#494 – The Two Person Rule

Chris Gammell: This is the AMFR Podcast. Released May 31st, 2020. Episode 494. The Two-Person Rule.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV Blog.

Chris Gammell: And I'm Chris Gammell of Contextual Electronics.

Dave Jones: What's up, nerd?

Chris Gammell: Well, isolation continues. Where do we start? Well, yeah. I've had lots of fun stuff in the past couple of weeks.

Dave Jones: Yep.

Chris Gammell: You're still in isolation? Still in isolation here. How about there?

Dave Jones: No, no. We're pretty much out of isolation now. I'm not going to say back to normal, but in my life, at least, apart from gyms being closed, everything's back to normal. Yep. So, yeah. Okay.

Chris Gammell: Well, you guys wear a face mask out or no?

Dave Jones: No, no, no. I've never worn a face mask out. No, no. Face masks are pretty rare here. Even during the peak, probably like a month ago or whatever, probably only a few percent of people wore masks. It's just not really a thing here.

Chris Gammell: Yeah, I got to say, other countries are doing it a lot better than the US.

Dave Jones: Oh, that's, yeah.

Chris Gammell: Yeah. We don't need to get into that. But, yeah, we're just trying to do what we can around here and stay safe. And, yeah. So, I'm working at home and I wear a face mask whenever I go out. It's not a big deal.

Dave Jones: Really? Is that common?

Chris Gammell: Around here, yeah. Yeah.

Dave Jones: Okay. Right. So, do you get strange looks if you don't have one?

Chris Gammell: If you go to a store, they won't let you in anymore.

Dave Jones: Oh, really? Wow. Okay. Yeah. No, that's never been the case here. It's like, no. If you come in with a mask, people, you know, go, are you sick? What's wrong with you? You know? Right. Like, yeah. Yeah. Anyway.

Chris Gammell: Yeah, it's different in all parts of the US. But around here, yep. Face mask. I'm cool with it. No. No big deal.

Dave Jones: No worries. You know. Yeah. She'll be right.

Chris Gammell: Good. Yep. Great. I'm glad you guys are doing well.

Dave Jones: That's enough of that rubbish.

Chris Gammell: Yeah. I will say I've been very dependent on overnight shipping. I've been using that a lot lately. Yeah. But that's actually worth doing. Like, I actually use it a lot anyways. And I have to say, I'm pretty much addicted at this point. I mean, you were addicted years ago to Amazon Prime. Prime, yeah. But this is more. That's not overnight. And that's not.

Dave Jones: I thought Prime was like same day in the US.

Chris Gammell: Some parts it is, yeah. Yeah. Yeah.

Dave Jones: You were bragging that you had a ladder within hours once. You know. You need a ladder. So he's ordered one on Amazon Prime. And it turned up like two hours later or something.

Chris Gammell: No. I don't remember that. Yeah. But I'm sure you're right. Yeah.

Dave Jones: It was on the show. Yeah.

Chris Gammell: Yeah. I'm sure you're right, Dave. You know. Yep. I'm always right. No. That no. Hard no.

Dave Jones: Come on.

Chris Gammell: Yeah. No. Oh, God. Yeah. So what's going on around there? I mean, I have a lot of issues that I could talk about in general terms this week. But before we do that, what's going on with your electronics life?

Dave Jones: I'm playing with vintage hardware. Oh, yeah. Kind of got a bit of a kick out of it. Oh, yeah. What kind of stuff? And a 1976 vintage Intel thing. Yeah.

Chris Gammell: I saw the title. I haven't seen the video yet. So it's on my list. Oh, okay.

Dave Jones: I was quite happy with that. That was fun. I wanted to build it. But I've got a mint in box kit. You know, it's like they're not. I'm not going to say they're ultra rare, but they're not common. And it's mint in box. It's like it's vacuum sealed all the parts, all the individual kit components from Intel. Because you used to buy it as a kit. And solder it yourself.

Chris Gammell: What's the value of that? What is the mint in box? And don't people usually say that for like resale value?

Dave Jones: Well, it probably could be worth like 500 bucks. Like it's not insignificant, right? It's pretty pricey. There are collectors out there who will pay a premium for that.

Chris Gammell: Okay, it's collectors. It's not like someone who's like, oh, I want to use this. It's just like they want to have early Intel. Oh, no, no, no. You can get built up.

Dave Jones: Yeah, that's why I found another one that was built up. That was already built up and hacked and untested. You know, so I bought it and I fingers crossed. And if anything, I can, you know, do a repair troubleshooting video. But no, it worked. So cool. I'm always disappointed when things work.

Chris Gammell: Well, let the kids at it, you know.

Dave Jones: Yeah, right.

Chris Gammell: There you go. Say again, have a screwdriver in a couple hours with this thing. See what happens.

Dave Jones: I've always thought about that. Like I could like get somebody to like put a random fault in a board. So then I have a repair video. But it's got to be like a double blind. Well, almost like a double blind thing, you know, really much. They've got to, you know.

Chris Gammell: It's like someone opens an envelope and then they hand an envelope to someone else. And that person actually does the thing.

Dave Jones: And then that does the, yeah, yeah. So you can't pass subtle clues, you know. I can't get.

Chris Gammell: Yeah, I think there is like puzzle finding type of stuff with like troubleshooting, right? You don't want to know the answer.

Dave Jones: No, no, no. Because that just, no, that just ruins the whole thing. Ruins the experience.

Chris Gammell: I've seen that for like dev board. I think one of the scopes that I've used in the past had like a board. I think maybe the MDO that was sent to me. It had like a test board where it's like, oh, we want to like generate these weird traces or these weird signals rather that you might see. And we want to showcase like what the scope can do. And so they sent to like a test board that like generated these like runt signals and stuff like that.

Dave Jones: Yeah, yeah. Yep. I've got quite a few of those. Yeah.

Chris Gammell: Yeah. So that's kind of like that. I mean, it's, you could, and I guess there's like PCB puzzles. Surely there's stuff out there like that. I guess that's what one of the DEF CON badges that the toy makers had made. DEF CON 27, maybe? Yeah. DEF CON 27. 26. 26. It had like a bunch of puzzles on it. So like one of the things was like, even though the chip was soldered on, onto the back, there's like a SOT 23-6. It was soldered on reversed. And like one of the puzzles was you had to figure out that it was reversed and then be like, oh, and then you reverse it. And then like some new function actually came online. So I thought that was cool.

Dave Jones: Oh, nice. Yep. Yep. That's pretty neat. How many times have you flipped apart like that? And then, yeah.

Chris Gammell: Yeah. Yeah. Yeah. Good fun. Yeah.

Dave Jones: All right.

Chris Gammell: I have not been flipping stuff back. I've been conformal coding stuff recently. That's been a new.

Dave Jones: Oh, conformal coding. Because you're now a high power man. You're now into high power stuff.

Chris Gammell: Not high. Yeah. Not super high power. But yeah. Oh, but you know. Dealing with some higher voltages. Not five volts. How about that? It's got a little juice behind it. So yeah. Tell us. I've been conformal coding. And I have the, I have the, actually have the case right here with me. It contains dimethyl ether, acetone, xylene, one methoxy, two propanol acetate, and ethyl benzene.

Chris Gammell: Gesundheit. Yeah. I can huff in this stuff. Quarantine's going by way faster for me, you know. Right. Huffing my way to freedom. Right.

Dave Jones: Right. So you've had some issues. Tell us in roundabout terms because this is. Sure. Yeah. Confidential stuff. Yeah.

Chris Gammell: Yeah. Some of the stuff. What was the first thing we were talking about? It was. One thing was like floating voltages. Another thing was like a bad board house documentation. But what was that first thing we were talking about? I can't remember. I can't remember. There's been so many. Expect me to remember.

Dave Jones: Come on.

Chris Gammell: Yeah. Dude.

Dave Jones: This is bad radio. We should come prepared. We should have wrote this down. We just talked about it like 10 minutes ago. And we've just totally forgotten. Jeez.

