#468 – The Tiny Lab Movement

Chris Gammell: This is the MFR Podcast. Released November 24th, 2019. Episode 468. The Tiny Lab Movement.

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

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

Dave Jones: You're in the office slash cubicle?

Chris Gammell: I'm in the loud parts of the office today. Loud parts of the office, right? We'll see. Maybe you get a little, you know, the hustle and bustle of a startup area. Oh, it's very exciting.

Dave Jones: Oh, those co-working spaces. Yeah, very hot. Yeah, WeWork is just killing it. Yeah, totally.

Chris Gammell: There are things being killed. Regarding WeWork, that's correct.

Dave Jones: And that's all the investors' money.

Chris Gammell: That's right, yeah.

Dave Jones: Should we talk about that? Because it's interesting. I mean, it's it. It kind of has relevance. So you've heard about this, obviously.

Chris Gammell: I used to work out of a WeWork, you know.

Dave Jones: Oh, you did?

Chris Gammell: Yeah, I hated it with a passion. I really, really did not. Really? Why?

Dave Jones: Why? Tell us why. Well, let's start out by telling people what WeWork is. You're better at that.

Chris Gammell: Yeah, it's like a co-working space, but it's meant to be all super hip and designy and all this stuff. But if you've ever, like, I remember I had been to, I didn't know where I was, but I knew the second time I was in one that I had been in one before. So I went to like an event, and it was at this place. And then I went to WeWork, my first day to WeWork for my former job. And I was like, oh, I have been to one of these before because every office is glass everywhere. It's all glass. And so you feel like you're in a fishbowl. It's just this very, oh, I do not like it.

Dave Jones: Right. So that was their style, was it?

Chris Gammell: Yeah.

Dave Jones: And is it like WeWork branded everywhere? Do they have like giant signs all over the place? Or is it more subtle than that?

Chris Gammell: It is not subtle. It's anything but subtle. So in the West Loop, which is this area here in Chicago, there's a huge picture window where they have a, it says on the front of the picture window, Hustle Harder. That's their branding. Hustle Harder.

Dave Jones: Oh, okay. Hustle Harder. Yeah. Hustle this. And if any people don't know, yeah, it was, they were about to go IPO. So they're about to, you know, a list on the exchange, and they're valued at what, $45 billion or something?

Chris Gammell: 47, yeah.

Dave Jones: For a company that's basically just a landlord. That's basically all they are, is they rent out these, they buy, well, they lease these properties. So what they do is they take out 15-year leases on massive swaths of real estate all through every city. Every city. All throughout the world. Are they only in the US?

Chris Gammell: No, no. They're all over the place. All over the place.

Dave Jones: They're all over the world. Okay. So they rent like 15-year leases on this place and then will rent you out like a six-month lease on a thing.

Chris Gammell: Oh, you can do month to month if you want to.

Dave Jones: Oh, you can do month to month. Yeah. So these short-term leases. So they're taking out all this long-term lease and then leasing out short-term. Hello, risk. Anyway, they're valued at $45 billion. They're about to go IPO public and then the wheels fell off the billy cart and people realized that they actually were kind of worthless. Yes, that's right. And they'll never make a profit. They will make a profit. $50,000. I think they will with the changes.

Chris Gammell: I really do. I mean, I think that there's a-

Dave Jones: Well, they've slashed everything. Yeah, exactly.

Chris Gammell: I think that once they're actually run like a business, that'll be- You know, it's an interesting look at like Nouveau, like Nouveau-Riche.

Dave Jones: Yeah, but they've got 15-year leases. That's the problem. Oh, sure, sure, sure. That's the problem. So anyway, everyone's talking about, if you watch the financial circles, everyone's talking about the collapse of WeWork and how it's the end of people trusting IPOs. It's just like we've had enough, like apparently the industry has changed now and this is one of the leading drivers. So yeah, it's a, yeah.

Chris Gammell: So to bring it back to hardware a little bit, it kind of highlights one of my frustrations with like the tech sector in general, where it's like in hardware, we're very much tied to reality. It's like, oh, well, there's a bomb cost of $50. Yeah, right. We can't, you know, like we could probably sell it for $60 for an amount of time based on, you know, just funding with investor cash. But at a certain point, someone's going to look at it and say, oh, $50, right? It's just harder to hide that, in my opinion. You know, it's harder to hide like the cost versus value kind of thing. And it's like in the tech industry when there's like, you know, they always talk about like if you watch Silicon Valley, the documentary Silicon Valley now in its last final season. Yeah, yeah, right. But, you know, it's just harder to schmooze and, you know, gloss over these things. So they talk about like, oh, if you have revenue, it's bad because then you, you know, you can't. Heaven forbid you have revenue. Right. Then you have something to basically compare against, right? I mean, that's kind of the idea. Oh, yeah, true. Yeah, yeah. This has always been my frustration. I think that software gets away with it a little bit. I think they're getting away with it less, like you're saying, a lot of the, there's just less.

Dave Jones: Well, software is just labor costs, basically. It's fairly easy to itemize. Whereas, yeah, a real estate business like this, it's like, yeah. Anyway, they were able to hide it quite well until they did their IPO. And then when you have to open your books, it's like, holy crap, this company is not going to.

Chris Gammell: Look at that. And that. Oh, and that. Oh, my. Adam Neumann is a D-bag.

Dave Jones: Adam Neumann, if you don't know, is the charismatic CEO who was famously going to be worth a trillion dollars. So he claimed. Yeah. And now he's, yeah, been booted out. Yeah. Yeah.

Chris Gammell: Just like the other snake oil salesmen are pretty similar in that way.

Dave Jones: So what's your opinion on the WeWork? Well, so. So the offices, you didn't like them, right?

Chris Gammell: I did not. And so, like, I am at my current space and I actually continue to pay rent for myself to work out of here instead of WeWork because, you know, I wanted to be around hardware and stuff like that. And, you know, this is a co-working space too. It's not like it's all that different.

Dave Jones: Yeah, but you've got actual hardware in there, right? Yeah, exactly. You've got CNC machines.

Chris Gammell: Right, right. And I think that, you know, there's, you know, some of it is, you know, what I'm saying, some of it is fluffy and it's like, oh, well, I want to be around, you know, people working on hardware stuff. That doesn't really, yeah, it's not that valuable. But like Dave's saying, it's more valuable to have access to, like, gear and stuff like that. And I've been thinking about that recently.

Dave Jones: Really? See, I would have thought you were the opposite because you're a people person.

Chris Gammell: I am a people person, but there's people.

Dave Jones: You're a contact collector, you are.

Chris Gammell: Sure, sure, sure. But there was people. There was tons of people at WeWork. That wasn't what I disliked about it. I disliked that it was, like, this sterile, like, you know, I don't know. You know, my former co-workers, I liked a lot of them. It was just that, like, the location was bad.

Dave Jones: But, and you just wish you would, like, be able to retreat to your cubicle. Yeah, exactly. Like, and hide from everyone sometimes. Right, right, right. Right.

