WEBVTT

00:00:00.000 --> 00:00:07.080
Pretty much everyone and their dog has a side hustle these days and dog walking or uber driving might work for some people

00:00:07.080 --> 00:00:10.440
But I'm better with tech than I am with organisms

00:00:10.880 --> 00:00:14.760
What's a socially awkward penguin to do? Hang on? I've got it

00:00:14.800 --> 00:00:20.160
There's all this broken tech all over sites like eBay and Craigslist

00:00:20.760 --> 00:00:24.760
Maybe I could fix it up and flip it for profit

00:00:25.480 --> 00:00:30.560
And where to start I mean I used to be pretty good at fixing bent pins

00:00:30.960 --> 00:00:38.320
So here's what we're gonna do. We're gonna buy some four parts am4 processors that don't look too far gone

00:00:38.480 --> 00:00:46.160
We're gonna set a timer fix them and find out how much we can pay ourselves to do this kind of work with minimal tools

00:00:47.440 --> 00:00:52.760
How about this one this one this one this one this one this one this one this one and this one

00:00:54.840 --> 00:01:03.520
Well, I mean they're not here yet But while we wait, how about we do this segue to our sponsor before we could try to fix anything

00:01:03.840 --> 00:01:10.120
Something already went wrong. We ordered a total of six CPUs, but as you can see, I'm holding only five

00:01:10.640 --> 00:01:14.720
Thankfully the one that didn't show up was canceled and refunded by the seller

00:01:14.720 --> 00:01:18.120
So we haven't lost any money at least not yet

00:01:18.560 --> 00:01:23.120
Let's give everybody their assignments. I've never unbent CPU pins before

00:01:23.120 --> 00:01:27.360
But I think I'm a good candidate for this because I'm a fairly handy guy. I'm a fairly DIY guy

00:01:27.360 --> 00:01:31.120
Like I've ran out of kitchen before I can do things. I've never done this

00:01:31.440 --> 00:01:38.800
It's been a while since I've done this, but I have repaired pins on like older CPUs before and like CPU sockets and motherboards, so

00:01:39.520 --> 00:01:43.520
I've got some experience at least just not super recent. I have fixed bent pins

00:01:43.880 --> 00:01:49.600
All right, that's about as far as I've gotten. I've never soldered pins. My experience level is pretty low

00:01:49.880 --> 00:01:52.760
But of the probably I'd say

00:01:53.200 --> 00:01:59.200
10 to 30 CPUs that I fixed over the years. I have above a 90% success rate

00:01:59.200 --> 00:02:03.960
All right, so I get a 5700 X and how many problem pins are there?

00:02:04.320 --> 00:02:10.760
Okay, first thing I'll say is these are really small. I could like make it I this but it would be a lot harder

00:02:10.760 --> 00:02:16.360
I'm grateful for this thing. This is actually a good way to check because you can see they're all at orderly and you can see gaps

00:02:16.360 --> 00:02:21.520
Between each row, so if there's anything that's wonked over you would see it fill that gap

00:02:21.600 --> 00:02:28.640
The only bent pins are on the corners over here and over here. This is a game changer having it on this thing

00:02:28.680 --> 00:02:31.000
There they are three amigos

00:02:32.480 --> 00:02:37.920
This microscope is so sweet. This is gonna make it so much easier. Oh, there's more bent pins than I thought on this corner

00:02:37.920 --> 00:02:44.220
I thought it was one, but it's four maybe even six two things. I'm worried about with this task is

00:02:44.580 --> 00:02:49.500
Number one, I don't know how many times you can bend it this way and that before it just snaps right off

00:02:49.540 --> 00:02:54.220
Number two I don't know if these are solid gold or just plated and I worry that if I were to grab it with like a plier

00:02:54.300 --> 00:02:58.100
That I would strip off a coating and then it would no longer function as it's supposed to

00:02:58.780 --> 00:03:06.700
Do you know the answer to these questions? I do not I'm still out of four confidence, but I don't know if I can do it in an hour, but maybe I can do it in 10 minutes

00:03:06.700 --> 00:03:11.100
I really don't know I'm gonna start with a spudger and the microscope and let's go. Oh my god

00:03:11.220 --> 00:03:18.460
The spudger is like wider than the gap between rows. I may not be able to use this tool. Okay. I might use a razor blade instead

00:03:18.980 --> 00:03:24.260
Okay The other thing I don't know is how straight do they need to be? What's the margin for error?