Chris Gammell: Well, I'll talk about the one thing that was. So I was chasing down a problem recently. And I have been using my VNA with the help of Mr. Jeff Mighty Young Kaiser. Yes. Yes. And it turned out. So actually it was a mistake on my part, partially, is that I didn't. I made an assumption. Someone else was ordering boards for me. And maybe you've made this mistake before. Hopefully not. But someone else was ordering boards for me. And I didn't send a drawing because they just told me what the stack up would be and all this other stuff. And I was trying to get controlled impedance. And yeah. Come on, folks. Yeah. And it was just a completely different board house. And so I was like, oh, okay. Well, and the stack up was completely different too. But let me tell you, after like two days of troubleshooting, a very simple board and just like, you know. It was basically me assuming that I was wrong with the VNA and Jeff was helping me kind of work through what I was doing with it. Finally, I was like, I looked at this. I looked at this and I was like, there's nothing else that it could be. And I looked at the stack up finally. I'm like, oh my God. It was completely different to the point where the controlled impedance was off by 100%. Wow.

Dave Jones: Right. So it was like, it was supposed to be like, like you ended up reading like 50, it was 25. Okay. Right. So you designed them. So you laid out this test coupon board and we'll explain test coupons in a minute. So you laid out this test coupon board expecting with the stack up from a particular manufacturer, right? A particular PCB manufacturer. But the person that got the board made used a different manufacturer than what you specified.

Chris Gammell: Yeah. Right.

Dave Jones: Thinking that, oh, it's just a PCB, you know.

Chris Gammell: I don't think it was. Yeah. There was no mouse at all. I think it was just, it was just really. Oh, no, no, no. It was just that. And then on my fault, it was that I didn't send a drawing through, right? So drawing then would have been like, then the manufacturer would have been like, oh, we're not doing what you're saying here.

Dave Jones: This is why you want me to tell you how to avoid this problem. Sure. You bake in the stack up onto your Gerbers. So you have a specific layer. This is why you have a mechanical manufacturing layer in your PCB file, right? This is what the mechanical layer is for. The mechanical layer is not just for doing your board outline. It's also for putting your manufacturing information around that. So most, you know, or big companies, they'll have like a company template. It'll be like a, you know, an A4 letter size sheet, you know, template with the grid thing, you know, and all that. And all this information stuff down the side. You know, look, we need this IPC standard board. We need this stack up. We need this.

Chris Gammell: Must conform to this. Must conform to that. Must conform to this.

Dave Jones: And I've even put like little pictures of the stack up. So I actually build, like I actually draw the stack up and go, this has to be, you know, 10 thou, you know, pre-prig. And this has to be, you know, one ounce, a half ounce copper and this, you know, et cetera, et cetera. And so you specify the stack up and which you can take from their website, right? Yeah, exactly. That's a great starting point.

Chris Gammell: Then you know they can do it.

Dave Jones: Yeah. So then if that Gerber, if there's somebody else in the loop has those Gerbers, those Gerbers will get sent to another manufacturer that you didn't specify. And they've got an entirely different stack up. Any good manufacturer, most will, right? They will see this and go, oh, you've got a stack up in here, but this doesn't match what we have. Right. So they will actually question it and they'll ping back to you and, you know, it's all sorted before the board's actually made.

Speaker ?: Yeah, exactly.

Dave Jones: It's like a catch.

Chris Gammell: It's like a documentation catch, right? Yeah. Yeah, exactly. Exactly.

Dave Jones: But it's got a bit, because if you embed it in a separate file, you know, might have a text file or a document file or something like that. They can miss that. If you embed it in the Gerbers, it shows up on their screen. That's a good point. When the people are doing, when you send your Gerbers to the manufacturing house, they have a team that just specifically scans all the incoming manufacturing Gerbers to make sure they're manufacturable, right? And, you know, then they'll flag any errors and things like that. So they will see that and they will read. They'll just pop up on their screen. Automated. They don't have to open some other document file.

Chris Gammell: Right. It's all there. It's in the CAM program.

Dave Jones: It's in the CAM program. It shows up and they will read that stack up and they'll go, you know, yep. Oops. This doesn't work. We can't do this. So we can, but we have to order all these different pre-preg materials and this will cost you a fortune, blah, blah. Yeah. And then you can sort it out before it's made. And you spend a week troubleshooting the damn thing, thinking that your measurement thing is wrong.

Chris Gammell: Two days. Two days. Two days. Very valuable. Very valuable. Valuable two days. Yeah.

Dave Jones: Yeah. That's great. Yep. You want the first and you won't be the last. No, definitely not.

Chris Gammell: I think that's a big switch over there too. Like, so I remember like when I was getting started too, it was kind of like, I think this is going to be more common because there's so many board houses that are just like, yeah, you just get what you get. It's a PCB. What's the problem? You know, like, uh, as I was coming to the industry or like starting to make boards and stuff like that, there was more and more of the online where you didn't, you know, like you, you basically just got what the stack up was at like an OSBAR or advanced PCB or advanced. And circuits or whatever. And so you didn't, you didn't think about it. And, you know, for a long time, it didn't. And, uh, and, uh, and for, and for like low frequency stuff that people might be doing.

Dave Jones: So if people are out there like doing, which is almost, which is 90% of jobs, right? Which is 90% people get, you can go your entire career and never have to worry about stack up.

Chris Gammell: Right. Right. Yeah. And like the, I think like any, you definitely need it for RF. You definitely need it for high speed signaling. Uh, you definitely need it. If you need to like a different board thickness, you probably want to play with it then. Uh, yeah. But like outside of that, like what else? I mean, is there anything else? I don't, I don't know. Like when I was in my analog days, I don't, I didn't care at all.

Dave Jones: You know, like, Oh, well, no, no, for the high power stuff that you're talking about, copper thickness, for example, you might need, okay. One internal layer has to be four ounce copper or two ounce copper or something like that. Yeah. That's a good point. You know?

Chris Gammell: Yep. Yep. Yep. Yeah. Yeah. And it's really anything outside of the, you know, like if you're used to using like a JLC or a, you know, a PCB way or anyone like that, you know, you're just, again, you're just getting what you get. Maybe you have different thicknesses or whatever. And they give you, they tell you what the stack ups are, but. Oh yeah. Yeah.

Dave Jones: They do. And you can work around those, which is fine, but you just got to know what they are.

Chris Gammell: Right. And anytime you change something, it's going to go up in cost. Right. Yeah. You're not getting that $5 PCB for custom.

Dave Jones: And they could change that on a whim, you know, because these are just prototype services, right? This isn't their professional service, right? These companies, they, you get your board for a dollar each for a reason. It's because they can change it at a whim. Right. Yeah. Don't, don't expect them to keep consistency for a dollar a ball. You know, it's just crazy. Right. So, yep.

Chris Gammell: Yeah. And I remember Ted Yappo, he gave a talk at Supercon about, he was just, you know, basically he's making like a crazy fast scope. I think it's like an eight giga sample scope or something like that. Right. Yeah.

Dave Jones: I think I've seen the one. Yeah.

Chris Gammell: Yeah. So Ted's been on the show. Yeah. He, and I think he talked about that a little bit. Did he talk about that on the show? I'm like looking back at the notes, but basically he, he, he did a bunch of the same kind of thing, like a test coupon. I remember him showing me the, like the Oshpark test coupon I did and I was like looking at it and I was like, what, why, why are you doing that? But it's the same reason he's testing basically controlled impedance. And I think Oshpark uses a FR 408 instead of FR four. And there's some benefits of doing that kind of thing a little bit more, you know, it's all about consistency and stack heights and process stack height versus.

Dave Jones: Oh, it's tolerance and all sorts of stuff. That's why you do, you know, you might use some Rogers material or something like that because they've got, you know, a hundred different types of material and all that. They have various specifications. Some are much tighter than others, you know, some, some really critical applications. You can't tolerate plus minus 5% tolerance on your controlled impedance, right? You've, you've got to be tighter than that. So you've got to, you know, you've got to pay a lot more for a, a raw material, a board raw material that, um, that has a greater controlled impedance and, and it changes with moisture and all sorts of stuff. You know, it's really, you know, if you're on the bleeding edge, it really matters. That sort of stuff really matters. You can't just use, you know, generic, you know, FR four, whatever you get from your, you know, Asian PCB house, Asian proto house. It's like, no, right. No, you, you're going to specifically, um, specify a particular material and if they don't have it in stock, which they often don't, they'll just, they'll have to order it in. So yeah.

