#478 – Optimization Beast

Chris Gammell: This is The Amp Hour Podcast. Release February 9th, 2020. Episode 478. Optimization Beast.

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: Two weeks in a row. This is a record.

Chris Gammell: Yeah, it's been a while, man. What shall we talk about? How about some electronics? Something different. Do we have electronics? I don't know. I mean, we've got chips we're going to talk about. I've got firmware I can talk about. I've got, you know, you've got, you know, moving your lab, maybe.

Dave Jones: I've got keypad design to talk about.

Chris Gammell: Yeah, yeah, keypad that came out recently. So, yeah, I think we're going to, maybe we should start to recap beforehand, you know, like get people interested in this thing.

Dave Jones: We still have no, like, plan or notes for the show, you know. We're just inherently lazy, aren't we? We're more post-hoc, you know. Right, okay. And a bit of a news from last time. We did announce that we're going to get an editor for the Amp Hour. We don't have one yet, but you're talking to someone tomorrow. Yeah, yeah. Who has bona fides in the industry. So, you know. Yes, so you did have, you did get some responses to the request. So, on the EV blog forum. And, oh, can we start, before I forget, can we, I know we don't just want to rehash stuff I've, you know, done on my videos and stuff like that. But it's inevitable, I guess.

Chris Gammell: Yeah, I mean, it's what you're doing. So, I think that's fine.

Dave Jones: Yeah. Can we talk about my impersonation scam?

Chris Gammell: Oh, I didn't watch that one, but.

Dave Jones: No, you didn't watch it. Okay.

Chris Gammell: So, is it someone like actually trying to impersonate you? Somebody impersonated me. I've tried to do your accent on here. You have. It's gone very poorly.

Dave Jones: Oh, okay. And, well, SO. Okay. Here's the TLD. Did I get that right? Yeah, you did. Too long, didn't read. Yes, I did. That's right. Oh, I'm hip and down with the kiddies. Lingo. Yeah. All right. Yeah. Anyway, somebody who I know who it is now, they impersonated me. They set up a Gmail account, dj.eevblog at gmail.com. So, you know, to look similar to my, you know, to make it look like it's me kind of thing. And they faked a footer as like an email footer as well, you know, saying, you know, David Jones from evblog linked in the website and all that sort of jazz. And they contacted a company who have, who shall remain nameless, who I've dealt with before. And they requested confidential technical information from the company. And the company, believing it was me, sent it to them in good faith. Right? Oh, my. So this is, yeah, this is a fraud to, you know, get, you know, confidential company information.

Chris Gammell: Oh, my. Yeah.

Dave Jones: And they got away with it. The company sent it. And I won't say what it is. But anyway, it's, yeah, they got away with it. It's actually rather clever. It's one of the oldest scams in the book, you know, impersonating someone. It's social hacking, you know, kind of thing. Yeah.

Chris Gammell: Yeah. I mean, same thing happens for phishing too, right? It's like, it looks like it's legit enough. And then.

Dave Jones: Well, it's, it's a variation of spear phishing is the term, you know, anyway.

Chris Gammell: Yeah.

Dave Jones: So yeah, pretended to be that. And if the company simply, you know, Google search that email address, they would have known that that doesn't exist. And I've never used it anywhere. So that should have, you know, rung some alarm bells. But yeah, it did it. And it got through quite a few people in the company. And, and they dug up the, they got the information and they, and they sent it to them. The only reason I found out about this is because somebody in the email chain, I had multiple people copied in my, my real email. And so I got copied in the middle of this email exchange and it's like, Whoa, what's this?

Chris Gammell: Yeah. That's pretty crazy.

Dave Jones: And yeah. And they, and they're talking to me and they go, Oh day, we've got a new product coming out. Would you like to have a look at it? You know, all that sort of stuff. So.

Chris Gammell: Yes, I would. Yes, I would.

Dave Jones: And yeah. So as it turns out, I had a hunch about something given my middle name is Sherlock Ohms. Huh. Get it? Ohms.

Chris Gammell: Yep. Yep.

Dave Jones: Anyway, I'm here, I'm here all week. And my hunch turned out to be correct. And I was able to track down this person and hilariously. So I found out their real email address. So I sent them an email and within the hour of sending that email, I was watching in real time as they tried to erase the evidence. So this is, you know, this is elite level hackery here for me, you know. Sure. Anyway, yes. So it's rather hilarious. And then I got another email and then I sent another email to them saying, Hey, I'm watching. And then they sent an email back going, Oh, my Gmail account was hacked. And like the oldest excuse in the book, my Gmail account was hacked. And I'm sorry for the inconvenience. It's like bullshit. Anyway, that's rather hilarious. So, um, yeah, this is a real thing.

Chris Gammell: That's well, I'm glad you, glad you got copied on there. That would have been bad otherwise.

Dave Jones: Yeah, no, it's, um, so I could see this having big implications in, I'm sure, you know, there are companies that big companies that have been hit by this by competitors and whatnot, or from people who want to get advanced information on products and things like that, you know?

Chris Gammell: Sure. Sure.

Dave Jones: Like I use the example in the video of like, if, you know, Apple's releasing a new iPhone and they pretend to be, you know, a 20 million subscriber, YouTuber, phone, phone review or something. And, and they get all the advanced information and all that sort of jazz. And so, yeah, you know, that can be, that can be a big coup.

Chris Gammell: So what I'm hearing here is crying, crying pays, huh, Dave?

Dave Jones: It could potentially pay. Anyway, I, so I did a video just warning companies actually be aware of this because this is, you know.

Chris Gammell: How often do you look at a email address like specifically if it just shows up as like from Dave Jones or.

Dave Jones: Yeah.

Chris Gammell: Well, they don't care about email from me. Right.

Dave Jones: And, and, and they're quite clever in that they actually did some research to find out who in the company I would have contacted and stuff like that. And they just dropped their name in the email, in the original email, like to make it small, persuasive that it's actually me. Right. So, you know, yeah. So it was rather clever. Anyway, I'm watching. It's just absolutely hilarious. I was able to track them down. So enough of that.

Chris Gammell: Well, how goes the, so you're moving labs maybe too.

Dave Jones: I potentially move in labs because I get the keys to my old lab back this weekend. So it is now vacant again. The, the, the lawyers who were inhabiting that have now gone. They didn't last very long. They flew the coop. And my lease here finishes in seven months or something. Finishes in October or something like that. So I'm just deciding what to do because David's left now. So I've got this hundred square meter lab, which is like pretty big for one person.

Chris Gammell: Yeah.

Dave Jones: You know. Right. And, and I would, I would love to keep it, but it's expensive, you know, and I'm trying to keep business expenses down as much as possible. So, you know, so it's one of those, you know, trade off things. I mean, who, who wouldn't want a bigger lab, right? There's, there's no such thing as too big a lab.

Chris Gammell: Yeah.

Dave Jones: Never. Right. Said that, said nobody ever.

Chris Gammell: Right. I think there's the, you know, a, it's like a volume of a fluid will, you know, fill the container. Yes. Yeah, exactly.

Dave Jones: The crap will expand to fill the space available. That's right. Yeah. Yeah. Yeah. Right. Which, which is the case here. I still, even with a hundred square, I still have that, that old fashioned snake in walking trail through the lab. So everything's on the floor and I just have like a track. They just pushed aside and I have a trail going.

Chris Gammell: Yeah. I can't do that. I, anything on the floor really, no, like floor stuff really bugs me. I like, that's like my first cleanup move. Anytime I'm like cleaning anything up, it's like, just get it off the floor. Even if it's just because it's psychological and it like goes up on a counter or something. Yeah. Because if I get restrained and like where I can walk and it's just, uh, that really, yeah. That part really bugs me for some reason. I don't know. I don't know what, cause I'm not a clean person, you know?

Dave Jones: Right.

Chris Gammell: Yeah.

Dave Jones: Oh, I cleaned up the bunker on the weekend.

Chris Gammell: Oh yeah. Okay.

Dave Jones: Yeah. Yeah. I, I thought bugger it. I'm going to, well, I kind of, I started throwing things out. Um, and I, I, I had, I had 15 monitors in there. 15 computer monitors that I'd got from the dumpster and I went, no, this is a sickness.