Chris Gammell: And some of it was, like, having equipment around, too. And so I'm actually dealing with this now. I'm trying to decide. So, like, you know, I said I had had that VNA. That was, like, a portable VNA. But now I'm trying to figure out if I should get something like that that's more portable. Or what I've actually been looking at is a piece of tonnage. And I think I might be. Boat anchor, yeah. Yeah, I might be buying that boat anchor that I had talked about.

Dave Jones: You might be getting the boat anchor. Is that going to fit in your little space?

Chris Gammell: My rollaway cart? I think so. Yeah, this is one of the frustrations about a place like this. Or really just any shared space. It's like, you know, I miss my basement, you know.

Dave Jones: Yeah, yeah, right.

Chris Gammell: But I do actually. One thing that I was thinking about the last time I was on the bench is that I do actually like the having to clean up at the end of the day. Which I think you've talked about before, too. And we've talked about, like, storing projects in, like, boxes and stuff like that. It is tough to break things down. But I think that there's a certain. There's obviously cleanness to it. But then there's also, like, compartmentalization that is required. Because you're literally putting stuff into compartments. And, you know, at the end of the day, you have to put stuff away. And, you know, it's tough. Like, if there's anything where you're doing a long-term experiment, that would be very detrimental, I think. But in terms of, like, my mindset throughout the day, I actually rather like that of, like, okay, no, stuff's done. The gear's put away. My project's put away. Okay, I'm ready to go home and, like, leave it there. I don't know if you're, like, I guess you leave stuff on your bench, right?

Dave Jones: I do leave a lot of stuff on my bench as a visual reminder that, oh, I need to continue to work on that.

Chris Gammell: Yeah.

Dave Jones: But the downside to that is, yeah, the place just looks like an absolute mess.

Chris Gammell: Yeah.

Dave Jones: And, yeah. And it's interesting you mentioned.

Chris Gammell: Well, it used to be, like, I would just leave it out forever, though. That's kind of the point I was getting at. Right, yeah, yeah, of course.

Dave Jones: Yeah, no, I've had stuff left out forever on the bench. That's the whole idea of having the benches is that I can leave, like, things in progress sitting there so that, oh, yeah, every time I walk past, there's a visual reminder. Oh, yeah, that project.

Chris Gammell: Yeah, right, right, right.

Dave Jones: You know, so I don't actually forget about it because if it goes in a tub and goes away, it's going to be forgotten about.

Chris Gammell: Yeah, yeah, yeah.

Dave Jones: So it doesn't mean I get back to it, but at least it's there to, you know, give me a visual kick. At least every six months when the cleaning happens, you know.

Chris Gammell: Yeah, yeah, exactly. Yeah. Well. Yeah.

Dave Jones: Yeah, it's interesting that you mentioned the, you know, the small spaces. I'm, you know, because I think this new space is just too big for me. Yeah. Well, there's no such thing as too big, but it just leads to just putting stuff anywhere. It leads to laziness. Yeah, right.

Chris Gammell: It's like a liquid or rather a fluid will fill any surface that it's, or any space that it's given, right?

Dave Jones: The chunk will expand to fill the space available. Yeah, that's weird. And like in a year's time, once my lease is up here, there's a good chance I might move back to my old 50 square meter lab, which is precisely half the size of here. Yeah, yeah. Because I was talking to Mrs. EEV blog the other day and I mentioned, I don't know what I'm going to mention, it's kind of noble to like have everything fit in a small space.

Chris Gammell: Oh, man. Dave's going to go back to the garage eventually. He's just going to keep shrinking back down. Yeah, exactly. Shrinking back down.

Dave Jones: Yeah, go back to five square meters. So go from 100 to 50 to five.

Chris Gammell: It's like the tiny house movement for labs, you know?

Dave Jones: You're right. Yeah, yeah, exactly.

Chris Gammell: The tiny lab movement.

Dave Jones: I just watched one of those last night. This girl's living in Tokyo in an eight square meter space.

Chris Gammell: I mean, the things people do with that.

Dave Jones: Her entire house. Yeah, I know.

Chris Gammell: I used to like really like put that up on a pedestal and then like I think, I may have mentioned on the show too, but like I had this like epiphany one day when I was like flying in an airplane. Yes, I remember. And in a bathroom. Yeah, I was just like, oh, why am I thinking about this right now? You know, like this is the worst scenario ever. Yeah.

Dave Jones: Right. Live in a tiny house. Yeah. Right. I can remember when you were going to, yep, hit the road in a tiny house. That's right. That's right. Yep. Yep. Ah, those flights of fancy. That's right. Right.

Chris Gammell: My young and heady days of being, living in, you know, a couple. Yep. Tens of squares of meters. Yeah.

Dave Jones: But yeah, you know. Yeah. So I thought, oh, yeah, there's something noble about making. Noble. Of course of a small. Noble.

Chris Gammell: That's an interesting word.

Dave Jones: Noble going all the way. Yeah. Okay. What about it as noble? I don't know. Like there's something like to be admired, so to speak, about somebody who can work, you know, do a lot in a small space rather than, oh, I got this gigantic warehouse. Look at me. Yeah. You know. Yeah. And it's just too easy once you've got a large amount of space. Yeah. To work in a small amount of space.

Chris Gammell: I think some of it might be the focus aspect too, right? Just like you really can't. Right. Yeah. You literally can't do more than one thing at a time. Right. And then, you know, it just kind of helps you to, you have to focus on one thing or else.

Dave Jones: You have to focus on one thing. And it's, yeah, I think the bigger the lab gets, maybe the less focus perhaps. I don't know. So, peeps out there, tell us. Yeah. Let us know. Do you, you know. Yeah.

Chris Gammell: I mean, when I had a lot of space in Cleveland, I definitely spread out. My old basement was a hot mess.

Speaker ?: But.

Dave Jones: Well, my old garage lab, if you remember the video of where I, I needed more bench space. So, I put a bench over my car.

Chris Gammell: I remember that. Yeah.

Dave Jones: Like, my car would drive under the bench. Like, the bonnet of the car would be under the bench. Yeah. And I thought that was genius, you know. And I gained a whole, a huge extra bench in this five square meter lab, you know. Yeah.

Chris Gammell: Yeah. And it was just, yeah. Efficiency of space and stuff like that. Yeah. You know, I think that like with your stuff too, it like had like the compactness, you know, behind you and stuff like that. Yeah. Some of the visual effect, of course, as well. But. Yes. Yeah. I don't know. Yeah.

Dave Jones: There's other YouTubers like that. Hister Coulson's lab comes to mind. Oh, my God.

Chris Gammell: Yeah. Yeah.

Dave Jones: If you've seen, you know, seen him, it's like this, like, it's a, it's a, it's a large-ish room, but it's not, it's not exactly, well, it's not huge, right? But it's small.

Chris Gammell: He's more like a vintage scope hoarder, though, you know.

