WEBVTT

00:00:00.440 --> 00:00:07.439
this is my car this is my wife's car and this right here is the power supply

00:00:05.359 --> 00:00:13.519
tester I just bought from chroma system solutions that cost more than both of

00:00:09.719 --> 00:00:15.240
them combined why am I smiling because

00:00:13.519 --> 00:00:19.199
with this I'm going to be able to separate the wheat from the chaff the

00:00:17.520 --> 00:00:23.680
good from the bad we're going to be able to take power supplies like this and

00:00:21.320 --> 00:00:27.800
blow them the freak up but of course in order to do that we're going to need to

00:00:25.840 --> 00:00:31.960
open it up aren't we which is going to involve some uh heavy

00:00:32.440 --> 00:00:37.760
tools oh I'm not supposed to do that you

00:00:35.840 --> 00:00:41.680
can see he's genuinely upset I've been 6 months researching this what are you

00:00:39.320 --> 00:00:46.399
doing man we're going to have some fun today and it's going to be sponsored by

00:00:44.480 --> 00:00:52.280
Future lus here didn't get the memo till we were done filming but apparently this

00:00:48.199 --> 00:00:56.000
video is sponsored by C Sonic imagine

00:00:52.280 --> 00:00:57.960
having the stones the Swagger to sponsor

00:00:56.000 --> 00:01:01.640
content about a piece of testing equipment that has the potential to to

00:01:00.079 --> 00:01:05.600
make your products look bad compared to the competition clearly they've got a

00:01:04.000 --> 00:01:09.960
lot of confidence in their products and you can check out seic power supplies at

00:01:07.439 --> 00:01:09.960
the link down

00:01:16.560 --> 00:01:23.560
below you swear on your mother's life that this is the good shiz right that's

00:01:20.759 --> 00:01:27.040
what C Sonic said they help me spec this and then I basically just got their spec

00:01:25.159 --> 00:01:30.880
just the newer stuff the newer stuff so we have better power supply testing

00:01:29.119 --> 00:01:35.119
equipment than a Tier 1 power supply manufacturer I'm assuming they've

00:01:32.200 --> 00:01:38.280
upgraded since we talked so maybe how strong are your wrists pretty good I

00:01:37.079 --> 00:01:41.920
didn't bring an impact driver and there's a lot of screws up there L LTT

00:01:40.399 --> 00:01:47.799
store.com you don't even have like a power anything no just use your guns man

00:01:45.640 --> 00:01:53.079
oh my God I love our screwdriver but these are a little long to do by hand

00:01:49.680 --> 00:01:54.920
there's like 25 of them it's all about

00:01:53.079 --> 00:01:59.159
right tool for the job right how's your ARM man I almost beat you I got eight of

00:01:56.840 --> 00:02:02.920
them out already dude we couldn't find impact drive is it a drill let me load

00:02:01.320 --> 00:02:06.560
up an impact driver and then start working on these side

00:02:04.399 --> 00:02:12.680
ones oh that didn't work waa this is looking real scientific

00:02:10.280 --> 00:02:18.760
holy okay come here this is really cool now it's all fine and good to put

00:02:15.120 --> 00:02:21.040
stickers on your package fragile this

00:02:18.760 --> 00:02:26.720
side up we all know what couriers do they see that they go let's play point

00:02:23.920 --> 00:02:31.200
the arrow down right well this guarantees they never did it if this

00:02:28.840 --> 00:02:35.959
indicator is red make a notation of delivery receipt blah blah blah see

00:02:33.160 --> 00:02:40.720
watch boom perar red that means it stayed right side up the whole time it

00:02:38.239 --> 00:02:44.640
was in transit super cool right now that he's crumpled up the serial number in

00:02:42.599 --> 00:02:47.360
theory these have serial numbers on them so that you know they weren't switched

00:02:45.800 --> 00:02:52.840
in transit is this open now hold on hold on

00:02:50.760 --> 00:03:01.319
hold on wait okay no it's okay it's fine we got no we got to get the moment no I

00:02:54.560 --> 00:03:01.319
know I know it's one two 3 Let's Go

00:03:01.920 --> 00:03:06.440
come on well fudge here we

00:03:09.440 --> 00:03:16.599
go well that was a little anticlimactic check it out there's another one inside