00:03:24.380 --> 00:03:30.540
Looking at the socket each pin has a hole So if it's off a bit, it should just get snapped into place

00:03:30.540 --> 00:03:37.780
That's the theory not snapped in the bad way snapped in the good way You don't have to force it, but like it's got a tiny bit of give at least almost looks like I'm done here

00:03:38.500 --> 00:03:42.340
But then when you turn it this way like oh, yeah those I need a helping hand

00:03:42.340 --> 00:03:49.220
So because like I can I use my nose to adjust the focus? Okay, so I previously was working on these ones and now I'm gonna go over this one

00:03:49.220 --> 00:03:52.540
So this is a mechanical pencil and the tip is a hollow cylinder

00:03:52.540 --> 00:03:56.740
And apparently you can actually fit the pins in that hollow cylinder and then just manipulate them that way

00:03:56.740 --> 00:04:00.500
I think you're gonna give that a shot because I love the idea of just being able to use a tool

00:04:00.500 --> 00:04:05.660
You might have lying around the house you got her on there now The question is do I give up having a feedback to like know what's happening?

00:04:05.660 --> 00:04:08.100
That looks really straight up and down. We'll pull it off

00:04:09.100 --> 00:04:16.780
Did we get any progress? Oh, yeah, not bad. It's not perfect yet. That one is still wonky that one right on the corner

00:04:16.780 --> 00:04:21.660
I'm I'm digging this pencil trick, but I don't know if I can get these ones because they're so leaned over

00:04:21.660 --> 00:04:25.180
They're touching the neighbor. Yeah, I don't think this pencil is gonna work for this application

00:04:25.420 --> 00:04:30.940
All right, where's my trusty razor blade? This is like when Luke just uses a force instead of that tracking system

00:04:31.820 --> 00:04:36.380
I Think I'm I'm pretty good here guys going into this

00:04:36.380 --> 00:04:40.140
I thought the definition of bent pin was like really leaned over and now I'm realizing

00:04:40.500 --> 00:04:45.780
It might just be like a little leaned over and that is worrisome. Aha. Oh

00:04:46.740 --> 00:04:49.060
She's in all right. Let's send it

00:04:50.540 --> 00:04:57.540
Let's go 21 and a half minutes CPU information AMD Ryzen 5700x 8 core processor

00:05:01.500 --> 00:05:04.260
Ploof that's the wind condition you see Ploof you win

00:05:04.780 --> 00:05:10.300
Cinebench is running fine. All the numbers are within expectations. So I'd say we're done. We made the money

00:05:10.300 --> 00:05:12.580
I hope the other guys are doing at least this well, but

00:05:14.220 --> 00:05:16.220
Who knows

00:05:20.860 --> 00:05:25.300
All right, what do we got It's it looks like a couple of rows are bent

00:05:26.020 --> 00:05:29.620
But I think maybe I can get them all at the same time with like a razor or something

00:05:29.620 --> 00:05:34.700
I like to stick it on to Something a little more substantial to hold on to just to make it easier

00:05:34.700 --> 00:05:39.180
I've got some double-sided tape on a cork here probably better than a half dozen bent pins, but

00:05:39.780 --> 00:05:42.180
They're not very bent. This is almost like

00:05:43.340 --> 00:05:47.820
Not bent enough like I don't think I would have tried to sell this on eBay. I would have tried to fix it

00:05:47.980 --> 00:05:49.980
I'm pretty happy with that to be honest. I

00:05:51.300 --> 00:06:00.220
Think that's pretty good. I Think this is a broken chip that someone just kind of creatively bent because I would have tried to fix that myself

