WEBVTT 00:00:00.080 --> 00:00:08.080 It all started with a bold vision to heat my pool using waste heat from my 00:00:05.680 --> 00:00:12.240 server rack at home. In practice, it's less about heating the pool and more 00:00:10.080 --> 00:00:17.840 about cooling the computers. But hey, that was fine, too, until one day it 00:00:14.799 --> 00:00:20.560 wasn't. I came home to a terrible noise 00:00:17.840 --> 00:00:24.800 and water all over the floor, which was caused by corrosion. After cleaning up 00:00:22.640 --> 00:00:28.800 the mess, grounding all the systems, and adding this isolation transformer thing 00:00:26.800 --> 00:00:34.399 that I'm pretty sure still isn't hooked up properly, we refilled the loop with 00:00:30.880 --> 00:00:36.719 some kind of purple BMW fluid. And here 00:00:34.399 --> 00:00:41.440 we are a few months later and I've got organic shoo growing in both of the 00:00:39.440 --> 00:00:45.360 reservoirs. If that wasn't enough evidence that I had a problem, I cracked 00:00:43.520 --> 00:00:49.440 open one of the cooling blocks from one of the systems and the smell was worse 00:00:47.600 --> 00:00:54.079 than day four of a land party in the middle of August. There is good news, 00:00:51.440 --> 00:00:57.840 though. I didn't find any new corrosion. However, it's clear that there's a lot 00:00:56.079 --> 00:01:02.000 of work to be done to get the system back to peak performance. So, here's 00:00:59.680 --> 00:01:05.680 what I'm thinking. We pull all of these out of the rack and clean the green 00:01:03.680 --> 00:01:09.600 stuff out of the blocks. We find an additive that actually prevents organic 00:01:07.840 --> 00:01:14.240 growth, and we come up with a way to filter out all of the gunk that's 00:01:11.439 --> 00:01:18.320 already in the loops. Now, and we'll also need a segue to our sponsor, 00:01:16.080 --> 00:01:22.560 Hetner. They now have three new server types available. General performance for 00:01:20.080 --> 00:01:26.479 steady CPU loads and databases, regular performance for a good balance of 00:01:24.080 --> 00:01:31.600 things, and cost optimized for maximum value. Get $20 off HNER Cloud Server 00:01:28.799 --> 00:01:35.000 with code LTT25 by clicking our link in the video description. 00:01:41.759 --> 00:01:47.759 Step one is to assemble my team of Avengers and get them started cracking 00:01:46.159 --> 00:01:52.079 all of these systems open. They're probably going to need some instructions 00:01:49.360 --> 00:01:55.360 though because uh these systems are a little bit on the special side. We'll 00:01:53.840 --> 00:02:01.439 check in with these guys in a minute. Step two is to channel me inner 00:01:57.840 --> 00:02:03.759 Blackbeard and bury me treasure. 00:02:01.439 --> 00:02:07.439 Actually, in all seriousness, I used to keep a silver coil in my water cooling 00:02:05.600 --> 00:02:12.959 loop back in the old days because I heard that silver ions were actually 00:02:09.599 --> 00:02:14.959 great for unaliving microbes. Now, I was 00:02:12.959 --> 00:02:19.200 led to believe that I wouldn't need any silver with this particular coolant, but 00:02:17.120 --> 00:02:23.520 either it's not as antimicrobial as we thought, or the concentration isn't high 00:02:20.959 --> 00:02:28.000 enough. Either way, I'm hoping this will solve the problem without causing a new 00:02:26.160 --> 00:02:32.160 interaction with our coolant because I don't know exactly what's in it. So, 00:02:30.080 --> 00:02:35.760 good luck everybody. Now, it's time to have a look at our 00:02:33.440 --> 00:02:40.480 filters. Both the inside loop, which goes between this heat exchanger and the 00:02:37.599 --> 00:02:44.800 rack of PCs, and the outside loop, which goes between this heat exchanger and the 00:02:42.480 --> 00:02:49.760 walls of the pool, have filter canisters plumbed into them. But, uh, only one of 