Chris Gammell: I wasn't gonna, I wasn't going to do an intervention or anything, but you know, I don't know what your plan is for all that stuff, but it's like, you're not harvesting components off of them. Yeah, I know. All of your dumpster diving is, it has consequences.

Dave Jones: It does have consequences. And I realized that after the, all these monitors, 15 of them, so I decided to bugger it. I'm just going to give them away. So I gave them away to a worthy cause who, uh, who needed these things. Yep. So they came and picked them up and, uh, yep. Gave them all away so that we will be, uh, recycled and used. Great. Um, for, uh, kids actually. So yeah. Even better. Yeah. Yeah. Yeah. It's great. Yeah. So yeah, I, that, well, that was my first, that was my first move in cleaning up the lab and I tossed a few other things out and I've got, I've got more to go, of course, but you know, so that's, uh, that's part of my process. But anyway, um, the, the audience out there, tell us how you, uh, is your lab just, you know, organized chaos or do you can't stand it being like that? Do you have to have everything labeled and categorized?

Chris Gammell: And yeah, I would, I would be willing to bet that that's, uh, uh, part and parcel of like how much space there is too. I think that when you have more space, you know, you feel a little bit more comfortable letting things go like, Oh, of course. Yeah. Yeah. I had like my tiny lab in my old apartment and like any amount of clutter, well, it was already pretty cluttered, but any like extra clutter past the organized clutter. And it was just like, you couldn't do anything then, you know, it was like literally, you couldn't move. No, that's right. Yeah.

Dave Jones: Well, I used to be like this when I was in the garage, right? I was working for a couple of years out of the garage and, and longer than that were for my, you know, um, hobby and contract, um, stuff as well. Yeah. I, I had like a five square meter lab, right? Five square, I don't know what that is in feet. Sorry. But yeah, the five square meters is tiny. Like it is one, right. It is one 10th of the size of my old lab, which you've been in. Um, so yeah, which is 50 square meters. So it's one 10th that size. And yeah, you, you had no choice but to organize things and, you know, keep it clean. Um, yeah, exactly.

Chris Gammell: Right. And it's, I think there's some benefit to that. I mean, I, sometimes, some of it's just me convincing myself these days. Um, but, uh, you know, like the, the need to like clean up at the end of the day and stuff like that too. I think we've talked about that on here before, you know, like there is at least a forcing function there where you have to kind of like close things down. I think the only downside to that really is when you're doing like a long-term test, if you're doing like a battery, battery drain test or something.

Dave Jones: That's why I need quite a few benches. I've got like six benches because like it's where you put your work in progress stuff.

Chris Gammell: Yeah. But I think, I think that the, uh, you know, it's easy for me to be like, oh, everything's a work in progress. Exactly. Yeah. Yeah. That's right. I know.

Dave Jones: It can get carried away, but.

Chris Gammell: Yeah. I saw a really sweet, um, modular setup recently, uh, where it was like a build plate. Um, I can't share too much about it, but it was like a build plate. And then it was like, they had prefab like cables that were like on their bench. And then every build plate they did for a project came over to the bench. It was like, yeah. Oh, right. Yeah. Yeah. Got it. Really nice. Yeah.

Dave Jones: I've done that in companies where, you know, everything's like really formalized and stuff like that. And a lot of companies will have a clean desk policy at, at the end of every day, your desk must be clean.

Chris Gammell: Wow.

Dave Jones: Yeah. Everything has to be put away. It's, you know, it's, it's once again, it's that forcing function, you know? So yeah. So please let us know in the audience.

Chris Gammell: And some of that, some of that security, I think too, right?

Dave Jones: Oh yeah. It can be. Yeah. Well, I've, I've worked in that where you work on a secret classified stuff. Right. And you like, it has to go in the locked safe at the end of every day. You know, if you're not using it. Yeah.

Chris Gammell: I've never, I've never worked on that. What's, what is that like in general?

Dave Jones: Oh, it's, you don't have to do it that often because you don't have to reference the classified material, you know, every day. But when you do, you can't leave it on the desk overnight. You know, you can't leave it in your drawer. It's got to go back into the locked security cabinet, you know, with, with the combination and the whole works. Right. So.

Chris Gammell: I read the, in that Skunk Works book, they actually talk, I mean, the, the guy who wrote it was like the former director of the lab. I don't think it was the stealth program. Maybe it was the SR 71. But the, he specifically was pushing for not having it top secret because like the amount of like overhead that they had to have. And like, yeah. Like the same thing you need, like security officers, you need to be like audited all the time. Of course. And it was just like, he was pushing really, really hard. And if I have not promoted that book enough on here, it is a fantastic book.

Dave Jones: You were, you were promoting it to me again personally the other day.

Chris Gammell: You would, you would love it, man. Like that is like, in terms of like nonfiction, it's like perfectly aligns with like, like nerdy, but like historical and like, I don't know. And like the SR 71 is just like the most amazing machine. Of course. Yeah. The fact that it was built in the sixties, like it's just so unbelievable, like that they built that thing. And yeah, it was, and you know, the interesting, the interesting factoid about that, right? What? It was actually the RS 71. Yes. That was the official name. And then a general put it backwards on a piece of paper. Oh, I didn't know they put it back. They changed everything because of that one mess up because they didn't want to make the general look bad.

Dave Jones: Okay. I didn't know that.

Chris Gammell: Isn't that amazing? Like, oh my God. That's pretty good. It's so bureaucratic and ridiculous.

Dave Jones: Great.

Chris Gammell: Yeah. Yeah.

Dave Jones: We have some news. Straight off teletype. Sad news, unfortunately.

Chris Gammell: Unfortunate. Yeah.

Dave Jones: We do have another passing of an analog legend. Barry Gilbert from analog devices fame, among other things. The Gilbert cell. What is the Gilbert cell, Dave? The mixer. I'm not an expert. Come on. It's a mixer. You know, it's a, it's a mixery thing. I'm not, you know. I asked actually quicker because I was like, I don't know what the hell that is. It's, it's, it's an analogy mixer used in chip design and stuff like that. Oh yeah. Well, you could do it discreet. I'm sure. But yeah. Yeah. Yeah. Anyway. Um, yes. Uh, just passed away. Aged 82, unfortunately. So we'll link in a, uh, a couple of videos. I've got another video for you here. Another, uh, presentation. Somebody put it on the, uh, forum as well.

Chris Gammell: Mm-hmm.

Dave Jones: But we've got a couple of videos. You can watch him talking about, uh, translinear 55 minute lecture on translinear circuits for all you translinear fanboys out there. I'm sure there's a lot, you know, it's, it's, it's a really nice, um, really nice presentation. So yeah, that, that's on the, uh, they've hosted that on the computer history museum, which is a good, um, source of stuff. If you don't subscribe to the computer history museum, they have, they, they release tons of stuff. It's just amazing. Tons of interviews.

Chris Gammell: So we had talked about, um, when Tom Lee was on the show, um, and it was just me and Tom, but the Tom is out doing a lot of interviews with like people that are getting older in the industry and, you know, just to make sure he preserves knowledge. This is actually something that the computer history museum as well does like Dave saying. And it's just like, it's, it's historical archiving where you're not going to get that knowledge otherwise, you know, just like, of course. Yeah. You know, like there's just so much, so much latent knowledge out there that is unfortunately passing away. And so it's really good. I mean, let's support the computer history museum if you can. It's great to watch those old videos too. I think that was one, I think Jim Williams and Bob Dobkin did a video set on there too before Jim passed away, you know? And it's just like a lot of these, a lot of these older engineers were like, why, why would I be interviewed? You know, like it's like, uh, you know, they have, they obviously have stories, but I think that these are not people who have the limelight. You know, they're not like, they're not, they're not egocentric like Dave and I, and they actually were talented. Exactly. Unlike us talentless social media people. Yeah. Well, you know, um, Dave, we get stuff in the limelight. We need to highlight things, but, uh, but yeah, that's what that's really great about these, uh, these interview series. And so I'm really excited about hearing from, from Tom as well. Once, once that goes up, um, he's going to try and talk to all the old tech, uh, tectronics engineers and stuff like that. Nice.

