WEBVTT 00:00:00.240 --> 00:00:07.040 How do you fix a problem you can't even 00:00:04.080 --> 00:00:12.000 see? In smaller systems, it's not that hard. Air leak on your tire, just throw 00:00:09.920 --> 00:00:17.359 some soapy water on it and find the bubbles. Easy. But in a larger system, 00:00:15.599 --> 00:00:24.880 like our workshop's compressed air hoses, it could take a fluke to find a 00:00:20.880 --> 00:00:28.080 tiny leak somewhere. A Fluke II 915 00:00:24.880 --> 00:00:30.000 acoustic imager, to be exact. This 00:00:28.080 --> 00:00:35.360 $25,000 handheld camera does not work like a 00:00:32.960 --> 00:00:41.120 normal camera. You know, taking in light and transforming it into internet cloud. 00:00:37.520 --> 00:00:42.879 Instead, it takes in sound and uses it 00:00:41.120 --> 00:00:47.680 to find leaks and equipment malfunctions, even detecting noises 00:00:45.280 --> 00:00:52.239 beyond the range of human hearing. It then represents that audio like a heat 00:00:50.239 --> 00:00:57.199 map. So, let's see if we can use it to find some really useful stuff like the 00:00:54.559 --> 00:01:02.640 leak in our system and maybe even a segue to our sponsor. 00:01:00.719 --> 00:01:07.119 Shut your trap, computer man. It's okay. I got this. Look at me. You're thinking 00:01:04.960 --> 00:01:10.720 I need more action in my life. And I agree. Bam. Look at that. Thousands of 00:01:08.960 --> 00:01:14.720 planes, choppers, tanks, and ships on our gently used War Thunder vehicles 00:01:12.799 --> 00:01:18.320 lot. Trust me, you're going to click the link below and try War Thunder, and 00:01:16.720 --> 00:01:23.159 you're going to like it. Ask me how I know. Later though, I'm busy. 00:01:32.479 --> 00:01:42.240 Like so many of our videos, this one was prompted by an actual problems 00:01:39.920 --> 00:01:46.320 that obviously it helps a little that we shoot most of our workshop videos on the 00:01:44.560 --> 00:01:51.200 other side of this wall, but it's still been a major source of disruptions over 00:01:48.880 --> 00:01:56.399 the last few years. Now, an obvious solution would be to just not have a 00:01:53.840 --> 00:02:01.439 compressed air system, but that's not really an option. We use it for 00:01:57.840 --> 00:02:03.920 everything from cleaning things off to 00:02:01.439 --> 00:02:08.319 cutting and polishing to hacking things apart. Excuse me. And I've done a fair 00:02:06.560 --> 00:02:12.319 bit of sand blasting with it, too. Mostly for my personal motorbike 00:02:09.840 --> 00:02:16.400 project. So, okay, we need an air compressor. Deal with it, right? Well, 00:02:14.879 --> 00:02:21.520 see, it wouldn't be that big of a problem. Except that somewhere in the 00:02:18.640 --> 00:02:27.200 system is a leak. So, the compressor goes off way more often, disrupting us 00:02:24.720 --> 00:02:31.599 way more often than it needs to. That's where Fluke's acoustic imager comes in. 00:02:29.280 --> 00:02:36.080 Theoretically, with this, in a matter of minutes, I should be able to track down 00:02:33.360 --> 00:02:40.080 the leak, saving us both time and money. Theoretically, I'm going to put that to 00:02:38.160 --> 00:02:44.800 the test. I've never actually used this before, so I don't know how I'm going to tell the difference between our HVAC 00:02:42.959 --> 00:02:50.800 system and air leaking out of a hose, but presumably, it's pretty easy. Wow, 00:02:48.000 --> 00:02:54.000 there's something in this wall. Oh, there's something bouncing off this 00:02:53.120 --> 00:02:57.360 wall. I think >> you'll get a lot of sounds reflecting 00:02:56.239 --> 00:03:03.040 from that. Sometimes >> that makes sense. >> Keep an eye on the right side and you'll 00:03:01.040 --> 00:03:06.319 see the range of frequency band that it's looking for. 00:03:04.640 --> 00:03:13.680 >> Okay, I feel like I could move pretty quickly at this point. Oh, hello, 00:03:09.519 --> 00:03:15.920 Captain. I think we've found something 00:03:13.680 --> 00:03:21.120 right there. Okay, but would I even be able to hear this? Oh, I can. I would 00:03:18.720 --> 00:03:26.599 have to be within like 10 inches of it though. One to go. And that took like 00:03:23.280 --> 00:03:26.599 two minutes. 