00:06:00.540 --> 00:06:03.260
All right, here goes nothing 3,700 X

00:06:04.620 --> 00:06:09.100
Ring 3,600 megahertz faulted is a good CPU looks good

00:06:10.140 --> 00:06:12.140
Let's see if we Windows. There we go

00:06:14.140 --> 00:06:19.660
That guy again everything looks okay to me. Let's see a Cinebench. Yeah, not too bad

00:06:24.300 --> 00:06:32.220
I Let's take a look at your chip. Okay, this one. I probably the one labeled me three pins on the very corner

00:06:32.220 --> 00:06:37.780
It seems oh, and I see maybe one that's bent in the bottom right corner by the triangle

00:06:37.780 --> 00:06:41.260
So three missing pins and a slightly bent pin

00:06:41.980 --> 00:06:48.580
But three pins all beside each other on the edge there and like the most open edge of the corner. I'm probably I think I'm

00:06:50.180 --> 00:06:56.520
We'll find out Jumping all over the place, right? Like oh all right

00:07:00.780 --> 00:07:08.660
All right, I am not having fun I think even if my life depended on this I would die. I like genuinely would not be able to get this done

00:07:13.220 --> 00:07:19.100
I think it's dust. Yeah, I think I got to call it. I'm sorry. No worries man. You just give it a good try

00:07:23.180 --> 00:07:27.580
I Hear you need some help a lot. You had some trouble with

00:07:28.020 --> 00:07:34.660
Soldering on some missing pins there. I have tons of soldering experience and working on small electronics. So I figured it might be able to do some

00:07:34.780 --> 00:07:40.340
All right, so I'm gonna give a shot. So it's yellow one to five then how confident are you? Probably about six

00:07:40.900 --> 00:07:44.900
Okay, that's off the confident. All right, take a look. All right

00:07:45.500 --> 00:07:50.300
Yeah, yeah, he managed that one pretty good a see for me. It's not the

00:07:50.740 --> 00:07:56.260
Getting them off part. That's gonna be difficult. It's the getting them back on part is gonna be fun

00:07:56.260 --> 00:08:02.220
First things first probably try just straightening out the ones that are still there. Are you still at a six?

00:08:03.100 --> 00:08:05.100
steadily dropping down to a four

00:08:06.220 --> 00:08:09.340
So this isn't like a for realsies thing sound and fired

00:08:10.180 --> 00:08:14.780
No, all done. All done. Got one. That's time

00:08:17.220 --> 00:08:19.220
So how many more do you have to do?

00:08:20.420 --> 00:08:23.560
Okay, that's not too bad though. You think it's a viable side hustle

00:08:24.540 --> 00:08:30.820
Not replacing pins unbending them sure, but if you're replacing them, there's no feasible way to make a dollar off this

00:08:37.940 --> 00:08:45.860
Now it's my turn. Oh Dude, oh, there's a bad one. I don't want any hints. I don't want to know how oh dude. Oh, there's two like that

00:08:46.620 --> 00:08:53.940
Dang those are both pretty mangled and there's a fair number of other ones with minor bends all along the edges here

00:08:53.940 --> 00:08:55.940
What did they do to you?

00:08:57.900 --> 00:09:03.420
Check it out. Just like the scribe driver, but a mechanical pencil. That is awesome. Yeah, where do I get it?

00:09:04.140 --> 00:09:10.260
Nowhere yet. Yeah, well you started already. Yeah, geez Jordan. Hey, you were poking at pins before well

00:09:10.260 --> 00:09:17.820
I was just I was looking at it I was looking at it what I want to do is I want to really get under that pin and see if I can get it to

00:09:18.660 --> 00:09:22.180
slightly less bent over without

00:09:22.180 --> 00:09:26.220
Ripping it off if I rip it off. I'm gonna have to put a new pin on it

00:09:26.220 --> 00:09:31.940
Which means I need a donor CPU and I need a lot more time. It's hard to get under this guy