Dave Jones: He's a vintage scope hoarder, yeah. And it's just like, oh, everything's neatly packed in behind him, like 10 foot high. Yeah. And shelves. And it's just, yeah, it just looks, looks magnificent. Yeah.

Chris Gammell: One thing that we didn't know about, about Paul is that he, he does like private Patreon videos, right? So if you're a member of his Patreon, he does like.

Dave Jones: Oh, I did know that. Okay. Yeah.

Chris Gammell: Tutorials and stuff like that. So like, he did like a whole soldering lesson for like an hour the other day. It was great. Right. So yeah, he's doing, he's doing really good stuff and he's, he's a good educator. So if people are interested in supporting him, it's definitely, it has some stuff hidden behind the curtain that is, it is worthwhile, I find. So it's good stuff. Yep. Yeah. Cool. Cool.

Dave Jones: I, I, I still haven't done the hidden behind the curtain thing. Yeah. All of my stuff is still out there. Virtually all of it is in the public eye. Yep. Yep. I don't know. Still haven't needed to move into that area.

Chris Gammell: Yeah. Yeah. Well, you know, we appreciate the stuff that's out there. Speaking of, you've been working on flex PCBs. So what's up with that?

Dave Jones: I have been working on a flex PCB. This came about, um, and coincidentally, I did that video about, uh, you know, all the PCB manufacturing options. And I, and I asked him, I deliberately said, look, I'm not going to go into flex PCBs because that's a whole hour video in its own right. And I asked people, would they like to see a flex PCB, uh, video, you know, tutorial video and everyone's like, yeah, yeah, yeah, yeah, yeah. I want to see it. So I will eventually do that. But then like a week after that, um, it turns out that we, uh, well, we didn't really goof the controller on the, we underestimated what we, the amount of space we will require in our microcontroller on the micro supply on the USB side of things. We thought, oh, 32 K that'll be plenty.

Chris Gammell: Oh yeah.

Dave Jones: Right. So that, you know, all we're doing is a USB power delivery thing and a USB HDI and a serial interface and, you know, a little bit of housekeeping and stuff like that. It'll easily fit in the 32 K. Wah, wah, wah, wah. No. Turns out that the menu, all the manufacturers libraries for USP PD are huge.

Chris Gammell: Right. Right.

Dave Jones: So yeah, I think, have we discussed this before on here?

Chris Gammell: You've talked about a little bit. I was thinking more because you mentioned in the video, you talked about that in your latest flex PCB video.

Dave Jones: So yeah, this was the whole background to it is that, yeah, the micro was so like we thought it was plenty, but it turns out, no, it couldn't even run the USB PD, let alone the other stuff that we needed. So we had to change the micro to a, like 128 K version and it's not footprint compatible. So of course we had to, I had to make an adapter board that went from the little T-SOP package, Julian Rowe T-SOP to the quad, larger quad flat pack package. And because of the tight real dimensions in there, you can't do this as a rigid FR4 board.

Chris Gammell: Right. Right.

Dave Jones: So I thought, aha, I'll just do it as a flex. And bingo, people wanted this flex video. So I, you know. That's great. So that came together very quickly. I just lashed together this tiny little board, which literally had one chip on it. And I turned it into a 50 minute video. Well, you know, that's fine. As I do. But it wasn't really a flex tutorial. I did mention some things in there about flex design and stuff like that, but it was just the obvious solution to making this little adapter, this little adapter board for the micro supply. So yeah. Anyway, there. Yeah. I got it manufactured. They're on the way. They're on the boat. They said it would be six to seven days, five, five to seven days manufacturing time. But no, they got it done in like two or something before I got the shipping announcement. Yeah. Yeah.

Chris Gammell: I'm actually, I'm for the first time, I am, I'm doing overseas assembly. I've not done that yet, actually. Like overseas, like using one of the, the low cost board houses and actually using assembly. China, obviously. Yes. Yes.

Dave Jones: Are you using the, like, is it a turnkey thing? Like you're getting them to do the board and everything?

Chris Gammell: Like to make the actual bare board? That's right. Yes. Yeah. Same, same company. Yeah. Yeah. Yeah. It's. Yeah.

Dave Jones: There's, there's quite a few of them do that PCB way, JLC PCB. They're all trying to do it. Yeah. Yeah. Yeah.

Dave Jones: Yeah. They're all trying to do these turnkey things.

Chris Gammell: They all want your business. It was basically an led like breakout board. And so it was, but I need to make enough of them that it wasn't been really painful to.

Dave Jones: Well, that, that, that sounds like a pretty simple build.

Chris Gammell: Yeah. Oh yeah.

Dave Jones: For like, in terms of like, there's nothing tricky on it.

Chris Gammell: Yeah, exactly. There's no necessary, like as long as they get the led side, right. Which I'm sure they will. Yeah. It'll be fine. And honestly, it's so few things that would, if we went the other direction, it's not the big deal either. So, so yeah, it's a, it'll be interesting. I mean, uh, the, the lead times were like a little bit, uh, longer than I thought they would be, but then I kind of compared them to a lot of the domestic, you know, online ones too, like all the macro fabs and circuit hubs. And, you know, if I don't want to pay through the teeth, like, uh, uh, yeah, for a quick job, like a tempo automation, we'll do it for like in three days. And so we'll screaming circuits and a couple others, but like, that's going to be like three grand on top of the actual parts and everything else. So it's. Yeah.

Dave Jones: Because they bump other people's jobs. That's right. Put yours on. So they'll charge you 10 times the cost. Yeah. Yeah.

Chris Gammell: They can do it. It's totally fine. Like I totally understand that. So, um, yeah. Yeah. That's, you know, I've never really seen, do you know if there's like a list of all of the ones that are available online? Cause that would be really useful.

Dave Jones: Not that I'm aware of the ones that do complete turnkey. Yeah.

Chris Gammell: Well, just, yeah. Like really, here's the real thing, Dave. There's a lot of CMs out there. I don't want to talk to a human.

Dave Jones: Yeah. Right. Okay. Yeah. You just want to upload your file and get something back. Yeah.

Chris Gammell: I mean, like, I know that that's kind of, uh, uh, a bit antisocial. Well, yeah. And I think even just kind of hopeful too. Like, I don't think that that's actually how it goes a lot of the times.

Dave Jones: Well, it's, it's okay for a lead driver board that has a couple of driver chips and a few bypass caps and some connectors and some leads, you know, but something like the micro supply or something like my multimeter, for example, much, you know, more complex. You've got to go into, you know, various things with them. Yeah.

Chris Gammell: Once you get your thing like done, that would be an interesting video too, to like send it out to a bunch of them and see, like do a compare contrast on like, cause you have a single, you know, like doing a single assembly and like sending out to a bunch of these online services would be interesting. I think.

Dave Jones: That'd be interesting. Like instead of a P a bare PCB manufacturing shootout, it'd be a turnkey.

Chris Gammell: Yeah. Right.

Dave Jones: Cause you could literally, you could literally. But then it's obsolete six months later, you know. Oh yeah. Totally.