00:02:47.920 --> 00:02:54.800 them actually has a filter in it because, I don't know, we intended to 00:02:52.560 --> 00:02:59.120 put one in and then we never did. So, what we're going to do is we're going to 00:02:56.160 --> 00:03:03.599 put new filters in both of them. I got a five pack of filters. So, the idea being 00:03:01.120 --> 00:03:07.920 that I can just throw fresh ones in here to hopefully catch the bulk of the junk 00:03:06.080 --> 00:03:14.239 that's been floating around in the loops and then once that happens, I can throw 00:03:10.319 --> 00:03:17.200 in new ones for longerterm maintenance. 00:03:14.239 --> 00:03:20.480 Oh god. Oh, it smells. >> Just don't breathe. 00:03:18.080 --> 00:03:26.080 >> Filters are just rope. Anywh who, the new ones are 10 micron 00:03:22.800 --> 00:03:27.920 filters from Lafuy, so they should still 00:03:26.080 --> 00:03:32.080 allow a decent flow rate while catching whatever we manage to un alive with our 00:03:30.080 --> 00:03:37.120 silver ions. I guess it just goes in like this. >> Just looks like a spool. 00:03:34.400 --> 00:03:40.879 >> Yeah. How is this a filter? I don't Yeah, sure. Whatever. This next one's a 00:03:38.959 --> 00:03:45.920 little trickier because I think there's more loop to drain into it, but much 00:03:43.360 --> 00:03:51.840 less room for buckets. Oh, this one has a valve on it. That's actually way 00:03:47.680 --> 00:03:53.360 better. I recommend the Aquaflow 26069 00:03:51.840 --> 00:03:58.319 over the other one that I have. That's super handy. This is the one that has no 00:03:55.440 --> 00:04:02.640 filter in it whatsoever. Thanks, Jordan. Uh 00:04:00.400 --> 00:04:08.400 >> oh, gross. I see you guys got started without me. That's good. There's three 00:04:06.080 --> 00:04:12.720 main tasks for each of the systems. One is to drain the systems. Second is to 00:04:10.640 --> 00:04:16.720 clean the GPU block. Third is to clean the CPU block. And then also check the 00:04:14.959 --> 00:04:21.440 orientation because as it turns out, it's a 50/50 shot whether the water is 00:04:19.519 --> 00:04:23.680 going through those blocks in the right direction. 00:04:22.560 --> 00:04:28.000 >> All right, >> we labeled them wrong. >> So, this has not been drained yet. 00:04:26.400 --> 00:04:35.199 >> I had found that the easiest way to drain them was just to pull the GPU off 00:04:31.600 --> 00:04:37.199 and then just open it up over a bucket. 00:04:35.199 --> 00:04:39.360 That may not be correct. I see you removed the quick connect fittings at 00:04:38.560 --> 00:04:47.280 the back of the system. >> That seems smarter to me, but you know, >> you know what? Sure. [screaming] Ah, 00:04:43.600 --> 00:04:48.479 these don't have wrench point spells, 00:04:47.280 --> 00:04:52.080 >> do they? >> If you guys want to grab your systems 00:04:50.400 --> 00:04:56.639 and then we'll we'll all kind of do the same thing. We'll go through together. 00:04:53.440 --> 00:04:58.560 This tubing is a nightmare to get off. 00:04:56.639 --> 00:05:03.600 It really bites on there on these fittings. Oh, how did I do this? Okay, 00:05:02.000 --> 00:05:06.160 Jordan, new plan. Okay, >> we're just going to open up the block. 00:05:05.199 --> 00:05:11.440 >> That's one way to do it. >> Yep. One of the big problems today is all of these systems were built by 00:05:09.280 --> 00:05:15.919 different people. After the fact, I went through and I corrected the two to five 00:05:13.680 --> 00:05:20.080 errors that were in all of the systems. But there are still differences that are 00:05:17.919 --> 00:05:24.880 not down to being necessarily errors, but just philosophically different 00:05:22.800 --> 00:05:30.320 judgment. Like this one, for instance, has the GPU velcroed in place, while 00:05:27.440 --> 00:05:33.720 this one does not. Everyone get ready for the smell. 