Dave Jones: Yeah.

Chris Gammell: Yeah. A lot of them are, uh, what, what is, what is like the top tier usually for like, uh, it's like staff scientists, but then it's like emeritus effectively. Like a lot of them are like, yeah, I think maybe even Barry was listed as this at one point, but they're like, you know, like lifetime appointment of like awesomest. Yeah. Awesomest, uh, engineering kind of thing. So that's how you usually find them. Usually find them at companies listed like that. They're like, we just want their name on our webpage and they still come in once in a while.

Dave Jones: Yeah. Yeah. Of course. That's right. Give them an honorary title.

Chris Gammell: And yeah. Give them a cube, let them come in, you know, like keep them, keep them telling stories and, and, uh, showcasing what they know. And yeah. Cool.

Dave Jones: Yeah. I think, um, analog built a dedicated R and D factory just for Barry. I think just so that it could, it could lure him back or something like that. I think I read that it could, they, they wanted to lure him back and they, so they devoted a whole R and D facility.

Chris Gammell: It never, it never like, uh, ceases to amaze me what companies will do when it's like, when someone's important enough. I saw, uh, you know, a smaller version of this today, but it was like a, it was on Twitter. It was like a bus, the bus schedule for like how far people get bused into the Bay area. And it was like, you know, people like three, four hours outside, you know, they're like, Oh yeah. I mean, just because I think it's a combination of people that are living that far because they want to be in nature, but also because it's cheaper housing. Uh, but then there's these buses that you get on and you're just like, you know, you can do work on them. They're comfortable places.

Dave Jones: Yeah. Yeah. Right.

Chris Gammell: But Holy crap. You know, like that's insane.

Dave Jones: I know that that's a big chunkier day.

Chris Gammell: Just traveling. That's insane. And so that's one thing, but then, you know, you'll see this as well. It's like there, Oh, there's a, there's a chip design center in like, I don't know, like middle of nowhere, Montana. And it's like, well, what, where did that come from? It's like, Oh, it's a one person office. Oh, right. Okay. Right.

Dave Jones: Cause that one person that they couldn't dig, you know, exactly.

Chris Gammell: They're not working from home. No, no, no. They're not working from home. They, they are working at a company office that just happens to have one employee in the middle of nowhere, you know? Like, yeah, it's, and I think that like, that's how a lot of companies get, you know, when you think about, I think about some of the people I've worked with and they're just like, they're just so hard to replace. And then, you know, so the companies are like, well, what's our, what's our option here? We can either pay them as much money as they want. And it's sometimes, and then eventually money stops having its luster, you know, like pay someone a couple, you know, a lot of money. And then eventually they're like, okay, well, what else can we do to keep them around? And it's like, well, maybe they can live somewhere else or, you know, and it's, um, yeah,

Dave Jones: that's that, that reminds me my mate, Doug Ford, who's like an analog guru, right? He, he used to be the head designer at Rode microphones, right? He used to design all their mics and stuff. And of course, you know, that's a pretty important position, right? Yeah, right. You know, like it's, it's the important position, right? He was, he was the number one person there that designed the, designed the products. And he decided to, uh, Lee, you know, move on and, uh, and, and leave. And they were just so desperate to, you know, to keep him on. It's like, we'll give you anything. What do you want? And he said, nah, nah, I'm done. You know, I just want to move on, you know? And yeah, but they would like, yeah, anything you want. Doesn't matter. Yeah.

Chris Gammell: Yeah.

Dave Jones: And nah, nah.

Chris Gammell: I mean, sometimes that's how it goes, right? But it's like, but, but, but that like upends, that upends like an entire like structure of like, uh, I don't know if, if I think about like what executives should be doing, I would think you'd be analyzing and trying to like figure out if there is anyone like that in your organization and at least having them train, train someone else at the very least, right? Like 20 hours a week is to train three other engineers or something like that. And it's never going to be the same thing, but you need, you need some kind of redundancy. Yeah.

Dave Jones: I mean, that, that, that could have set a company like that back a year or two in terms of R and D product development, right? Because how many world-class microphone designers are there in the world? Exactly. You know, like, yeah. Count them on one hand, I'm sure.

Chris Gammell: And even if it's like, you know, you're just, so Doug's example, if he was just training, even a technician just to like know how to tune a mic or, you know, maybe it's not for design. You can maybe like survive by just making the same thing for the next five years while you find that next, you know, mic engineer.

Dave Jones: That's might've what they've done, you know? And you can like hire like a really top, you can hire top design engineers, but they're still not going to have, there's still a huge ramp for them to come up to speed, you know, in terms of, you know, getting that specific product knowledge, for example.

Chris Gammell: Yeah. And we were talking about Doug last week. Cause I, I had, I had confounded his name with another Doug. Who was the other Doug that I? Doug Self. Doug Self. That's right. So it's Doug Ford, Doug Self. Doug Self, but he does audio books though, right? He, he does audio books and stuff. That's right. Right. So different people. And Dave knows the Ford, not the Self. That's right. I do not know Doug Self.

Dave Jones: Yeah. And he has a website.

Chris Gammell: Who does? Both, both probably do.

Dave Jones: No, Doug Ford doesn't know. He's a bit of a computer Luddite. Oh, okay. Okay. Yeah. I got it. No, but Self does.

Chris Gammell: Yeah. Yeah.

Dave Jones: Like Self does. You know, the design of active crossovers out now.

Chris Gammell: Yeah. Yeah. Yeah.

Dave Jones: All that audio stuff. So yeah. Maybe we can get some audio people out because there's a lot of interest out there for audio design stuff. Yeah.

Chris Gammell: We've had a couple recently. So we had Nash who works at Sonos. He was on here recently. He was more on the digital side. Right. But, and then Kendall was also kind of, kind of audio. He does, you know, filters and stuff like that. So the Kendall Kester Perry was, was two weeks ago. So yeah. But I think, I think generally the reason I had also been looking for audio people is because we don't have a ton of coverage. And I think there's a ton in the, you know, somewhere in the analog realm, you know, like depends on where you are. It's like if you're making somewhere from microphones to tube amps. So it's just such a big industry that it's kind of crazy how much stuff's out there. But if people have suggestions on people they'd like to see, we do have a suggestion form on our website for guest episodes and they can, you can throw stuff in there. So it's under the for you link, which is probably not very well named, but there's a.

Dave Jones: No, I like that how it's categorized for you, for us.

Chris Gammell: You know, I guess that is under for us. It's under suggestions. So yeah, more audio though, maybe in the future. It's a good idea. What else has been going on? Well, you may have heard about a conference that I ran last year.

Dave Jones: I heard about that. That's so that's big news. Yeah.

Chris Gammell: So KaiCon was a smashing success last year is by all accounts, you know, as the person running it, of course, I'm going to say that.

Dave Jones: Well, yeah, I'd say so. How many people?

Chris Gammell: So last year is 220, 215 people last year. And then we started talking about that. I was like, look, I can't do it.

Dave Jones: You need 2020 people this year, don't you?

Chris Gammell: Yeah, right, right, right. But I was like, I don't want to do this. I can't do it again this year.

Dave Jones: No, you couldn't? Okay.

Chris Gammell: It was just it was very, you know, it was very difficult last year. It was a lot of time and stress. So where was the conference last year? Was it in Chicago? Chicago. Yeah.

Dave Jones: Yeah. Right. It was in Chicago.

Chris Gammell: Yeah. It was at my workspace. Yeah.

Dave Jones: I forgot to mention is where the imposter is from.

Chris Gammell: Oh, yes.

Dave Jones: I'm watching you, Chris. The flat thickets. Maybe it was me. Yeah.

Chris Gammell: It was my evil twin, Chris. All right. With a K. Okay. Yeah. So it went well last year, but we started talking about, we talked about it at the time. We're like, oh, maybe we could do it at CERN next year. And then some of the developers started talking about it. They're like, we really want to do it again this year. And we talked to the CERN folks and they're like, yeah. So we are going to be having KICON 2019, 2020 rather, at CERN in Geneva, Switzerland, September 11th to the 13th of 2020. So it's a little bit later, but that's because we needed to get all the facilities there. They actually have like.