Chris Gammell: Totally. Yeah.

Dave Jones: Yeah. It's not.

Chris Gammell: Yeah. I should mention too, um, uh, past guests of the show. I don't think I brought it, maybe I brought it up once, but past guests of the show, uh, Dan Hinch, uh, he does a thing called manufacturing reports now. Um, and he does a, uh, he does comparisons between PCB fabs.

Dave Jones: Oh, nice. Yeah.

Chris Gammell: So that's a, that's a.

Dave Jones: Is it a blog?

Chris Gammell: Is it a, what is it? Yeah. It's kind of more like a wire cutter, you know, like, so it's actually like a qualitative. Wire cutter? Yeah. Oh, wire cutter is like an online tech review site. That's, that's, um, I like, I like a lot. So it's, it's more qualitative. So it's like, um, you know, he has like a standard test coupon. He sends to a bunch of them. He gets them manufactured without telling them and gets it back and does comparisons. And he'll do like a pull test on like a, on a USB, um, micro connector. Oh yeah. Right. Which is great. Yep. Um, he was on, he was on, um, the macro fab podcast talking about it too. I'll link that in.

Dave Jones: Okay. So that's a, that's an interesting test. So if people don't know a test like that, what you're testing for is the quality of the, uh, of the glue essentially used on the PCB that holds the copper onto the pre pre laminate. Yeah. Um, so that, you know, it's at how easily will these pads lift off with mechanical pressure. And then you could also repeat the same test at an elevated temperature as well. So you can test, you know, how easily when you rework pads, will they come off? But a lot of that, but a lot of the thermal side of that has to do with the rating. So if you get a TG 130 board, for example, it means it's rated to 130, the fiberglass in it is rated to 130 degrees and up above 130 degrees, the fiberglass stops being fiberglass and it goes all gooey. And it's basically, so really a top quality one will be say 180. So it'll be a TG 180 board. So if you go into the, when you get your board manufactured, if you see a dropdown option that says TG 180 or TG 130, that's what it's about. A higher quality board that should stand more physical abuse, more thermal abuse, stuff like that. So it's just higher quality.

Chris Gammell: Yeah, no, that's good. And it's, it's interesting too, like what, um, you know, what reality is versus what they spec and stuff like that too. And it's, I, I don't usually attribute it to malice. You know, I think it's more just, uh, you know, they're going fast. They're low cost. Like, so like the JLC stuff, it's all one TG 130. It's like, you know, as cheap as you can get, but for most people, it's like, it's

Dave Jones: like, it's practically free.

Chris Gammell: Totally. Yeah.

Dave Jones: I mean, what do you want?

Chris Gammell: Yeah, no, I agree. And I think that that's, I think it's important to point that out to people. Cause I didn't know about that until like a couple months ago. I didn't know the TG 130 versus 170, 180, whatever. So, um, you know, like it's good to just kind of know what that knob is really, you know, like, so like, like when you were going through the, the online calculator. That's why I did my video. Yeah. That's great. That was a good video. Um, yeah.

Dave Jones: And hour long, but yeah, you know, whatever, but there's a lot of stuff there.

Chris Gammell: Yeah.

Dave Jones: But that's how long it takes to go through. Yeah. There's like, you know, 30 options in there.

Chris Gammell: You're usually very concise. Otherwise, uh, so there was a really good talk. I just got back from Supercon, um, uh, Hackaday Superconference that happened out in LA and, uh, Ted Yapo, who had been on the show a couple episodes ago, he did, he did actually a, uh, a test where he was testing. So he's making an eight gigahertz sampling oscilloscope on, uh, FR408 from Oshpark. Right. So interesting stuff, but he did the thing where he was actually comparing, uh, he was doing some calculations on like line widths and, uh, characteristic impedance. And he tested that at all these different like stepwise functions. And so I'll link his talk and his talk's actually already up. Um, but, uh, it was basically like showing that like the spec impedance was different, uh, different than expected and different than spec'd. Uh, so. Right. Yeah. It was interesting.

Dave Jones: Yeah. But that's because you're not paying for a tight tolerance on that. Yeah. Where they say the, um, they, they, they say it's, you know, 4.3 dielectric constant, but it's like, it can be 4.3 plus minus.

Chris Gammell: Right.

Dave Jones: You know, 0.4 or something. Right. Right. You know, it's like, like it's all over the shop. That's why you pay more for the Rogers material, which specifies that to 0.01%. Yep.

Chris Gammell: Right.

Dave Jones: It's like, that's the difference. Right. Right.

Chris Gammell: I think that's another good thing to know though too. Like that's what, when you see Rogers on a, on a thing, you're paying for really, really high quality.

Dave Jones: Yep.

Chris Gammell: Yeah. So.

Dave Jones: And, and just dimensional tolerance as well, like in terms of the pre-preg thicknesses they will use. It's not just the dielectric material. It's like, you know, when you order your prototype board, well, is it 1.6 millimeters? Well, yeah. Maybe. You know? Ish. Exactly. Yeah. Yeah. And, um, yeah. And that's something we, um, had a problem with, with the 121 multimeter, for example, is the thickness of the PCB. It came in under, nobody, you know, nobody, uh, checked. And the boards were manufactured and they were, they were too small. So we had a few range switch issues.

Chris Gammell: Oh, cause it was, it was, it was depending on that, that thickness and the, the actual,

Dave Jones: yeah. The actual pressure of the contacts on the range, which depends upon the thickness of the control thickness of the PCB. And it came in at 1.35 millimeters instead of 1.6 millimeters. And it was like, yeah, that really, like really screwed it. Yeah.

Chris Gammell: Yep. Yep.

Dave Jones: Yep. So yeah. Yep. It's a real issue. So yeah, if you don't, you know, if you, if you're not paying for control, if you're not paying for controlled specs, then expect them to be anything. Right. Like, you know, it may come in good. Like, you know, like you're saying that guy was doing the tests across all the fab houses, right? Well, one week he may test it and it's fine. And the next week from the same fab house, it may be something different because they're getting their material from somewhere else. Oh, that's a good point.

Chris Gammell: Yeah. Yeah.

Dave Jones: Because when you're paying a dollar for a PCB, like, you know, don't expect the same material and the same process every time. That's right. Right. Yeah. Yeah. Yeah. Yeah.

Chris Gammell: I think that's interesting too. Like from a DFM kind of perspective, like, you know, this is what incoming inspection and type stuff is like high quality operations are doing that kind of thing anyways. So yeah. Understanding what's going into your boards and stuff like that. I was surprised.

Dave Jones: Anyway. So I'm, yeah.

Chris Gammell: Oh, I was just surprised that we were talking to, I was talking to Ted about, after his talk, he was talking about, um, some of the test equipment he has and he thought he had actually a line based, uh, measurement, uh, PC test gear that was like on the line measuring, uh, measuring some RF characteristics and then actually like being matched after the fact. So like characterizing antennas and similar things. Um, cause he had, I think he said one of the pieces of gear he had, what had like 80,000 hours on it or something. Oh, wow. Yeah. So, yeah, that's a lot of.

