:44.569 --> 00:18:49.467 we can actually see the halogen capsule resting in the middle of the fluorescent spiral. 00:18:49.467 --> 00:18:54.707 This behaves identically to the larger bulb but this is only a 65 watt equivalent 00:18:54.707 --> 00:19:01.110 so it uses less power both in the hybrid startup condition and the CFL-only operation. 00:19:01.110 --> 00:19:04.186 And if you’re wondering whether they did this for cool-white bulbs, 00:19:04.186 --> 00:19:06.242 the answer is yes! 00:19:06.242 --> 00:19:08.882 And it’s really freaking weird! 00:19:08.882 --> 00:19:17.141 When first powered up, the CFL is hardly contributing so it appears only slightly cool, like a very intense halogen lamp. 00:19:17.141 --> 00:19:21.135 But as it warms up, the light output shifts cooler and cooler, 00:19:21.135 --> 00:19:24.787 and then the halogen lamp shuts off with a very jarring 00:19:24.787 --> 00:19:25.799 BLAM 00:19:25.799 --> 00:19:29.597 and the light appears… well like that. 00:19:29.597 --> 00:19:33.520 I know some of you keep saying you like this but I’ll never understand it. 00:19:33.520 --> 00:19:41.979 And that jarring transition from a crisp white to a very clinical dank cold white reveals the worst aspect of these bulbs: 00:19:41.979 --> 00:19:43.372 Which is... 00:19:43.372 --> 00:19:45.636 GE half-assed this. 00:19:45.636 --> 00:19:52.475 Like, the idea is simple enough so perhaps justly they chose to perform a simple execution of the idea. 00:19:52.475 --> 00:19:56.070 But in my opinion, it’s much too simple. 00:19:56.070 --> 00:20:02.041 The halogen capsule is simply toggled off with no thought given to a seamless transition. 00:20:02.041 --> 00:20:06.127 So even with the warm-white bulbs where the transition is much less jarring, 00:20:06.127 --> 00:20:10.908 you still notice a pronounced light drop-off when the halogen light goes out. 00:20:10.908 --> 00:20:18.779 Of course, cramming in the circuitry required to slowly dim the halogen light’s output as the CFL warmed up would add complexity and cost 00:20:18.779 --> 00:20:21.436 so it’s easy to see why it didn’t happen. 00:20:21.436 --> 00:20:24.953 But that’s not my biggest problem with this design. 00:20:24.953 --> 00:20:28.409 Consider the idea of temperature compensation. 00:20:28.409 --> 00:20:31.808 The halogen capsule doesn’t need to operate if, say, 00:20:31.808 --> 00:20:35.783 the light had only been switched off for a few moments then switched back on. 00:20:35.783 --> 00:20:42.240 In that case, the fluorescent tube will still be plenty warm and it'll output full brightness right away. 00:20:42.240 --> 00:20:47.480 If you test this, you’ll see that GE appears to have built this in. 00:20:47.480 --> 00:20:51.377 The halogen capsule illuminates very briefly but then goes out, 00:20:51.377 --> 00:20:55.411 which suggests the bulb is aware of the ambient temperature. 00:20:56.090 --> 00:20:59.059 Alas, they faked that. 00:20:59.059 --> 00:21:03.765 The temperature compensation circuit in here is much like the one in modern toasters - 00:21:03.765 --> 00:21:07.506 likely using a capacitor which holds onto its charge for a while, 00:21:07.506 --> 00:21:14.888 the runtime of the halogen capsule is influenced by how long it has been since the last time it was powered on. 00:21:14.888 --> 00:21:20.803 If it was just on, that capacitor still has a decent charge so the halogen capsule runs only briefly. 00:21:20.803 --> 00:21:27.230 But if it’s been a while, that capacitor has little charge in it so the halogen capsule runs for the full minute. 00:21:27.230 --> 00:21:33.721 This works well enough but, ironically, it fails when the lamp most needs the help. 00:21:33.721 --> 00:21:38.230 Consider what happens when you use this outdoors in cold weather. 00:21:38.230 --> 00:21:41.580 This one’s been in the freezer so the tube is very cold. 00:21:41.580 --> 00:21:48.960 Upon power-up, you wouldn’t really know that because the halogen capsule is doing its job and providing plenty of light. 