Dave Jones: They've got conference centers and stuff that handles 2,000 people?

Chris Gammell: They have, not 2,000 people. A lot of people though. They do a ton of conferences there and they are, you know, because they're like publicly funded. So they have like a charter to be like a resource for a lot of the scientific community. Yeah, of course. And you can look at all the, oh my goodness, like if you look up the conferences that go there. Oh, man. Oh, sure. You know, it's some smarty pants people that are going there because it's obviously at CERN. There's the Large Hadron Collider there.

Dave Jones: And all of them collect spores, molds, and fungi.

Chris Gammell: Yeah. Yeah, for funsies. Yeah. Yeah. But they, because of that, they had to, you know, the resources. So they have, they actually have like a hostel on site. They have a restaurant on site that like does all the food and stuff like that. So there we go. Analog computing, past, present, and future, February 14th. That's coming up, Dave.

Dave Jones: Wow.

Chris Gammell: Wow.

Dave Jones: What would be the average age of the attendee there?

Chris Gammell: I think they said future. Come on.

Dave Jones: Future analog computing. Okay. Yeah. Yeah. But that, that doesn't involve like quantum computing and stuff like that. So.

Chris Gammell: I think it's like multiple states kind of thing. I think that's probably what they're, you know.

Dave Jones: Oh, okay. So they'll be given lectures on qubits and. Yeah. Probably stuff like that. Yeah.

Speaker ?: Right.

Dave Jones: Yeah. Yeah. Yeah. Yeah.

Dave Jones: Fun stuff. Awesome. See, I would, I would love to go, but you know, it's like. Oh, that's.

Chris Gammell: I think that's pretty, even further than here. So yeah.

Dave Jones: It is further than. Yeah. Chicago. Yeah. Yeah. So.

Chris Gammell: So yeah, I'll, I'll be talking about as we get closer. I'll probably try and, you know, fly in somewhere other than Geneva and take a train over just to have the experience. And so maybe flying to Paris or similar. If people are in, you know, the Europe area, you know, Europe's not that small, right? But yeah, if people are going, there's a, there's a mailing list on the site, which is 2020.kicad-kicon.com, which is not a great website name, but whatever. We'll link it as well. So.

Dave Jones: And the biggest frequently asked questions is, can I get a tour? Can I get a tour?

Chris Gammell: Yes. Yeah. There will be tours. It'll be tied to the ticket though, I think, because I think there'll be limited numbers. So sign up early. Well, right now you can just sign up for the email address. But, but yeah, that's, I mean, like, I definitely want to go for that. Like, holy crap. You know, like, I don't know what parts we'll be able to see. There's, you know, there's obviously lots of parts.

Dave Jones: Well, everyone wants to see the main, the main receiver. You know. Atlas. The Atlas receiver.

Chris Gammell: Is the picture that everybody sees. I don't think. That is the money shot.

Dave Jones: Everyone wants a selfie with the Atlas.

Chris Gammell: Yeah, I don't think that's, that's probably not going to happen. Because I think they will be in shutdown. Because they're reworking all of CERN right now.

Dave Jones: Oh, I like. Okay.

Chris Gammell: Yeah, they're like, they're like souping up all the magnets in the, in the loop and stuff like that. And just like reading some of the specs. I actually had a chance to talk to someone out here who works at Fermi. So Fermi's out here in Chicago. They have, so I have a accelerator. And just like the, the power numbers they deal with and the speeds of, you know, capture they have to do on these things. And just how much general technology is in there is just so bonkers. Yep. You know. And you just, I, I guess when you're, whenever you're doing like a scientific thing, you're on the edge of, of what's capable. So it's, it's very impressive engineering. I'm very excited to, to see it. But I think also just to talk to people that are, are nearby it. Um, I still haven't done a tour of Fermi lab. I need to do that. Um, but it's like way the hell out. There's actually like, uh, they have like this like monolithic, like building structure. If you look up the, the, the photo online and then they have a herd, a herd of bison that are around there. What? Okay. You know, you know, the bison, you know, bison. Yeah. I know. Yeah. So they, they were like almost extinct in the, in the U S but they have a herd that they protect there. And, uh, yeah, it's, it's really cool. So like on all the branding, there's like bison on everything. So it's, it's fun. It's fun.

Dave Jones: Got it.

Chris Gammell: So, yeah, my goal is to pair that trip. When I go out to Fermi here in Chicago is to do that, pair that up with a, uh, skydiving. I've never been skydiving and there's like a bunch of, because it's just a big flat area out there, you know, like go skydiving and then go to Fermi lab. I don't know which would be more exciting.

Dave Jones: I thought you were, you were terrified of heights.

Chris Gammell: Oh, I am. But you got to jump out of a plane once in your life, right? Dave.

Dave Jones: Right. Okay. Fair enough.

Chris Gammell: I'm not going to do what I'm, yeah, Dave's talking about. Cause when I went to Australia, I was afraid to do everything.

Dave Jones: Oh dude, we were climbing down some ladders on the side of a cliff and you went, this is my limit. This is my limit.

Chris Gammell: I was actually looking at video of that the other day. It was, uh, yeah, it was, it was pretty, I'm not, I'm not a fan of heights. I don't. Jeez.

Dave Jones: I'm sorry. I don't know. I'm very surprised. You seem rather blasé about going skydiving.

Chris Gammell: I mean, you know, if it doesn't work, it's a great way to go, right? Right.

Dave Jones: Yeah. Okay. Fair call. Yeah.

Chris Gammell: Just a nice solid splat, you know?

Dave Jones: So it even works or it doesn't, you know?

Chris Gammell: That's right. That's right. Yeah.

Dave Jones: Well, there is a middle ground there where, you know, you land really hard and you break your spine or something.

Chris Gammell: Sure. Sure. Well, you know, how likely is that though? Speaking of things going up in the air, uh, one of our former, two of our former guests actually worked at this company that's been just, just announced its existence to the world. Todd Bailey, who was on the show a while back and Jerry Ellsworth both worked on this, uh, small, it's a small rocket maker is running a different kind of space race. So basically, I didn't know Jerry worked on it. Yeah. She worked there. I don't think she worked there towards the, I think towards the beginning of when this was around. I think she was doing some of the control electronic stuff and Todd also was working on that. And, uh, yeah, so it's an interesting idea.

Dave Jones: It's like, hang on. Where'd she find time for that? Because I thought she went straight from straight into one VR, one AR startup to the next.

Chris Gammell: I think when the thing shut down, I think that's when she was.

Dave Jones: Oh, right. Okay. There was some downtime there. Yes.

Chris Gammell: I was referring people to that episode. If you haven't listened to that episode, I list that as one of our top episodes in terms of like gut busting laughter of like dark humor. Uh, I think it was, I forget what it was. It was like Jerry returns to talk about business or something. Oh, and the demise of Cast AR. So we'll link that as well. It's episode 394. That's definitely like, uh, it's basically our small version of the show Silicon Valley. It's, it's basically Jerry, Jerry, uh, spilling the beans. Spilling the beans. Yeah. Yeah. Yeah. Yeah. So yeah, so Todd, Todd was working there too at the, uh, at this, uh, this startup called Astra, um, which is like a play on ad Astra, I believe. Right. Oh, okay. All right. It's like stars. Right. Isn't that what Astra means?

Dave Jones: Yeah. Yeah. Okay. Fair enough.

Chris Gammell: And it's like a smaller version, you know, so like SpaceX, I just actually just finished listening to the, uh, the Elon Musk biography, which is like five years old at this point, but holy crap, man, that's the story, you know, like just listening to the depths of the story and everything. It's a, it's a, it's an interesting listen. Uh, but the, uh, but the size difference is really interesting. So it's like a, so it's like a 40 foot or like a 13 meter tall rocket instead of a 220 foot or what? A hundred meter rocket. No, that's not right. Uh, 80 meter rocket. So, you know, pretty, pretty different in size, but in terms of like payload as well.