00:05:34.560 --> 00:05:40.639 Sorry. Ew, [screaming] groy. 00:05:39.039 --> 00:05:46.880 It's amazing that these things were still performing. Okay. Oh, ew. Oh, 00:05:43.759 --> 00:05:48.560 gross. But this is really good news. As 00:05:46.880 --> 00:05:53.759 I saw with the first system that I disassembled, there is no sign 00:05:51.120 --> 00:05:57.759 whatsoever of any kind of corrosion in here. It's all just organic. 00:05:56.639 --> 00:06:02.400 >> Oh, gross. >> What do you got? >> Check that out. >> Oh my goodness. Yours is even worse. 00:06:01.919 --> 00:06:08.240 That is >> That is terrible. >> That is literally disgusting. Yep. >> Cool. Whose system was this? 00:06:06.319 --> 00:06:11.520 >> Yep. Oh, this is my youngest daughter's kids always making messes. You know, I'd 00:06:10.240 --> 00:06:16.720 like to make it clear if it wasn't already that this is not normal. Maintenance is part of having a water 00:06:14.720 --> 00:06:20.560 cooled system. But if you use the proper coolants, you should never have organic 00:06:18.639 --> 00:06:24.000 growth or corrosion in your loop. While Jordan buttons up that GPU, let's talk 00:06:22.400 --> 00:06:28.160 about what happened with these CPU blocks. See, we thought that the 00:06:25.919 --> 00:06:32.160 internals of them were symmetrical. So, when we labeled the inlets and outlets 00:06:30.160 --> 00:06:35.520 on both sides, we did them such that they were the same regardless of which 00:06:34.000 --> 00:06:39.680 way we put these stainless steel plates on. As it turns out, that is not the 00:06:37.600 --> 00:06:46.400 case, which we're about to find out when I crack open this block. Get out. 00:06:43.680 --> 00:06:53.520 What is holding it on right now? Oh, it's just the O-ring was tight. Okay, 00:06:49.440 --> 00:06:56.000 that is nasty. See, the two inlets are 00:06:53.520 --> 00:06:59.759 here and here. They go into here, through the accelerator jet, and then 00:06:57.600 --> 00:07:05.840 into the fins. But for whatever reason, we had it labeled as in here and in 00:07:02.319 --> 00:07:08.400 here. But this clearly is what gathers 00:07:05.840 --> 00:07:14.560 the water that comes out of the jets and comes out here, way over here. It looks 00:07:11.520 --> 00:07:16.960 like there's also a more efficient flow 00:07:14.560 --> 00:07:20.400 path depending on sort of which way you have it oriented. It looks like this 00:07:18.560 --> 00:07:25.280 side would be the more efficient side cuz it comes straight in here and then 00:07:22.960 --> 00:07:29.199 comes straight out here instead of being pulled in way over here to come out 00:07:26.800 --> 00:07:34.000 here. Okay, that's great to know. spent the last five minutes looking for the 00:07:31.680 --> 00:07:37.759 screws to go into the back of this GPU. Okay, Jordan, thank you for spotting 00:07:35.840 --> 00:07:39.759 them. Also, curse you for putting them on [laughter] this rag. Jordan's like 00:07:39.440 --> 00:07:44.319 beer. >> I was sorting. >> He's the solution to and cause of all of 00:07:43.120 --> 00:07:47.520 life's problems. >> Jeez, >> it's a Simpsons quote. 00:07:46.880 --> 00:07:54.960 >> No duh. >> Do you need help with that? Okay. Oh jeez. Yeah, I got it. 00:07:52.400 --> 00:07:58.000 >> Jeez, Jordan. Oh my god, there's water on the power supply. 00:07:56.720 --> 00:08:01.680 >> A little bit. Yeah, >> mostly algae. It's okay. >> How's it going, newbies? 00:08:00.080 --> 00:08:07.039 >> So far so good. I haven't broken anything >> that you know. If anyone's wondering, 00:08:05.280 --> 00:08:11.680 no, I would not recommend this configuration for your home. I doubt 00:08:08.960 --> 00:08:14.240 anyone was asking, but now it's answered. 