Chris Gammell: Yeah.

Dave Jones: Anyway, test coupons. We, we didn't actually tell, I said we'd explain test coupons. Oh, sure. Sure. What, what test coupons are. If you don't know, test coupons are just little strips that you put on this, usually strips that you put on the outside of your board. Like if you're making a panel, you've got your PCB in there and then you've got your manufacturing strips top and bottom. So with the socket holes in them that allow your pick and place machine to take your board through the, you know, I've done, this is all part of DFM. Yeah.

Chris Gammell: Panelization video you made. Yep.

Dave Jones: Panelization videos. I've done several videos on that.

Chris Gammell: Yep.

Dave Jones: How to do that sort of stuff and test coupons. And there's, so there's all this space on the side of the panel, which you're paying for anyway. So it often pays to put a test coupon, which might just have a, a couple of test traces on it of a certain, of a certain known width, um, label it on the board, by the way, this is like 10 thou width or whatever. So just so you know later, so you don't have to look it up. Right.

Chris Gammell: You know, like pulling up a micrometer or something like that.

Dave Jones: Yeah, yeah, exactly. And, um, and so you'll put specific, especially controlled impedance traces. And, uh, as you said, you might put like a, an SMA connector footprint at each end. So you can, you know, so it just allows you to hook it up to your VNA or your, your, your network analyzer. Or TDR as well. TDR. So that you can, yeah. So that you can measure the characteristics of that trace. Because as I said, there's a tolerance variance when they manufacture these things. Is it 50 ohms or is it 55 ohms or is it 45 ohms? You don't know. You want to measure it. Right. So when you get bored back from the test house and if you're a large company, you will have, um, incoming inspection team, which will actually do these measurements for you. You know, if it's that, uh, critical, if your company's big enough and this is important enough, then you'll probably have someone dedicated to this job. That's right.

Chris Gammell: Yeah, yeah. Yeah, exactly. They're just, they're just hooking up a VNA all day long and just, or TDR all day long.

Dave Jones: Hooking up a VNA all day long and their TDR measuring the, you know, the impedance of these traces, making sure that the PCB house hasn't screwed up because, uh, you know, I've worked on boards that cost hundreds of thousands of dollars each, right? Each board.

Chris Gammell: That's nuts, man.

Dave Jones: Right? Yeah, that's crazy. So if you, if you just get the board from the PCB house, assume it's okay, populate it with a hundred thousand dollars worth of parts, right? Right. And then go, and then you go, you know, power it all up. Oh, why doesn't this work? Ooh, oops. It's because they, they screwed up the stack up at the house, all the tolerances out and, and it doesn't work. Cause there's nothing you can do about it. You scrap a hundred thousand dollar board, right?

Chris Gammell: Or you've got the desoldering gun and you're really, really going to town.

Dave Jones: No, no. Because it's, well, yeah, you're that desperate, but no.

Chris Gammell: No. Yeah. Right.

Dave Jones: It's yeah. Yeah. Like you can scrap, it can cost you time and a huge amount of money as well. Okay. Dave. On high end boards.

Chris Gammell: Well now, now, now you're on the hot seat here. Have you scrapped a hundred thousand dollar board before?

Dave Jones: I didn't know. I've probably scrapped 10, I've scrapped boards in the order of 10,000. 10,000 dollar boards? 10,000, 10s of thousands. No. So five digit boards, not six digit boards. Holy crap. That's still, that's, that's insane. I know. It's yeah. But this is like, yeah, it happens. But yeah, this is for projects which have multi-million dollar budgets. Of course. Right. Of course. But still, like, yeah. Yeah. Yeah. Yeah.

Chris Gammell: It's tough. It's tough when. Yeah. No, tens of thousands of dollars. Or it's go boom or, you know, like it's.

Dave Jones: Well, this is why I mentioned this before. It, it, it became an instant, doesn't matter who you were, it became an instant sackable offense within the company. They would sack you if you entered the room without proper ESD and you didn't test your ESD strap. If you went, it doesn't matter. You're a manager or whatever. You're not allowed in that room without your ESD coat, your ESD strap, and you have to test it and you have to sign the book that you've tested it and you're allowed to go in and then you're allowed to play with the board. Right. Because these boards were so critical to the project that we could lose a $20 million project.

Chris Gammell: Contract. Yeah.

Dave Jones: A $20 million contract because somebody blew up the board. Right. It's like, yeah.

Chris Gammell: You know, it's interesting too. I think about this now. I think about like, like the anxiety I felt with Jujutsu saying that, like that is like really anxiety inducing. It's huge. But then I think about like, I've been in similar situations. I've been, I've been handling like hundreds of thousands of dollars of wafers in the past and been around really expensive. Right. I think the thing is you normalize to it. It's like, kind of like you, you know, you get into it. Obviously there's going to be people that are like, I've been thinking about team size as well. I'm working on very small teams right now. And like thinking, I've been like looking, I like drove past a Tesla today and just thinking about like the amount of integration that's in that. And I'm like, man, there are thousands of engineers. It's like, yeah, Chris, you're not supposed to build a Tesla. You can maybe build one board in a Tesla. Don't be so hard on yourself. You know, like there are thousands of engineers working on that stuff and there's testing. And I was just thinking like one, one board review is, you know, like one, one board in a, it like might have three engineers working on it in a Tesla that does like the windshield controller or something like that. And, and it's got 16 board reviews or, you know, PCB reviews. It's like, yeah, yeah. Right. It's just thinking about team size. I've been thinking about team sizes a lot lately because I've been doing a lot of system integration stuff with small teams and it's been very stressful. And, uh, I'm not sure I want to work on big teams at any time in the near future. Cause I like the, the, the hands on for a lot of different things, but at the same time, like big teams can build big things and. Oh yeah. Yeah. That's right.

Dave Jones: You can fly to the moon. Yeah.

Chris Gammell: That's right. Yeah. Yeah, exactly. Or, you know, or you can have a weather delay, unfortunately. Uh, like, yeah, like look, obviously the space X launch, like looking at that stuff. I was so anxious yesterday, like watching just them get ready for stuff. And that's just, that's just the launch. That wasn't even the pre-launch, I suppose. You know, it's not even like all of the pre-work and the testing and the iteration stuff. I don't know. I've just been thinking about big teams a lot lately and, uh, and it makes me a little anxious. I'm also locked inside. So I'm anxious about a lot of things. Yeah.

Dave Jones: Have you ever had the, uh, two person rule for board power-ups? I have not. What's that? No. See, once again, these expensive boards, right? Uh-huh. You know, a hundred thousand dollar board with a hundred thousand dollars worth of parts on it, you know, $20 million contract on the line, right? And you've got that proto it's built up. You want to power it up for the first time. They don't let just, you know, Joe Schmoe just go hook it up to the lab power supply. No, you know, no, you've got to have, it's a two person rule for valuable boards like this. You've got to have somebody double check your settings on the power supply. Yeah. Do double, just in case you goof it, right? Cause you could be having a bad day. You know, you wake up on the wrong side of the bed and you know, Murphy's going to get you.

Chris Gammell: You haven't had enough coffee, whatever it is, right?

Dave Jones: Yeah. And yes, there's somebody who's got to sign off and you sign off before you power up the board. And it's like, yes, I have checked. All these said, you know, you would have a checklist, you'd have a power up checklist and somebody would co-sign it. Seriously. Yeah.

Chris Gammell: I like that a lot. I've talked about, I wrote about checklists a while ago and there was, I'd asked for, like, there are some electronics checklists out there, but the problem is that they're just so generic or they're so broad.

Dave Jones: Generic ones can't ever work. They, they, they, they can never work. No.

Chris Gammell: It's a really good idea for like a pre, like a pre-flight pre-test kind of checklist. That's a great idea.

Dave Jones: Yeah. You know, of course, if you get your board from, you know, your $2 board from JLC piece of and it's got 10 parts on it worth a dollar, you know, like, yeah, like, but, but, well,

Chris Gammell: sometimes, but if you're on a, I mean, if there's a, I feel like it's, if there's a tight deadline, yeah, no deadline.