Dave Jones: And so they're targeting for those who plan long at home, they're targeting 165 to 450 pound payload. So that's, you know, that's like, that's not a human person. Well, that's basically, you know, barely a person. It's a satellite.

Chris Gammell: I mean, that's really what it is. It's a, it's a true commercial play and for satellite type stuff.

Dave Jones: And their cost per launch claiming 2.5 mil.

Chris Gammell: Right. Which is much less than the claim than the average for, I think, I think they said SpaceX is about 60 now.

Dave Jones: I think they're trying to get it down to $60 million. Yeah.

Chris Gammell: They're trying to push that down to 30, I think.

Dave Jones: Right. But still, that's an order of magnitude greater than this one's claiming. So. Right. Right. But also then historically, I mean, I think it's like, it's always like cost per payload.

Chris Gammell: Yep. And so, or what is it? Price per pound or price per kilogram. You know, like that's, that's ultimately the measure I think that they do. Because then there's also the cost of the, there's like, that includes just like fuel costs and operational and everything like that. So.

Dave Jones: I was wondering who they got their funding from, but it turns out Airbus is one of the funders.

Chris Gammell: Mm-hmm.

Dave Jones: Mm-hmm. A Salesforce co-founder and various other companies. Yeah. Rich people. Rich people love space. Yeah. Let me tell you. Oh, yeah. If I had billions of dollars, yeah, I'd probably be throwing, you know, 50 million to some space company just, just so I could go there and hang out. Yeah. Right. And, and just, yeah. And then just be involved. Yeah. Of course. Yeah. Totally. Yeah. I do.

Chris Gammell: I mean, I, I've not as into space as Dave is, as people probably know from listening to this show in the past, but I do wonder just about the logistics of like, like there's just a lot of stuff going up and like.

Dave Jones: It's rocket science, dude.

Chris Gammell: No, no. I get that. I'm saying actually like stuff in orbit though. Like. Oh, yeah. You know, like, like.

Dave Jones: And it's just exploded with Starlink. Yeah. Yeah, exactly. So it's just like insane levels now.

Chris Gammell: Yeah. I wonder, someone actually, a couple of people wrote in about this. The, we talked about that last week on the show. Look at this, we get to do a follow-up the week after. How different is this? Uh, they wrote in, uh, the, the difference in timing is because fiber is going through glass, which is like much, much lower speed than going through, through free space. And so that is, that is one of the biggest things. So there's a really good video of explaining it with like the guy did like homemade simulations too. So Paul had written about this and I forget who else had written in about this, but they're actually like these homemade simulations where he's showing all of the Starlink satellites when they're operational and like if they're going laser link or after, if they're going down to the ground and back up, if people don't remember Starlink is the, the, um, SpaceX version of space internet and, uh, and just the math, the math around it is insane. But, but like a, a New York, New York to Seattle link could be as, as much as half, half the time, um, of what it is currently. And so that's what I think is going to pay for it, but then it'll also give a lot more access to people around the world, which is pretty, pretty crazy. I have a search going on Twitter and it's, it's, it's a mix of people that are like, uh, it's like, I'd say it's in thirds. It's like a third of the people are like speculative investors. A third of them are like, just like rah, rah cheerleaders of SpaceX. And a third of them are pissed off astronomers. Let me tell you.

Dave Jones: Yeah, exactly.

Chris Gammell: They are very upset because, uh, there's actually a site you can go to, which I think was sent when someone emailed us too, but you can go to and actually see when these are going to be going over a little simulate when they're going to go over your, your, um, your part of the world. And so you can see the, the train of, of, of Starlink satellites. But, uh, yeah.

Dave Jones: Tesla had to change, change the design of them. They had to paint them matte black so they wouldn't be reflective as, as reflective. Yeah. As reflective. Yeah, exactly. Which was, you know, I can't believe like you can like literally do rocket science, get, you know, thousands of satellites design and get them up there and then not think of that. Right. I mean, it's just like, it's nuts. Right. Yeah. I mean. How can you screw that up? I don't know. Anyway. Yeah. So anyway, so I, I am watching the, um, the ping animation thing now, I guess you could call it. Is it a ping animation?

Chris Gammell: I think so. Yeah. Yeah. Yeah. Yeah.

Dave Jones: Something like that. And, um, yeah, going from one side of the U S to the other. And, uh, yeah, it's, it's rather fascinating. Yeah. So, so they jumped from one satellite to the next.

Chris Gammell: Well, no. So the right now they don't have the laser link to go satellite to satellite. That's what they're, they're claiming they're going to do. Okay. So the first simulation that you watch on that video, which I said, we'll link, um, is, uh, it goes down to base stations. So it's basically, and then the idea being that you could maybe even do peering where instead of having specific like SpaceX base stations, you start to use people's modems in their homes to kind of piggyback traffic and stuff like that. Got it. So I just amazed at like the idea of like, like the bandwidth and throughput, you know,

Dave Jones: a distributed computer thing. Yeah. In theory.

Chris Gammell: I think, I think right now they're just pieces of things, you know, they're just in test mode. So I don't think there is anything. I don't know if there's a plan released for that, but.

Dave Jones: Yeah. I, I haven't heard of any services that you can buy yet. Yeah.

Chris Gammell: Yeah. There's, oh, there's definitely not. Yeah.

Dave Jones: All right. So will, will these satellites be, I mean, what's, what's the, um, lifespan of these because they're so low orbit, I believe the lifespan is much lower. So what's the, do you know what the lifespan is and will these ones actually be used for services or are these just test ones that are.

Chris Gammell: No, I think these will be, I mean, I think they'll, they'll be on a much, I think they have the capability to replace them a lot easier. You know, they could decommission and, and put them into one of the different orbits. I was really surprised reading that other book about the, um, the iridium satellites, like, I always just assumed that like the, it was just like, yeah, you just burn them up. You know, like there's just definitely going to burn up. And yet there was like a ton of that book was dedicated to like the Motorola lawyers being worried about like someone getting hit by this piece of space junk or like, I just, I guess I never really thought about it. I just, I always thought that they always just talk about things burning up. I just assume things burn up, but like, yeah, I guess some stuff can make it through.

Dave Jones: Little chunks actually survive, you know, chunks of titanium, other metals and things, they survive and just, yeah, fall to the ground at terminal velocity. Yeah.

Chris Gammell: That's, that's not optimal.

Dave Jones: No, that's suboptimal.

Chris Gammell: That is suboptimal. Yeah. Oh boy.

Dave Jones: Yeah. Anyway. Yes. Yes. It's a mess. These, yeah. And, and there's, it's not just SpaceX doing it. What are they? At least two other companies who plan a similar size system? Like the, this, the SpaceX one, Starlink is bad enough.

Chris Gammell: Right. And that's like 2,500 to 5,000 satellites or something like that. It's like a pretty.

Dave Jones: They, they, they plan like tens of thousands.

Chris Gammell: Yeah. Yeah.

Dave Jones: Like, and, and there's other companies do, and Amazon want to do it as well, don't they? And, or someone. Anyway, yeah. Multiple companies. Yeah.

Chris Gammell: I mean, I imagine, I'd imagine like to get like coverage, like I was just thinking about the bandwidth you'd need, you know, like, so you're beaming down to this, to the earth, fine, whatever. But like the amount of bandwidth you'd need, if, if you have a dense area, I don't think it'd be as usable. Right. Oh, no, no. It's going to, yeah. Maybe you'd have to like go from like, if you were in like New York proper, New York City or something, you'd have to like fan out to like the Catskills and then like hop to there and then maybe that sends it up or something. Because just thinking about how loaded down the satellite over top of New York would be, you know, or any really population center. Oh, of course. Yeah, yeah. You know? So how much could, how much capacity could they really get out of that then? I'm sure that they've thought about all this stuff, but it's just like, it's, it's interesting to think about. Yep. We're trying a new type of ad here at the Amp Hour, which is going to be focused on teaching you something while also hearing from our sponsors. This past week, I talked to Daniel from Keysight, who wanted to tell us about ESD and the inputs to your test equipment.

