#432 – Check The Dummy Box

Chris Gammell: This is The Amp Hour Podcast. Released March 3rd, 2019. Episode 432. Check the dummy box.

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: Hey, Ned.

Chris Gammell: Hey, man. Welcome back.

Dave Jones: Thank you very much.

Chris Gammell: It's been a couple weeks, but always good to talk to you.

Speaker ?: Yep.

Chris Gammell: Boy, do I have something to say to you.

Dave Jones: What?

Chris Gammell: You should come to KiCon.

Dave Jones: All right. It's all happening.

Chris Gammell: How do you feel about flying for 50 hours to see me?

Dave Jones: Oh, it's about 40. Don't exaggerate.

Chris Gammell: Oh, sorry. Oh, sorry.

Dave Jones: No, no, no. Where is it again?

Chris Gammell: Chicago.

Dave Jones: Oh, Chicago. No, it's probably 50. Yeah.

Chris Gammell: It's like 15 to LA or whatever. Or Dallas. Yeah, another 40 here. Yeah, right. Yeah.

Dave Jones: Yep. Yep. With layovers. Yeah, it's good.

Chris Gammell: Not a good time. Not a good time. When is it? April 26th and 27th in Chicago, Illinois. And tickets are on sale now. And I don't know. Did you see the talks that were announced?

Dave Jones: Yes, I did. I had a look at the list. It's very good.

Chris Gammell: Yeah, it's pretty good, actually, right?

Dave Jones: There's a lot. Yeah, it's tons.

Chris Gammell: I'm excited. I think the thing that I'm excited about is that, you know, I doubt that people are going to come if they're not KiCAD users, but I think there's enough stuff there. If someone wasn't, they were just interested in like PCB design and manufacturing.

Dave Jones: I was going to say, there seemed to be other stuff that it was, would be interesting that wasn't sort of KiCAD specific, really. Right. But I'm sure they'll all.

Chris Gammell: Well, yeah, because if it was just, if it was just about the program, it would be pretty dull, I think. Right. Yeah. So. Yeah, totally. So I think there's enough other stuff. And I think the people listening might be interested in it. I mean, I think there's going to be a lot of hooks into KiCAD, but like the things like, so like one of the top talks here by Andre is applying CI and CD. I don't know if you know what that is. Continuous integration.

Dave Jones: Oh, right. Yeah. Yeah.

Chris Gammell: Yeah. It's like the software automated stuff. And I think just like automating PCB design, regardless of how you do it, it's kind of interesting. So like tying into GitHub, tying into similar stuff like that. I'm excited about that just because the things that we do over and over again, some of that could be automated. And for a long time, I had no capability to do it. Now I benefit from other people using software people coming in. I appreciate them. And plus, you know, hang out with me if you want. You know, that's the cool part, I guess. And we're going to party around Chicago. So we'll go do Chicago-y things on Sunday. So I'm excited. I'm nervous, but excited.

Dave Jones: Come on. You're a conference veteran.

Chris Gammell: Yeah. Yes, I am. Yeah. I get nervous about those too. Right. Okay.

Dave Jones: Why? What do you think people aren't going to show up? What are the fears that go through your mind?

Chris Gammell: So like imagine everything that can go wrong. Yep. And then double that. And that's like usually what goes through my brain.

Dave Jones: Right. Okay. So it's like nobody turns up. Then they call in a bomb threat. And then. Yeah, exactly.

Chris Gammell: And then the food isn't there. And the food isn't there. And the food isn't there. Food poisoning. Yeah, yeah, yeah.

Dave Jones: And then the projector doesn't work. And then the mics don't work. And yeah, yeah.

Chris Gammell: The badges all explode. Yeah, right. Yeah. Yeah. So pretty much everything that I could think of.

Dave Jones: Right.

Chris Gammell: I got that covered.

Dave Jones: Has anything gone wrong at conferences you've organized? Like are things, you know, broken down and you've had to make do or?

Chris Gammell: No, not to date. I've sufficiently worried about stuff I think beforehand.

Dave Jones: Right. But that you've sort of had backups or plans in place. So yeah, I'm looking like, you know, KICAD PCB design for rapid prototyping. You know, it's like, yeah, okay. It's kind of, yeah, you can do it in KICAD, but I'm sure there's, you know, you can do it in any tool, right? It's just like, yeah. And the heroine schematics, the heroine of electronics design, you know, like that's really, you know, is it really about KICAD?

Chris Gammell: That one actually, that's a Python tool. So that's, I mean, that's about using software to build hardware, which is kind of interesting. Right. But I think that stuff could be extensible too.

Dave Jones: Conversion control, you know, designing with decentralized groups, you know, top five footprint mistakes, you know, like, yeah, like this is all generic stuff, you know. So that's great.

Chris Gammell: Yeah. Yeah, yeah, yeah. Yeah. So it's a lot of great sponsors and badges and all that stuff.

Dave Jones: So, yeah, it'd be boring if everyone turned up and like there was a keynote about, oh, isn't KICAD wonderful? And everyone just sits around going, oh, yeah, KICAD's great. And, you know, it's, oh, it's the best thing ever. Yeah, fist bump. Right. You know, like, and that was it. Yeah.

Chris Gammell: Right. Well, when I helped out with SuperCon too, it was just like, I don't know, like I remember going to conferences where I wanted to watch certain things or like, and there was always stuff like kind of missing. And that's what I was trying to get out of people is like, be like, give a talk about that. I want to see this, you know. Right. That's, I don't know if there's like, we've struggled and we've talked about in the past too of like, like you have electronics down there and that helps, but there's not the talks as much as it's a trade show floor, right?

Dave Jones: No, no. It's yeah. No, there are no. Well, there's, you know, occasionally like there's a PCB conference associated with it, but it's all just, you know.

Chris Gammell: Oh, I didn't know that. Yeah.

Dave Jones: Yeah. They, they sort of run at the same time. Yeah. Like in the same, like they share the same venue and stuff like that. You know, they're sort of intertwined, but yeah, but that's, you know, something you've got to pay huge, huge money for. It's a, you know, it's a big thing. It's not, you know, it's not really part of the conference. It's not really part of the exhibition. So. Yeah. Yeah.

Chris Gammell: Cool. Yeah.

Dave Jones: So how do they, how does it work? Like, are they just one after the other over a couple of days or is it, or do they like on at the same time in separate rooms and you have to pick and choose?

Chris Gammell: Yeah. Yeah. It's going to be pick and choose. So there's two rooms. Right. Well, there's three total rooms. Two of them are going to actually have talks. Got it. And then they'll both get recorded though. I mean, basically I'm just copying Supercon. I'm going to be completely honest. Right. Okay. And, you know, for like as cheaply as I can.

Dave Jones: Right.

Chris Gammell: I think that's the best way to see. Cause otherwise, like if you think about the logistics of it, say you get like four slots in the morning, four slots in the afternoon.

Dave Jones: Yeah.

Chris Gammell: If you have a two day conference, you only have 16 speakers then.

Dave Jones: Right.

Chris Gammell: That's kind of crazy. So we're doubling that and, you know, it'll be squeezing in more where we can. But there's, you know, a lot of people that want to give talks. So that's great.

Dave Jones: So do you, do you have to pony up the money up front or you're relying on the sales to your, the ticket sales to, you know, you paid a small deposit yourself. And how does that work financially?

Chris Gammell: Financially. So like it's out of the space here that I work in, M-Hub. So we're just going to come here. Oh, okay.

Dave Jones: That's right. M-Hub are donating. Okay. Right.

Chris Gammell: Yeah. So they, well, they're not donating the whole space, but I, you know, they gave me very nice payment terms. And then there's some great sponsors too. And that stuff together is, yeah, I'm, I think I don't think I would have a confidence to do it unless, unless some of the sponsors stepped up. So that was really helpful.