00:08:12.319 --> 00:08:20.080 >> This block is in terrible shape. >> Oh, okay. So, that was a thing. Because 00:08:17.120 --> 00:08:24.319 of all of the corrosion and crap that was in them, we actually like took a 00:08:22.080 --> 00:08:27.199 razor blade and like scraped off the nickel plate. >> So, you broke it? 00:08:25.759 --> 00:08:33.200 >> Yeah. Yeah. No, this is fine. This is fine for performance. This is totally totally fine. With the CPU block clean, 00:08:31.039 --> 00:08:38.320 now it's time to put it back together properly this time. Wait, no. Because 00:08:36.560 --> 00:08:42.880 before we do this, we've got to actually check your loop order. Remember, I said 00:08:40.399 --> 00:08:45.279 that it's a 50/50 shot whether the coolant is flowing through the CPU block 00:08:44.640 --> 00:08:51.440 correctly. >> Right. >> Right. Because we either assembled them 00:08:48.240 --> 00:08:52.320 this way or this way and we labeled them 00:08:51.440 --> 00:08:59.440 randomly. >> Okay. So, let's have a look. In goes to 00:08:57.600 --> 00:09:04.800 out. This one's wrong. >> We were close. >> We were close. So, that's bad news 00:09:02.959 --> 00:09:08.240 because what that means is that we need to get those tubes off. And these tubes 00:09:07.279 --> 00:09:13.120 are a nightmare. >> Yep. Fun time. >> I'm going to go clean my block. Here's 00:09:10.880 --> 00:09:15.360 an idea. Why don't we just take all the screws out of here? 00:09:14.080 --> 00:09:18.640 >> I've lost my screw for >> No screw for you. >> Oh boy. 00:09:17.200 --> 00:09:22.560 >> Don't feel bad. It happens to me all the time. The more of these systems I get 00:09:20.320 --> 00:09:27.360 apart, the more sure I am that we don't have any issues with corrosion. None of 00:09:25.600 --> 00:09:31.120 the same symptoms from last time. So, whatever it is that we did has clearly 00:09:29.440 --> 00:09:34.720 addressed that. This looks great. There is a bit of blue color in some spots 00:09:33.040 --> 00:09:38.720 that could be attributed to copper corrosion, but it comes off very easily 00:09:37.200 --> 00:09:43.360 with the fingernail, which tells me it's probably just dye from the coolant. Now, it's time to figure out how the heck I 00:09:41.839 --> 00:09:50.240 managed to get these tubes off on the one that I did do it on. No, I remember 00:09:46.080 --> 00:09:53.760 I used two flatheads. So, I levered this 00:09:50.240 --> 00:09:56.720 up and then I held it somehow and then I 00:09:53.760 --> 00:10:03.839 got a second one up in there and then I got kind of on it and then got to Ah, 00:10:00.080 --> 00:10:05.680 yes. Fuel line hoses recommended, 00:10:03.839 --> 00:10:09.360 I guess. I mean, they don't come off. Well, I may have savaged the tube a 00:10:08.080 --> 00:10:13.519 little bit, though, so we may end up having to cut them anyway. Uh, but not 00:10:11.440 --> 00:10:16.800 much. This is really important. Don't underestimate this. When you're building 00:10:14.800 --> 00:10:21.519 a water cooled system, the orientation of the tube as you slide it onto the 00:10:19.360 --> 00:10:26.079 barb matters because if you have it twisted one way, then over time it will 00:10:24.240 --> 00:10:29.279 apply more pressure to the fitting. If you have it twisted the other way, 00:10:27.497 --> 00:10:33.519 [music] then it can apply a loosening pressure to the fitting. So, what we 00:10:31.120 --> 00:10:39.079 want to do is we want to run this here and make sure that there is no tension 00:10:35.600 --> 00:10:39.079 on it whatsoever. 00:10:40.240 --> 00:10:45.680 Okay, I don't think that's going anywhere. Man, I am running out of how 00:10:44.000 --> 00:10:49.200 many times I can snip a little bit off of these. >> Yeah, we're getting a little iffy here. 00:10:47.519 --> 00:10:52.320 >> Yep. Still good today, though. And that's all I need today. Uh 00:10:51.200 --> 00:10:56.640 >> oh. >> How did these ones end up upside down? 00:10:54.640 --> 00:11:00.320 Uh guys, on the GPU, make sure you when you put the worm drive clamps back on, 00:10:58.160 --> 00:11:03.760 they're right side up. >> Yeah, 00:11:01.440 --> 00:11:09.440 >> that's fine. Whatever. It's just as much work to take it out now and fix it and 00:11:06.320 --> 00:11:12.560 put it back as it will be next time. So, 00:11:09.440 --> 00:11:13.120 that's a future Linus problem. Wait a 00:11:12.560 --> 00:11:17.120 minute. >> This was your computer. >> Which one? >> You did this one. 00:11:16.079 --> 00:11:24.120 >> I did the GPU. >> Yeah, >> but I didn't I didn't loosen those fittings, though. 