00:03:26.720 --> 00:03:32.959 Oh jeez. Okay. Once you know what you're looking for, it stands out like a 00:03:30.879 --> 00:03:39.040 beacon. Oh yeah, this one's a lot more obvious. 00:03:36.400 --> 00:03:42.560 So between these two, this one was a pretty substantial one. We found both of 00:03:41.040 --> 00:03:46.400 them within just a couple minutes. >> This is a lot like the video Adam Savage 00:03:44.159 --> 00:03:50.720 did with this thing. Okay, two cool people can have the same cool idea at 00:03:48.080 --> 00:03:56.640 the same time. Adam Savage and Electroboom and also me, just slower. I 00:03:54.799 --> 00:04:01.760 want to show you guys separate from the video mode, you can put it in leak Q 00:03:58.720 --> 00:04:04.239 mode. This allows you to take a picture 00:04:01.760 --> 00:04:09.360 of a leak. Then according to your settings, so gas price, compression, 00:04:07.040 --> 00:04:14.239 energy cost, as well as your currency, it will tell you how much that leak is 00:04:11.120 --> 00:04:16.400 going to cost you here. I had no idea 00:04:14.239 --> 00:04:21.359 that just a little air leak could cost $45 a year. So, replacing the fitting is 00:04:19.280 --> 00:04:24.639 a no-brainer. That assumes it's running all year. I mean, it is running all 00:04:23.120 --> 00:04:28.560 year. We never turn it off. Like, that's the whole thing with a leak. It's like a 00:04:26.400 --> 00:04:32.880 faucet, right? Like a every once in a while drip will waste like more gallons 00:04:31.040 --> 00:04:36.720 in a month than you could possibly fathom. Now, obviously, if the only 00:04:34.720 --> 00:04:39.919 gases you're working with are compressed air, it would take a little while for an 00:04:38.320 --> 00:04:44.080 investment like this to pay for itself. But in certain industries, this could be 00:04:41.919 --> 00:04:47.759 worth its weight in gold in cost savings. Now, while the workshop team 00:04:46.240 --> 00:04:53.040 gets those fittings sorted out, I'm going to go see what else this thing can do. The magic of this device is really 00:04:51.199 --> 00:04:58.000 in its algorithm and precise calibration. The hardware itself is 00:04:55.680 --> 00:05:02.560 pretty simple. This array of 64 microphones is arranged in a logarithmic 00:05:00.160 --> 00:05:07.360 spiral to capture sound, and it uses beam forming to determine the source of 00:05:04.400 --> 00:05:12.960 the sound. Then it overlays that on the camera image as what Fluke calls a sound 00:05:10.240 --> 00:05:16.880 map. In the default mode, a sound map looks an awful lot like a heat map. All 00:05:15.120 --> 00:05:22.000 of this is possible because the microphones are all in known locations 00:05:19.440 --> 00:05:26.080 relative to both each other as well as to the camera right in the center. So 00:05:23.919 --> 00:05:29.680 sound hits the array and with a little bit of maths, the source can be 00:05:27.680 --> 00:05:33.280 determined by analyzing the reception delays. There are some limitations 00:05:31.520 --> 00:05:39.440 though. Its microphones pick up sound between 2 kHz and 100 kHz, which is well 00:05:37.440 --> 00:05:43.680 above the range of human hearing. So if you wanted to find something like a 00:05:41.039 --> 00:05:49.280 small liquid leak, say in our water cooling loop here, the frequency of a of 00:05:46.400 --> 00:05:54.160 a drip drip drip might actually be too low for the fluke to pick up. 00:05:51.199 --> 00:05:59.400 Fortunately, we have an idea for that. Why not