Dave Jones: Okay. Yeah. Yeah. Cause that'd be a huge risk if you had to, you know, pony up a huge deposit yourself. Yeah. Yeah. Exactly.

Chris Gammell: Cool. Yeah, man. So I'm excited. We'll see how it all works out. I'll be stressed for the next eight weeks or so, but after that I'll be fine.

Dave Jones: I, I, I went to a meetup last night.

Chris Gammell: Yeah. How was it?

Dave Jones: I, it was great. I won't say who. Yeah. Former employees.

Chris Gammell: It was technical in nature. It sounded like you had a great time.

Dave Jones: Very technical in nature. Former, uh, uh, employees of, uh, a company who shall remain nameless. Right. And, um, yeah, they happened to be in the country for a, uh, new startup they're doing. And, uh, we met up and yeah, it was awesome. Yeah, man. Talking about all sorts of crap.

Chris Gammell: Look at this. Dave's, Dave's finally social. When there's, when there's smart enough people in the room, Dave will be social. It can be said. How is that? Uh, how's that EV blog meetup number three going? Is that going to be a once a year thing? Like the EV blog holiday party?

Dave Jones: I, I, when was the last one? Shit.

Chris Gammell: You've done two now. One when I was there. And then the second one was, uh, two or three months ago, I think.

Dave Jones: Yeah. Yeah. No, like I, I was taking a straw poll there while I was there and everyone said, no, six months at a minimum. Yeah. Like don't, you know, make it more than once or twice a year. You know.

Chris Gammell: I think there's, I think there's benefit to that actually. Cause then people have time to build up projects. There's something to look forward to.

Dave Jones: Exactly. They have something new to show off, new to talk about and yeah. Yeah. Things like that. But that, yeah, that works pretty well. You know, I've toyed with the idea of having a more conference like thing where people could give, you know, everyone sits down and people give a talk kind of thing. But I, I don't know. I, yeah.

Chris Gammell: That's not a bar then. That's for sure. Like it's gotta be a different venue. No, no.

Dave Jones: It's gotta be, yeah. It's gotta be a conference center. And yeah, I'd have to. Yeah. And just like you, I'd have to get sponsors. And then money's involved. Right. I'd have to get sponsors and, you know, crap like that. And yeah. And, uh, one of the interesting things we were talking about last night was that, um, like, uh, what are the differences between like the scene here in Australia and like overseas? Like, and like they seem quite surprised that I said, well, if I was, you know, single and I had no ties, I wouldn't be doing the blog from here. You know, I'd be in the U S right. Cause that's just where all the action is. Yeah.

Chris Gammell: What were they surprised about though? I mean, it was for the surprise, not the other stuff.

Dave Jones: They thought, you know, that there was so much stuff happening here and I'm going, well, you know, just not like at a vastly smaller scale.

Chris Gammell: Yeah. Vastly smaller. If they have any software inclination though too, they might be like, oh, well, Sydney does have a ton of software stuff.

Dave Jones: Oh yeah. No, of course. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah. And you know, we might, like there's a hard, there's a hardware startup. I just posted on the forum. There's like a, like a startup-y hardware conference. It's up in Queensland. It's next month or something. Um, you know, but it's kind of, yeah, it's not, it's, it's not really saying like we don't have the, any conferences here, like apart from the, um, yearly electronics exhibition, it's not even a conference, you know, we, we, we don't have any conferences here. We don't have any of that. Whereas you guys will have, you know, half a dozen. Right.

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

Dave Jones: A year easy. Right. Oh, sure. I mean like even like the independent. Yeah.

Chris Gammell: Right. I was going to say like the independent ones I could think of are like Supercon, Teardown just got announced, which is the, uh, that's the crowd supply based one. So it's kind of like general hardware. Yeah. Yep.

Dave Jones: Yeah. We have nothing like that. And you guys like have probably two dozen. I'm sure if you went and researched them and listed them, you'd have a couple of dozen a year. Sure. Yeah. I think, I think you're right. Yeah.

Chris Gammell: And it depends on which specialization you go into as well. Oh yeah. Of course. Like you can go real deep. If you go in the security world, they've got their own stuff. If you go into like the more industrial, like right now as we're recording, this is Wednesday, right? We're recording this the 27th. Um, there's, uh, uh, what's it called? Embedded worlds out in Nuremberg in Germany. I guess that's not the U S but like, you know, like there's just always something going on.

Dave Jones: Yeah. There's something going on. Exactly. And that's where all the big names are and that's where all the big networking happens. You know, if you're based here in Sydney, in Australia, like you're so far from networking with people, it's just not funny.

Chris Gammell: Right. You know? Right. Right. Right. And yeah, it's spaced out too. I think, I mean, it's pretty sprawling there. So. Yeah. But that's, that's tough. And I mean, I, like I know some people that live, you know, in more remote areas of the U S and if they want to get together with people, they got to go, they have to centralize at a conference or something.

Dave Jones: Yeah. They have to. Yeah. Yeah. They've got to do their six hour flight, you know, to.

Chris Gammell: Exactly.

Dave Jones: Yep. But we don't really, you know, even have the option here. Look, I could turn up to this Queensland startup. Yeah. I could fly up. It's only maybe an hour and a bit, you know, to fly up to Queensland, but yeah, you know, it's, it's, I don't know. It's just not the same scale.

Chris Gammell: Maybe. I mean, it's not the same scale. I agree with that. Yeah. Yeah. But yeah, you gotta, you basically gotta work a lot harder if you want to meet the same number of people. Right. Whereas if you go to like the Bay area, you can go to like two meetups and probably meet the equivalent amount that you'd meet up. Exactly. At that larger conference there. Yeah. So. And I think the other thing too, is like when I think about the kind of people that I look to meet, you know, like basically the, I like meeting all people that they, to be completely honest, but like when I would like geek out with people, it's like usually a pretty, it's a pretty simple, pretty similar kind of crowd, right? It's like the EV blog meetup that you did in, when I was there, like that crowd was all hardware nerds who were the, from the forum and watch your videos and listen to the show and like stuff like that. I think that's a very specific kind of thing. Sure. You go to a conference, you're also gonna get like, oh, I'm in business development or I'm in software and like all these other things. And it's like, I like meeting them. And it's like, let me say, it's very important to meet those people, especially if you're starting a business or something, but, but it's, it's a different, you know, you're not gonna geek out with them as much in my opinion. Not everyone's going to bring a project to that meetup. How about that? No, that's right. Yeah. So yeah.

Dave Jones: Yeah. Maybe, maybe I will go. Maybe I'll. I think you should, man. Yeah. Like it didn't sound that exciting. There's only so many options. It was more like to, like if you're doing a startup, you know, it was more like, you know, try and maybe like meet with, you know, investors, you know, stuff like that.

Chris Gammell: See, I think the other thing though, too, is to be what I call a colonizer. I tried to do this in Cleveland, Ohio. Let me tell you, there's not many harder things to do than being a colonizer in Cleveland, Ohio, where you say like, hey, you should come here. Right. That's like a hard thing to do. But like, but very easily you could be like you, you, Dave Jones could be like, I'm going to this, this thing in Queensland. It might not be the most relevant thing for everyone, but let's all just gather there and make like a mini conference, women conference.

Dave Jones: I was, that's what I was thinking about. I was just, that just ran through my mind a minute ago. I thought, oh yeah, look, if I'm going to this, I'm not just going to go hang out at that event. Yeah, exactly. Like I'll advertise it and have like, we'll. Bring your posse. You know, we'll rent another room in that hotel or whatever, you know, and. Yeah.

Chris Gammell: Or even just be at the same conference. Yeah. Yeah. Yeah. Of course. Or hang at a bar at some point, you know. Yeah.