00:11:20.320 --> 00:11:24.120 >> I don't know what to tell you. 00:11:24.480 --> 00:11:30.079 >> Future Linus problem. While we find out if this system is going to leak by 00:11:27.680 --> 00:11:34.640 pressurizing it and then waiting to see if the pressure is going to drop, let's 00:11:31.760 --> 00:11:38.320 go have a look at another strategy that we're going to be employing for 00:11:36.079 --> 00:11:42.320 antimicrobial treatment for the loop. For our microorganism death rays, we 00:11:40.320 --> 00:11:46.480 thought about using the ultraviolet or UV lights from that time that Dbrand 00:11:44.480 --> 00:11:51.440 made my car glow in the dark. Unfortunately, it turns out these put 00:11:48.880 --> 00:11:56.720 out the wrong kind of UV. Uh, it's great for making your t-shirt glow. But UVA, 00:11:55.040 --> 00:12:01.200 which is at the upper end of the UV spectrum, is just below the light that 00:11:58.959 --> 00:12:06.800 we humans can see. And if we want to actually kill microorganisms, we need to 00:12:03.360 --> 00:12:10.079 go deeper past UVB all the way to the 00:12:06.800 --> 00:12:14.000 sweet spot at a wavelength of about 253 00:12:10.079 --> 00:12:15.680 nmters down in UVC territory. This is 00:12:14.000 --> 00:12:20.000 what's going to help us disinfect the coolant by penetrating all the 00:12:17.600 --> 00:12:23.519 microorganisms and damaging their DNA and RNA. Also not great for your skin or 00:12:21.839 --> 00:12:26.959 eyes, though, so take appropriate precautions if you're following along at 00:12:25.120 --> 00:12:31.920 home with your pool to server rack cooling loop. You can buy UVC filters 00:12:29.920 --> 00:12:36.399 for aquariums and pools that plum into your loop. But I don't know if my plan 00:12:34.480 --> 00:12:40.800 today is going to work at all, and we've already got a ton to do. So, we opted 00:12:38.399 --> 00:12:45.600 for these drop-in cleaning systems. Basically, what they do is they have a 00:12:43.040 --> 00:12:50.079 little pump built into them. They slurp up coolant, expose it to UVC, and then 00:12:48.399 --> 00:12:53.279 just spit it back out into the reservoir. They're even fully enclosed, 00:12:52.000 --> 00:12:57.360 so we don't have to worry about any stray UVC doing any damage to the 00:12:55.440 --> 00:13:00.800 plastics in our loops. Got a little quick connect thing here. So, we could 00:12:58.639 --> 00:13:04.959 maybe feed it through the lid somehow. >> Yeah. Our plan was actually to replace 00:13:03.279 --> 00:13:08.160 the lids and just put like a little cover on it cuz this is sort of a 00:13:06.320 --> 00:13:11.360 short-term experiment. But maybe I could just jam this in there. 00:13:09.839 --> 00:13:14.079 >> It might work. Yeah. >> Here. Let's Let's see. All right. So, I 00:13:13.040 --> 00:13:18.399 just chuck it in there, right? >> Yeah, I guess so. >> Okay. >> Good luck, everyone. 00:13:17.120 --> 00:13:21.120 >> I mean, whatever. I've been touching this stuff all day. I guess what difference does it make? 00:13:20.240 --> 00:13:25.440 >> Not dead yet. >> Yeah. >> Where' that paper towel go? 00:13:23.440 --> 00:13:29.680 >> Uh oh. It's up there. Uh there's uh there's some like loose. No, no. Right. 00:13:27.920 --> 00:13:33.200 Wait, where'd it go? Oh, here it is. It fell behind me. 00:13:31.040 --> 00:13:37.360 >> Oh jeez. Okay. Thought I was crazy. >> No, no, no, no. Well, maybe, but not 00:13:36.320 --> 00:13:40.000 from that. >> My thinking is you can just plug this in with your pumps. 