00:09:33.260 --> 00:09:40.380
You can really see boys. This is the smallest flathead in the iFixit kit and it's still kind of hard to get it under there wait

00:09:41.740 --> 00:09:47.380
I figured out what they did. Oh, I think they put it into the socket the wrong way Jordan

00:09:47.380 --> 00:09:51.700
I wonder if they like had it the wrong had it oriented the wrong way or something dude

00:09:53.060 --> 00:10:01.020
I think I may have nailed it and the thing is it doesn't have to be completely straight. Yeah, it just has to go in

00:10:01.740 --> 00:10:04.060
Hey, hey, hey, she's in now

00:10:05.180 --> 00:10:08.740
Just because she's in the socket doesn't mean I want to fire up the board yet

00:10:08.740 --> 00:10:14.900
I want to see if I like really mashed one down because if I fire it up while I've got a pin mashed down

00:10:15.860 --> 00:10:21.060
Who knows what's gonna happen, right? At least that way I get another shot at it

00:10:22.700 --> 00:10:27.980
Don't get one shot one opportunity. Okay. Okay. I'm ready. It's my okay. So it took me

00:10:28.780 --> 00:10:35.260
15 minutes Ryzen five thirty five hundred. How much I pay for this thing 30 bucks 30 bucks

00:10:36.380 --> 00:10:40.140
dude 15 minutes

00:10:44.820 --> 00:10:52.100
I did it faster But not the same CPU the point is that aside from some overhead here

00:10:52.100 --> 00:10:59.660
I can afford to pay myself 80 bucks an hour to fix this CPU dude. I'm feeling good challenge number two for Mr

00:10:59.660 --> 00:11:04.620
Linus, I get a bonus CPU. He gets to because neither of them is worth that much. That's why they call me the boner

00:11:05.620 --> 00:11:10.340
Oh dear because I get a bonus CPU and bonus tech tips

00:11:12.500 --> 00:11:14.500
Okay, sorry, what are we talking about?

00:11:15.860 --> 00:11:22.300
This time I get a thirty five hundred X. This also seems like it's not in very bad shape

00:11:22.860 --> 00:11:27.140
Wait, is this just fine? They said it's missing a pin. I couldn't find it now

00:11:27.380 --> 00:11:32.540
Occasionally you're gonna buy stuff that's for parts and it's just gonna be dead. Okay, stop the clock

00:11:32.540 --> 00:11:35.660
I think I'm ready to test this thing because I do not see a

00:11:36.140 --> 00:11:42.740
Missing or bent pin on it. Well, she went in the socket easily. Yeah, nothing was bent on it before this fires up or doesn't

00:11:42.740 --> 00:11:50.420
What's my potential profit here? 15 20 bucks. Okay. Oh, I just saw the keyboard light up. Holy crap. Did we just get a for part CPU?

00:11:50.420 --> 00:11:53.180
That's just perfectly fine. No way

00:11:55.340 --> 00:12:00.420
Six core processor all six cores there. They freaking go. She's turbo in baby

00:12:00.420 --> 00:12:03.420
Yeah, and she seems like she's working just fine

00:12:09.140 --> 00:12:15.500
Now what if I told you that actually is missing a pin a missing pin doesn't necessarily mean a broken CPU though

00:12:16.020 --> 00:12:20.820
Sometimes they have unused pins or ground pins that are extra

00:12:21.700 --> 00:12:23.700
vestigial pins if you will

00:12:24.700 --> 00:12:29.380
Dude, I feel bad about missing it though Naughty Linus. I was getting old man

00:12:30.900 --> 00:12:37.340
Here pop it out. We'll take a look and there she is. Yep. How'd everybody do so James took about 20 minutes

00:12:37.340 --> 00:12:40.380
How'd yours go took me about eight minutes and I got it working fine

00:12:40.380 --> 00:12:47.580
How about Elijah Elijah had a bad time did we sabotage Elijah has never soldered before and he had one was missing three pins in the corner

00:12:47.580 --> 00:12:52.860
And it didn't go out Justin ended up cleaning it up and getting one pin back onto it in the time allowed