Dave Jones: Yeah.

Chris Gammell: Deadlines are, oof. Yeah. Time is money. I've got a couple. Yeah. Yeah. So, I mean, and like, you know, like as much as I love overnight shipping, even waiting 20 hours, you know, like it's pretty, pretty tough. So.

Dave Jones: And, and it's not just, and it's not just the time wasted. It could be face as well. Like it could be your, your, your reputation. If you're a contractor, like, like you are, for example, and you goof up and like blow up your board. Oh, sorry. You know, you have to admit to him. Oh, sorry. Look, the board blew up. You know, um, you might make up some lame excuse, but who cares? Right. The fact that, you know, you didn't, you didn't deliver means you'll miss the next contract and the next one. And, you know, all of a sudden you could be out of business because, you know, because you did this $2 board didn't work. Right. Yeah. So. Yeah.

Chris Gammell: That's a good point. That's good. I, I, I didn't have anxiety about the rest of my business before Dave, but now I definitely do. Oh my God.

Dave Jones: Oh, just. Yeah. Think about it next time. Think about everything that can go wrong the next time. It's going to be fine. It's going to be fine. Yeah. It's like, you know, us engineers flying on planes, you know, we can think of 10,000 ways that this plane can, you know, things that can go wrong.

Chris Gammell: That's right.

Dave Jones: It's like, if you think about it, you would just go nuts and you'd never fly.

Chris Gammell: That's true. It's true. Yeah. You just got to be like, well, you know, yeah, the last flight went okay. So I guess this flight will go fine.

Dave Jones: She'll be right.

Chris Gammell: What could go wrong?

Dave Jones: Come on. Somebody's followed some checklist somewhere, hopefully.

Chris Gammell: Yeah. Yeah. I, I just, uh, I was, I was kind of lukewarm on the book, but there's a book called humble pie that I listened to.

Dave Jones: Oh yes. Yeah. That's, um, uh, Matt's Parker.

Chris Gammell: Yeah, that's right. Yeah. Yeah. Yeah. The book.

Dave Jones: I haven't read it yet. Yeah.

Chris Gammell: It's tough to listen to. I think, I think I would have preferred to read it and the people that I, it was in a book club thing. And so the people that did read it seemed to like it a lot more than the people that listened to it. Cause like he was reading out strings of digits, but a lot of very familiar engineering things like Tacoma Narrows Bridge and the Millennium Bridge and London, stuff like that. And there's a really great story about the checklists and like a bolt inside of a airplane windshield and stuff like that. And just like how tolerancing matters. That actually was a great story. So I will recommend it at least for that chapter. There's, and there's a bunch of stories in there. People, like I said, if you're, if you went through engineering school, you're going to recognize a lot of the, uh, a lot of the stories or a lot of the things. There is a lot of programming stuff in there as well. So yeah, it's like I said, it was lukewarm, but it's, it's worth, it's probably worth reading.

Dave Jones: Right. Uh, how do you know who that is? Matt, what is his name? Matt, Matt Parker. I follow him on a Twitter. I follow him on, he's a, uh, YouTuber. He's a math YouTuber. He's one of the maths, uh, YouTubers. Got it. Got it. Yeah. Yeah. For you Yanks.

Chris Gammell: That's right.

Dave Jones: Mathematics as opposed to maths. I, I, I just find it, even though I'm an Aussie and I'm supposed to say maths, maths, I got my, it's just, it's too hard for me. I know. It's a bit of a mouthful. Yeah. Yeah. Yeah. I used to be able to do it as a kid, but for some reason I've been conditioned probably through all this Yankee talk, um, into saying math now because it's just easier. Just having an impact. Yeah. I know. I know. Yeah. Bastard. Yeah. Oh boy. Anyway.

Chris Gammell: Yeah.

Dave Jones: That reminds me of the engineering, the famous engineering fail. I'm going to have to pull up the video for it. Um, uh, walk, walk.

Chris Gammell: What is the subject of the fail while you're, while you're looking at it?

Dave Jones: It's a, it's, it's a, ah, here it is.

Chris Gammell: Oh, no. St. Louis.

Dave Jones: No, the, uh, Hyatt Regency collapse. The Hyatt Regency collapse.

Chris Gammell: Somewhere in the Midwest, Kansas, St. Louis, something like that. It's a Tom Scott video. That's the statics problem. That's the statics problem. So this is the one, if you're in statics class and it's like, you have a beam and it could support a certain amount of weight, but if you split it in half and have a support in the middle, it can no longer support that amount of weight. That's all. Like, that's pretty much all I remember from statics class that and like pulleys and how much I frigging hate statics class.

Dave Jones: Yep.

Chris Gammell: Hated, hated, hated. Oh my God. So glad I don't build bridges. Yep. And other people should be too. It's really just. Yeah.

Dave Jones: No, people died for this thing. Like, yeah. If you don't know, it was like they were, they had this suspended walkway, like, you know, in a big shopping center or something. No, yeah. The Hyatt Regency hotel. It was a hotel. Yeah. Yeah. A big, big suspended. What? It was a two level walk suspended walkway. And they had these big rods coming down from the ceiling that supported both walkways, one on top of the other. Right. And the, the design drawings were correct. And they showed you must use one long rod, which goes from the ceiling and then goes through a hole in the top bridge, like the top walkway goes through. And the same rod goes down and connects to and anchors to the bottom walkway so that both walkways are suspended to the same rod going up to hold it up. Right. Yeah. But somebody who made it went, oh, it's exactly the same. If I have the rods going down to the first walkway and then I'll do you separate rods that to then connect from the second walkway down to the bottom walkway. That's the same. It's the same weight. Right. It's the, it's the same load because it's, you know, when the thing, the same load on the ceiling. Right. And yeah. And no, um, there's.

Chris Gammell: It's very, very different. Yeah. Very specific reasons why that fails. If you look at it. Intuitively. It looks. Yeah. It looks fine. It looks right. But it's definitely not.

Dave Jones: It's definitely not. Which I learned to statics. Yeah. Yeah. Yeah. And it's, it's same with, um, abseiling. It's, it's vector loads. Right. I'm not using the terms correctly because I'm not a mechanical engineer. Right. But, uh, if you, um, like if you're anchored, your rope is anchored into a rock. Right. And you've got two. Let's say you've got two anchors. Right. For redundancy in case one pulls out. Right. Okay. So you've got these ring bolts embedded into the rock. Right. They actually, uh, glue them in, into the rock, but you've got two of them because you feel safer. Right. Well, it's, uh, if you put them far enough apart so that it's like a triangle. Right. So it's a lot like a Y shape. Right. So you've got a sling coming from one, say they're a meter apart and you've got a sling coming down to one point and, and it's joined. And that's where you put your rope into. Right. Then, then that Y that if you can visualize it, that, that Y spread, the angle of that spread means you're going to put more force on those anchors. And if you put them wide enough apart, even though you're only like, you know, 70 kilos or something. Right. So it's the same weight on the rope, but the force on the bolts multiplies the further apart you put them. Right. And it's like, you were safer off just using one bolt. Right. It's because of the vector forces. Yeah. It's like, does that exist in electronics?

Chris Gammell: You think, is that like something where like the math looks, I mean, what's the analogy?

Dave Jones: Oh, what's an analogy? Yeah.

Chris Gammell: I don't know what it would be.

Dave Jones: Oh, leave it in the comments down below.

Chris Gammell: I would think, I would think maybe like, uh, you know, like parallel resistors maybe are like unintuitive, well, they still are kind of intuitive, but I don't know.

Dave Jones: Something that multiplies, like, you know, that there's not unintuitive that, you know, you would think, oh, it's the weight. It's the weight. All that matters is the weight on the rope. That's 70 kilo person. Right. Right. So therefore there must be, oh, the, the weight must be split, you know, 45 kilos on each, you know, or sorry, 35 kilos on each bolt. Right. No, that's not, you can get the force multiplies because of the vector angle or whatever.

Chris Gammell: I feel like something with like, like resonance of like LC resonance feels like the same kind of like unintuitive weirdness, you know, I don't know. I don't know what it would be for electronics.