00:13:39.680 --> 00:13:46.720 >> Yeah, >> this will just stick onto the >> Sure. >> reservoir here. >> Love it. This is quite tight. Let's do 00:13:45.920 --> 00:13:50.959 another one. >> What about second one? >> Yeah. 00:13:48.959 --> 00:13:53.360 >> Oh, why did I even dry my arm? >> Uh, that was stupid. 00:13:52.320 --> 00:13:57.120 >> Because stupid. >> Yeah. Well, it is worth noting that 00:13:55.440 --> 00:14:00.240 these are only designed to last for about a year at a time. So, I'm going to 00:13:58.639 --> 00:14:04.880 have to set an alarm for myself in a year to um put in new ones. Oh, it 00:14:03.120 --> 00:14:07.839 stinks. This is the stinkier one for sure. Well, I'm going to unplug the 00:14:06.320 --> 00:14:13.000 pumps for now and then we can turn these on and see if they I don't know, do 00:14:09.760 --> 00:14:13.000 anything interesting. 00:14:13.360 --> 00:14:20.000 12hour. Do I have to manually run these every time? >> No, you're good. It's 24-hour cycle. 00:14:18.000 --> 00:14:23.519 >> Okay, so this will turn on and off every 12 then. That seems reasonable, right? I 00:14:22.079 --> 00:14:27.440 don't know. I'm not a scientist. >> We got the filters and we got the 00:14:25.760 --> 00:14:34.519 silver. So, theoretically, these are overkill. >> There will never be a problem again. 00:14:30.399 --> 00:14:34.519 >> Why would you say that? 00:14:35.040 --> 00:14:40.880 >> Let's go see how the systems are doing. >> Oh, I got a leak. >> What? >> Video card. 00:14:38.720 --> 00:14:42.639 >> That explains why this fitting was not oriented correctly. 00:14:42.079 --> 00:14:48.560 >> Right. >> It [music] got twisted at some point. This might be one of those cases of that 00:14:46.480 --> 00:14:52.639 thing I was talking about where there's tension on the tube. Okay, so now it's 00:14:50.560 --> 00:14:56.880 going to loosen at this end. A crud. I think what we need to do is we need to 00:14:54.959 --> 00:15:00.639 loosen the worm drives on both sides, tighten them both, and then put the worm 00:14:58.560 --> 00:15:03.199 drives back on so that the tubing can spin. I can't spin the tube. Oh, yes, I 00:15:02.480 --> 00:15:07.199 can. >> Oh, shoot. But I'm spinning it not the way that I wanted. I can't tell if I'm 00:15:06.000 --> 00:15:13.440 spinning the tube or the fitting or both. >> Yes. >> Yeah. Okay, there we go. Now she's 00:15:10.480 --> 00:15:18.720 tight. Okay, that's tight. That's tight. That's tight. I think we've got one test 00:15:15.600 --> 00:15:20.399 subject ready for [music] subjecting to 00:15:18.720 --> 00:15:25.120 tests. Sorry. >> Nice. Where's the screws for the top 00:15:23.360 --> 00:15:29.519 here? There are four. There are only supposed to be three. >> There should be three. I only gave you 00:15:28.000 --> 00:15:32.639 three screws. Did you invent another screw? >> 1 2 3 4. 00:15:31.040 --> 00:15:35.040 >> No, the force wasn't in the same bin. >> Oh, okay. >> That's for this one. 00:15:34.240 --> 00:15:40.800 >> Okay. Okay. >> Don't be adding to the thing. >> I'm not trying to add to the thing. 00:15:37.680 --> 00:15:44.880 Tempers were running high. It was hour 00:15:40.800 --> 00:15:47.440 four of fixing computers. Four. 00:15:44.880 --> 00:15:50.880 >> Okay. I'm confident. I'm so confident I'm putting the screws in. 00:15:48.560 --> 00:15:54.399 >> Yeah. Yeah. It's not a good idea. >> Actually, this is a cool opportunity to 00:15:52.399 --> 00:15:58.160 show you guys a bench that I made for this off camera. Back when we checked 00:15:56.079 --> 00:16:03.360 out this external radiator pump system from Bixie, we found that it had a bit 00:16:00.480 --> 00:16:09.120 of a fatal flaw. It was darn near impossible to fill. So, what I did