Dave Jones: They were doing what, uh, select on test components or something, were they?

Chris Gammell: Yeah. Something like that. I think. Yeah.

Dave Jones: So if people don't know and a common, you know, not an uncommon thing in, you know, like precision RF stuff and things like that is that they will measure the board performance of the board after it's assembled and then they'll select the component value to put in and they'll tweak it to, you know, that's really extreme cases, but it certainly it does happen. Yeah. It's not uncommon.

Chris Gammell: Yeah. Yeah. Yeah. That's, uh, yeah, that's something I gave my talk again about RF and that's the, uh, pie filter. I think is the, the thing that they would probably mess with at the end there. They'd have like some matching caps and inductors and stuff like that. So I can, I guess I could, I could, I could link it in if you want. I mean, I guess like I could, you know, put my talk up again if you want me to, you know,

Dave Jones: why not plug your own stuff? Where else are you going to plug it?

Chris Gammell: That's true. That's true. That's true, Dave. If not here, then where? You can't plug it on your own show. Exactly. Then where? You know? Oh, dearie.

Dave Jones: So anyway, yeah, I'm waiting on my 100. I ordered a hundred flex PCBs cause they're so small. They're like 23 millimeters by, which is under an inch. Like it's, it's less than an inch by an inch. So it's absolutely tiny. So there's no difference between ordering 10 and ordering a hundred. I think the next step up was like 150 would cost me a bit more. Yeah.

Chris Gammell: Cause obviously tooling and stuff like that.

Dave Jones: Yeah. Yeah. Yeah. Yeah. It's all tooling. So it's, it's much more expensive. If you want to know, it costs me 77 us dollars for a hundred of these plus, um, you know, plus delivery and stuff like that. That's yeah. That's pretty good. But considering it's a flex PCB, then yeah. Yeah.

Chris Gammell: I think the real, the real question is like, so five, 10 years ago, how much would that have cost?

Dave Jones: Oh, that would be so specialized that, you know, you, you, you didn't bother asking the price. Got it. You just bent over and said, yes, sir. I'll have them please. You know, and then hand it over your checkbook. That's right. Okay. Okay.

Chris Gammell: Yeah. Yeah. If you're, if you're in that spot, you're saying you're, you're, it'd be cheaper to just re-spin the whole board. Yeah. Yeah. Yeah.

Dave Jones: Even though like flex, flex has been around, like I showed in my video, like an example from the eighties. Right. It's like a common in consumer gear, you know, even seventies consumer gear had, you know, flex PCB material. Yeah. True. True. Not, not as advanced rigid flex as we have now where they're, you know, embedded in the layers and, and stuff like that, that they just stick it on the bottom and stick it as, as a bottom layer and things. It was a bit cruder, you know, the manufacturing processes are a lot better now, but still, yeah, it's, you know, it's not a new, people think it's a new tech. Right. Right.

Chris Gammell: Right.

Dave Jones: And open any bit of Sony gear from, you know, the 1970s, 1980s, you'll find flex PCB everywhere. Yeah.

Chris Gammell: Yeah. And it's still, it still will actually. Oh yeah. Open a camera and it's pretty much. No, more so now. Although this is how it is. Yeah.

Speaker ?: Yeah.

Dave Jones: But, but the fact that you can do this as a shopping cart thing now.

Chris Gammell: Yeah. Right.

Dave Jones: Like it's the only difference though, I noted with, um, this, although I haven't used this company before, um, to get a board made, I don't think I have maybe donkey shoes ago, but I forget. Anyway, um, I couldn't just upload the board. I couldn't actually pay for it until they had and complete the checkout process until they had done their engineering evaluation of it.

Chris Gammell: That's just a PCB way thing. Not all of them do that. Is it? Yeah. Yeah. Yeah.

Dave Jones: It's, it's, it's just a PCB way thing. Okay. Right. I thought it was right. Yeah. Because I had chosen flex. Right. Okay. Yeah. Right. It's normal. All right.

Chris Gammell: It's normal. Yeah. Cool.

Dave Jones: Yep. Yeah. So it, it happened within hours. Like somebody had evaluated within a couple of hours and then I was allowed to complete the checkout. Yep. Um, and that's, yeah, that's fine. And there were no questions. There were no, you know, because mine's not a complicated, mine's not a rigid flex PCB. There's a big difference between rigid flex and flex. Yeah. Um, adjust a flex PCB on its own. And as I said, in my video, there is basically virtually no difference between designing and getting a normal board made and getting a flex board made. Mm-hmm. You just say, I want flex, please. Instead of, and I'll give you a 0.1 millimeter flex board instead of a 1.6 millimeter fiberglass. That's it. Yep. That's, that, that is the, literally the only difference. Mm-hmm. Um, it, it only gets more complicated when you want to combine the flex into the rigid layers and stuff like that. Right. So that's what I will eventually do a tutorial video about. Okay.

Chris Gammell: I've actually never done, I've never done that before.

Dave Jones: It's actually quite, quite a lot of effort. Okay. Yeah. To do a video for that. Takes a lot of prep work and stuff like that. And, and I don't really have any examples here. Um. Right. So.

Chris Gammell: You wouldn't have been doing it for hobby projects. You wouldn't have been doing it for work, right? No, no. Exactly.

Dave Jones: It would have been dance work stuff. Yeah. And I don't have them anymore. Right. Right. Right. So, yeah. So I will just have to show like online examples and things like that. Yeah. Unfortunately. Yeah. So that, that's a bit of a cop out, but you know, I can.

Chris Gammell: There's still time, man. You can still make one, you know?

Dave Jones: Oh, well, I can, I can design one and get one made, but then I can't release the video until I. Until it's back.

Chris Gammell: Yeah.

Dave Jones: Until it's back. And then I can go, hey, look, here's the final product. Right. And then I can't put that in the front of the video. It's like, nah. You know.

Chris Gammell: Yeah.

Dave Jones: So anyway. Yeah.

Chris Gammell: Yeah.

Dave Jones: All right. Although I could take someone's existing, like Altium have like demo files. Yeah. So I can take one of their files and just simply send that out for manufacture. Oh, true. You know, and, you know, rather than design my own from scratch, which would, you know, I guess it's, you know, it's, it's a bit of a cop out to use someone else's example, but you know, I'm, I'm, I'm not going to spend a week making one video. Right. Right. Right. Right. You know, it's like designing something just for one video. It's like. Yeah. Yeah. I don't have the luxury of that really. So. Yep. Gonna have a cop out. Oh, well. Please excuse my cop outs. That's all right. Up front. That's all right. All right. What else have we got?

Chris Gammell: Let's see.

Dave Jones: Oh, I love this. I, I, I have to do this. Yeah. I, I don't know who sent me this, but it's a robots. We do love robots here on the.

Chris Gammell: We do.