Dave Jones: Right. Right. Because you can get a multiplied voltage due to resonance. Yeah, exactly. You can, you're like a tank circuit, right?

Chris Gammell: Yep. Yep. Yep.

Dave Jones: Yep. In, in theory, you can go to infinite volts with a tank circuit.

Chris Gammell: Right. Right.

Dave Jones: If your Q factor is high enough. Right. Well, it has to be infinite. So anyway. Right. Yeah. If you've got a good analogy to what we're talking about here, hopefully people can visualize this over the radio. They can visualize what we're talking about, but yeah. Anyway, we'll link in the video. We'll link in the video.

Chris Gammell: Yeah. The videos. Yep.

Dave Jones: The video to the, the Tom Scott video.

Chris Gammell: Lots of, lots of flashbacks to, uh, intro to engineering classes. I'm sure.

Dave Jones: Right. I've, I've forgotten all that stuff. I just, yep. Yep. Yeah. In one ear out the other. If it wasn't electronics, I didn't care about it. You know, it was like, oh God. Yeah. Yep. Yep. Do I have to learn this? Why? I can do management classes. How many management classes did you have? Um, do you have any management classes? Was that part of, you know, you learn the, you know, laissez faire system and the, you know, all the different management styles.

Chris Gammell: Engineering. No, not, not like management classes. There was a lot of like engineering communications and stuff like that. And then there was an investment class at the management school that I signed up for, but that's.

Dave Jones: Oh, okay. Oh, but that was an optional thing. That was just for fun.

Chris Gammell: Yeah.

Dave Jones: Right. Okay. Yeah. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah.

Chris Gammell: What did you, what did you learn at the management classes, Dave?

Dave Jones: Have you learned how to be a manager? Yeah. I learned I didn't like it because it wasn't electronics. I didn't give a shit.

Chris Gammell: Yeah. You know, so I have, I've, I've been working with like subcontractors on stuff and it's been really quite positive as an experience, but like, it's kind of like taking me more towards the management side of things. And, and that has been, and, and, and not on a daily basis, more like just like on a one-off kind of like once in a while kind of thing. Mm-hmm. And it has, it has me kind of like, it has me feeling empathy for my past bosses. Uh, you know, just because like the things you have to deal with of like, uh, tracking other people's time and yeah, I don't know, like, you know, giving direction even just like, well, what should we do next? I don't know. Like someone has to decide that stuff. And, and when you're in a purely technical role, sometimes it's just like heads down, just go for it. You know, just, you have that one thing to do and you get to spend all of your, all of your time kind of doing that one thing. Uh, it's made me realize that I don't really like management. Uh, but you know, there's some, sometimes you just have bigger projects. You need to have some other people on board, but, um, yeah.

Dave Jones: Well, this, this came up on the EAV blog forum, uh, last night, actually about software management, managing, um, software projects and stuff like that.

Chris Gammell: And, uh, what did you think of that, Dave?

Dave Jones: Well, cause it was directed at me actually. Um, I, I won't go into details. It's over on the EAV blog forum, but it's that sometimes a management decisions are only, uh, obvious with hindsight, right? After the project's finished. Oh yeah. We shouldn't have gone that direction because, you know, we, we had just problem after problem after problem after problem. Right. And, but at the time I won't mention the project or what's involved, but like the.

Chris Gammell: Microsupply.

Dave Jones: At the time. But at the time it's like, oh, it's just one, one hurdle. Okay. Well just, you know, it, it seems like that, that that's, that's the last hurdle to overcome. Right.

Chris Gammell: Right. Right. But software is like bottomless, you know, like you can always keep coming.

Dave Jones: It just keeps coming. Yeah. Yeah. Yeah. You can always keep coming.

Chris Gammell: Right. And it's interesting too. I think that, well, I think that, you know, like a bunch of your products have software and the stuff that i've been working on has software too and like and it's it's amazing that it really makes the product a hundred times more extensible like i can't say what i'm working on but i can say that i'm replacing a device that has much less of what i've been doing right so much less software in general i could say that much and just like the extensibility of it like oh my god like the the the features we can add in and like being able to yeah access things remotely and you know just it's it's amazing it really is and it makes me think again about this is kind of what made me think about the team sizes too of like some of what the modern software enables is having smaller teams in general you just couldn't do this in the past you would have had teams that were either would have had a much longer timeline or you would have had a team that was five times as big because you just need more people to work on all the different pieces or you have lots and lots less features right but now it's just like you can do a software update you can add new features in you could fix bugs you can do all the stuff that you know that are still stressors of course right having bugs in general are stressors and having more complex products and systems and software are going to introduce more bugs um regardless of what jack cancel says uh right and unless you hire him you should hire him uh uh yeah and like all that stuff though it enables i'm not saying anything that anyone doesn't understand but like it's it i i marvel at it still you know it's just that that the that how much how much software enables these days so well and and this was the argument

Dave Jones: on the forum it's like at what point of at what complexity of a project do you like take extensibility and and modularity and all sorts of stuff so serious that you move from c to c plus plus and all sorts of fancy you know programming techniques which are great if you're developing microsoft windows or something like that you know um right just saying you know writing some right or if you're gonna be

Chris Gammell: building if if you're a company and you're like hey look this is a 20-year product i'm gonna be building eight different versions of this then it's like yeah that's right best of course but that's of course you don't know that like how would you how would you know up front to do it yeah

Dave Jones: and and it was like so we would compare what we're doing on the component forum is we're comparing what what i now call dave c just like dave dave pad right there's there's dave c how how how dave this missing mythical person called dave writes he his c code and um how somebody else would write it in an extensible thing so i posted an example from this mythical person called dave and uh right and it was like seven lines of code that did a box car average in filter right yeah so you know i posted the example of the code like it was you know seven lines of code because it just needed to do one simple job and that was it and somebody went oh yeah you know this is how i would write it as an extensible thing and it was like i don't know it was like a hundred two hundred lines long or something and it was yeah you can add all these and you can change the filter types and you can do all sorts of you know things like that whereas yeah okay great if that's what you wanted to do in your product right right no it was

Chris Gammell: yeah i know exactly what you're talking about right now because yeah it's basically like like dave c is much like chris c i'm sure where it's right i'm sure i'm sure we're very us hardware people doing software when you move to the next project and you're like hey i need to reuse this code uh you know we'll call him bob whoever wrote that hundred lines of code bob is sitting pretty right bob c or c plus plus probably is like he drops it in it works great it's got an api it you know it's tested the top level and like it just works right because it's written in that way and yet i have to go chris c has to go and copy paste and and copy paste it and retest it and like and just make sure it's working at the top level yeah and it's like it's called you know yeah exactly and it's like it's um it's of course it's better the other way but it's but it's also an investment and i don't know it's an upfront investment it's an upfront investment and you've

Dave Jones: got to weigh it up you know and also it's a you could also argue that it's an investment in the level of software engineer that you need right because if you're you know really pushing the boundaries of what you know c plus plus can do doing all this obscure wacky stuff right which which might be perfectly legitimate in many projects right but it's like oh gee is that legitimate because you want to flash a lead like do you need all that you know i'm boiling it down to a ridiculous level

Chris Gammell: right no of course but i think what what we were talking about before though too is like and when is it due you know what i mean like right yeah that stuff is great yeah yeah but a lot of dave c and christy is written because it's due and it needs to get out the door or it needs to get tested and it's like yeah that's right that's what like technical debt is of course that's what every software engineer knows that one but like sometimes technical debt is just going to build up and it's just i don't know and like it doesn't need to be maybe it doesn't need to be that crazy you know what i mean like i don't know right i'm sure we're gonna get a lot of hate mail for this like i feel like i feel it in my bones you know what i mean like i got like that ache you know what i mean like we're gonna get some like angry like because a lot of software people listen and i'd like to apologize beforehand but they're just shaking their head and yeah yeah i don't care really i mean like i know you're right you're definitely right of course but so are we yeah in certain circumstances here's what i will say i so when people write in because you know you're gonna write in the thing i would like to know about is uh a lot of the resources online are uh about writing like really high level stuff i'd love to get better at writing better c right and i know that there's some stuff out there about this but like it doesn't feel all that practical and like i think what i really love i would love to have like a workbook of like really well written c and like people like oh just go and go and get github and just like browse around no i don't want that i want a book of like like really nicely written c with like chris and dave c written next to it and be like no this is bad and this is good