00:03:14.239 --> 00:03:21.560
come on David check so this indicates not only did it stay right side up but

00:03:19.519 --> 00:03:25.959
it didn't tip because you can see the sand will like go to one side or

00:03:24.120 --> 00:03:33.439
whatever test instrumentation and automated test systems power conversion

00:03:28.840 --> 00:03:34.959
wow free us b key uh no no that's I

00:03:33.439 --> 00:03:38.959
think that's the software key sofware yeah I don't know if it if it's true in

00:03:37.439 --> 00:03:42.640
the new software we're getting but the old software that's the license key if

00:03:40.760 --> 00:03:46.400
you lose that it's 70 grand replace it 70 grand I'm pulling the number up my

00:03:44.439 --> 00:03:51.040
butt but it's it's expens don't lose that no no no no no no put put it back

00:03:48.560 --> 00:03:57.720
how much power can this whole thing draw or does it the whole thing draw so

00:03:54.560 --> 00:04:00.760
208 3phase at 50 amps is what it's rated

00:03:57.720 --> 00:04:02.959
for what the that's what

00:04:00.760 --> 00:04:09.159
Watt that's what the ca's rated for this is rated for 10,000 Watts uh maybe um I

00:04:07.560 --> 00:04:13.280
haven't done the conversion in my head but we can only draw the maximum that

00:04:10.760 --> 00:04:17.840
the source will allow and that's 3,500 Watts 3,500 Watts that means that we'll

00:04:15.480 --> 00:04:21.880
be able to test basically any power supply on the market because anything

00:04:19.799 --> 00:04:26.840
over about 2,000 watts is going to trip a North American breaker and anything

00:04:24.280 --> 00:04:30.080
over about what is it like 2500 3,000 WTS that's going to trip a European

00:04:28.400 --> 00:04:34.479
breaker anyway we're on the limit of what Europe can do oh this is really

00:04:32.240 --> 00:04:41.840
cool you can see from the plug we've got a ground connection to this gigantic

00:04:38.759 --> 00:04:44.560
bare metal strip here and then unlike a

00:04:41.840 --> 00:04:48.560
normal server shelf these are bare metal as well and they're screwed in like this

00:04:45.880 --> 00:04:54.520
what that means is that every single shelf every piece of metal in this

00:04:50.600 --> 00:04:56.479
entire enclosure is grounded so in a

00:04:54.520 --> 00:05:01.960
production environment that's important because well if a if a live Le would is

00:05:00.039 --> 00:05:05.479
short to the chassis yeah and an operator was touching it you want to

00:05:03.240 --> 00:05:10.039
make sure the operator isn't the lowest path to ground right so we'll have all

00:05:07.560 --> 00:05:13.680
of these green wires everywhere right and even on the powder coated stuff

00:05:11.639 --> 00:05:17.000
it'll go straight through they'll powder coat afterwards and make sure the stud

00:05:15.039 --> 00:05:22.039
touches the actual metal that's also why the whole thing's enclosed locked safety

00:05:19.960 --> 00:05:26.240
labels the whole shebang okay there's shall we take off

00:05:24.680 --> 00:05:29.520
there's a monitor yeah there's a computer there's a computer on here

00:05:27.680 --> 00:05:34.520
comes with a little shelf and everything yeah look at that I asked for the Shelf

00:05:31.960 --> 00:05:38.440
oh okay how much did I pay for the Shelf $130,000 no no I mean just I have no

00:05:36.720 --> 00:05:42.080
idea you didn't even ask what the different options are worth or anything

00:05:40.280 --> 00:05:46.840
like to decide if we could have provided our own shelf maybe like how much should

00:05:43.600 --> 00:05:48.840
I pay for this Logitech mouse oh not

00:05:46.840 --> 00:05:53.600
even a rechargeable battery in there brother is there a computer did I pay

00:05:51.360 --> 00:05:57.680
for a computer yes it's industrial grade Dell computer oh my god oh look it comes

00:05:55.840 --> 00:06:02.360
with the key what is with all of these discs why do we have discs you even