Chris Gammell: What we're seeing is our support teams are getting these calls and they're getting these boxes come back and they're damaged. And yet the user, the engineer is like, why, what happened? Everything's normal. And then all of a sudden it was just gone. I don't, I don't know what happened. We did everything right.

Chris Gammell: So when I was talking to him, I figured it was going to be some kind of large spark or similar thing.

Chris Gammell: The most dramatic obviously is kind of that wool socks on shed carpet, like an arc. And then the signal just doesn't work. It's actually a lot more subtle than that. And that's what tricks people. So we all know, like if you touch a board or you touch a scope input and you feel a shock, that's really bad, right? Like that should make you nervous right away. But that's a couple KV to make that arc. And you can even build up and damage equipment with just like two, 300 volts. Unless you make contact with something and it transfers and transfers that charge from your body to the equipment, you won't feel it because it's not arcing. It's relatively low for what we're feeling, but the equipment is not ready for that level of voltage coming in. So, or power coming in. But more often than not, it's subtle. So you start to see more of your tests fail or you start to see just some weird performance under obscure conditions or even more normal conditions. Part of identifying it is knowing how things are supposed to look and even comparing it to other pieces of equipment and saying, here's what it looks like on this. Here's what it looks like on that. Oh, there's a 10% difference. Is it my probes? Is it something else? Or is it actually? So there's that level of calibration and being aware of what things should look like and understanding when it's your equipment that's wrong versus your device that's wrong. The other one is noise, like baseband noise on a scope or even a RF gear. If you see your noise force come up, that's an indicator that something could be wrong.

Chris Gammell: And so this knowledge is part of an app note series and some checklists and also a video series where they showed you how this was looking on the bench.

Chris Gammell: We walk through, here's what's actually happening on these boards. So we charge them up with a cart or packing material and we'll take a non-contact voltage sensor and kind of hover it over the board and you can actually see how much charge is on it. We've also done, we did a measurement on there where we hooked a current probe around a paperclip that was connected to ground, charged up and connected to the paperclip. And you can see how much amperage is going through that connection using that current probe.

Chris Gammell: Starting in March, they're going to be bringing back a contest that listeners of the Amp Hour will also remember from past years.

Chris Gammell: It's called Wave. And we actually have a link where anyone who signs up using the Amp Hour link gets an extra entry. You can also go to wavekeysight.com and sign up there, but using the Amp Hour link will get you an extra entry. So it's a good thing.

Chris Gammell: Sign up starting today for drawings that will start from March 2nd all the way to March 13th. They're going to be giving away over 100 different pieces of test gear. If you want to learn more about what you heard here, you can check out the links in the show notes. And now we're back to the program. Okay, let's switch away from internet. Let's switch away to, I bet some electronics. You've got a chip. You've got a chip that you like. Chip of the week. Chip of the week. Two in a row, man.

Dave Jones: Exactly. It is. Let me pull it up.

Chris Gammell: Let me just say up to the outside. This thing is ugly.

Dave Jones: No, it is. It is ugly, Dave. It's something that a mother loves.

Chris Gammell: Yeah, right. Only a mother loves, yeah.

Dave Jones: I'm sure its mother loved it. It is from Fairchild. Oh, yeah. One of the originals. All right. It is the FSBB20. And there's various variants of that. But it's the FSBB20. CH60 for those playing along at home. And it got semi-famous on Twitter because somebody posted a photo of its butt-ugly package. Yeah. And it's an evil package. And it's just, yep.

Chris Gammell: It looks like, you know, if this was going to be like a dead bug, you know, if you were going to dead bug this thing, it would look like a centipede. I mean, it's just got so many legs and weird configurations.

Dave Jones: It's different types of pins. Yeah. It's a dual-in-line package. It's a dip package. But it's a combination of a staggered pin, wide spacing.

Chris Gammell: If it does that, like, doesn't it have to be in line then? Like, is this technically a dip?

Dave Jones: It is technically a dip because it's a plastic package. So that's the P part of dip, dual-in-line plastic package. And it's got pins down both sides. But it's a staggered row one, just like you got back in the 70s and 80s when pin spacing, because they didn't have surface mount rubbish back then. So they, you know, but they wanted more pins on the chip. And dip packaging was the thing. So they went, well, we can't fit them. The pin pitch is just getting too small. We can't have through-hole, you know, pads. It just doesn't work. So they staggered the pins. So you'd have the zip. You know, I can't remember the exact acronym for it, but you would have the pins. It'd be staggered. So one would be in, one would be out, and they'd alternate back and forth. So it'd be the zigzag staggered pin arrangement. It's still a dip package. It's still a dual-in-line package.

Chris Gammell: I just always thought the line, I thought in-line was like that they were.

Dave Jones: Well, it is. That's not its correct name, but it's still dip-ish.

Chris Gammell: Yeah, totally. Yeah. I mean, like, yeah. I think if you said dip to someone and you pointed it in a lineup, it'd be like, oh, yeah, it's that thing. Right? Yeah, yeah, of course. I was just wondering if there was like a definition of it, you know.

Dave Jones: But there probably is.

Chris Gammell: But we don't know what it is. That's what it was coming down to.

Dave Jones: Offhand, we don't know. Right, right. Anyway, it's absolutely gorgeous. I think it's gorgeous. I think because it is, right, the person who, people who designed this just went, bugger it. We are not going to compromise on the pin, on the performance of this thing. We're going to get the pin spacing we need. We're going to get the clearance we need. We're going to get the pin current we need. Because not only does it have staggered pin arrangements, but it has different width pins on it and different spacings as well as staggering. That's right. So it's like the quadruple whammy. And it's got a nice copper heat pad on the top, too. Yeah. So, you know, anyway. So what is it for? What is it in? It is a 600-volt, 20-amp, three-phase IGBT, that's insulated gate bipolar transistor for those playing along at home, inverter with integral gate drivers and protection.

Chris Gammell: That's right. So basically, you can hook a micro up to this thing and drive it with, I think, 5-volt logic. Yeah, 5-volt logic, it looks like.

Dave Jones: Yeah, yeah. It's very cool. And it's, yeah, 25. It's actually the isolation rate. It's 600-volt rated transistors, but its isolation rate is 2,500 volts RMS minimum. Right. Because of the, you know, the large pin spacing on this thing. So it's just, you know, it's one of these dedicated system type chips, you know. And if you look inside the block diagram, we've basically got six insulated gate bipolar transistors, IGBTs, of course, with back diodes in them, and various drivers and stuff like that. So it's, you know, if you're into your motory and control drivers, I'm sure you are. Yeah, three-phase type stuff. Yeah, yeah, yeah. Yeah, yeah. Yeah, I use this all the time. Yep. So, and they're short circuit rated as well. So even if you short circuit them, they, you know, they're not going to snot themselves.

Chris Gammell: Yep, yep, definitely. I was trying to explain to someone. So I did some voltage isolation for a product I'm working on right now. And, like, I was like, well, look, I don't want to have the thing that's out there blow up the thing that's in here. And so that's why I put all this, you know, I've got optos, and I've got an isolated power supply and stuff like that. And so I was, like, trying to explain this to someone, and they were a little bit confused about it. And I was like, well, it's like they just might be at different potentials. Or, you know, if you have a spike, again, you might just want to have, you know, you can have a ground bounce or similar kind of thing, and you want to have protection against that because you don't want your whole system to be swinging around like that. I don't know if you would explain it differently.

Dave Jones: But transients are one of the big things. You know, you've got to design. A lot of people don't design for transients. They'll design, you know, if they design a mains product, for example, okay, they design it for the mains. Okay, it's 240-volt mains. I'm protected for that. They don't protect against transients on the mains. And that's where you can come a gutter because you didn't, you know, you don't think that somebody switches on a motor in the office next to you, you know, some big compressor motor or something. Boom, there's this huge, you know, transient on the mains. And if you don't, if you can't absorb that energy in your product, you know, you can really be screwed. So you can release the magic smoke.

Chris Gammell: That's right.