Dave Jones: And then like afterwards spill over to something else. Exactly. Actually, I think it's like a couple of day thing. Oh, perfect. Yeah. You know. Yeah. So anyway. Which is interesting that it's in Queensland and it's not in Melbourne or Sydney. You know, it's like Sydney's usually very slack. Like there's more stuff in Melbourne or, you know, even Adelaide or Western Australia or Queensland. You know, it's really. I don't know what it is about Sydney. Sydney has like its software ones. Right.

Chris Gammell: It's the price. Y'all ain't cheap. Y'all ain't cheap.

Dave Jones: Prices are plummeting, you know.

Chris Gammell: Oh, yeah. Oh, good. Yeah. Yeah.

Dave Jones: It's slightly under a million bucks now, you know. Yeah. Right, right, right. So affordable. Yeah. So, yeah. Yep. Yep. Let me investigate that.

Chris Gammell: Okay. All right. Well, future announcements to come from the EEV blog meetup in Queensland. Yep. Why not?

Dave Jones: Yeah. I need to get out more.

Chris Gammell: Yeah, exactly.

Dave Jones: I've got one. Well, it's not scheduled yet, but I'm doing a meetup with someone just down the road. They're in the local business park and they want to, they're doing like an in-house talk at their company. It's a huge company, which I remain nameless at this stage. And they, like, they're like an expert in manufacturing yield and stuff like that. You know, how to, like, they, you know, he reckons that they've got their failure rates on their boards down to, like, virtually zero. Wow. And there's a whole bunch of mechanisms. Six Sigma type stuff. Yeah. All that sort of jazz. Yeah. Yeah. And, yeah. And he's like, and he gave, like, gave, like, an internal talk about it and wants to now do it, like, kind of external. So he has invited me in to sort of, you know, we can discuss that. Maybe it can be an Amp Air episode. Maybe I'll just take my little handheld recorder. And we'll just do an hour talking about manufacturing, production, optimization, you know?

Chris Gammell: You know, I've been thinking about that recently because I've been doing, I've been making some boards lately. And, you know, it's nothing groundbreaking. Right. But I keep thinking about, like, the things that I'm doing are very obviously, and all the mistakes that I make are very obviously the same mistakes a lot of other people make. Sure. You know, like, and for, like, specifically to manufacturing, I was thinking about the manufacturing side of it, nobody runs into that stuff until they, like, they run smack into it. Until they run smack into it. I'm about to do 10,000 units. Yeah, yeah. Exactly. And it's like, now I have a problem. And it's the same problem that everybody else had, of course. Yeah. But how would you know about it until you get there? And so you kind of have to, like, sample, like, what you're talking about. You have to, like, kind of sample all these different disciplines so that you at least know to be, like, like, at least so you see this thing and your brain goes, uh-oh, there's danger here. Yeah. You know, so you're not running towards it. Maybe you walk towards it or you crawl towards it or this potential problem. No, it's always head on. But you're at least thinking about it.

Dave Jones: No, it's always last minute. Oh, man. You know, and yeah.

Chris Gammell: Same thing happens for, like, accounting for me. You know, like, it's like, I don't know. I have no idea what I'm doing. But I'm running into the same exact problems that everybody else does. And sometimes it's amazing to me that there aren't guides out there. You know, like, obviously it's not, no one's going to have the exact same situation that I do.

Dave Jones: That's it. You can't. It's hard to do a generic guide. There'd be so much things. It's like the last time we were on, we looked at that list, that PCB checklist. And it's practically infinite. Yeah, that's true. You know? You can't cater for it.

Chris Gammell: But still, I think that having something where you can skim it and, like, be like, you're right. Like, you're not going to get everything because, like, it's always the sub note in the data sheet on page, you know, 58, you know, underneath the diagram that says, this feature does not work. Or whatever it is. You know, it's like, it's the last thing that you read. And then you read it and you go, oh, crap.

Dave Jones: Yeah.

Chris Gammell: But at least having the list, you might catch one or two, you know? Right. Yeah. I don't know.

Dave Jones: And it's like, if you're really lucky, you might get a manufacturer who might, like an assembler or something, you know, who might notice, like, know that you're going to run into a problem with this. Yeah, that's true. And they might tell you up front, but a lot of them won't, even if they do know.

Chris Gammell: Yeah, because they're also, like, busy anyways, right? Yeah, exactly. They just expect that you've done it before. Yeah, yeah. You know, it's almost like there should be, like, a checkbox for, like, I'm the new person. You know?

Dave Jones: Right. I'm a dummy. Yes, please tell me everything I'm doing wrong. The dummy box. That's what we need. The dummy box. All right.

Chris Gammell: Yeah.

Dave Jones: Yeah, because they don't want to, yeah, like, because you're not paying them to, you know, to, as a, as a engineering, like, advice service. You're paying them to do what you tell them to do. Right, right, right. And if you tell them to do this, and they know it's stupid, and it's not going to work, they're just going to go, well, we'll do it anyway. It's their money, you know? Yeah. Right. Or it might work, and we'll get, you know, a 50% yield, and well, tough luck. That's not my problem.

Chris Gammell: That means they've got to pay for more services. That's it. Yeah. Yeah.

Speaker ?: Yeah.

Chris Gammell: Yeah. How's your, how are your boards going these days? You still, micro supplies, chugging along?

Dave Jones: It's chugging along. It's, no, it's, or something always new always comes up, you know? It's, like, we just had to change a few parts for a couple of reasons, and well, no, it's unstable again now, you know? Yeah. Yeah, I've got to go through the whole. Like oscillation stuff? Yeah. Yeah, I've got to go through the whole process again, and yeah. Yeah.

Chris Gammell: Yeah, source meter days are coming back to me. I was never good with that stuff, but yeah, capacitive loads and all that jazz.

Dave Jones: And just dealing with the software side of things, because the software part of it is so abstracted and high level in the bleeding edge C++, you know? Oh, yeah. Yeah.

Chris Gammell: Right, so you're not digging into registers as much as you're, like, doing, like, object-oriented stuff.

Dave Jones: Yeah, it's doing, you know, the highest-end object-oriented, you know, professional-type approach to it, and we're finding bugs in the GCC compiler, you know? No, really? They're showstoppers. Yeah, they're showstoppers. Wow. And we have to wait for the GCC developers to come back and go, oh, yeah, that kind of doesn't work, you know?

Chris Gammell: Is this a new part or an older part?

Dave Jones: It's a ST part. It's, you know, it's been around for a while, but I'm sure... Yeah, and it's got bugs. It's got silicon bugs, right? There's a silicon bug in the ADC, and of course, we're using the ADC, and there's this weird silicon bug to do with the low bit resolution, or, like, the... I don't know. I can't remember. But, yeah, like, it's, you know...

Chris Gammell: But it's the kind of thing where it's, like, it's buried on page 58, and it's like, yeah, sorry, this doesn't work.

Dave Jones: No, no, it's not even on there. Oh, no? And sure enough, you go on the ST forums, and it's like, you know, the odd person has had this problem, and ST have gone, oh, yeah, we think there's an issue there we're working on it, you know? So, kind of, okay, well, you know, it's a bit... We rely on the accuracy of that ADC, right? And screw in with the accuracy, right? It actually, like, it's out by, you know, four least significant bits or something. It's like, you know, there's some weird-ass bug in it, and it's a pain in the ass, and we're thinking it's part of our circuit, right? We're thinking it's the tolerance of our parts, our reference, and all sorts of things. No, it turns out it's the... There's something wrong with the analog-to-digital converter, you know, at the silicon level.

Chris Gammell: Wow.

Dave Jones: So, and they...

Chris Gammell: Is that not a showstopper, though?