00:06:00.120 --> 00:06:08.560
included a USB key are they just all the same thing or what like oh my God Kyle

00:06:06.680 --> 00:06:14.039
yes they don't expect us to build our own bloody cable harnesses here do they

00:06:10.520 --> 00:06:16.680
those are like 24 pin ATX pins and

00:06:14.039 --> 00:06:20.880
here's a 24 pin ATX connector do I have to like build my own DIY power supply

00:06:19.400 --> 00:06:25.560
here or what why are you telling me this like you're the one who's actually going to be building this stuff well do you

00:06:24.199 --> 00:06:30.080
understand how configurable this machine is trust me you want to build youring

00:06:27.919 --> 00:06:35.240
cables you see all of these ones yeah you see how it says load 2 3 4 5 6

00:06:33.039 --> 00:06:41.599
7even yeah those are these guys okay so if these like overcurrent or something

00:06:37.840 --> 00:06:43.800
then would they blow their load no

00:06:41.599 --> 00:06:47.919
no you know what we'll go through it once we actually do a rundown of each of

00:06:45.639 --> 00:06:51.720
the modules um and I'll explain why those are important yeah check it out I

00:06:50.080 --> 00:06:56.080
made a whole thing oh they actually connect together that's Prett cool yeah

00:06:53.680 --> 00:07:02.000
that's not in but yeah that's what she said I think we should go bottom to top

00:07:00.080 --> 00:07:08.560
reason being because that's where power comes in so this is a

00:07:04.759 --> 00:07:11.240
3,000 volt amp AC so we'll be able to

00:07:08.560 --> 00:07:15.720
test any PC power supply on the market and realistically just about any other

00:07:13.560 --> 00:07:19.879
power supply as well so this is also a three-phase power supply it can do both

00:07:17.680 --> 00:07:23.199
single and three-phase but we only have it hooked up for single phase right now

00:07:21.560 --> 00:07:27.360
because we're only targeting adx power supplies eventually we might do service

00:07:25.199 --> 00:07:32.919
stuff but that's in the future what's cool about this is it can do any

00:07:29.560 --> 00:07:36.240
anything with the power so it can do 240

00:07:32.919 --> 00:07:38.080
volts 50 HZ for European Brothers it can

00:07:36.240 --> 00:07:43.479
do 11060 or 1260 for North America it can also start

00:07:41.440 --> 00:07:47.840
injecting like yeah can we simulate dirty power dirty power power like bad

00:07:45.879 --> 00:07:53.120
power so we can we can Brown out the power we can overvolt the power we can

00:07:50.599 --> 00:07:56.919
inject harmonics into the actual AC sine wave like if you had a blender or some

00:07:54.879 --> 00:08:02.520
crap running Elsewhere on the circuit like a very noisy like motor somewhere

00:08:00.080 --> 00:08:06.199
that like feeding very like dirty a blender yeah or microwave or

00:08:04.720 --> 00:08:11.360
something on the there's no motor in a microwave except it's been the turntable that's probably not the problem yeah

00:08:08.720 --> 00:08:16.319
probably the worst thing to compare it to but it it it can do that and it can

00:08:14.199 --> 00:08:20.840
do that all automatic thing what does this do all right so that's an onoff

00:08:18.759 --> 00:08:25.479
controller it turns stuff on and off um oh my god really that's all it does uh

00:08:23.520 --> 00:08:30.039
yeah what's this de ShortCircuit protection and ovp tester so basically

00:08:28.199 --> 00:08:34.719
that will short out the power Supply the most fun test ever as well as it will

00:08:32.599 --> 00:08:39.440
overvolt the power supply so it takes this is one of our likely fire starters

00:08:36.839 --> 00:08:44.159
then right here maybe so what it does is it'll take voltage from the DC source

00:08:41.399 --> 00:08:47.800
and add it to the output from the power supply so if you're getting 12 volts out

00:08:46.160 --> 00:08:51.920
it will just like add a couple voltage to that and see if the power supply holy

00:08:50.760 --> 00:08:55.760
crap this is not supposed to be happening and and cut and cuts it yeah