00:21:48.960 --> 00:21:55.120 But the circuitry in control of the halogen lamp has no idea that the bulb is so cold. 00:21:55.120 --> 00:21:58.000 It just knows it hasn’t been used in a while. 00:21:58.000 --> 00:22:01.834 So the halogen capsule shuts off near the one minute mark like always, 00:22:01.834 --> 00:22:07.809 and the fluorescent tube is still much too cold to produce meaningful light on its own. 00:22:07.809 --> 00:22:14.735 So, using one of these outdoors in a cold climate gives you a light which is bright *at* the start, 00:22:14.735 --> 00:22:19.840 but then suddenly gets very dim and needs a few more minutes to reach full brightness. 00:22:20.193 --> 00:22:23.540 I mean I guess that’s better than not having the halogen capsule in there - 00:22:23.540 --> 00:22:29.746 especially since its heat could theoretically help the CFL to start if it’s in extreme cold. 00:22:29.746 --> 00:22:31.616 But it’s far from ideal. 00:22:31.616 --> 00:22:39.009 And even more ironically, they only claim these will start reliably down to freezing point which isn’t impressive at all. 00:22:39.009 --> 00:22:44.927 Lots of CFLs claimed reliable starts down to 5 degrees Fahrenheit, which is - 00:22:44.927 --> 00:22:45.750 ya know what? 00:22:45.750 --> 00:22:48.677 Ask a search engine for what that is in Celsius. 00:22:48.677 --> 00:22:54.147 You can do that every time one of us silly Americans uses our silly numbers which, as a bonus, 00:22:54.147 --> 00:22:57.699 is a lot faster than complaining about it in the comments. 00:22:57.699 --> 00:23:00.188 But I will tell you it’s quite cold. 00:23:00.188 --> 00:23:08.043 Uh, anyway, the reason I find this particularly ironic is that, when we’re talking about bulbs which are being used indoors, 00:23:08.043 --> 00:23:15.730 arguably the only reason this feature needed to exist in the first place is that the tube is enclosed. 00:23:15.730 --> 00:23:20.879 If you don't care about that, an open-air CFL produces useful brightness right away. 00:23:20.879 --> 00:23:28.834 But either kind - enclosed or exposed tube - experiences terrible cold starts in cold weather. 00:23:28.834 --> 00:23:33.265 So if they had designed this with actual temperature compensation 00:23:33.265 --> 00:23:38.127 which forced the halogen lamp to stay on longer in freezing conditions, 00:23:38.127 --> 00:23:41.830 they could have actually fixed a real problem with the technology - 00:23:41.830 --> 00:23:51.557 especially if they went the extra mile and disabled the CFL circuit until the tube got enough heat from the halogen capsule to actually start reliably. 00:23:51.557 --> 00:23:55.781 That would have been a real game changer - if you remember the days of CFLs 00:23:55.781 --> 00:24:00.314 and experienced a cold porch light being useless when you needed it, 00:24:00.314 --> 00:24:05.349 an actually well-thought-out implementation of this idea would have been great! 00:24:05.349 --> 00:24:10.790 But these bulbs will only give you one minute of full brightness no matter what. 00:24:10.790 --> 00:24:14.460 Which isn't the most helpful thing in winters like ours. 00:24:14.460 --> 00:24:18.880 Still, I very much admire the spirit shown here. 00:24:18.880 --> 00:24:23.200 You may remember a video I made some years back on LED traffic lights 00:24:23.200 --> 00:24:29.593 and the fixation some people have with their inability to melt snow, despite all their other advantages. 00:24:30.000 --> 00:24:34.427 That “but sometimes!” way of thinking could be dealt with 00:24:34.427 --> 00:24:41.469 either by refusing to progress as some rather annoyingly like to advocate loudly for these days 00:24:41.469 --> 00:24:49.975 or by applying one of my favorite parts of the human spirit and actually making the effort to fix the sometimes. 