Dave Jones: So anyway, yeah. So this is a great package because it means that with the large pin spacing on here, 2,500-volt isolation, of course, you wouldn't get that with just creepage. You wouldn't get 2,500-volt isolation creepage. Creepage. You would get clearance. Now, for those who haven't heard me explain the difference between creepage and clearance before, clearance is a physical air gap. So it's just, it's between one pin and another pin. Or it's, you know, it's between one wire and another wire through free air. Okay. But, and so that's clearance. And you typically will have a voltage rating for clearance. But you also have, in terms of PCB design, but you also have another spec for creepage. And that is the path that goes on.

Chris Gammell: How creepy a thing is. Yeah, that's right.

Dave Jones: How it creeps along the PCB.

Chris Gammell: So this part is very creepy. It looks like a dead bug. And it's very creepy. So this is high creepage. Is that right?

Dave Jones: Only when you populate it, only when you put it on a board does it become creepy.

Chris Gammell: That's right.

Dave Jones: Seriously. It's because it's, you know, it's clear as a day before you put it on the board. Right. It's clearance. Right, right. But it becomes creepy once you install it. Right. So, yeah. So that's why, you know, you always see me in my videos talk about, you know, slot cutouts in PCBs. That's how you convert creepage across the board because you can get contamination on the board. You can get, you know, moisture on the board. You can get dust and grime and dirt and fingerprints and all sorts of stuff. And that will actually lower your rating compared to the clearance if you just had the pins separately isolated by the air. So that's what a slot out. So the difference between the pin spacing on these pins allows you to put a nice couple of millimeter. I don't know the exact dimensions, but a nice couple of millimeter slot between all the pins. And then you turn your creepage into clearance. And that's where you get your greater potential. And if you want even more than that, you'll see this in multimeter designs. You'll actually have molded into your plastic case of a multimeter, especially the high-end Cat 4 ones, you know, 1000-volt Cat 4 jobbies, for example. You'll have molded into your case little plastic walls that stick up through the slots in your PCB.

Chris Gammell: Yep, yep.

Dave Jones: So you've gone to level three, right? So level one is to get rid of your creepage by putting in the slot. And then the next one is having clearance. And then, not only that, you put a physical barrier, a plastic barrier between the two pins. So you've gone from air clearance to, you know, it's got to go through plastic. Although it sort of like turns into creepage, then it turns into arc creepage around. So it depends on how tall the plastic wall is, for example, and how high the arc has to get over. So it sort of turns back into a bit of a creepage problem once you put in plastic walls between pins and stuff like that. So, you know, or it's a combination because it's got to get clearance. It's got a high-voltage spark. It's got to jump across the clearance gap between the pin and the wall. Then it's got to make its way up and over the wall. And then from the wall to the pin over another clearance gap. But it can happen, you know.

Chris Gammell: Yeah, definitely.

Dave Jones: Everything breaks down given high enough voltage.

Chris Gammell: That's right.

Dave Jones: So, yeah.

Chris Gammell: I was trying to explain to someone. They're like, well, why would creepage ever be larger? Or when would creepage ever be a bigger thing than clearance, right? Because isn't the shortest point always between two lines? And I said, actually, yeah, that's true, right? In free space, that's true. But if you had a chip on one side of the board and a chip on the other side of the board, and you had to have them separated by some distance that met the clearance spec, the creepage would actually maybe be shorter because it would be going around the edge of the board. Of course. And that's how I help people kind of visualize it. It's like if it's kind of slinking along the board itself, instead of going, obviously, in air, you'd have to go kind of like just along the edge of the board. Or rather, you'd be kind of like surfing just above it. And so the creepage would be a little bit closer. That's kind of how I think about it.

Dave Jones: And another trap for young players with creepage and clearance, it changes with altitude.

Chris Gammell: Oh, yeah.

Dave Jones: So if you're designing products in, say, the IPC spec for PCB clearance and creepage, for example, you will have a limit to the altitude. You know, above 10,000 feet is a different figure again. You know, like it doesn't just magically become different at 10,000 feet or whatever. Right, right, right. But you've got to have a line. But, yeah, they put it somewhere. Yeah, they put it somewhere. But, yeah, it does change with altitude. So there you go. Yeah. So that's why you'll have altitude specs in multimeter manuals and stuff like that. You'll go, oh, it works up to, you know, X,000 meters or something. Right.

Chris Gammell: That's why you never take it on a plane, you know. Right. That and the pointy probes, you know. They'll take away your pointy probes at the security check and then they'll probe you instead.

Dave Jones: Anyway, so I love, like, no-compromised designs like this. They just went bugger it. But, you know, we're designing this thing to a spec and the pins will be what they are. You know, I just, yeah, beautiful. Thing of beauty is a joy forever.

Chris Gammell: Yep. Yep.

Dave Jones: Yep. Brings a tear to the eye. Yeah, definitely. So I don't think it's ugly.

Chris Gammell: I think it is still. Sorry. No, no. It's gorgeous. But it's well-designed.

Dave Jones: Gorgeous package.

Chris Gammell: Yeah.

Dave Jones: And people go, oh, what a pain in the ass. But, like, it literally takes you, like, less than five minutes to do a new footprint for that in your CAD package. If it takes you longer than that, you're too slow.

Chris Gammell: I think it's mostly a pain in the ass. It's, like, already through a hole. So, like, you know, it's already a pain in the ass, you know? Why?

Dave Jones: What's wrong with through hole?

Chris Gammell: Well, from an assembly perspective, it's just not as easy. But if you're doing high-power stuff, you're already in the realm where you're, like, yeah, there's going to be some through hole components. You know, like getting even the biggest, you know, DPAC type of thing is it's, like, sometimes they just aren't big enough, you know? But I think that what you're talking about with, like, creepage and clearance and then also if you kind of roll in, like, the, you know, the power specs of things, all that stuff together, kind of, it helps you kind of take a new view on why certain packages are the way they are, you know? Or, like, even seeing, like, sometimes in certain pin spacings are weird, you know, maybe, like, underneath an LGA or something like that. Sometimes it's due to that kind of stuff as well. It's just, like, you need to hit these certain specs and that's why you've got a pad over here and a big pad over there. And it's just kind of like this. It looks like there's no logic to it, but there very much was.

Dave Jones: So that is chip of the week? Yeah. And you might think this is obscure, but no, DigiKey have a stock of 83,000 of these. Oh, sorry, 78 in stock. At the factory, there's 83,000. Sorry, I got all excited.

Chris Gammell: Yeah. Well, you know, I'm sure these things blow up quite often. Yeah. Right. It seems like the kind of product that might, you know.

Dave Jones: Yeah. But it's not to order. You know, that factory stock, you will get that in a week or two. You know, they'll just order it. If you place the order, you know, yeah, I need 80,000 of these, please, DigiKey. Yeah. And it's at, sorry, at, what is it, 21 bucks a pop, you know, for 500 plus quantity. Right. But yeah, it's probably the only thing that does the job.

Chris Gammell: Yeah.

Dave Jones: Yeah. Yeah, it's very possible. I mean, it's like, oh yeah, you can do it with discrete IGBT transistors and other drivers and stuff. But like, sometimes like price doesn't matter for parts. You know, in low volume or specific, like in very high margin products and things like that, you aren't optimizing your bomb cost. Right. You know, you're, you're always thinking about it. Engineers are always thinking about it because we can't bloody well help ourselves, you know, because we're, we're, we're just optimization beasts. That's right.

Chris Gammell: Yep. So that's what my, that's what my dating profile used to say. I'm an optimization beast.

Dave Jones: Optimization beast. That's got to be the title of this episode. Optimization beast.

Chris Gammell: Oh boy. Yep. Yeah. Do it. Do it. Fine. All right. All right.

Dave Jones: You can, and you can update your Tinder profile. Yeah. Well, yeah, no. Optimization. That's great. Anyway.

Chris Gammell: I was asking today on Twitter, I was, I was thinking along the lines of being an optimization beast. I've been thinking about this recently of like, like the time I spend at different parts of my office and stuff like that too. And like, I was thinking like, I'm not really spending that much time at the bench these days. And I was trying to think of like, what is my mix of like time at the bench or doing a layout or like, like programming or writing docs and stuff like that. And so I was asking on Twitter and a bunch of people responded, but it's kind of all over the place. It really is pretty, it is obviously very project dependent. I was wondering if you had numbers for yourself, though. If you, if you think about it like a standardized product that you've worked on, what is your mix of bench time versus layout time versus programming or docs or meetings or whatever?

