WEBVTT 00:00:00.000 --> 00:00:06.960 Hey MTV, welcome to my crib. Oh, I mean, sorry, prong video. Hey Flow Play, this is Jake here in the lab, and today 00:00:06.960 --> 00:00:13.640 I'm doing a deep dive on our new Raymond EMC supplied RF chamber. Raymond EMC, these guys huge shout out once again 00:00:13.920 --> 00:00:18.800 They provided this whole thing for us, and they were a huge help with getting some of the equipment that we need 00:00:19.200 --> 00:00:22.360 And getting everything started. So without further ado 00:00:24.240 --> 00:00:26.580 The most satisfying door in this place 00:00:27.340 --> 00:00:34.220 This is the Raymond EMC quiet door, and I know Linus showed this off before 00:00:34.580 --> 00:00:39.300 But this is a knife edge that goes into some brilliant copper finger stock 00:00:39.900 --> 00:00:43.700 And I've seen a lot of people asking questions around, you know 00:00:43.900 --> 00:00:52.860 What's the maintenance like on on this kind of thing? Especially because we reference that one of the screws is actually out of torque and that caused a leakage at 18 gigahertz 00:00:53.020 --> 00:00:55.980 So with this we actually have a duty cycle 00:00:57.140 --> 00:01:02.660 Every hundred times this door is open and closed. We've got to wipe all the contact surfaces around the outside 00:01:03.380 --> 00:01:08.380 Every thousand cycles we have to inspect and lubricate all the moving parts and every 5,000 cycles 00:01:08.380 --> 00:01:12.260 We actually have to polish all the contact surfaces and check the door alignment 00:01:12.900 --> 00:01:18.980 Raymond EMC is actually we coming back out to film a maintenance video for this that they'll have on their channel and 00:01:19.940 --> 00:01:24.500 To do this basically we use methyl spirits and a brillo pad. So it's nothing too crazy 00:01:24.980 --> 00:01:31.060 But it does present a bit of a challenge for Height challenged individuals like Linus to get up to the top of the door 00:01:31.340 --> 00:01:38.900 so Another thing that people have been asking is, you know, what are we doing with the electromagnetic interference and EMI off the cameras? 00:01:39.180 --> 00:01:44.660 Etc. And even the lighting so one thing we're not going to be filming the tests 00:01:44.780 --> 00:01:50.080 We will film recreations afterwards, but we're going to use the actual clean test data. So with the lighting 00:01:50.560 --> 00:01:57.160 EMI is actually a big consideration. Thankfully Constellation lighting down in the States. They make these link 00:01:57.800 --> 00:02:00.600 LED systems and these are mil spec 00:02:00.960 --> 00:02:05.960 So they're military standard 461g tested for zero EMI emission 00:02:06.400 --> 00:02:13.920 They're 65 watt 7,000 lumen and we can actually position on two axes so we can kind of adjust lighting in here 00:02:14.640 --> 00:02:18.520 That makes it a lot easier to see because when the lights are out and the doors close 00:02:19.520 --> 00:02:27.240 It doesn't work so well Over here, these are our negatives for our floor panels. They're actually 00:02:28.160 --> 00:02:33.880 Really stable. So we're gonna have our test set up over here. We're gonna be broadcasting on our fire cell 00:02:34.320 --> 00:02:37.520 Lab kit that's gonna send signal over in that direction 00:02:37.520 --> 00:02:43.760 And we're gonna have our device under test on a pedestal there. Ignore the big blue box. It will not exist and we're all ready 00:02:45.320 --> 00:02:50.000 effectively One of the considerations that we have when we're broadcasting 00:02:50.400 --> 00:02:56.720 RF is that we can't broadcast too close to the device under test and we can't broadcast it to a high of a power rating 00:02:57.080 --> 00:03:05.360 Mainly because you can actually burn out equipment that way So with the lab kit, we're gonna have it positioned about a meter off the wall 00:03:05.440 --> 00:03:13.440 And we're gonna have the device under test But the same and the reason for that is if your wavelength is too long so too low of a frequency 00:03:14.040 --> 00:03:16.840 If you're too close, it can cause weirdness 00:03:17.320 --> 00:03:22.040 They don't call RF black magic for nothing. So we're gonna have it positioned off the wall 00:03:22.040 --> 00:03:24.840 We're gonna have about three to four meters in between 00:03:25.240 --> 00:03:28.440 And then we'll be able to measure different transmission power levels 00:03:28.440 --> 00:03:33.440 I know for for your people Arjun, and I'm gonna say that 00:03:34.200 --> 00:03:36.200 in Australia 