Dave Jones: It's not a showstopper. I think they've... ST have provided some sort of, like, software workaround or something, and there's other things you can do to do workarounds for it. But, yeah, it's like... It's like... This sort of stuff takes weeks to just... These little issues, and bingo, you've wasted another two weeks, another two weeks, another, you know, all these little issues that pop up. Yeah. And, yep. It's just... Yeah. It's a pain. But we are thinking about an external ADC now, and it's like, you know, because this thing's just given us, you know, a lot of grief, and there's not much you can do. When you hit those limits and you're reliant upon the developer of some open source software to fix it, or you're relying on, you know, the manufacturer to fix a silicon bug or come up with a workaround and investigate it and, you know, give you a workaround app note or something, then, yeah. It's...

Chris Gammell: Yeah. Yeah.

Dave Jones: It just kills your development time.

Chris Gammell: Yeah. This is the darker side of open source, I think. You know, you basically... Mm-hmm. You have to... Sometimes you either got to dig in and start changing the code yourself, which is not great. No, right, yeah. Or... Yeah, no, we don't know. Or you got to wait around and just hope that you can either pay a developer or find a developer.

Dave Jones: We can go in and fix it ourselves, right, in theory, right? Apart from the silicon, we can go in and fix the GCC compiler ourselves, but, geez, you know, like, you don't want to be doing that. The other thing we're working on is I've got a... In my hand right now, I've got a custom planar transformer. We needed the efficiency across a wider range and especially at lower standby currents and stuff like that. And, yeah, we worked out the parameters that we required for this transformer. We went to a custom planar transformer manufacturer. That's all they do. And, yeah, and, you know, they weren't particularly cheap to get some prototypes made, but they actually designed this job and it's really quite sexy.

Chris Gammell: Is this for their isolation or is this for switch mode?

Dave Jones: This is for an isolated switch mode.

Chris Gammell: Oh, there you go. Might have both.

Dave Jones: And, yeah, and it's cool. Like, it's basically a, like, an eight-layer PCB with a ferrite, you know, actually through the board and going around. And it uses the internal layers to make the coils. And, yeah. And it's...

Chris Gammell: So what's the upside? I mean, upside is really low profile, right?

Dave Jones: It's lower profile. That's the advantage of planar transformers usually. And traditionally, technically, this kind of isn't a planar transformer because a planar transformer is that the coils are the... Well, the coils aren't stacked on top of each other. They actually fan out flat. So a planar transformer is technically a big, flat, wide transformer. And the coils just radiate outwards, if you know what I'm talking about. Like bigger and bigger circles.

Chris Gammell: It looks like a spiral, basically.

Dave Jones: It looks like a spiral from the top down, yeah. But whereas this one is not that. It's actually using a PCB to drop down to the next layer. One thick trace of the... A via to drop down to the next layer and go around. So it's, you know. But it's kind of like the planar manufacturing type. So, you know, you call it a planar transformer.

Chris Gammell: I think, yeah. PCB transformer is another probably way to say it, right? Or something like that.

Speaker ?: Maybe.

Dave Jones: Something like that. But, yeah, planar is the...

Speaker ?: Yeah.

Dave Jones: And there's advantage, you know, as always, there's pros and cons. Otherwise, everyone would be doing planar transformers, right? Right. If they're so fantastic. You know, they're fantastic at getting better coupling, but losses at X frequency. I can't remember the trade-offs. But there's, you know, sure you can go look at the details. Right.

Chris Gammell: Because you're basically, now you're not using air. You're using like a different dielectric. You're using the PCB dielectric.

Dave Jones: You're using a different dielectric and stuff like that. But, yeah, it's, you know, it's pretty cool. And it took them, like, apparently our requirements were pretty tough for them. Oh, cool. And they had to go through, you know, it actually took some time for them to do multiple revisions and stuff like that. And I was, yeah. So, yeah. I'm not sure if they have, like, modeling tools for it. You would think, like, a planar transformer manufacturer, that's all they do, would have some whiz-bang custom tool to, you know, you whack in the parameters and out pops this final design. But apparently that's not how it works. Someone strokes their beard and goes, you know.

Chris Gammell: Yeah, it didn't have a volume, though.

Dave Jones: Yeah, it actually took some time to get this made. And they said, yeah, they had to go through quite a few revisions of this thing to actually meet our specs. So, yep. Anyway, I think it's pretty sexy. And it's cheaper. Because, you know, you might think transformers are cheap, but good transformers aren't cheap. So, you know, so we are saving by going to this planar transformer in terms of, like, a regular custom wound transformer, you know.

Chris Gammell: Right. You'll also save in, like, weight and, you know, I guess you'll have...

Dave Jones: Weight and the efficiency's better and everything else. So, yeah, if you want a really efficient traditional wound one cheap, yeah, you can do it. But it's probably going to be much, you know, it's going to be much bigger. That's interesting.

Chris Gammell: So, you're giving up some efficiency on the power side.

Dave Jones: No, this one's definitely more efficient. It's more efficient and smaller. It's better in almost every way. For our particular niche purpose.

Chris Gammell: Interesting. Yeah. But is it the battery is powering the transformer that then pushes it to the isolated side?

Dave Jones: I'm not saying.

Chris Gammell: Secretive. Yeah, so, all right, fine. You do your thing. I'll wait a little while. All right. That's cool. I was curious, though, like, how do you... So, like, how do you think about that stuff, too? Because I... So, I've been talking on the consulting forum. And, again, I guess I have to be a little secretive here. But one of the questions I think about is, so, between consulting and doing a product, like, it's almost like you're... If you're doing a product, you're the client and you're the designer. Yes. Yes, that's right. So, like, do you... I know you don't really bill for hours, but, like, how do you think about it? Like, is it just like, well, we want to get this thing done, so we just get it done?

Dave Jones: That's the difference between David, too, and I. You know, he likes, you know, ridiculously rigid specs and we must stick to the specs. And we must, you know, it's like there's a deadline and there's a, you know, specific specs to meet. Right, right, right. I'm more like, meh, you know, like, I'm going, he goes, what spec do we need? And I'm going, meh, just make it decent, you know. And, like, to me, that's a valid approach because this is, you know, I don't... Like, you can get caught up by trying to meet a spec, right? And, like, let's say you want 0.1% accuracy, for example, right? Well, at what voltage? You know, do you want 0.1% at full scale? Do you still want 0.1% down at, you know, 10% of full scale? Because if you do, that's going to be much harder. You end up stepping up, paying an extra $2 for a better reference just to meet that. Whereas, like, you might not care about, you know, that at lower... But you might argue, hey, well, just make your spec better, right? As in more, like, more requirements in your spec. Okay, at 10%, it's X spec and at, you know, 50%. Yeah, but that's tough, too, when there's not, like, a... I know, it's silly.

Chris Gammell: You don't have a product group yet. You know, like, after there's something in the market for, you know, 10 years, and it's like, well, there's all these different offerings. And if we, you know, if we slaughter our specs in between, you know, the number two and number three producer out there, then we can definitely hit all the...

Dave Jones: That's right, that's when it pays. But if it's like... But when you're developing your own thing for... Yeah. That nobody really has a hard requirement for, it's just, you know, it's... Right.

Chris Gammell: I'd say almost like it's like, you know, the spec is the, what are people willing to pay for at the beginning? Yeah. And then make the category.

Dave Jones: Price matters, yeah, of course. Yeah, yeah. And there's some other things you spec, you know, like, yeah, but it's like, you can't just make everything rigid because you might see, oh, well, look, if I loosen my spec by 0.1%, I can use this part that's 10 times cheaper, you know, and it's like, oh, that spec's going to be good enough, and it's going to be, or we can get the user to calibrate it themselves or something, or, you know, things like that. Right, yeah.

Chris Gammell: Yeah, yeah, yeah.

Dave Jones: Or we can have a calibration process in production instead of designing in the calibration by buying expensive parts. Oh, like a self-test. Like, yeah, and doing all that sort of stuff, you know, so, yeah. It's, yeah, it's difficult.