turn our liquid leak into 00:05:59.520 --> 00:06:07.680 an air leak? All we need to do is pressurize our loop. 00:06:05.120 --> 00:06:11.759 Well, that was easy. Oh, you put a hole in the tube. Usually the leaks are a lot 00:06:10.080 --> 00:06:15.440 smaller than that in these water cooling systems. I kind of want to know if it 00:06:13.280 --> 00:06:18.400 can handle a much smaller leak, like uh an, you know, an O-ring issue or 00:06:17.199 --> 00:06:22.960 something. Every once in a while, especially on hardline fittings, these 00:06:20.639 --> 00:06:27.360 O-rings will go missing and the system can look like it's all sealed up, but 00:06:25.919 --> 00:06:32.960 and when that drip gets on the back of your GPU, h you're going to have a 00:06:29.520 --> 00:06:35.600 pretty bad time. Put this up in the blue 00:06:32.960 --> 00:06:40.639 zone. You can see even though there's no O-ring in there, it's sealed right now. 00:06:38.960 --> 00:06:47.199 But let's see if our fluke can pick it up. Oh, dude. 00:06:44.479 --> 00:06:53.919 Easy. And that is a much much smaller leak. 00:06:51.440 --> 00:06:58.080 What a trouble saver that would be. Now, it is worth noting that our brilliant 00:06:55.520 --> 00:07:02.319 strategy of just emptying the loop and pumping it up isn't going to work in 00:07:00.479 --> 00:07:05.840 every situation. Clean rooms, for instance, need to maintain strict 00:07:04.000 --> 00:07:09.120 control of airborne contaminants. And one of the ways that they do this is by 00:07:07.360 --> 00:07:13.039 building them to be fairly airtight and to maintain positive pressure inside. 00:07:11.199 --> 00:07:17.440 Not a ton of pressure cuz like humans need to work in there, uh, but some 00:07:14.880 --> 00:07:22.639 pressure. So, how would they track down then a low pressure leak? Turns out the 00:07:19.759 --> 00:07:27.280 answer is Fluke's Beacon. This little device emits sound at 40,000 hertz, well 00:07:25.520 --> 00:07:32.720 beyond what would bother any human. You just pop it into your system, seal it 00:07:29.680 --> 00:07:34.479 up, tune it to 40,000 hertz, and 00:07:32.720 --> 00:07:38.160 theoretically, we should be able to find our leak. I can definitely detect it 00:07:36.160 --> 00:07:44.479 through the acrylic now, but our system may be too complex for such a short 00:07:41.599 --> 00:07:48.240 wavelength to travel through like that. It's probably just because this knows 00:07:46.720 --> 00:07:51.840 there's sound coming from here or somewhere off the side and so it's not going to pick up the same frequency 00:07:51.360 --> 00:07:55.680 somewhere else. >> Hold it up over there then. >> Oh, dude. Bingo. 00:07:54.560 --> 00:07:59.199 >> You know what? This would be a great opportunity to check the multiple source 00:07:58.319 --> 00:08:05.759 feature here. >> Ooh, yeah. So, in acoustics, we can 00:08:01.759 --> 00:08:08.319 click multiple sources and then Oh, 00:08:05.759 --> 00:08:12.479 dude, that's so cool. It'll show the different levels. However, those still 00:08:10.960 --> 00:08:16.400 totally overpower it. >> Really? >> Yeah. that really doesn't care about 00:08:14.000 --> 00:08:21.520 this leak unless it's the only one on screen. Very interesting. Wonder what 00:08:19.680 --> 00:08:25.840 else we could use this for. A really cool application for acoustic imagers is 00:08:23.919 --> 00:08:31.360 identifying problems with transmission towers at a distance, say from the 00:08:28.879 --> 00:08:37.440 ground. It apparently works at up to 120 m or 393 ft and can be used to detect 00:08:34.880 --> 00:08:40.880 partial discharge or arcing, the sort of thing that could lead to a forest fire. 00:08:39.279 --> 00:08:46.800 Now, we don't have anything like that to test, but we do have some whiny 00:08:43.839 --> 00:08:50.399 electronics like this GPU that has some coil wine. Now, coil