00:03:38.040 --> 00:03:42.840 In Australia these LTE band 28 which is 700 megahertz. It's a fairly low band 00:03:43.480 --> 00:03:47.000 And we've had a couple people ask if we're gonna include them testing 00:03:47.000 --> 00:03:51.920 I am gonna go through every LTE frequency band. So there will be data on that. Oh, yeah 00:03:52.400 --> 00:03:58.200 And one of the really cool things to consider is okay. Well, if you're on a network that's utilizing a 00:03:59.000 --> 00:04:03.600 Lower end of the frequency spectrum. You can actually get away with smaller batteries 00:04:04.480 --> 00:04:07.720 Just for the simple fact that doesn't draw as much power 00:04:07.840 --> 00:04:11.160 So yeah, we're we're gonna be doing some really cool tests on that 00:04:11.560 --> 00:04:17.240 I'm meeting with some reps from a couple of the different manufacturers of the spectrum analyzers 00:04:17.920 --> 00:04:25.480 So we're looking into Possibly a rodent Schwartz like an FSW 43 or maybe one of the Gauss 00:04:27.320 --> 00:04:36.480 One of the G line probably But basically we need a 43 gigahertz spectrum analyzer and that's gonna come down to you know cost 00:04:37.760 --> 00:04:40.160 line ass or 00:04:41.160 --> 00:04:45.360 In general ease of use, right? I want to get everything running as quickly as possible 00:04:45.760 --> 00:04:48.880 So we're gonna try and hit the ground running once everything comes in 00:04:49.320 --> 00:04:54.560 In the meantime before we get that though, we do have our lab kit fire cell. It has arrived 00:04:54.560 --> 00:04:56.560 It's actually outside if you guys want to go 00:04:58.520 --> 00:05:04.760 So I know it doesn't look like much but in these three boxes here, we actually have our 5g and 00:05:05.160 --> 00:05:09.080 Our 4g radio as well as the actual main compute unit 00:05:10.320 --> 00:05:14.760 These are are really cool These are 00:05:14.760 --> 00:05:16.760 this is the actual 00:05:18.040 --> 00:05:24.040 Network broadcasts or the SDR. So this is the edis research. It's a usrp b210 00:05:24.160 --> 00:05:26.600 So it's 70 megahertz to six gigahertz 00:05:27.280 --> 00:05:32.080 Right now. We can't do millimeter wave That upgrade is coming out early next year 00:05:32.680 --> 00:05:37.200 In the meantime, however, this is gonna get us kind of 80% of the way there 00:05:37.520 --> 00:05:42.800 So we've got our 5g. We have our 4g in here. So we're getting the upgraded millimeter wave 5g 00:05:43.480 --> 00:05:49.680 Early next year for our lab kit. We might actually add some equipment in the meantime that will kind of bridge that 00:05:50.240 --> 00:05:53.640 I'm actually talking to to one of our viewers. Hopefully next week 00:05:53.920 --> 00:05:57.680 Who is in this space and and has some advice for us? 00:05:57.680 --> 00:06:00.400 So we're gonna listen and make sure that you know, we're not barking up the wrong tree 00:06:00.960 --> 00:06:06.760 and then outside of that You know, this is running our network, but we also need to you know 00:06:07.200 --> 00:06:14.560 Have something in there to record tests. So for that, we went with this sira7. It's a completely silent fanless 00:06:15.040 --> 00:06:17.600 solid-state PC. It's totally passively cooled 00:06:18.560 --> 00:06:21.680 You can tell because it's basically a giant aluminium heatsink 00:06:22.880 --> 00:06:28.960 And the reason we went with a solid-state PC is just because there's no fans and fans can generate EMI 00:06:29.120 --> 00:06:33.440 So we want to try and minimize that as much as possible. So that's why we wanted that 00:06:34.160 --> 00:06:39.120 It's a decent little AMD Ryzen 5. It's a 5000 series 00:06:39.840 --> 00:06:44.960 CPU in there Nothing crazy, but we don't really need anything too high spec. So yeah 00:06:45.280 --> 00:06:50.560 The other really cool thing is about how we get our network inside the box because obviously can't use wi-fi 00:06:52.000 --> 00:06:55.760 And it's this wave guide now one of the really really interesting things with this 00:06:56.480 --> 00:07:03.680 Is that this is actually a plug And you notice it has this really really handy late warning label 00:07:04.400 --> 00:07:07.280 saying Tempest hazard with plug removed 00:07:08.560 --> 00:07:11.760 Now for those of you that don't know what a tempest attack is 00:07:13.040 --> 00:07:18.880 It's effectively a way to use rf to eavesdrop on displays. It's really cool 00:07:19.440 --> 00:07:24.960 It's not new But there's not really a huge amount of people kind of in the public space that talk about it 