Chris Gammell: No, it's a tough balance, and I think, honestly, I think sometimes it's harder to design something when there are no specs just because, like, what do you do? Like, how do you make these decisions? You know, like, if you have a complete greenfield, you don't have constraints, and engineering is all about constraints, so then, like... It is. You basically have to impose constraints, and you have to hope that your constraints aren't ridiculous, you know?

Dave Jones: Well, see, I can see it from David's point of view because it's my project, right? Yeah. But he's the one doing most of the, you know, most of the design work on it. Right. And it's like, so to me, it's more personal. To him, it's just, oh, it's just an engineering task that needs to be completed, you know? Yeah. And I would be in exactly the same position, right, when I had a real job. And, yeah, I would want those specs. It's like, well, don't give me open specs. I want something to work to. Otherwise, I'm just going to, you know, make something up on my own.

Chris Gammell: How do you know you're done otherwise? Yeah. You know, like, that's a big thing. Yeah.

Dave Jones: How do you know you meet the deadline for this month so you can show it off at the design review meeting, and you can, you know, the Gantt chart, you can tick off that section of the Gantt chart, and it's done.

Chris Gammell: Yeah, that's true.

Dave Jones: And all that sort of stuff. Whereas, you know, me, because this is more of a personal project for me rather than, you know, a hard and fast, you know, professionally, like, you know, sort of like a, like a, like, it's just a product and I don't care, right? It's, yeah, it's more personal to me. So that's when I don't like specs because I like being able to just breeze on through and go, oh, I'll change the direction this way because that's more fun or, you know, this will be cooler or, you know, I don't, like, it's just, it's very different to when you're designing for a client or whether you're not, you know, whether you're designing for a company and, you know, you're just an engineer on the team or whatever. It's very different.

Chris Gammell: Yeah.

Dave Jones: Thank you. So, yeah. It, it'd be different if my job was not making videos and my job was an, an engineering company designing products, right? If that was my, that'd be very different.

Chris Gammell: I was going to say too, if you, so like you said, you're designing it for you, right? Yeah, right. And that's, I think that is actually a good way to do it sometimes because. Sure. In a wide open field where there isn't really a thing in this cat, in this category, it's like, okay, well then yeah, pick one person and why not you? Right. But I think if you're, like you're saying, if it was a company, then what you do, you would do is then you would be the product manager and you'd go out and you'd be like, all right, I'm going to find a hundred people who are likely buyers and then figure out what they need and capture all their things. It's like boring, boring, boring, boring, boring. Yeah, boring, boring, boring. And, you know, what is, what is the willing, what are you willing to pay for this right now? It's like, oh, well, you know. Yeah. Yep. So that's fun. Yeah. At least you don't have to do that piece. At least you're like, you just have to like scratch your head a little bit and be like, yeah, I think I'd like that. It's okay. Right.

Dave Jones: Oh, dear. Anyway.

Chris Gammell: Yeah. Well, get them specs, man. Get them specs. Look at that. David's texting you right now and he's just like, Dave, I need, I need, I need the specs.

Dave Jones: That's my meetup last night. He's doing, he's, he's actually working on this, um, uh, new startup product. And he doesn't like, it's, it's, you know, it's reasonably groundbreaking. And he didn't tell me how he did it. Right. So he's, he's, he's like, there's this specific technical thing that they're doing. And it's like, nobody's done it this way before. And, you know, so we're, you know, I'm throwing a few ideas at the, uh, at the, when we met up and he's like, nah, it's not really that. Nah. Keep on thinking about it. You know, and this morning I finally, aha, this is the only way he can do it. And I think I'm, I think I'm on the money. So he's now, he's now texting me. So I texted him and it's like, it's on the money. I think it's on the money. I think I'm, I think I've figured it out. It was like a, you know, a process of elimination. No, they can't use this. There's only, you know, several techniques you can use to do this thing. And yeah, I, yep. That's great. Yep. Anyway. So yeah, I think I've got him. I like it.

Chris Gammell: Well, another product I was going to mention, um, is one I've been playing with a little bit is the JewelScope. Have you seen that one?

Dave Jones: What's a JewelScope?

Chris Gammell: It's on the, it's on the subreddit. Uh, so basically this is not a necessarily a new concept, but I, I like the interface and everything too. So it's Matt Liberty who's doing this. He was just on, uh, embedded actually. Right. Um, uh, two weeks ago, one week ago. And it's got a Kickstarter. Basically. Oh, he's. Yep. It's a Kickstarter.

Dave Jones: He has met his goal. He's met his goal. Yep.

Chris Gammell: I, I think he actually has a, uh, a JewelScope in your mailbag right now, but he's, I don't think he'll make it in time. So that's okay. Oh, really? What's sitting here?

Dave Jones: Apparently. Yeah. Oh, okay.

Chris Gammell: You could probably like open it up while we're going. No. Oh, geez. Okay.

Dave Jones: And, uh, it's got, it's got another 21 days left. So, you know.

Chris Gammell: Yeah. Perfect. Uh, it's cool though. So it's like a low power measurement thing. And, you know, it's a hard thing. He was trying to explain it to Alicia and Chris on embedded. And the, basically it's like the software switching of ranging. And basically it's, you know, it's a competitor to, to, um, you know, somewhat competitor to the, uh, the microcurrent, but it's, you know, it's got digital range switching and stuff like that. And, uh, but, but like the same thing that the microcurrent does, basically it's for people that are trying to measure super low currents and then it can range up if you have like, so say you have like a wifi device, it's in sleep mode, then it turns on.

Dave Jones: It's, you know, it's, it's, it's, it's, it's, it looks like I'm working on the exact same thing. So. Oh yeah. Okay. Yep. Okay. I'm what?

Chris Gammell: Micro supplies like that or no?

Dave Jones: No, no, no, no. I follow up to the microcurrent. Yeah.

Chris Gammell: Got it. Got it. Yeah.

Dave Jones: Yeah. Yeah.

Chris Gammell: Yeah. So it's cool though. It's, uh, and you know, and so there's other people that are doing this, um, uh, what was it called? There is another one. Uh, I'll look it up in a, in a second there, you know, so ST, ST actually has like a dev board that does this kind of thing as well. So it's not like you have to do this kind of thing. And this is definitely a little, little up more up market. Yeah. There's, there's one that's a cheap, it's like, it's meant to be, maybe it was NXP. It was basically one of the board manufacturers put one of these kind of in line. They use a bunch of op amps and they put it in line for doing power monitoring ostensibly to, you know, show, Hey, we're also making these low current parts. Why don't you measure the low current while we go? So, um, and now I'm trying to remember the other one. It was like, it had a, I think they advertised on your site at one point. Um, it was like another current monitor. It was out of Sweden. Um, I don't think they're advertising. I don't recall. Yeah. But there was another one that was like that. And they were, they were at, uh, they, they were doing similar things as well where they basically could do current ranging, you know, from low levels up to high levels. And this is, you know, like, see, I think you've made videos about this before doing, you know, sleep versus. Yeah.

Dave Jones: It's hard because it's a dynamic range issue and, and range switching isn't as trivial as, as it is because you're, when you're a, say you're in a, um, when you're measuring a sleep current, for example, and when that processor wakes up, you've got to have, well, to measure that low current, you've got to have a series shunt resistor that is a high enough value so that you can actually read the, you know, it's, you know, taken, you know, half a microamp or something. Right. So you've got to have a high enough value, but then if it instantly wakes up, then that resistor starves the processor of actual current before it has time to actually switch into, to detect that and switch ranges. So it's, you know, there's a trade-off and the way you can get around that is by adding extra capacitance there, um, to sort of like hold the charge long enough so it detects it and then can, you know, and then it can, has time to range, uh, to a different, to a lower value shunt resistor so that then the processor won't, you know, shut itself off because the voltage has dropped too far and it's, yeah, it's a, it's a, it's really tricky business. It is tough.