Dave Jones: and here's how you can get better at it you know what i mean and like yeah but but but but there's no it's it's like saying here's a good pcb layout right everyone does it differently yeah right i i don't think there is you there's no one textbook just like there's no everyone asked me what's the best textbook there is no one best textbook no i don't want a textbook that teaches things the right way no i'm saying you know i'm using an analogy sure sure right but i i just want a collated version

Chris Gammell: so like for pcbs what i want is i want like it's hard to get like a good set of pcbs to look at too like like i like when you do reviews and you you showcase like hey this is a great format or this is a good thing to do like i want that like boiled down over and over and over again right hey do this thing oh like and basically it's like we've talked about on the show before of like building a mental library i want to build my mental library and i want to do that for c and i would do that for pcbs and i just want like a bag of tricks that i could keep throwing stuff into and hopefully have some kind of recall at some point um you know what i mean yeah no of course like some of that is just called experiment experience you know like that's just experience

Dave Jones: there is no one best way to write software and if you think there is you're you're kidding yourself and i'm sure the software people would agree but no they'll probably think their way is the best way

Chris Gammell: right well there's many hours spent arguing with people on the internet if that is their their true

Dave Jones: thought you know yeah honey i can't come to bed somebody's wrong on the internet that's right that's right oh boy classic yep it is a classic ah so there you go okay so i finally remembered what i was

Chris Gammell: trying to talk about earlier yeah okay here's what it was it was being excited when someone else yes

Dave Jones: that's it that's right that's right yes do tell do tell us a summary yes the summary is uh

Chris Gammell: i was dealing with a problem uh with uh a customer and not like not like a problem with the customer but we were working on the problem problem together and basically my stuff was potentially at fault and i was very worried about this and they were very worried about this and everybody

Dave Jones: was worried about this did you assume it was your thing did you have the spidey sense it was your thing at fault or was that just the fear thing oh like i had a fear thing it wasn't a spidey

Chris Gammell: but the thing was i trusted i trusted the person that was doing the testing they're very good troubleshooter right and that is that is worth 90 in my book because like if they're taking all the right steps and they're telling me the right steps i'm like oh boy this is it's like this like kind of creeping suspicion it's getting worse and worse i'm like oh yeah it's not that oh it's not that oh boy oh boy and i'm like throwing out suggestions and like they were they were super helpful and and they tried a lot of things that i asked them to and like they were awesome and it was just looking worse and worse for you and it was really looking bad today and then so finally i uh i was there and i asked that they swap out one final board and the exact same it was not my board anymore and the exact same condition happened and i was just i i kind of celebrated a little bit and i probably shouldn't have yeah because i wasn't happy i wasn't happy that it wasn't my stuff i wasn't like haha it's on you now see ya right yeah but it was i was just really happy to like we at least had like isolating the problem is so important and we had finally isolated it it was this one unit that was had this problem and it yeah so i i felt a little bad and i i explained myself i was so the symptoms manifest

Dave Jones: itself back to point towards your board did it exactly and then right yeah yes exactly and because

Chris Gammell: it was isolated in that one unit and and what we were doing is basically saying okay well there's other vintage of the the board like there's my board and then another board and we were trying on different units uh the other board was always working and uh oh one of those faults oh i'm sure

Dave Jones: there's a lot of people at home now nodding going yep yeah exactly and so it was basically like hey the

Chris Gammell: old one works fine and the new one which i made was not working great and everything we could find we could not replicate the problem with any of the old ones and finally it was a swapperoo with a different with my board with a different part of it and finally we verified okay it actually was a different part of the system than mine yeah geez i've been there yeah and luckily it had no consequences because it was just like one unit of many that could be fixed later so the deadline was made but excellent

Dave Jones: you had a secret celebration inside you had no i had i had an outward explanation but i definitely

Chris Gammell: explained myself because i didn't want to i felt like i was like i was like heck yeah and then i was like sucked in it's not my fault yeah yeah uh so it was that's not a good i explained that it was not like i was not like happy that there was a failing unit i was happy that we had found and isolated the problem and i think you had said you have also done something like this before no guy many times i don't

Dave Jones: know if i can remember struggling to pull a specific instance out i can just i i know i can remember the feeling oh yeah that's all i remember is the feeling of going oh it wasn't me oh yeah right

Chris Gammell: exactly yeah exactly yeah and it's yeah well especially on a ten thousand dollar board like

Dave Jones: you're talking about yeah it's something will come to mind but no i don't know about you but i like i

Chris Gammell: i usually assume worst case that it is me and that has gotten me in trouble a couple times but it also i think is the right frame to like it then chase towards the problem and say like hey if this is me i think it could be this this this and this and if not then i'm starting to clear my my plate of you know responsibility of course here i'm still going to help but like but you have to like you have to narrow down and isolate different elements of a system in order to tell what the actual problem was

Dave Jones: and right yeah i don't know oh huge relief i'm sure lots of people are nodding going yep yep yeah

Chris Gammell: i don't know i think i think troubleshooting is is such a quintessential skill and i i wish there was a better way to teach it i don't know what that would be i think some of it is just like you're talking about like getting it's not really you've got to do it yeah you've got you've got to do it and

Dave Jones: each experience is different each experience adds a troubleshooting part of your tool adds a

Chris Gammell: part of your troubleshooting toolkit yeah you know yeah yeah totally right and it's about i think i think it's about rigor as well you know like being able to like track your progress right if you have a really long troubleshooting streak you have to write stuff down i buy i personally yeah you know

Dave Jones: and oh i've worked on stuff that's taken weeks to troubleshoot you know yeah weeks and weeks like yeah you can't as you say oh did did i test that last week like you just don't exactly exactly you just don't remember i'm sure i've gone through and tested things like 10 times right because i just simply didn't remember that i had done it before and it's yeah and you kind of get desperate going oh you know it's look it's always pointing back to that fault so you go and test it for the 10th time and it's you know it's like yeah it's not that but but everything points towards that it's some other weird combination of symptoms somewhere that's causing some obscure fault somewhere you know i feel like

Chris Gammell: something it's always like it's it's always good to try i mean not always good but it's it's important to like list out your assumptions as well i remember a specific time at a past company where someone was working on a board is a weird like had a bad fault condition it was like i think it was like a uh oscillating op amp or something like that right and for some reason someone had a heat gun in their hand at the time and they're like oh whenever whenever the heat gun hits this part right right about here it it goes it starts oscillating and then if i get out the cold spray and i spray it here and it stops oscillating and so we would go back and forth hot cold hot cold hot cold hot cold and it's like yeah that is actually a really good way to start to like figure out things that are like temperature

Dave Jones: sensitive you've got to try and find patterns in in in the failure modes yeah but it was a complete

Chris Gammell: red herring and so like like how do you know that right it's like because your your entire like you're saying you check it a 10 times and you're trying the same thing 10 times when do you stop and think like oh i should you know try this other thing over here like it ended up being a component halfway across the board it was like a you know a diode or something like that

Dave Jones: or it ended up being the heat gun ended up being the noise from the heat gun actually in yeah exactly right it's microphonics or something yeah or the emf from the heat gun inducing something somewhere else that's and you think it's the heat but it's not it's the fact that you've turned on this giant magnet you know this giant coil like inches away from your board right right right yeah i've i've had that where i've worked weeks troubleshooting a board it turned out to be uh well the cause of the um problem wasn't real but it was um the thing that we were chasing down was caused by the air con switching on at night the uh but you know the things at night so we'd leave it leave it you know set up the trigger overnight i'm sure i've told the story before set up the you know scope trigger set up all the probes and the logic analyze and everything you know set up the trigger overnight because it was a very intermittent right you know fault that only came with time and we found yeah it was being triggered by you know at 2 a.m in the morning when the some system switched on you know oh my

Chris Gammell: god some yeah so yeah it's anyway that's terrible yeah yeah no intermittent ones are the worst for