00:07:25.520 --> 00:07:30.160 Effectively you could be in another room or on the other side of a hotel wall 00:07:31.200 --> 00:07:35.200 Broadcast a signal and you can pick up what's going on 00:07:35.760 --> 00:07:40.400 On a display and you can set one of these attacks up for like 300 to 400 dollars 00:07:40.960 --> 00:07:45.440 If you plus a laptop So one of the things that I really want to do 00:07:45.840 --> 00:07:52.720 Possibly for a flow plane exclusive. We'll see is actually run a tempest attack online us in the chamber 00:07:53.600 --> 00:08:01.120 Come on we can leak the stream keys again. We can we can find out his emails. We you know everything 00:08:01.120 --> 00:08:05.760 It'll be great his life easier, right? So yeah, so if that plug is removed 00:08:06.880 --> 00:08:11.280 Boom, it's wide open Yeah, so I think that'll be pretty cool 00:08:11.600 --> 00:08:14.960 And finally one of the things that I've seen quite a few people ask is 00:08:15.440 --> 00:08:19.760 You know if one screw or bolt out of torque can cause a leak 00:08:20.320 --> 00:08:29.040 You know how a how do we maintain that? Um and b why didn't we just weld everything together? Well, these chambers are actually designed to be modular 00:08:29.120 --> 00:08:31.920 We can extend this if we need to say we want to 00:08:32.480 --> 00:08:36.480 Turn it into a three meter chamber and do full iso compliance tests 00:08:37.120 --> 00:08:39.600 We can do that. We get raymond emc to come back out 00:08:40.160 --> 00:08:43.760 We re-level this area of the floor and they can extend this as much as we want 00:08:44.320 --> 00:08:47.600 So it's designed to be modular and then 00:08:47.600 --> 00:08:52.960 Secondly as far as testing to make sure that you know everything's within spec that comes down to the duty cycle 00:08:53.280 --> 00:08:58.160 When we're inspecting any of the moving parts and things And we can actually do signal generator 00:08:59.040 --> 00:09:04.480 Via spectrum analyzer tests so we can have it broadcasting inside and we can go around and we can measure the outside 00:09:05.120 --> 00:09:12.400 It's not something we're going to have to do often But we'll probably end up doing it probably once every four to six months 00:09:12.960 --> 00:09:17.680 Just to make sure everything's working well I might even do a spot check once a month 00:09:18.320 --> 00:09:23.360 Just in a couple different places mainly around like the moving parts make sure the door has good contact 00:09:24.400 --> 00:09:28.400 Near the electrical maybe but yeah, it's not really a huge concern 00:09:28.640 --> 00:09:34.800 But thank you guys for bringing it up Now traditionally a lot of the foam in here would be held on by adhesive 00:09:35.360 --> 00:09:42.560 It's a bit of a pain Not just installed but you know someone falls against the foam or tip breaks off and you want to replace it 00:09:42.960 --> 00:09:45.680 So tdk came up with something really cool 00:09:50.480 --> 00:09:57.280 Their magnets we can actually pull off these panels if we wanted to introduce reflection 00:09:57.920 --> 00:10:01.600 So if we wanted to you know get the signal bouncing real good in here 00:10:02.080 --> 00:10:06.400 Take off a row of these and yeah, you can introduce some really interesting artifacts 00:10:07.200 --> 00:10:10.880 If we need to replace them, we just unscrew the magnets at the back put the new foam 00:10:11.120 --> 00:10:15.680 And pop it back in so yeah, it's a it's a really ingenious 00:10:16.640 --> 00:10:22.960 Installation process and it makes life a lot easier to maintain. That's so satisfying. Yeah 00:10:24.000 --> 00:10:26.800 the only the big pain in the ass with that though 00:10:28.880 --> 00:10:34.560 Was that When the electrician was on the roof wiring everything up 00:10:36.640 --> 00:10:40.160 The thud of him knocked one of these panels 00:10:41.840 --> 00:10:44.240 Up here off right 00:10:45.600 --> 00:10:47.600 Well, I'm six three 00:10:48.720 --> 00:10:52.640 And I can't reach yeah, and I can't put a ladder in here. Yeah 00:10:54.720 --> 00:11:00.960 So, uh, yeah unless it completely falls off you can't no it did. Oh, I did fall. Okay. Okay. It's stuck 00:11:00.960 --> 00:11:04.720 Oh, I guess you still can't be into the answer. Yeah. Yeah, I had to grab a broom 00:11:06.160 --> 00:11:09.840 and I had to grab a mop like a 00:11:09.840 --> 00:11:12.000 One of the mobs that has like the removable 00:11:13.120 --> 00:11:19.120 She or like cloth over it and I had to use two of those to like push it back up onto the ceiling