Chris Gammell: It's tough. I mean, that's, I mean, QFly too, but we weren't even doing the visualization piece. So it was like most of the time. And then I think that's the thing that a lot of people associate with. If you're looking at it, say you have like a nice DMM and you are, and it is great at auto ranging, even in the fastest auto ranging, are you going to catch what it looks like? You know, if you have a spike in current, you know, if you just see that the display blips, you pretty much need to be looking at like on a, on a time-based scope in order to like, to catch like, oh, there was an event there, you know, you need to see that spike and that kind of thing. So, and I found the other, the other one I was thinking of is OT, O-T-I-I by Koitek. And that's, that's another similar product. So they used to advertise on your site. If you didn't know. Oh, Koitek, yes. Yeah, yeah, yeah.

Dave Jones: It's, what is it? The O-T-I-I?

Chris Gammell: That's the name of the actual product. The OT is the product.

Dave Jones: Oh, yes. Oh, of course. Yes, yes, yes. Yeah.

Chris Gammell: Yeah. So I've used that too. That one was a little tougher for me. I think the voltage ranges were different on that one. Like that was only 3.7 and below. Yeah.

Dave Jones: No, it didn't, that didn't go down low. It was for higher power. Was it from memory or something? I don't know. I'd have to.

Chris Gammell: I don't know. But it did, I mean, it did similar things too. And it, you know, it, it's still out there. So, but I think that, like, basically these days in order to get, you know, the measurements you need like that, you know, so I'm doing like Laura stuff and cellular stuff and wifi stuff. And basically when you're doing these things, most of the time you're going to be asleep. So you got to be ready for that when that slug of current comes in. Because surprisingly, even with Laura stuff, I don't know if you know, like even Laura stuff when it's in like a transmit mode, it's, it's dropping 20 milliamps, you know, or more. Yeah. So. Yeah, for sure. It's up there, you know.

Dave Jones: No, the, the Koitek one is only 4k samples per second and it's only one, one microamps to five amps. So it doesn't do the really low stuff. That's. Oh, it doesn't do really low stuff. That's right. Yeah. Yeah. Yeah. Yep. Got it. But it's cool. So, you know. Yeah. Yeah. They're great. I mean, they're both. But no, this, this new, this new one, this dual scope looks all right.

Chris Gammell: Yeah. Yeah. Yeah.

Dave Jones: It's, it's pretty much, it's, it's not too dissimilar to what I'm working on at the moment, actually. So.

Chris Gammell: Well, there we go. Competition. Yeah.

Dave Jones: Yeah. Totally. Cool. So. Hey man, competition is good.

Chris Gammell: That means it's market. So that's good. That's good. Now, yeah. Mine was going to be high. Now you have the specs to compete against.

Dave Jones: Right. Now I have the specs to, right. I've got a price and I've got a spec. Right. I can beat that. Exactly.

Chris Gammell: Yep. Yep. Yep. Yep. That's great. That's great. Yeah. Yep. I was going to mention the, I didn't actually put this on the thing anywhere. I, I've been playing around with the NRF 91. I don't, did I mention that in the show at all yet? No. I'm really excited about that too. So like I had mentioned that, that, what was it? The Sam R34. That was that Laura chip I mentioned a couple of weeks back. Yeah. That's a cool looking little chip. But this is a module that's from Nordic. So Nordic Semiconductor got into the cellular business. I knew about this a while ago, but I wasn't allowed to talk about it, but now it's released. And, and so it's a modem, but it's got, so it can do LTE, LTE, LTE, which is like the low power cellular service. It's rolled out a bunch of places. It can, it can't do, it can't do a lot of other stuff though. It can only pretty much do low power stuff. It's going to have NB IOT, which is like the, the other that's, you know, CAT M1 and CAT M2 are the two cellular, low power cellular things. But, but the crazy thing is it can do GPS as well. So in like this package, it's probably like 20 millimeters by 10 millimeters. I think it's pretty tiny. It can do GPS and cellular for like, I think the list price is like 15 bucks. So, I mean, it's still not like super cheap, but you know.

Dave Jones: No, no, no. It's not dime a dozen, but it's yeah.

Chris Gammell: Right.

Dave Jones: No, to have that sort of functionality a few years ago would have been super expensive. Oh my God.

Chris Gammell: Yeah. Yeah, exactly. I mean, so like, and I was designing with like Ublox modems and similar, like in other ones, there's just, you know, they're power hungry. You know, this is going to be low power. And so having access to cellular stuff too, I just think I'm, I'm really excited. So I got a dev board and I'm really excited to try it out or I've been playing with a little bit, but like, you know, try it out in a, in my own design. And, and yeah, it's a, it's a crazy time we're living in, man. Like the chips that are available these days, everything's just like getting combined into everything else, you know? Yeah. But that's, you know, then there's also, I think like a lot of these new things too. So like the NRF 91, I know it uses Zephyr, which is a RTOS. And a lot of the things are like, you know, basically there's these higher level things like you're talking about at the beginning of the show, right? You're so abstracted from everything, which is good at the beginning because, you know, you and I are software dummies, but bad when something goes wrong, you know, basically you have to find someone to call when there's something wrong because digging down into the stacks and stacks of software, it's like, boy, I hope, I hope I can find something, you know?

Dave Jones: Mm-hmm. Yeah. You're just, you're just bumming around really. Right. Exactly. If you just try and do it yourself.

Chris Gammell: Yep. Yep. Yeah. I am. So speaking of RTOS as well, I don't know if you've seen the AWS thing too, there was on the list. Um, so yesterday they announced, so like this stuff's all, yeah. So Amazon web services, that's like the, uh, yeah.

Dave Jones: That's what I thought. Yeah.

Chris Gammell: Yeah.

Dave Jones: So they do, um, so yeah, they have been doing one.

Chris Gammell: So they've been doing one for a while. This is basically AWS IOT is like one of the offerings where they're trying to like suck in your data. It's one of their million offerings. Yeah, exactly. Yeah. So it's like using MQTT servers and like a lot of like standard stuff like that. Basically like these verified endpoints. So basically you have a device somewhere in a field and it's talking to the internet and it needs to dump its data somewhere. And so basically it has to certify itself with the AWS network and then it can say, then it'll basically dump data to this effectively public, but not actually public web address. Right. And that's what the certificates are for.

Dave Jones: So, um, I did, I did say years ago, didn't I, that, um, Amazon, um, internet of things would actually be the winner. Would actually rule them all. I think, yeah, I remember talking about that. Yeah. The smaller players would, you know, ultimately fall by the wayside.

Chris Gammell: I don't agree with this. Well, maybe the smaller players, but like Google's in this and Microsoft Azure, like it's definitely like the big, the big players that are, are. Yeah.

Dave Jones: Well, look, if I was, yeah, we mentioned this before. If I was designing a product that needed, you know, like to work for the next 10 years, I'm not going to be using, you know, some little tin pot service. I'm going to be using the Amazon internet of things service. Right. Yeah. It's just, it's just more guaranteed to be there longterm.

Chris Gammell: Yes. I generally agree.

Dave Jones: Maybe not a Google one. Cause Google have a habit of, you know, canceling their projects and stuff like that.

Chris Gammell: I think on the server side stuff, they're usually a little better. Maybe the consumer stuff. I wouldn't trust with my life. I wouldn't, you know, I wouldn't bet my Google glass on it. How about that?

Dave Jones: Still can't sell it.

Chris Gammell: You know, at this point, Dave, I'm just cherishing the memories.

Dave Jones: It'll be like the equivalent of like long hair and sideburns in the 1980s. You know, it's like, you know, the, the, the big hair thing, you'll look back and I can't believe I wore that. You know, I can't believe I had hair like that. I can't believe I wore Google glass, you know? Yeah. Yeah.