Dave Jones: Amp Hour. And, um, I, I, I posted, I tweeted, um, a video of this insane, a hundred million dollar robotic warehouse.

Chris Gammell: For groceries.

Dave Jones: And. For groceries. I feel like I didn't get that the first time.

Chris Gammell: Yeah. I didn't, I didn't understand it.

Dave Jones: Yeah. Yeah. It's actually because it's some, yeah, it's in Europe somewhere. And the name of the company didn't ring a bell and they, yeah, they distribute groceries. So they built this hundred million dollar robotic warehouse and, and we'll have to link it in because the video is just so, especially if you speed it up, if you play it at speed, like I did in one of my videos. Yeah. My, um, yeah. Anyway, um, basically they've built this warehouse up to where, to right to the roof. And these robots run around near the roof on this X, Y grid of, um. Yeah. Like rails. Like train rails. Train rails. Yeah. Yeah. Yeah. Train rails. And they, right. And so, so it's complete like square. It's like a grid square thing of like a hundred by a hundred squares or something that fill up the entire warehouse. And all these little robots run back and forth across this grid at the top at like five, five, four meters per second. Oh, wow. Which is really, which is really hauling us. Right. And there's only a five millimeter gap actually between the robots and they can do that because they're on rails. They're not fully autonomous. Right. So they, and they've got sensors that they don't bang into each other. And then, and then actually below the robots trying to build a visual thing here. So you understand what I'm talking about is that actually below them are these baskets of all the goods. Right. So they contain the fruit or the cereal boxes or whatever it is, you know. Right. I think about it. Like, I think they're a food company.

Chris Gammell: It's probably like a, I think it was like the robots, like maybe a meter, meter and a half across in like X and Y.

Dave Jones: Yeah. Yeah. Yeah. Yeah. So the meter is about, yeah. A robot's probably a meter by a meter. Yeah. Maybe something.

Chris Gammell: So then the bucket is the same width basically. That's like, it's the opening between those rails. So that's kind of the way. Yes.

Dave Jones: It's the opening between the rails. So I guess they've got like a little claw that goes down there, a little camera that looks down and they pick it up and then they can shoot that little, that, you know, serial box over to a bin, which then on the side, which has that customer's order. Right.

Chris Gammell: I think, I think it's the other way around. I think that the, the bin that they're, I think they're dropping stuff into the bin. That is the way they, that is the customer's order. Into the bin. Yeah, I think so.

Dave Jones: Oh, that is the customer's order. Okay. Right. Different. Okay. So the stocks on the side, I didn't see the stock on the side though.

Chris Gammell: Yeah. They don't really show it, but I think that's, yeah, I think because that would make sense, right? If you needed to have all the delivery to happen, you'd want it to have an essential location, you know?

Dave Jones: Yeah. And, and, and these baskets actually, and, and they're, they're like stacked like 10 high under each other. Yeah. Yeah. So I guess they would, you know, so it's a huge, uh, buffer. Essentially it's a huge order buffer. Mm-hmm. The, the robots drop and let's assume they do drop stuff into the bucket and that's the bucket's order. Then the bucket eventually gets lower and lower and lower until it drops out the bottom. And I guess it might physically drop onto a truck or drops into a packing and shipping. Yeah.

Chris Gammell: Yeah.

Dave Jones: You know, a packing place. And then they haul us out of there. Anyway, um, it's absolutely fascinating to watch the video of this and it caught on fire and burned down. Apparently this was a big thing in Europe and like, and it was one of the, you can just visualize this. Apparently one of the robots caught on fire and then it just kept on going. Oh yeah. So like, like it actually kept on doing its job. So it was going picking up stuff, but it was on fire. So it spread the fire all over.

Chris Gammell: Yep.

Dave Jones: I would, please do they have CCT footage of this robot running around with its head on fire. Yeah. Um, there was something wrong with the charger or whatever. So the, you know, lithium ion battery just, yeah. Anyway, but it kept on going. Good on it. So that was good, redundant design there. Right. Right. And it kept operating when it was on fire.

Chris Gammell: Got to put a heat sensor on there though, you know.

Dave Jones: Yeah. A flame sensor. Anyway, that was, and it burnt down. Yeah. The whole place was gone. A hundred million dollar warehouse trashed.

Chris Gammell: Yep.

Dave Jones: Apparently it was big news in Europe for like a week cause they couldn't put the fire out. Yeah.

Chris Gammell: I mean, lithium does not want to stop going. No, no.

Dave Jones: It doesn't want to go out easily. And, um, and then all of that. Anyway, I just can, I can just picture it. Anyway. Okay. And so I guess that let that be a lesson. Don't use an exothermic battery technology.

Chris Gammell: Right. That's the lesson. Okay.

Dave Jones: Now, well, wouldn't it be?

Chris Gammell: I don't know. I mean, I guess. I mean, what are your options? You can use a non-exothermic battery technology. Do you like a lead acid or something?

Dave Jones: Well, no. There's other lithium, Holly put the kettle on battery materials that aren't exothermic. Got it. Got it. Right. That are inherently safe. You don't have to use lithium ion. There's other lithium poly things, you know. So. Something safer. Something safer. Yeah. Something safer. Yeah. So, oops. Yeah. I just think that's hilarious. Anyway, it was, I was just mesmerized when I saw that video. Yeah. I thought I've seen the future and well, yeah. Anyway, that was great. Yeah.

Chris Gammell: It's great. It's great.

Dave Jones: What else we got? What's this assembly language versus machine code rubbish?

Chris Gammell: So, that is former guest Ben Eater. It's another. Yep. He's building up another computer. And it's basically, he's kind of walking you through how assembly language becomes machine code. That's kind of the main idea there.

Dave Jones: Oh, right. So, he's not advocating for using machine code. Oh, God. That would be. No, right. That would be a truly hardcore thing. That's pretty die hard. Yeah. Well, there are some. There are some out there. Yeah.

Chris Gammell: I've got my hex keypad and I'm just going to do it myself, man. Yeah. Yeah. Yeah. That's not for me.

Dave Jones: Yeah. See, for me, I think that level of detail is pretty redundant. It's like once you've done, like the whole idea of assembly is that you execute instructions register by, you know, instruction by instruction. You're instructing the process of what to do. The fact that it's in a higher level thing, you don't, it makes no practical, you don't gain any knowledge by knowing, oh, that converts it into 8D in hex. Right, right, right. You know, like it's, you're already at the instruction level. You don't, there's no lower, there's really no reason to go lower than that. Yeah. In fact, you can argue that you can't. You're, you're, you're already on the floor and you're just trying to explain what the floor looks like.

Chris Gammell: I mean, I think the reason, the reason that maybe you'd want to go lower and it might be relevant to Ben's thing is that like, you know, you need to get to machine code because I always think about it like, like, uh, uh, uh, instruction that's going to a processor is effectively, you know, it's, it's a series of ones and zeros, right? That's every processor. Yeah, of course. But that is effectively then activating gates that opens, you know, trap doors and that's where data is going, you know, it's basically switching between different things. So if you're trying to construct the gates that are in the ALU or something like that, then yeah, you do need to know.