00:24:50.165 --> 00:24:57.950 And much like LED traffic lights with heaters built-in or special covers which reduce snow buildup in the first place are now available, 00:24:57.950 --> 00:25:04.301 GE (and some other manufacturers from what I can tell) decided to fix the sometimes of the CFL 00:25:04.301 --> 00:25:10.202 by cleverly integrating a bit of old tech to quite literally fill in the gap. 00:25:10.202 --> 00:25:17.107 They didn’t quite do it with as much finesse as I think they should have, but the idea at its core was pretty great. 00:25:17.107 --> 00:25:22.642 But of course now that LED bulbs have gotten so cheap that you can pick them up for about a buck each, 00:25:22.642 --> 00:25:26.319 that sometimes isn’t even relevant in most applications. 00:25:26.319 --> 00:25:29.529 These don’t care how cold it is - they’ll just work right away. 00:25:30.262 --> 00:25:33.148 Oh, uh and carmakers? 00:25:33.148 --> 00:25:38.409 People keep sending me links to articles about LED headlights not melting snow. 00:25:38.409 --> 00:25:43.976 Before I have to make another video, add some defroster wires to the headlight housings why don’t ya? 00:25:43.976 --> 00:25:46.679 You’re already doing it on the rear glass. 00:25:46.679 --> 00:25:48.174 You can have that one for free. 00:25:48.337 --> 00:25:51.910 And stop it already with the red rear turn signals! 00:25:51.910 --> 00:25:52.774 Gosh! 00:25:52.774 --> 00:25:55.566 How is that not an obvious problem to you? 00:25:55.566 --> 00:25:59.487 And don’t just do a Stellantis and make the lamp color change - 00:25:59.487 --> 00:26:04.671 then you still lose the brake light on whatever side you're using the turn signal on and that’s hardly any better! 00:26:04.671 --> 00:26:06.860 Why do I have to think of this crap? 00:26:07.401 --> 00:26:09.620 It’s because I have too much time on my hands. 00:26:09.620 --> 00:26:13.493 Anyway, these light bulbs show that we can make things better 00:26:13.493 --> 00:26:17.850 even when new technologies present us with new limitations and new problems. 00:26:17.850 --> 00:26:23.263 And oftentimes one way to do that is to put a twist on the tech of the past. 00:26:23.263 --> 00:26:27.740 I mean, how do you think the CFL became a thing in the first place, amiright? 00:26:27.984 --> 00:26:30.400 Ho boy, better just end this right now. 00:26:30.400 --> 00:26:31.534 Thanks for watching! 00:26:32.276 --> 00:26:34.750 ♫ initially smooth jazz ♫ 00:26:36.461 --> 00:26:41.040 Hey there, I’d like to announce that I will be at Open Sauce 2024. 00:26:41.040 --> 00:26:47.460 That’s happening June 15 and 16 in San Francisco, and if you’d like more info visit opensauce.com 00:26:47.460 --> 00:26:49.590 (there’s a link in the description). 00:26:49.590 --> 00:26:53.234 Oh and if you’re going, be sure to wear some flowers in your hair. 00:26:54.500 --> 00:26:59.260 But, in the years prior to the LED bulb… and I’ve run too fast! 00:26:59.260 --> 00:27:06.104 …you might just stuff it into a light bulb-shaped thing to make it look more like a standard light bulb or… 00:27:06.104 --> 00:27:09.298 you could lose your place in the teleprompter and need to back up. 00:27:09.298 --> 00:27:13.774 Lots of CFLs claimed relialblel ba. 00:27:13.774 --> 00:27:16.297 This is what happens when it’s a very long line! 00:27:16.297 --> 00:27:22.433 But then suddenly gets very dim and needs another few minutes to reach full brightness. 00:27:22.433 --> 00:27:25.579 I mean, I … another few? 00:27:28.864 --> 00:27:30.538 Oh don’t tell me you’re dead. 00:27:32.737 --> 00:27:33.751 Ha! 00:27:33.751 --> 00:27:36.809 Well hold on, then, I gotta get a different bulb. 00:27:36.809 --> 00:27:40.422 Some designs like that weird IKEA lamp… 00:27:40.422 --> 00:27:41.181 crap! 00:27:42.267 --> 00:27:46.218 So, I released this on April Fool's Day. 00:27:46.218 --> 00:27:49.733 Did you think this was an April Fools video? 00:27:49.733 --> 00:27:53.536 Because the topic does seem pretty off-the-wall. 00:27:53.536 --> 00:27:56.094 Unless you remember these, I guess. 00:27:56.094 --> 00:27:57.820 Pretty bright idea, though.