Dave Jones: sure yeah oh boy all the ones that take a long time to manifest yeah yep like that's even worse you know oh look you have to work to you know you have to test it for eight hours before it fails it's like oh god you know you only get one shot right you only get one you know you have to make one assumption wait your eight hours and see if that's the problem no then you're gonna do it again wait you know oh that that wasn't the reason yeah you're gonna have another guess that's when you get

Chris Gammell: good at uh using data logging and and uh scripting and similar things it was a important important

Dave Jones: skill to have wow anyway yeah all right have we got news we've got news this week the most upvoted thing on reddit this is where you can upvote things and we talk about them yep i guess because we've only got like 10 minutes left tsmc i built in a 12 billion dollar factory in arizona five

Chris Gammell: nanometers geez when i was a boy you know it was five microns uh yeah a little smaller uh i don't really have a good feel unlike because i feel like uh you know like all the news releases are about the the leading edge type stuff and i don't really have a good feel for like what's even out there you know like some things are still using 23 nanometers i know most things don't need five

Dave Jones: nanometers you know it's like oh you know like 200 nanometers is fine you know for most of the you

Chris Gammell: know most of the chips out there yeah well and that is the that is usually the gate that's usually the gate width of a single fet as well that's like the that's like the yeah right what the nanometers usually refers to so and it's usually high density stuff and you know tsmc does does stuff for fpga

Dave Jones: vendors and similar yeah right really bleeding edge it's interesting what i think some of what's driving

Chris Gammell: this is the i'm not sure why tsmc went for it but i i read something dude politics you don't know

Dave Jones: it's politics and it's tax breaks you can't you can't tell me that uh the uh that the state government or and or federal government aren't driving this right because they want to onshore this with the whole china situation right yeah i think it's that too and of course our tsmc is not a chinese company they're a taiwanese company hence the name that's right taiwan semiconductor manufacturing corporation or something that's right yeah t t stands for taiwan so this is a taiwanese company yeah yeah and uh yeah no this is this is a political play bet your bottom dollar and they're probably getting like they're probably like paying no tax or something like that they're probably getting huge benefits to build this plant i guarantee i was also wondering if maybe they were

Chris Gammell: like sometimes they also chase uh talent in a certain area and i think that intel had some arizona facilities as well and like you think about like you need

Dave Jones: process engineers in arizona yeah arizona microchip as in microchip they're in arizona

Chris Gammell: right right right microchip yeah yeah which is it's picks it's kind of weird too because like you know fabs need to be like temperature controlled and stuff like that but like damn is arizona hot like it's just you know uh well it's nothing you can't solve with you know

Dave Jones: oodles amount of power with compressors lots of compressors and free on compressors and

Chris Gammell: yeah yeah i guess if you have a really solid power source you're okay but oh yeah no you're fine

Dave Jones: yeah it's all good yeah so yeah a lot of you're going to see that this is going to be huge right with the whole china thing won't get into the politics of that but companies are realizing holy shit we're incredibly dependent upon china so they're going to

Chris Gammell: yeah no i mean i think it's it's interesting i mean we've talked about on here many many times about you know our own reliance on it right i mean like you and i buy of course every part comes from there

Dave Jones: right of course chinese new year every time chinese new year happens yeah of course world stops

Chris Gammell: right i mean it's just you know right i just wonder if it's if it's lip service or if there actually will be you know stuff outside of china so oh i hope it's not lip service because i want

Dave Jones: local manufacturing to come back in every country i would love a diversified supply chain oh my god yeah like every country should be self-sufficient on the basics it's like medicine and things like that it's like well okay india and china manufacture 90 of the world's medicines okay yeah all the raw materials for the medicines it's like okay not good like you know if you can't get basic stuff like that if you can't be independent when stuff like this happens right this whole crisis thing and you know international travel stops and politics are involved and you know um yeah right you know trade disputes and tariffs and all sorts of things going nuts being flinged around for political power then it's yeah you really need local manufacturing so like i said here in australia we do not have a well no there's one but i'm going to essentially say we do not have a single place to manufacture a pcb to assemble a pcb even a pcb yes we do wow yes no we do not there there is there's one house left in all of this country that i know of and they do high-end high-end like you know rf military boards right you can't buy your prototype yeah if you bought buy a prototype from them it's like a thousand dollars right they'll they'll make it for you they'll they'll make your little piddly you know two-layer or four-layer board for you but it'll be like a thousand dollars please sir you know yeah um yeah no they they specialize in like high-end stuff in the whole country

Chris Gammell: the entire country well a lot of them are shutting down in the u.s as well though because like you know

Dave Jones: margins are not great oh yeah well they're probably other than a high-margin business yeah will you get tax

Chris Gammell: breaks will you get you know oh maybe but i think it's also that they're i think like the the vintage of when they were started to like a lot of the uh it's like mom and pop type businesses they kind of grew over time and then there also weren't people that were willing to take over the family business so and and like the business didn't seem like it was that good looking forward you know there was a

Dave Jones: prototype famous prototype uh pcb house here in sydney um rcs radio right and they made like really you know bare bones like none of the solder mask rubbish right but if you wanted a double-sided board made right you know that's where you went you know 20 30 years ago right that that's who you went to to get like a cheap in quote marks um you know double-sided prototype how about affordable affordable yeah affordable board made and uh yeah you know you'd get your double-sided tin plate you know thank you very much and uh you know it'd be plated through but you get your double-sided tin plate board and you know you're happy and uh but it was basically just one guy one crusty old guy running this yeah um everyone knows who i'm talking about if you talk about rcs radio but uh yeah crusty old um guy running this thing it was you know like if if you laid out your board wrong he would insult you he would say no i'm not manufacturing that you put right angles on your traces piss off you know like he would he would tell you to piss off because you put right angles on your traces i'm not going to make that you know i'm not going to manufacture that bugger off and learn how to lay out a board you know and yeah like a really crusty old yeah sounds like dave's mentor or something you know and anyway he was famous for this and once he died that was my pop right yeah and and once he died like he tried to sell the business but like no you know they tried to sell the business and nobody wanted

Chris Gammell: it because nobody wanted to take over well no one want to come work with him too you know well no

Dave Jones: he's dead now you know no i know i'm just saying but no nobody wanted this messy business of making pcbs like it's just not glamorous like nobody like you know no one wants to get to their up to their armpits in chemicals and making boards and you know fiberglass dust going everywhere and you know

Chris Gammell: yeah so yeah no i mean dave i'll have i'll be honest i had i was sitting here soldering parts onto a milled pcb today because i needed something right yeah and i was thinking like this is terrible of course you know this is like soldering onto you know there's no there's no solder mask it's tough to do it's small i did small parts for some reason because i needed to be tight tolerant or you know small size whatever and i was just thinking like i'm okay with this it wasn't my favorite thing but i i was fine with it but i had the thought like this is why people don't want to do electronics in general right like why not just let someone else out of this together like it's just so much easier to do that i'll be the idea person right that's like the the phrase that has been uttered throughout the u.s for the past 20 years and you know i'll just ship it over overseas what's the big deal it's like damn it is a big deal you know like you got to be close to this stuff sometimes and it sucks it's manufacturing sucks i will i will i will die on that hill it sucks but you got to do it anyways

Dave Jones: you know and many and many a kickstarter has died on that hill too that's right well well yeah they raise their money and think oh yeah i'll just i'll just toss it over to china and back my magic product from the oompa loompa factory and right yeah oh boy garbage in garbage out present yep yep exactly exactly boy anything else on the list we should talk about well i do do like the tear down of the polaroid insta insta camera oh that was fun polaroid that was fun yeah that's

Chris Gammell: an element yeah seeing the old school the the wire harness on there it's uh oh just the through

Dave Jones: whole parts and the individual wiring and like ah this is beautiful it would have been a bugger to assemble at the factory you know they probably had hundreds of people just they're soldering on the line all these stripping these individual wires and you know manufacturing sucks come on yeah it does oh yep no there's not i i haven't watched the full thing but i don't think there was any flex

Chris Gammell: technology in there was there no i mean flexible wires yeah flexible wires yeah that is right

Dave Jones: flexible copper wires there's your flex sunny oh boy yep and and i just love i love i sorry i gotta mention the photo of that can we make this the photo for this video for this video for this podcast that that control room in that soviet factory