Dave Jones: Oh, at, at the chip level, it matters. At the chip level, yes. You're talking about chip level design. Oh yeah, of course, of course. Right. But from a programmer point of view.

Chris Gammell: Yeah, I think the rest is more academic and things like that.

Dave Jones: Please, I'm sure there will be hardcore people out there who will totally correct me on this, but you know. Mm-hmm. Yeah. Once you've done your assembly language, but, but there are, and like, there is high level assembly language. There's like macros and stuff like that, you know, like that's, it's kind of like higher level stuff, you know, and, and you can do loops and, you know, things like that. And, you know, that aren't necessarily, you know, that, that will substitute in the, you know, the jump command and things like that. So it's kind of, yeah, yeah, you know, but still, once you're down at assembly, I think you'd, you know, you're.

Chris Gammell: Yeah, you don't need to go any lower. Doing the traditional thing. Right, right.

Dave Jones: Mason, yeah, it'd be extreme example of where you'd need to go lower.

Chris Gammell: Right.

Dave Jones: So, I mean, it's interesting. It's fascinating. Yeah. Don't get me wrong.

Chris Gammell: Well, I had a, I had a rude reintroduction to the world of, of low level digital logic this past weekend. When I was.

Dave Jones: Oh, do you tell?

Chris Gammell: Well, so that the super conference badges wall was all, it was an FPGA and it had three processors on it. It had two risk fives and a pick 16, I think.

Dave Jones: Was it a hard, were these hard?

Chris Gammell: No, no, they're all, they're all soft core FPGAs. They're all soft core. Yeah, the two.

Dave Jones: In that case, they didn't have them. It's just an FPGA. You don't need to say any more.

Chris Gammell: It.

Dave Jones: You mean it came preloaded. Yes. Sorry. Soft core. And then like you can compile it. And FPGA is just an FPGA. Yes, that's right. Yes, yes, yes, yes.

Chris Gammell: Yeah. Yeah. No, it was, it was confusing as hell, man. I'll be honest. Right. I really, you know, so like I really appreciated being able to like try stuff out and, you know, get to play with some Verilog again, all these open tools that we've been talking about on here.

Dave Jones: Of course, but it's not easy stuff.

Chris Gammell: It is not easy stuff at all. So I was very, very, very impressed. I don't know how the video turned out, but like the badge hacking competition was at the end of the weekend. And, you know, there was people that were working on the FPGA and actually writing like custom logic for the badge hacking. Right. So some of them like wrote custom logic, wrote code for the RISC-V processors that then controlled the custom logic. And like it was just, yeah, it's like a masterclass. So I have this badge on my desk now and I'm just telling myself, don't put it away. Don't, you know, like it's like.

Dave Jones: Yeah, yeah, yeah.

Chris Gammell: I was talking to a friend earlier today. I was like the, this is a reference to a movie, I guess. But the idea of like someone saying, you know, I'll use the conference badge as a dev board is the equivalent of like the bridesmaid being like, oh, wear the bridesmaid dress again as a cocktail dress. You know, something like that. It's just like, it's just a complete lie. We all know it's a lie. It's never going to be used again. But, you know, they both look very nice. Right. But it had like a huge FPGA in it too. It had like a 45,000 LUT FPGA on it. So you could, even with all the stuff they had in there, it was. And so, I mean, and Sprite was on the show when I recorded at the conference. But with all that stuff in there, it was still 60% full after three processors, a display, driver, and everything else that was in there.

Dave Jones: Oh, yeah.

Chris Gammell: Yeah. No.

Dave Jones: Well, if people don't realize, just, you know, when you say like it's got 43,000 elements, right? Yeah. Logic elements, LEs or whatever you want to call them. Yeah, right, right. They go under different names for different brands. They're not equivalent between, not only between brands, but between different models. Yeah, yeah, yeah.

Chris Gammell: I remember.

Dave Jones: You know, stuff like that. So, you know.

Chris Gammell: Actel, MicroSemi, whoever's owns them now, they were six. I think there were six lookups in those and four design elements.

Dave Jones: Yeah, it depends how many lookups you've got, how many flip-flops you've got, what sort of architecture is inside and how they're connected, everything else. So, one design in one FPGA may take, you know, a thousand logic elements. Right. So to speak, or whatever you want to call them. And some other design may take 50. Right. You know, because it may be more efficient for that particular thing that you want.

Chris Gammell: Yeah.

Dave Jones: Actually to do. So, you know. Right. There's huge variability in that.

Chris Gammell: Well, and honestly, I think for a long time it wasn't a big deal because it was like, oh, you're comparing Xilinx to, you know, MicroSemi or Intel, Elterra, whoever the hell they are now. You know, it's like, you're never going to switch, right? But I think the crazy thing is now you actually, with the tools, you can actually switch. So that was the other piece of news that came out of there is that they have, so Tim Ansell was on the show a while back, but he gave a talk about how, oh, I forget the name of it. Oh, Project X-Ray is the project to get Xilinx parts, Series 7 parts, working with these open tools. And they're, I think, 85, 90% of the way there, which is kind of crazy. Wow. So now there are two viable options, two viable real, you know, chips that are, that have reverse engineer bit streams, which is kind of a big deal, you know. And especially because the Series 7 stuff, I think the Arctic stuff, the Arctic 7 has some higher end kind of things. And like the zinc is also potentially, that's the one with the hard processors in there.

Dave Jones: So unfortunately, I don't keep up with all the new FPGA family architectures these days. It's like, God, there's just so many.

Chris Gammell: I mean, they're still making Spartan 6 and 7. So that's like, you know, the majority of the market, I think.

Dave Jones: 6 and 7? Spartan 6 and 7? I can remember the original Spartan.

Chris Gammell: Oh, yeah, yeah, yeah. Sorry. I thought you had been maybe recent enough. Spartan 6 has been around for five, seven years or something like that.

Dave Jones: No, no, it's been, it literally is, it's been nine years since I worked on an FPGA project.

Chris Gammell: Oh, okay, okay. Yeah, well.

Dave Jones: That was post-Altium. I was still doing FPGA stuff post-Altium as a contract role. Oh, okay. But, yeah, since then, no. Yeah. No, I basically haven't, I haven't really done anything apart from, you know, the odd eval kit. I've looked at stuff like that and gone, oh, yeah, you know, they've now got this new Kintex whoop-de-doo, you know, architecture or something, you know.

Chris Gammell: Right. And it is incremental. A lot of it is incremental progress. I think a lot of the stuff that's been coming out is the, like, the faster and faster transceivers that are out there. But, yeah, I don't know.