Dave Jones: Oh, okay. Well, you have to be talking about company work instead of my own.

Chris Gammell: Well, if you, if it was your projects, it would be like 80% forum time. And then, you know, 5% Twitter and then, you know, whatever else.

Speaker ?: Yeah.

Dave Jones: Just bum it around, you know. Right. But no, when you, when you, when you're dedicated to a task, oh, it depends. And it depends on the company that I've worked at. Like I've spent like, I've spent days just like documenting a screw.

Chris Gammell: Yeah. Right. Right. Right. Right.

Dave Jones: Like when you're working on stuff for the military, like that's part and parcel of doing that.

Chris Gammell: Right. And your meetings are going to be much more intense and, you know, yeah, there's a lot more overhead and stuff like that.

Dave Jones: And then you work at, you know, more, less regimented companies like Altium, for example. And it's just like, yeah, it's just a hundred percent just doing stuff. There are no meetings. There is no paperwork. Right. There is no, you know, it's just nothing. Right. Documenting code. No, don't bother. That's a waste of time.

Chris Gammell: Right. Right. Yeah. Yeah. Like get stuff working and like high speed kind of stuff too. Yeah. I guess I think about that too. I guess the thing that I always like, I've been thinking about like a lot of like task switching type stuff and I've been, you know, it's like my, my days now are kind of more task switching like between clients and it is kind of hard to do that. But then I was thinking like, well, but in projects, usually you have these phases, you know, you're like, okay, I'm in hardware phase and I'm going to be doing, I'm looking at data sheets and I'm doing layout and it's all very hardware focused. And then maybe I go and build it myself and I'm time on the bench and then just bring up and bring up Scott, you know, maybe some code, maybe some looking back at the schematics and stuff like that. But then it's like, you know, then the firmware stuff and it's like, I just, I feel like right now I've been doing so much firmware. It's, which is interesting for me, but it's just been like switch to the switch back is like a pretty drastic thing. And like, you know, it takes some, some time to kind of get back into it and like, think about like how things are working and software workflows and all that stuff. I don't know. Yeah.

Dave Jones: Well, I forced my, when I'm getting a project done, I forced myself not to task switch because it just kills, it kills your productivity. If I'm laying out a board, do not talk to me for two days. Yeah.

Chris Gammell: Yeah.

Dave Jones: You know, like, or do not talk to me for a week.

Chris Gammell: Yeah.

Dave Jones: For example, it's like, no, I'm doing a board design bugger off.

Chris Gammell: Right. Right.

Dave Jones: Right. It's like, I don't want to know about anything else. I want to focus on being in the zone.

Chris Gammell: Yeah. Yeah.

Dave Jones: Same with coding as well.

Chris Gammell: Yeah. Yeah. I think those are similar in that, like, you're just very goal, goal oriented kind of thing. You know, it's like get this thing to the point where it's working and, or out the door to get it shipping and whatever else. Yeah. Yeah.

Dave Jones: No, it's a task switching kills productivity. Yeah. It really does. Just absolutely destroys it. But sometimes you've got no choice because you don't, you know, you're back and forth between, oh yeah, data sheets and I'll, oh, I'll add this footprint to the package and then I'll do this other thing and then I'll, you know, try and do some design calcs and then I'll do some other stuff and, you know, you're all over the sheet and, you know, you can't help it. Yeah. Once you're dedicated, catered to a large task, you should, you shouldn't task switch.

Chris Gammell: Yeah. Yep. Yep.

Dave Jones: Head down, bum up.

Chris Gammell: Definitely. Yeah. My, uh, the firmware thing I've been dealing with is like, I, uh, I have, I have two Uarts that I'm working with and, um, and I'm learning the hard way that like.

Dave Jones: Software, hardware, Uarts?

Chris Gammell: Uh, uh, software, the hardware is fine. It's, it's all, I verified it all works, but I was like.

Dave Jones: No, but, but you're using hardware Uarts.

Chris Gammell: Oh yeah. Sorry. The hardware Uarts. I'm not, I'm not doing that.

Dave Jones: Are these inside a micro?

Chris Gammell: That's right. Yep. And, um, but, uh, you know, I, I'd been like having the easy life just doing print F out of the first UART. Everything's going fine. Yeah.

Dave Jones: Yeah. You young whippersnappers. I know. I know. Hardware Uarts, you know.

Chris Gammell: Yeah. Yeah. Super easy. Right. And then, uh, and then I was like, oh, how do I, how do I write to that second one? And then that kind of like put me down the rabbit hole and like, I was like, oh, I, okay. Yeah. No, you have to actually, you know, it's not just this magical function that pops out data anymore. You know, it's like, yes. When you print after one, you can't print after the other. Well, actually you can, but. Yeah. Yeah. I mean, you have to like write a new function basically that switch allows you to switch between the two. And I was like, oh, okay. So that has me.

Dave Jones: Why wouldn't it be the one function and then you pass it the port you want to go to?

Chris Gammell: That's just, I would have to write, I would have to write that basically. And it's totally possible to do that.

Dave Jones: That's a given.

Chris Gammell: Well, no, I mean, with a standard IO, like you could just, with a standard IO, you could just like push it out to the, to the, you know, you basically tell it, hey, it's going to this port and it's fine. Right. But.

Dave Jones: Right.

Chris Gammell: Yeah. So.

Dave Jones: Oh, come on. How is this a complaint? I mean. It's not a complaint. No, I'm just saying this is what I've been.

Chris Gammell: How is this a whinge? This is where, this is what I've been dealing with. I've, you know, I'm, you know, doing firmware stuff more than I used to. And I'm like, oh, okay. You know, this is interesting kind of things. I don't know. Like the.

Dave Jones: The struggle is real.

Chris Gammell: The struggle is real. Yeah. I guess I'm having empathy for firmware people more than I used to. I don't know. Right. I was just on embedded. So I think after this, I will be cross posting the episode, but I was just on embedded talking to them. This. I talked to them. Complaining about it. On Saturday. Getting it to you up. Up and running. It was a little bit less complainy, but it was more about the general flow and learn, you know, learning and relearning firmware stuff. So that'll come out midweek. I think next Thursday we'll have that come out on the Amp Hour channel as well. So yeah, it was, it was a good chat, but it was not as whingy as Dave said, I guess, you know. Okay. Sorry. I'm sorry, Dave. You know, we never went wine on here otherwise.

Chris Gammell: No, no. Yeah.

Dave Jones: No, we're not known for that at all.

Chris Gammell: Right. Right.

Dave Jones: Yeah.

Chris Gammell: One last thing I wanted to bring up before the end of the show. A former guest of the show and friend of the show, Ron Kwan, has come out with his third book. And I don't think it was on the list last week when we talked.

Dave Jones: No, it wasn't. Oh, yeah, it was, but we didn't get around it.

Chris Gammell: We didn't talk about it. Okay. Yeah. So it is about troubleshooting circuits. So relevant. Very relevant to this crowd. For sure. Yeah. And so he's been working on it for like a year or two now. And I just ordered a copy. I'm very excited to try it out. But it's, yeah, his other books. So he has the, he had one about RF and one about, I forget his other one. Oh, Electronics from the Ground Up. That was the other one. So we'll also link to Ron's episode. But new book coming out. So if you're interested, check it out.

Dave Jones: Sweet.

Chris Gammell: Yeah.

Dave Jones: And that's it for our Amp Hour. We are trying to keep the episodes to an hour. Yes, that's right. That's some, including interviews and stuff like that. We'll try and keep some of the feedback. We'll see. If it's a really important guest, we'll, you know. Yeah. Let them go on for multiple hours. But people do like consistency. So, but please leave your thoughts in the comments about that.

Chris Gammell: Yeah. We do have a comment section. Who knew?

Dave Jones: Yeah, we do. Go figure.

Chris Gammell: All right. Catch you next time. See you later, you, wait, optimization beast.

Dave Jones: Bye.