Chris Gammell: oh i know oh my god that photo well it's multiple photos but yeah it's just like a bunch of different

Dave Jones: control rooms the first one is the one i'm talking about yeah it's just you know we don't we don't own

Chris Gammell: this photo but we'll see we'll see what the we'll see what the uh the license is on this photo oh i'm

Dave Jones: tweeting maybe old enough that it might be okay yeah yeah but it's oh it's just it's brilliant yeah yeah

Chris Gammell: so it's got i mean like and it's got the actual control diagrams like like stenciled on to with like the gauges it looks like it's probably like a boiler or something like that um but it's

Dave Jones: got all the gauges there some maybe it's some power it's some power plant some nuclear power probably something perhaps you know yeah and um yeah it's just like mid-century modern furniture and

Chris Gammell: everything like that too it's the guys are wearing butcher hats i don't understand a lot of what's

Dave Jones: going on here have you ever built a mimic panel no i haven't what's that oh mimic panels mimic panels are common in um all sorts of fields please leave it in the comments down below if you're uh if you you know see mimic panels mimic m-i like actually don't ask me to spell stuff i'm an engineer

Chris Gammell: i can't spell well i think like mimic like like to like to copy something yes mimic to copy a mimic

Dave Jones: panel is is like um look it's the one two three four five the fifth photo down in this one with all the colored diagrams right with all the color diagrams and the gauges oh yeah yeah yeah i see what you're

Chris Gammell: saying and people will if they go click on this link okay yeah and and so it's it's like in this

Dave Jones: case it has the you know the the system diagram there with all the lights and and switches and things like that and a mimic panel is generally not this is like a proper control panel the the real control switches are there and stuff like that but a mimic panel is something that like simulates the system it's usually just like lights you know leads back bulbs back in the old days and um yeah and it would just like simulate the system or start like the you'd have the silks as you said like silk screened overlay to the system then you'd have all these leads just mimicking what the controllers are doing and then it indicates so it's sort of like a remote uh visual display that it's you know that something's working it's like a uh fire panel in a in a uh building right you don't control

Chris Gammell: anything so it shows all the mimic shows the floors and stuff it shows all the floors and it where where the third floor has a fire fire alarm going yeah i'm in the toilet because somebody you

Dave Jones: know had a durry in the dunnies you know and of course right whatever that means durry in the dunny yep anyway uh aussie people mimic panels yeah mimic panels yes mimic panels yeah yeah that's cool yeah this kind of yeah i see i see how you're talking about

Chris Gammell: that too yeah it makes sense that it so like when you stencil the the thing on it it's kind of the same thing of yep yeah i just i don't really get the traditional drawings of uh yeah it's just like

Dave Jones: you can have this stuff on like a little lcd panel or something like that but it's not the same as as drawing the system in and then having the lights you know you can visually see where things flow

Chris Gammell: and actually you know dave are you looking at that what's actually really great about the detail of this photo too is that if you look in the upper left of the photo we're talking about so it has to explain what it is to everyone here it has all of the drawings it has all the valves that are

Dave Jones: looking like they're opening and closing whatever it's got like like pumps little diagrams of pumps

Chris Gammell: exactly yeah and then in the upper left there's a line drawing to and it says acc i'm guessing they accumulate maybe the output of that but they actually just wrote on the wall to make notes there's like multiple sharpie marks all over this wall to give to make notes of where there's

Dave Jones: different system problems yeah yeah terrific so it could be a whiteboard and you could write on it and then you could rub it out you know it's great anyway oh fantastic that's just oh it's just gorgeous photo it's just like oh the first one i just love it like i i just like artistically it's just like a the way just the way it's lit and the way you know this is this is dave's design sensibilities coming out oh it's just it's just i like and i love the guy sitting down at the table and he's got this little mimic panel um you know side panel down beside him on the floor kind of thing and it's just

Chris Gammell: oh that's the that's the mid-century modern furniture like i said oh right yeah that's like

Dave Jones: people look at this we'll let people look at this furniture yeah but that's it reminds me of the old phone table are you old enough to remember the phone table in the hallway i never i know what

Chris Gammell: you're talking about but i've never i never we never had one you never had so you do know what

Dave Jones: i'm talking about the old you know i do yep there was one phone outlet in in your house and it was always in the hallway and you bought a you know it was part of it was a piece of furniture you bought this you know phone table that had a seat you know a nice uh you sat at it and there's your rotary dial phone none of this push button rubbish and you sat there and you made your you know oh man you sounded pretty old i gotta say this i sounded pretty up you want me to sound even older

Chris Gammell: this for for a android or an iphone tab thing in your pocket come on man yeah no yeah not not i

Dave Jones: wouldn't trade it for totally yep i'm thinking about going back old school phone table you get you get

Chris Gammell: one of those you you let us know what mrs ev blog thinks about that oh absolutely you could get your polaroid camera you know your polaroid camera you know your grainy photos and uh you know your disposable flash bulbs and you know it costs 85 dollars for a pack of oh that's these days actually but anyways yeah no i don't think i don't i'm never i'm never going back nope not me right you can just steal some of the design elements here and then just make it modern that's what i'm saying right okay got it i agree it looks cool okay last two things uh there are two uh technical uh things in the subreddit one is a power topology handbook it's real it's got like every i forget i think someone on twitter pointed this out to me but it's got like every type of power topology which we've talked about the show before flyback and buck and boost and all that stuff but basically explanations this is basically like a mini book right to the point of full phase shifted full bridge converters and

Dave Jones: vinebergs and well that's this is almost identical to a video i was going to do like like doing an overview of or a summary of all the different types of converters and then i just thought about it for a while i just sat there thinking and i always thought of a new topology that went god how long would this video no it's like read through this man 20 different types of bloody you know like no it was just oh god no the yeah my eyes just started rolling in the back of my head at the thought of trying to summarize every different type of power converter topology it's like oh my god no yeah no it's has has this got a nice table summary in it of like the pros and cons of each one does it have that i do not it's got a nice brief set of formulas for each one i really somebody's gone how long did it who's who's the author of this marcus zahendena and matthias ullman matthias ullman yeah matthias ullman yeah yeah i'm an engineer i can't spell or pronounce german sound german yep anyway uh hats off like this this looks fantastic i'm just skimming through it now but yeah is there like a summary table of like pros and cons of which is best to because that was the idea of my video is that i'd listen you know that would be a good idea yeah yeah like what should you pull out of the the bag of tricks right when you need something right it's you know here's a you know it's it's a summary of course it's probably better as a it doesn't need to be a video really it could just be like a pdf like this that you know like like just simply a one-page table that lists 20 different types and the pros and cons like you know um you know is it isolated give up this give up this video business and just make pdfs there's a lot of money just make pdfs i could sell it yeah i could you know yeah just just like they sell those industry specific things for like you know two thousand dollars you can buy this report that's right yeah you know you can buy this industry report on who's the biggest 5g yeah or or who's the biggest uh microcontroller manufacturer and there's like 20 customers for this whole report and they charge like you know five thousand dollars yeah exactly yeah two thousand that's cheap yeah right anyway yeah all right last thing last thing

Chris Gammell: is uh a paper by past guest dr eric bogaton uh tdr and vna techniques for pcb characterization you may understand why i was looking at this so uh this is pretty short paper but it's nice uh way of comparison of tdrs and vnas so yeah if you're doing it if you're doing stuff like i was doing you'll be

Dave Jones: reading this thing oh boy oh i so want this to have a summary at the end then my video is done i'll

Chris Gammell: just copy this you're right please dave on video he's like page 85 now now kids turn to page 85 you could just do a live stream where you just do you just read it you just it's just yeah it's just

Dave Jones: by the fireplace that's right that's a smoking pipe exactly that's apple juice and uh right yeah oh my god this is just uh no there's no summary oh no oh there's a marketplace for it dave oh there's a marketplace i'm just mentally trying to figure out how much work it would be for me to go through and really give a definitive list of gee oh god no no no no too hard no i've got to put that in the too hard basket no sorry it's all right well we'll find out in two weeks whether dave followed

Chris Gammell: through on this tune in next time to learn more see you then catch you next time

Dave Jones: you