to 00:16:05.680 --> 00:16:10.560 fix it was I added another fill port. 00:16:09.120 --> 00:16:14.639 Okay, there's some pressure in it right now. That's interesting. I added another 00:16:12.720 --> 00:16:18.880 fill port so that when you need to top it up, you can just crack the other one 00:16:16.880 --> 00:16:22.160 and it will make it easier for the fluid that you're putting in to displace the 00:16:20.399 --> 00:16:25.279 air and it will come out and it bleeds super easy. Now, so this is here for 00:16:24.000 --> 00:16:34.959 anytime I need to do maintenance on these systems. And uh burp and burp. 00:16:32.240 --> 00:16:39.600 Okay, this is going to seem crazy because it is, but I'm just going to 00:16:36.880 --> 00:16:43.120 reuse the coolant from earlier. I mean, what we got filtration, we got bioside. 00:16:41.839 --> 00:16:48.160 >> Not in this system. You don't. >> What could go wrong? And and 00:16:46.560 --> 00:16:53.440 hopefully we didn't break any of these systems today. Hey, there we go. All 00:16:51.680 --> 00:16:56.800 right. Beauty. >> All right, we've got everything back in 00:16:54.720 --> 00:17:01.680 the rack here. We'll leave it for maybe a week or so. Then we'll come back and 00:16:58.240 --> 00:17:04.000 see how the filters are looking. 00:17:01.680 --> 00:17:10.720 >> Time to find out if any of that worked. Let's start with the indoor loop. Um, 00:17:08.319 --> 00:17:15.039 oh, like before, the indoor loop is clearer than the outdoor loop, but like 00:17:12.799 --> 00:17:20.559 before, both of them still have gunk in them. And like before, sometimes Lionus' 00:17:18.160 --> 00:17:25.280 ADD brain overlooks things. Jordan. >> Yes, sir. >> The bypasses are on. 00:17:22.799 --> 00:17:27.919 >> Well, why'd you do that? >> Because I forgot, Jordan. 00:17:26.720 --> 00:17:31.760 >> That was all for nothing, then? >> No, not at all. It just means that we 00:17:29.760 --> 00:17:35.360 weren't really actively cleaning the water in the reservoirs. the rest of the 00:17:34.000 --> 00:17:39.600 loops are still running through the filters. >> Okay, >> so we're just not getting the kind of 00:17:38.080 --> 00:17:42.640 water flow through these that it might take in order to actually clean them. 00:17:42.160 --> 00:17:48.960 >> Okay, >> it also means that well the silver ions are probably not circulating to quite 00:17:46.640 --> 00:17:52.480 the same degree [music] and the UV cleaning that we're doing is not going 00:17:50.960 --> 00:17:56.640 to be very thorough because the water isn't all cycling through the resz. 00:17:54.559 --> 00:18:02.120 >> All right, should we check the filters? >> I think we should just check in in 00:17:58.160 --> 00:18:02.120 another week. Okay, 00:18:02.720 --> 00:18:11.520 >> big moment. Oh, dude. Whoa, dude. She's clear, dude. She's 00:18:09.200 --> 00:18:15.360 clear, buddy. Like, there's still some shoo in the bottom. I kind of like wiped 00:18:13.760 --> 00:18:19.520 it out a little while ago when I couldn't really see anything, but wow. 00:18:17.679 --> 00:18:24.160 >> Is the pump thing still stuck to the side? >> Yep. >> Still in there. So, that stuck better 00:18:22.080 --> 00:18:29.039 once I gave it a little wipe down. Now, reservoir number two. Still a couple 00:18:26.720 --> 00:18:32.799 little floaties on the surface. I think this one might be more clear, too. Where 00:18:31.039 --> 00:18:37.200 are the coins? That's the easiest way to tell how well I can see to the bottom. 