wine isn't 00:08:48.720 --> 00:08:55.200 dangerous like those transmission tower problems. It's just vibrations in board 00:08:53.200 --> 00:08:59.279 components, specifically the coils, as electricity passes through. It's just it 00:08:57.519 --> 00:09:02.000 can be pretty flipping annoying. Now, you can see that I'm not picking 00:09:00.640 --> 00:09:06.560 anything up right now. That's because this GPU isn't under load. Let's just 00:09:04.399 --> 00:09:12.880 plug in our keyboard using a handydandy spec cable from ltstore.com and fire up 00:09:09.360 --> 00:09:14.320 a GPU inensive application. There it is. 00:09:12.880 --> 00:09:20.080 But where is it coming from specifically? There's a hot spot right 00:09:17.279 --> 00:09:24.240 where the 12vt 2x6 comes into the back of the board. Yep, it's right across the 00:09:22.320 --> 00:09:27.839 back of the card where all the VRM would be. Doesn't really help us solve the 00:09:25.920 --> 00:09:31.120 problem, but now we know. Actually, I had another situation very recently 00:09:29.680 --> 00:09:36.560 where coil wine was driving me absolutely nuts. I use these USB fiber 00:09:34.000 --> 00:09:41.040 optic docks in order to have my system downstairs, but my monitor and all of my 00:09:38.959 --> 00:09:46.080 peripherals upstairs. I was hearing this like high-pitched almost like a CRT TV 00:09:44.399 --> 00:09:51.680 in the next room whine. I was like, can anyone else hear that? Where's that coming from? And everyone's looking at 00:09:48.240 --> 00:09:53.040 me like I'm crazy. 00:09:51.680 --> 00:09:59.040 I'm having trouble getting it to go lower. >> Change the decibb from automatic to 00:09:57.120 --> 00:10:02.720 manual. It's trouble though when you get into the low frequencies because if you 00:10:01.279 --> 00:10:06.240 talk you can actually see it >> exactly when it's something that you can 00:10:04.320 --> 00:10:09.519 hear like go wine it's always going to be in the frequency that we can hear so 00:10:07.760 --> 00:10:14.480 it can be a little bit of trouble >> right at like 10,000 hertz. That's not 00:10:13.200 --> 00:10:20.399 even that high pitch. Why couldn't anybody else hear this? It was driving me freaking nuts. Like can everyone can 00:10:18.320 --> 00:10:23.600 hear 10,000 hertz, right? Some with hearing loss or significant age- related 00:10:22.160 --> 00:10:30.399 hearing decline might struggle to detect it. >> So you shouldn't have had any problems. Listen. Listen. It's really cool how 00:10:28.240 --> 00:10:35.279 adjusting the frequency band as well as the sensitivity can really help you dial 00:10:32.560 --> 00:10:38.240 it in to detect particular things. Of course, your next question is probably, 00:10:36.560 --> 00:10:42.560 "But what can I do about coil wine liners?" Some people have reported 00:10:40.000 --> 00:10:48.640 success putting over these vibrating components, but your mileage, not to 00:10:45.360 --> 00:10:50.480 mention your warranty, may vary. Is this 00:10:48.640 --> 00:10:55.440 kind of analysis a practical application for such an expensive device? Yeah, 00:10:53.200 --> 00:11:00.800 probably not. But there are more affordable options for this sort of 00:10:57.120 --> 00:11:03.200 thing. The Fluke III500 is about $85,000 00:11:00.800 --> 00:11:06.800 but is just limited to the leak Q functionality that we showed you guys 00:11:04.800 --> 00:11:12.240 earlier with a more limited frequency band. Our II15 has some added functions 00:11:09.760 --> 00:11:16.560 like mech, which detects things like a faulty bearing in an assembly line to 00:11:14.160 --> 00:11:20.320 help identify and fix problems during planned downtime rather than when they 00:11:18.320 --> 00:11:24.079 just come up unexpectedly. But if you want to go even cheaper, you totally 00:11:22.000 --> 00:11:28.959 can. Other options are out there