Chris Gammell: Moving on. Anyways.

Dave Jones: You were so excited about that. It was hilarious. Okay.

Chris Gammell: All right. Yes. We all think it's so frigging funny, Dave.

Dave Jones: It's great. Oh boy.

Chris Gammell: Yeah.

Dave Jones: Anyway.

Chris Gammell: Um, I've lost my train of thought. Oh yeah. So for free RTOS. So Amazon a while back, it was probably about two or three years ago. They started basically developing for free RTOS, having their own flavor of it. So basically they put a bunch of hooks in so they could hook into AWS easier. They developed, they've been developing for like expressive. So like the ESP32 and ESP8266. So there's a bunch of starter kits for that kind of stuff. Um, uh, yeah. And now they just announced that they're going to start supporting risk, risk five stuff as well. So obviously that's risk five is an ISA. It is not a chip, which means it's an instruction set architecture as we've learned over and over again, listen to past shows, uh, and all of the angry comments about it. But, uh, they're, they're supporting things like the site. Uh, no, sorry. The, uh, oh, what's in the article. Uh, but basically there's a couple of chips that are out that they're supporting at the beginning and basically they're going to start porting to other architectures over time. Right. I think that's going to be even more interesting, you know, is they, if they define things for broader architectures, right? So like, um, if they make a, like a device that could hook into that ISA and work with that, but I don't know how the hell that would work. Because the reason I say as much is because I think the risk five stuff that's really interesting is the people that are putting on an FPGA is making custom things. So if you can only use it on, you know, two or three chips in general anyways, it's less interesting, but, uh, but, uh, this is a good move. I think this is, this is great.

Dave Jones: So cool. Does it work? Do you know if the, uh, risk V works on like the super low end FPGAs?

Chris Gammell: It does. Yeah. So like, uh, Clifford, who's been on the show, uh, he actually has one, it's called like Pico LV, I think. Okay. Or Pico RV, I believe. Yeah. So Pico risk five, right? Right. Uh, Pico R5.

Dave Jones: So it's a special version of it. That's optimized for low.

Chris Gammell: Yeah.

Dave Jones: Um, yeah.

Chris Gammell: So that, I think that's going to be the thing that's interesting. We're going to see more and more people that are building their own processors based on the architecture. Like there, I'm not sure how many there are out there, but I keep seeing newer, newer and newer things out there where people are like, they're building entire processors around this architecture. Yeah.

Dave Jones: But it's only like a handful of people in this kind of like, you know, Joe Average is not going to be, they just want to get their product done. They just want to get the job done. They're not going to worry about rolling their own processor. Yes.

Chris Gammell: You know, I totally agree with that, but I'm just saying there's a lot more than there would have been. Oh, sure. So like, so like in the FPGA space, right? So for a long time, it's like, if you wanted a processor in there, you got to either use one of the open ones that I think the Cortex M0 maybe had like, they had like a license free things, but I'm just thinking about like the old FPGAs that I worked on. There was like some that you could put in there, but they were either the ones from Xilinx or Altera or like one or two other, maybe ARM like licensed some of the low end stuff. And then there's like an 8051. Yeah, exactly. There weren't like a ton, but now it feels like there's just a lot more available out there and people are doing open source licensing as well. So it's interesting, I think, you know. I don't know how that stuff works at all. Cause then you have to probably have to, I mean, there must be what they have to write like your own machine code to actually talk to that stuff. I'm not sure.

Dave Jones: I was going to say about that, you know, we've had enough issues with, you know, mainstream processes, you know, the ST ones in the compilers that like everyone's using. And yet we still run into issues. I mean, you know, you've got to wonder, got to be concerned with how many issues are hidden in these. Sure. Yeah. You know, like in the scheme of things, like hardly anyone's using, you know, these risk V ones yet. So it's like, you know, how many hidden issues, if you're going to start developing your product, you know, you might run into a, some weirdo issue. So, yep.

Chris Gammell: I think you're right. But I think that also goes back to like most people, most people listening, most people doing this stuff are not going to use that stuff. I think they're going to, you know, maybe if they pick a risk V one, they're going to, they're going to grab one off the shelf that there is like a, you know, consensus around, but I don't think that it's going to be, I think right now it's researchers and people tinkering around and, you know, the, the true, the true, the true crazies in the best of ways like Clifford. Uh, uh, so yeah, I, I think, uh, I think we're gonna move towards that in the future. So it would be interesting to see how it plays with like services like this though. Right. So like how then does RTOS get ported to a new processor like that and, and seeing how that stuff works in the future would be interesting. Um, yeah. So, oh, the reason I was looking at RTOS in the first, the free RTOS, um, this came up when I was looking at stuff a while back, I was looking at like, okay, so now I have a thing in the field, say it's cellular, say it's wifi, say whatever it is, it's connected somehow. Yeah. How do I actually manage? And I think maybe mentioned on the show, how do I manage to push new firmware to some of the stuff that I have out there? Right. So say I have a thousand units in the field, right? They're all connected. They're capable of receiving this stuff, but how do you actually manage like, okay, well

Dave Jones: you need a wireless kernel to, you need a wireless bootloader system. You've got to add the, you know, like everyone's used to like serial bootloaders and you know, USB bootloaders and stuff like that. But then you've got to add the Siri. Yeah. If you're going to, you've got to add the wifi stack and all this has to go in the bootloader section and some sections of micros, right. Only have limited bootloader space, physical space. It's not an arbitrary large, you know, it's not like 30, you know, if you've got a 64 K process, you can't make a 32 K bootloader, you know? Right. It's like some of them are actually physically constrained.

Chris Gammell: So, hmm. Yep. Yeah. Sounds tricky. And I think that that, so like the free RTOS has it built in basically. So like some of that stuff is, is it helps handle that, that kind of thing. There's another one out there called Mongo. So, and I, and I basically asked on Twitter, like I'm not like finding this and I'm like, Oh, I know everything. No, this is me asking on Twitter and a lot of people responded very nicely. Uh, there's another one called Mongoose OS and that basically handles, uh, some of that stuff as well. And, uh, and there was one or two other solutions. Cause I was basically like, okay, I, so now I make a completely custom device. I want to have this thing in the field. I want to be able to update it. What are my options? And it's, you know, you know, either you write your own bootloader, like you're talking about, or you grab someone off the shelf, like, you know, the Amazon free RTOS or Mongoose or something like that. And then you deal with all the other things that are missing or included with it. That's, that's always the thing when you're targeting for a specific feature, then you're stuck with all of the other warts that it has. You know, that's the, that's the problematic side of it. I've not made a decision yet personally, but I'd love to hear if people have, I mean, like it's just a hard problem to solve. Yeah. And you know, there are services too. So like particle is a good example. Like they basically handle a lot of this stuff. You're in their ecosystem, but you have to be like, you can't make anything custom out. You know, basically they have that module that like the hardware is locked down at that point and you can, you know, like interface with things, but, but then they handle that, that's that software piece. And so it's kind of, you know, either, you know, you pay to be in that ecosystem, which is fine.

Dave Jones: You've got to be locked into the, yeah. You become locked into the ecosystem. You've got to choose to buy into that.

Chris Gammell: Right. And, but, but, you know, like when you really look at the costs sometimes, you know, so I've looked at the costs and I'm like, oh shit, maybe it's just worth it to do that. You know, like there are balance points where, where sometimes it's worth it just to pay your way out of it. And, uh, oh, totally. So interesting stuff there as well. I mean, some of this stuff, we'll see if any of it ever takes off. You know, a lot of these, uh, projects that I work on, it's like, I'm trying to figure out if it'll work and can it work. And sometimes the answer is nope. So, or, or I, I, you know, I, I talk about, I talked with, I talk with people about what the actual answer is and here, here, you know, lay it all out there and they go, nope. So that's, that's fun.