00:12:52.860 --> 00:12:55.940
It's it's been through some something to watch for is

00:12:56.380 --> 00:13:04.380
Sometimes you can get away with sticking a little piece of metal into the socket as long as you can get a good contact here

00:13:04.380 --> 00:13:08.820
I actually had a CPU that I put into my parent system an old tricor

00:13:08.820 --> 00:13:13.340
I think it worked for about 10 years like that with a broken off pin on the CPU

00:13:13.340 --> 00:13:19.140
You're right. Yep, but with a little slightly over long piece of metal in the socket so it would make contact

00:13:20.140 --> 00:13:25.220
Should we try it? I was thinking maybe if that's memory channel B

00:13:25.220 --> 00:13:29.660
Maybe we could run it in single channel that CPU. I dropped that Xeon back in the day

00:13:29.660 --> 00:13:35.820
Yeah, where I dented the corner of it. It did still work. Yeah, just five of I believe six memory channels

00:13:37.220 --> 00:13:45.020
No way no way There could still be issues are all of those pins memory channel B. I'll double check

00:13:45.860 --> 00:13:53.860
because if they aren't Then it might still work, but maybe one of our PCIe lanes is not functioning properly, right?

00:13:53.980 --> 00:13:59.740
Who knows at all memory be dude. She's in Windows. She's in Windows, brother. I

00:14:00.340 --> 00:14:06.980
Cannot believe that let's run the numbers James should be able to sell this CPU for about a hundred thirty five US dollars

00:14:06.980 --> 00:14:10.820
And we're not able to completely ignore the cost of our tools

00:14:10.980 --> 00:14:18.100
But they are good for more than just unbending pins So let's take away a couple of bucks from his profit for that that leaves him with

00:14:18.780 --> 00:14:24.180
$64 of profit for just over 20 minutes of work a rate of a hundred seventy nine dollars an hour

00:14:24.460 --> 00:14:28.860
Not too shabby assuming he can keep that pace up as for team Elijah

00:14:29.220 --> 00:14:36.260
He didn't have such a good time after two hours his 5800 X only works with a single memory channel

00:14:36.300 --> 00:14:40.900
Now he might still be able to sell that be it at a reduced price

00:14:41.140 --> 00:14:47.180
But he wouldn't be making even close to minimum wage as for Jordan's 3700 X

00:14:47.180 --> 00:14:53.460
Well doesn't have quite the profit potential of James or Elijah's chips, but he made up for it by being quick

00:14:53.620 --> 00:14:57.980
He should be able to get close to 40 bucks after expenses or around

00:14:58.820 --> 00:15:04.140
285 an hour though. There's no way that that's a sustainable pace as for me

00:15:04.140 --> 00:15:10.300
I spent about 16 minutes on my 3500 and really I just used the precision screwdriver

00:15:10.300 --> 00:15:13.500
So I should see about $19 of profit coming back

00:15:14.260 --> 00:15:17.900
$71 an hour if you're playing along at home as for the 3500 X

00:15:18.060 --> 00:15:22.820
Well, I didn't spend any time fixing this one and I could get 20 bucks of profit there too

00:15:22.820 --> 00:15:27.300
And I'm only out the time it took to test it and ship it not too bad

00:15:27.380 --> 00:15:33.900
But is this a viable side hustle? Well, I think so but there are some caveats

00:15:34.060 --> 00:15:42.500
There is a lot of upfront risk I'd say it's important to make sure that you are picking chips with good potential profit picking ones where the seller

00:15:42.900 --> 00:15:49.220
Clearly lays out what's wrong with them so you can claim something back if the issue is something else and try to get chips with

00:15:49.220 --> 00:15:55.820
As you bent pins as possible So you're not wasting a ton of time before you can get right to this segue to our sponsor

00:15:55.820 --> 00:16:00.300
If you guys enjoyed this video, maybe you'll like watching us fix bent pins in a CPU socket

00:16:00.540 --> 00:16:03.060
That can be even more finicky than this