and there are even open- source acoustic 00:11:25.920 --> 00:11:31.120 imagers. The Centenna, a modular system, 00:11:28.959 --> 00:11:35.200 looks like a standout. Of course, that sort of thing means that the calibration 00:11:32.720 --> 00:11:37.839 is entirely up to the user. If you're a business, well, you might not want to 00:11:36.800 --> 00:11:43.519 deal with that. All right. Do you think they're done with the fixes in the shop? How'd all the fixes go? We're thinking 00:11:41.920 --> 00:11:47.040 the regulator was leaking. It may have been the hose, but we swapped both out. 00:11:45.440 --> 00:11:50.160 I wanted to replace the hoses anyway. The connection before was questionable 00:11:48.560 --> 00:11:52.959 at best. Tore the regulator down, rebuilt it, and no leaks at all. 00:11:51.680 --> 00:11:56.320 Sometimes you just have to take it apart and put it back together. 00:11:55.120 --> 00:12:01.120 This one was a little bit different. I knew that there was a leak in here cuz this wasn't tight enough, but I 00:11:59.360 --> 00:12:04.640 discovered that there was also an O-ring missing in her fasteners. Just 00:12:03.279 --> 00:12:08.079 completely missing. >> Yeah. I was actually quite surprised 00:12:06.480 --> 00:12:12.320 that it wasn't leaking worse than it was, but we ended up getting it put back 00:12:10.399 --> 00:12:16.720 in and now it's good and tight. >> So, that was never in there. 00:12:14.079 --> 00:12:21.440 >> No. This whole time? >> Yeah. 00:12:19.120 --> 00:12:26.160 >> So, it was only the two leaks. That was it. Well, then what is this that I'm 00:12:23.600 --> 00:12:31.440 picking up right now? I knew it. It's another segue to our 00:12:29.600 --> 00:12:38.000 sponsor. Wow. Wow. Kind of rude, but it's okay. I 00:12:35.920 --> 00:12:42.160 can take it. Hey, Johnny Thunder here from Johnny's War Machine Emporium. If 00:12:40.320 --> 00:12:45.680 you're looking for a gently used fighting vehicle that's seen actual 00:12:44.160 --> 00:12:49.839 combat in War Thunder, the most comprehensive vehicle combat game ever 00:12:47.200 --> 00:12:54.480 made, come on down and join our 95 million satisfied customers from Rancho 00:12:52.079 --> 00:13:00.240 Cucamonga to Calamazoo. We're talking choppers, whoppers, showstoppers, mom 00:12:58.000 --> 00:13:06.399 and poppers, ornithopters if they were real. Not yet. Oh. Oh, you want boats? 00:13:02.560 --> 00:13:09.120 How about goats? USS Alaska. Japanese 00:13:06.399 --> 00:13:12.720 Yamato. Hya. Biggest battleship ever built. War Thunder might be free to 00:13:10.800 --> 00:13:17.120 play, but this ain't no game to me, okay? This is my life now. Hey, hey, if 00:13:15.120 --> 00:13:20.800 you want silly little boxes filled with simple hit points, you can head right to 00:13:18.880 --> 00:13:25.680 Rusty Richard's Lot down the road, okay? And he'd be happy to see you. What I got 00:13:23.040 --> 00:13:29.839 here is premium pre-owned instruments of death, meticulously detailed inside and 00:13:27.920 --> 00:13:34.720 out. Every bit of damage affects components, not a big green bar. Mama 00:13:32.320 --> 00:13:38.560 used to say, "Damage is as damage does." Everyone seems to forget that. See, when 00:13:36.720 --> 00:13:42.079 you get behind one of Johnny's French Jaguar A's, you're going to be fully 00:13:40.639 --> 00:13:45.920 immersed. All right, these combat vehicles run on petrol, diesel, Xbox, 00:13:44.000 --> 00:13:49.680 PlayStation, PC, mobile, your mom, you name it. Use our link for extra 00:13:47.279 --> 00:13:54.440 vehicles, packs, and more. Or don't. It'd be your loss anyway. See if I care. 00:13:55.040 --> 00:14:02.399 If you guys enjoyed this video, why not check out the one we did on a Caterpillar branded smartphone that has 00:14:00.399 --> 00:14:05.120 a Fleer thermal camera built into it. That was pretty cool.