Dave Jones: But you're not thinking outside the box. Can't just say nope.

Chris Gammell: I'm not thinking, I mean, I can, I can say nope. I say nope. It's just, you know, I have to accept that the answer might be, we're not going to pay you anymore, Chris. Okay. How's the consulting gig going? It's going pretty good. Um, I was thinking about the other day, I think I'm averaging about two boards per month. Um, you know, I've got a bunch of different clients and, uh, it's still challenging. That's good.

Dave Jones: Is that board layout you're doing or are you designing systems or are you like, what are they calling on you for?

Chris Gammell: A little bit of both. I mean, like, um, so sometimes it's like, so for like a simple system I just started on, it's like, here's a problem. How do we do this? And that from that, you know, ended up like the solution was pretty simple actually. You know, it was just like needing to switch things on and off in certain ways. And okay. I, you know, I basically use some of my past experience to design a board, fit their form factor. That was like one thing. Right. Uh, and, and that's cool. Right. It's not going to be the most, I could have gone crazy. I could have been like, well, let's put an Intel processor in there. You know, we need a 20 watt power supply, you know, but like, that's not the right actual answer for solving a problem. Um, and sometimes it's more like, Hey, I've got a prototype. I need to make this a product. That's another one that I've been doing. And that's usually board layout and hardening. And, and that's interesting too, from its own perspectives.

Dave Jones: So do you think you'll end up finding yourself like not pushed, but sort of like find, like falling into some niche like, uh, Mike Harrison, for example, with his, you know, lead art things. Do you, do you think you'll end up like that? Or do you, or are you, uh, would you like that or will, or would you actively try and fight it if you have found yourself? Oh, no, I'm not going to take that job. Cause I did that last time. You know?

Chris Gammell: Um, I think there are some jobs that'll eventually will like, I would, you know, pass off. I wouldn't say no to, I would, you know, try and find someone else for him first. Right. Um, but I don't know. I think, I mean, I think that's more of like, honestly, it's, some of it is skill. Right. And I think that Mike is, I mean, obviously people haven't listened, go listen to shows where Mike's on. Holy crap. Is that guy skilled? Right. Um, and like, look at his projects. It's amazing. Uh, but I think that's also a function in networking as much as anything else. And that's one of the things that I've been learning is it's, you know, it's so like Mike definitely has a name in that space as well. So like the reason he keeps getting work like that is because he is one of the best in the air, you know, like it's just, and he takes jobs based on the projects he wants to, the problems he wants to solve. Um, and I think that that, you know, has been increasing in complexity over time. I'm speaking for him, of course. I don't know if that's actually the case. I think it's right. Um, yeah. And, uh, so I, I think it's going to come down to like, if I keep doing decent networking and I have clients that like me and they recommend me for other things, I could get pigeonholed, but I don't think that they're specific enough right now that that's a problem. Got it. So.

Dave Jones: Yep.

Chris Gammell: Cool.

Dave Jones: Yeah.

Chris Gammell: So we'll see. Um, but I, you know, I really do like it right now. I mean, there's some tough days, but. Right. You know, I, the, I will say again, uh, the forum that I started, the consulting forum, it's been great, great for support and asking questions and, you know, bitching about stuff like that's great. So if there's people that are interested in consulting or are consulting, there's a, there's a, um, a form for, uh, for applying and, uh, you can join us if you want.

Dave Jones: What's the address?

Chris Gammell: Uh, it's a Google forum. So I'll, I'll just put a link in the show notes.

Dave Jones: Oh, it's a good. Do Googles do forums? What?

Chris Gammell: It's a Google form. Sorry. Oh, form. Form and forum. Form. Yeah. It's a Google form to apply. Yeah. And then it's a, what do you mean? Yeah. So it's like the forums that we do for.

Dave Jones: Yeah. Yeah. Yeah. I know about the forums, but how do people chat and discuss things?

Chris Gammell: Oh, they don't. So they, they, that's what they use to apply. And then they, then I big give them access to the forum with a U, which is just a hidden part of the contextual electronics forum.

Dave Jones: Oh, okay. Right. It's a private section. Right. Okay. Got it. Yeah. Yeah. Got it.

Chris Gammell: Yeah.

Dave Jones: Yep.

Chris Gammell: Yep. Cool. So yeah, it's great. It's great. We are starting to run out of time. Are there any things on this list that you wanted to go over? There's some stuff I'd like to mention real quick. Then you mentioned them. So first off, LatchUp, another conference is happening. This one is-

Dave Jones: There are conferences everywhere that we don't have here. Let me guess. It's not in Australia.

Chris Gammell: It is not. It's in Portland, Oregon. It's a, it's about open silicon. So this might be some of the RISC-V people, RISC-V. So that's coming up in the week after KiCon, actually. So it's May 4th and 5th, I think. That's out in Portland. So that'll be an interesting one.

Dave Jones: It's a conference dedicated to free and open source silicon. Wow. That's a niche crowd. Yeah, it is.

Chris Gammell: And I think the last one was in Europe. It was, I think like 150 people.

Dave Jones: Wow. Okay. Cool. That was probably the entire community turned up, right? The entire- Yeah, I think it was a lot of people.

Chris Gammell: Open chip community. Yeah, I think a lot of people were working on similar stuff, right? Yeah. Were there. Right. But I think there's a lot more on the state side. So that kind of helps to, you know, if you make it at least on the right continent, people can travel a little easier.

Dave Jones: Yeah, of course.

Chris Gammell: Yeah. There's another link, Kickstarter and China manufacturing. Got it all wrong. It's a bit of a longer read. But basically, it's kind of a discussion about like people, you know, stuff we've talked about here where people are going to China with the Kickstarter stuff. They don't really know what they're getting into.

Dave Jones: And they think that they can just pay money and it happens magically. It magically happens.

Chris Gammell: Right. Right. Yeah. So that's a good post. I do recommend it. Let me guess. What else was on the list here?

Dave Jones: The answer is no. It doesn't just magically happen.

Chris Gammell: It does not just magically happen. That is correct. How did you know?

Dave Jones: I'm an Oracle.

Chris Gammell: Yeah. Let's see. Robert Ferenik, he made a video about cheaper PCBs. So past guest of the show. Definitely check out that video. Let's see. Any others here? Oh, this is an interesting one as well. Timpin is an open source hardware and software developer for hearing aid stuff. I thought you'd like that because you've done hearing aid stuff in the past, right?

Dave Jones: Yes, I've done work for Cochlear. Yep. It wasn't hearing aid.

Chris Gammell: It was the implant stuff. So this is kind of trying to... But yeah. Oh, okay. Implant, right. Yeah. So that's an interesting open source project that's coming up. And yeah. Sweet. I don't know. There's a lot of other stuff on the... Oh, there's the Grace Hopper video too. That one's good. Have you seen that one? I've seen that one.

Dave Jones: It does the rounds every couple of years, you know. Yeah. It's old.

Chris Gammell: Yeah, exactly.

Dave Jones: People post it and then it gets viral for a few days and then, yeah. Rinse and repeat every couple of years. I mean, because she's awesome. Yep. Yeah, it's cool. I've watched it several times.

Chris Gammell: That's about it. I mean, there's a bunch of other things on there. So yeah, check it out, I think.

Dave Jones: Cool bananas.

Chris Gammell: I think there's a lot of good... There's always a good post on... If you follow the subreddit, that's a good place to go. Those all get posted to the Amp Hour Twitter, so you can follow it there as well. And, you know, you can always follow myself and Dave. Dave is EEVblog on Twitter, and I'm Chris underscore Gamble.

Dave Jones: And you don't always get electronics-related stuff. No.

Chris Gammell: You don't. That's correct. Take it or leave it. That's right. Yep. All right, man. Well, we'll talk soon. Catch you next time. Bye.

Speaker ?: We'll be right back.