00:08:54.240 --> 00:08:59.120
right you claim this is where things get interesting as we were talking about

00:08:57.440 --> 00:09:03.600
before this is the Anderson connection panel these all hook into these loads

00:09:01.360 --> 00:09:09.160
which we'll get to in a minute and that's basically the output of the AC

00:09:07.200 --> 00:09:15.120
source so that's where we plug power supplies into got it okay so after after

00:09:12.600 --> 00:09:22.440
adapting it from that to normal plug right so our AC source is as clean or

00:09:17.959 --> 00:09:25.680
dirty as we Speck it to be and this is

00:09:22.440 --> 00:09:27.680
effectively your wall outlet at home but

00:09:25.680 --> 00:09:32.160
you know with an adap well so the idea with this is to actually have a test

00:09:29.519 --> 00:09:37.240
enclosure that we put down here and plug into this or we even make a back plane

00:09:34.560 --> 00:09:42.360
that just slots in right cuz these are all very precisely put in place right so

00:09:40.200 --> 00:09:47.399
we could easily have like a PCB with Anderson connectors the PCB Mount ones

00:09:44.680 --> 00:09:52.440
not regular ones sure that we could just slide in and there's power these are all

00:09:50.079 --> 00:09:58.720
your signal stuff that goes up to um this guy and this guy yeah and these are

00:09:55.240 --> 00:10:00.760
all these and um these channels I

00:09:58.720 --> 00:10:05.519
believe are temperature sensing channels that is a byproduct of this guy Okay

00:10:03.480 --> 00:10:10.680
cool so this is basically the business end where we plug the victims in and the

00:10:08.200 --> 00:10:14.480
idea is that by having an enclosure here we can control the environment that the

00:10:12.399 --> 00:10:20.320
power supply is operating in so that we could run any number of different

00:10:16.839 --> 00:10:23.279
permutations clean power dirty power

00:10:20.320 --> 00:10:28.480
cold hot there's a more practical thing for an enclosure so when

00:10:26.240 --> 00:10:33.760
power supplies blow up we don't yeah that too this this yeah what's this

00:10:30.600 --> 00:10:35.680
point uh so it's a timing noise analyzer

00:10:33.760 --> 00:10:41.079
the module has a whole bunch of slots that you can stick into it what these do

00:10:38.040 --> 00:10:44.680
is Monitor all the signals right so if

00:10:41.079 --> 00:10:47.519
you have 12 volts 5 Vols and 3.3 this is

00:10:44.680 --> 00:10:51.839
monitoring it from a past fail event so we don't actually get data out of this

00:10:49.600 --> 00:10:56.600
so you set limits and this will see if it breaks those limits got it well where

00:10:53.920 --> 00:11:01.320
do we get data from then uh this guy okay oh and that's missing that's the

00:10:58.600 --> 00:11:04.480
oscilloscope I see they allow a variety of third party oscilloscopes we haven't

00:11:02.839 --> 00:11:08.399
decided on one yet but that's going to go there a four Channel oscilloscope and

00:11:06.480 --> 00:11:12.200
that will give you pretty graphs and like actual things that people can

00:11:10.639 --> 00:11:15.680
understand where this will give you the past fail data now here's something I

00:11:14.040 --> 00:11:20.880
need to understand a little bit better if our were l oh wait no I get it oh my

00:11:18.959 --> 00:11:28.279
God why do we need so many of them okay cool I spent the majority of time like

00:11:24.600 --> 00:11:30.079
gaming the system to get us this the

00:11:28.279 --> 00:11:35.720
idea of these is we have 400 WTS 400 wats 400 watts

00:11:33.040 --> 00:11:44.200
400 wat 400 4 4 400 these are eight 400 WT loads and one dual 100

00:11:41.639 --> 00:11:48.600
watt load so these Can it can do 200 Watts total but 100 on each channel the

00:11:46.760 --> 00:11:53.279
idea is this is for your 5 volts and your 3.3 most power supplies are under

00:11:51.200 --> 00:11:57.160
100 Watts on those rails there are a few that are above only older stuff the

00:11:55.560 --> 00:12:02.440
newest stuff the 1600 watts can sometimes do 125 these loads can be