00:18:34.880 --> 00:18:41.600 How strong is this shelf? Just don't don't worry. Don't Don't worry about it. 00:18:39.200 --> 00:18:46.160 I wonder if the concentration of the coolant is just way higher on this one, 00:18:44.160 --> 00:18:49.919 though, because it is harder for me to see down to the bottom. Like, the light 00:18:47.760 --> 00:18:53.440 doesn't penetrate and and bounce around as much. >> It's a darker blue, 00:18:51.600 --> 00:18:56.080 >> but I can definitely see all the way down to the bottom. Nice. 00:18:55.280 --> 00:19:02.960 >> 100%. >> We could not do that before. So, >> like there's my coins now to see how 00:19:00.400 --> 00:19:07.200 much crap is in the filters. Other than just feeling kind of slimy. It doesn't 00:19:05.440 --> 00:19:11.679 seem that full of stuff. Should we replace it anyway? >> Might as well. We got more filters. 00:19:10.080 --> 00:19:16.240 >> Okay, let's do it. Like, tell me the difference between this and that. This just looks like it's soggier. 00:19:14.880 --> 00:19:19.520 >> Oh, that's interesting. There's [music] a big air bubble. >> Oh, yeah. It'll get stuck there for a 00:19:18.799 --> 00:19:24.960 while. >> Give it like a week and it'll be gone cuz the flow rate's not that high. 00:19:22.480 --> 00:19:31.039 >> Yeah. For the inside loop, I don't have a valve on it, so I just need to turn 00:19:26.880 --> 00:19:34.039 off the pump and then move fast. 00:19:31.039 --> 00:19:34.039 >> Okay, 00:19:34.480 --> 00:19:40.480 bypass. Wait, >> what just happened? Oh, yeah, right. 00:19:39.668 --> 00:19:46.720 [laughter] That explains why the other loop had a little flip out while I did that. Wait, 00:19:44.799 --> 00:19:49.919 wait, wait, wait. Glenn, you mentioned you used to change these filters on a 00:19:48.320 --> 00:19:53.600 pretty regular basis. Is this clean or dirty? That one looked 00:19:52.720 --> 00:20:00.320 middle >> in the middle. >> Quite yellow. >> Yeah. >> So, the outside loop was more gked, 00:19:58.640 --> 00:20:05.679 which kind of makes sense cuz it's a much larger loop. So, if it had stuff 00:20:02.880 --> 00:20:11.280 growing in it, then it had a lot more growing in it. Okay. Night and day. So, 00:20:08.320 --> 00:20:15.200 I just replaced a clean filter and also a dirty one. So, I could maybe even 00:20:13.440 --> 00:20:18.320 leave this one on here for a bit, give it like a month, and then maybe do it 00:20:16.880 --> 00:20:20.480 one more time. I'm getting better at losing less of my coolant when I do 00:20:20.080 --> 00:20:24.960 this. >> Time to declare victory. >> I think so. >> What was the solution? 00:20:23.760 --> 00:20:30.480 >> Don't do this at home. >> Well, right, but I just But was it the 00:20:27.760 --> 00:20:33.120 filtration, the silver, or the UV? >> Yes. 00:20:31.039 --> 00:20:37.360 >> Okay, fair enough. See you later after this segue to our sponsor, Squarespace. 00:20:36.080 --> 00:20:42.240 If your website is looking a little rough around the edges, you're probably not using Squarespace, but you should. 00:20:40.320 --> 00:20:46.240 They make bringing your dream site to life a breeze, thanks to a code-free, 00:20:44.400 --> 00:20:49.840 drag and drop approach to editing and customization. And if you don't know how 00:20:48.159 --> 00:20:53.760 you want to start, their professionally curated templates can jog some 00:20:51.360 --> 00:20:57.280 inspiration. Or you can just make things really easy with their AI tools. Just 00:20:55.600 --> 00:21:00.480 throw in a few prompts and watch your vision take shape. Speaking of shapes, 00:20:59.120 --> 00:21:04.000 if you're building a site for your manysided business, you'll get tools to 00:21:02.799 --> 00:21:10.080 help you run it directly from Squarespace, like email marketing, advanced analytics, and many more. Head 00:21:07.760 --> 00:21:14.640 over to squarespace.com/LTT to get started for free and save 10% off 00:21:12.720 --> 00:21:18.960 your first purchase of a website or a domain. If you guys enjoyed this video, 00:21:16.559 --> 00:21:22.880 maybe it'd be fun to go back to the very early concepts. It started with a 00:21:21.280 --> 00:21:26.480 drawing on a piece of cardboard and ended up 00:21:25.919 --> 00:21:30.080 here. >> It was a simpler time.