00:11:59.720 --> 00:12:06.560
connected in parallel right so you can set all of these guys to 12 vs the

00:12:05.000 --> 00:12:11.240
reason why I went with so many is something called slew rate so that's the

00:12:08.800 --> 00:12:17.040
rate at which these loads can change their demand right okay so each one of

00:12:14.199 --> 00:12:23.519
these loads can change the demand at 6 amps per microc so you take 6 amps you

00:12:19.560 --> 00:12:25.199
multiply by 12 you get uh 60 72 so 72

00:12:23.519 --> 00:12:30.480
watts each one of these can change their load at 72 watts per micros with all

00:12:28.160 --> 00:12:37.880
eight of these loads we can change the the the rate of power draw at 760 watts

00:12:34.560 --> 00:12:41.360
per microc okay the idea was to

00:12:37.880 --> 00:12:44.079
electronically simulate 3090 kicking on

00:12:41.360 --> 00:12:48.600
but then NVIDIA released 4090 and I was like I don't have enough space so the

00:12:46.720 --> 00:12:51.320
future goal is just to have a test bench right next to it where you can just

00:12:50.079 --> 00:12:56.440
literally hook the power supply instead of hooking into this hook it into an actual computer and simulate a real

00:12:54.880 --> 00:12:59.480
world example well think about that won't be as repeatable though no but

00:12:58.360 --> 00:13:05.240
think about it this way someone's complaining about a certain power supply blowing up under certain architecture

00:13:02.880 --> 00:13:09.480
you can build that architecture get the power supply put the power supply down

00:13:07.600 --> 00:13:13.920
and monitor the voltages as it's and then we could create a simulated load

00:13:12.079 --> 00:13:17.399
based on what we monitored and find out if anything will blow up you hide a guy

00:13:15.480 --> 00:13:22.320
to automate GPU testing just get him to get a script that cranks it to like 150%

00:13:20.480 --> 00:13:26.199
and just sends it as we get more experience using this I'm assuming we're

00:13:24.160 --> 00:13:31.399
going to come up with our own stuff cool and then what do these do so uh single

00:13:28.880 --> 00:13:34.880
phase power meter Yeah so basically efficiency so if we want if we want to

00:13:33.760 --> 00:13:40.440
test uh 80 whatever or maybe something more

00:13:36.920 --> 00:13:42.920
meaningful like 90 plus yeah sure okay

00:13:40.440 --> 00:13:46.360
so that that just monitors your power through and then this just looks like a

00:13:44.560 --> 00:13:50.240
basic ass multimeter yeah digital multimeter so it's got It's got a um if

00:13:48.440 --> 00:13:54.000
we have an oscilloscope do we need this there's a reason why we need it I've

00:13:51.639 --> 00:13:56.360
forgotten it um I'm going to blame I'm going to blame food again you know what

00:13:55.480 --> 00:14:02.560
I've forgotten our sponsor see Sonic man

00:14:00.160 --> 00:14:06.680
those guys never cease to amaze me great people great power supplies I mean

00:14:05.199 --> 00:14:11.240
clearly because if they didn't have the confidence to know this thing is going to make them look good rather than bad

00:14:10.000 --> 00:14:14.880
why would they have gotten involved in helping us Speck it this is like I said

00:14:13.360 --> 00:14:19.320
very similar to the unit they have although ours is a little bit newer

00:14:17.079 --> 00:14:22.959
unless they upgrade which means that if they're doing a great job of Designing

00:14:20.800 --> 00:14:27.320
their power supplies there shouldn't be any surprises so go check them out at

00:14:25.440 --> 00:14:32.079
the link down in the video description they've got long warranties great

00:14:29.120 --> 00:14:35.480
efficiency just they're quiet what what else can you ask for if you guys think

00:14:33.959 --> 00:14:40.320
Kyle and I are brutal with each other you wouldn't believe what doesn't make it into the videos we just uploaded an

00:14:38.399 --> 00:14:47.560
exclusive on Floatplane.com of me and Jake and it gets pretty spicy BMWs are

00:14:43.519 --> 00:14:47.560
bad cars Jake come at me