Dave Jones: Well, that's one end of what FPGAs do. You can, like, specific architectures, they will target high-speed I.O. Yeah. And that's what, so they're doing the 100-gig-bit Ethernet interfaces, right? Right, yeah. That's what they'll have hardware cores in there that handle that sort of stuff, whereas other architectures, they won't have that. They'll be optimized for, you know, soft processors and, you know, other stuff, and they won't contain any high-speed, you know, serial stuff around the outside. Right, right. Yeah. So.

Chris Gammell: Yeah, it's interesting, too. Like, so part of my thought process, you know, not very well thought out, as you'll be able to tell soon. But, like, you know, I was thinking through, like, okay, so this part is on the board. The one that they're using is, like, 45,000 LUTs. There's another one that's, like, I think 12,000 LUTs. Again, you know, whatever that's worth. It's at least.

Dave Jones: Lookup table, logic elements, flip-flops.

Chris Gammell: No, no, no, I just mean, like, you know, what the difference between 12,000 and 45,000 effectively enables, you know. And, but, like, the one that's, like, 12,000 is, like, $5, $6, you know, at decently low quantities. And so you start to think, like, okay, so now I can get an FPGA. I could put a soft processor in there. And, you know, and then you have flexibility around it. And it's, like, I don't think that that's a great, I think it's much more an academic exercise. Like, right now, like, if I put that into a client product, it'd be, like, okay, if I'm really, really future-proofing something, maybe I'd be doing that. But I don't think that there's, there's definitely not a cost basis to do any of this stuff. But there might be reasons to do this.

Dave Jones: No, and this was, well, of course. But this, it's always going to be niche. And this was the whole vision of Altium for a decade was that the future would be FPGAs. And you wouldn't be using regular processors anymore. Everything, why? Why the hell would you use rigid hardware when you can, you know, use these FPGAs that can just reprogram to do anything you damn well like? It's, like, and it just never worked out. It just, yeah. Right. Because of economics. It was in the detail.

Chris Gammell: Economics, too, of, like, yeah.

Dave Jones: Economics of silicon space. Yeah, it was just never. But they thought, you know, that was the vision, that FPGAs would be, you know, you could buy one for 10 cents.

Chris Gammell: Oh, not that cheap.

Dave Jones: And, you know, you wouldn't need to buy anything else. Yeah, exactly. It's, like.

Chris Gammell: Although, maybe those new Go-In ones, the Chinese ones, you know. We'll see how those go.

Dave Jones: No, right. Yeah, well, you know.

Chris Gammell: But I think that, like, this is, I think this is partially why I'm excited about the, like, the FPGA stuff, too, of, like, the swapability. I know we've talked about that on here before, and your blood pressure just rose a little bit. But, you know, the tool chain kind of stuff. I think that it's interesting, you know, especially with all the RISC-V stuff that's coming out, too. So, it's, it'll be interesting to see where it all goes. And, you know, I might consider it for, you know, if I had, like, a client that was not super, you know, that they were worried about their future, but they weren't necessarily, like, super stringent on what they needed cost-wise or stuff like that. Yeah, maybe it would make sense, you know. So, at least enough that I'd like to put this in the toolbox at some point. You know, I definitely didn't, didn't hit my goal for the weekend.

Dave Jones: Right. It's pretty rare that you've got to worry about the future in quote marks when it comes to choosing a processor, though, for a product. You know, it's like, it's, it's pretty niche circumstances where that's going to be an issue. Like, if you choose an ST ARM processor or you choose a microchip pick or something like that, right, you're, like, they're pretty safe darn choices. Oh, yeah. No, no, that's true. I meant more like, it's like.

Chris Gammell: So, so I've had, I've had clients that, like, you know, they were working through, we're working through multiple iterations. And, basically, we needed to enable different functionality. And just having, like, flexible pins and stuff like that. Like, you know, there's many ways to skin this CAD, of course, but, like, it would have been, I could think of at least one scenario where it would have been useful in the past. Now, there's chips that are out there that even microcontrollers are, like, the NRF chipset, like, you can switch a lot of those I.O. around, too. So, it's not, it's not like it's that huge of a deal, you know, like, if it's just flexible I.O. But just thinking through, like, you know, customization and similar things, you could, you could start to use it for, if you got in a bind, I think it would be useful. It's probably not a reason to design it in. But it's, it might be nice to have at some point.

Dave Jones: Just like, I mean, there's some microcontrollers like that. Like, the pick, the higher end picks are an example. They've got configurable pins. So, you don't have to hardwire things into, you know, like, if you've got a UART, for example, you don't have to put it to the UART pins. You can put it to almost any I.O. pin and then route it and route the hardware module inside. They've actually got FPGA-like routing inside.

Chris Gammell: It's like a cross-tatch. That's super, super convenient, wherever that exists. But, yeah, that's, like I said, that's not the reason to design it in FPGA, for sure. But it's one of, that same scenario is one where I thought it could have been useful.

Dave Jones: And it's just not a longevity thing. It's like just a, you run into a silicon bug issue. I used to pick 24F series chip once. And I was designing into a product. And the UART pins were back to front.

Chris Gammell: On the chip?

Dave Jones: And it was not me. Wow. On the chip. On the chip. In the silicon, it was documented wrongly. So I laid out my board based on the data sheet pinout. And my design was correct. I didn't goof it. It turns out that the next version of the data sheet had, oh, yeah, by the way, here's the e-ratter of the chip. And, yeah, the UART pins are swapped. Oh, that would be the worst.

Chris Gammell: Because, like, you know, like when you're like, oh, please don't let it be this. Please don't let it be this. You know, like when you're like on the bench, something's not working. You go check this data sheet. You're like, oh, it's not this. And then it, wait, what is it then? And in your case, it was that the silicon was wrong or the documentation was wrong. Oh, my God.

Dave Jones: And that's one of the things where a very common goof, I've done it many times, where I've made the mistake of swapping TX and RX, right? I mean, I've done it probably too many times to count. So I assumed that I had goofed. You know, when I was debugging this thing, it didn't work. It's the first thing you check, right? The first thing you check is going, oh, my God, I've goofed. I've goofed TX and RX, right? Everyone does it. It's a rite of passage. And you do it more than once. Oh, yeah.

Chris Gammell: I started doing it where I do from, you know, device to device. That's like how I label it. So like I say, like RX from micro TX to sensor. Right.

Dave Jones: You're very descriptive. That's the only way to do it. Yeah.

Chris Gammell: But I then hooked it up to the wrong pin when I actually, the label was right. And then I still hooked it to the wrong damn pin on the screen. Yeah. So it's not a panacea.

Dave Jones: Yeah. So I can vividly remember going through the debugging process and going, oh, I've goofed it. Easy. You know, I'll swap around. Right, right.

Chris Gammell: Yeah.

Dave Jones: And like, and I really didn't like, and I thought, so I looked through and I thought, oh, it looks okay, but no, I must be interpreting it wrong. So I physically got in there and swapped the tracks over. You know, I bodged it all in. It was hard because it was my little micro cow. Oh, yeah. Right. So it was a little ultra micro, ultra thin, you know, like, so I was paying the ass to do physically, but I managed to swap the pins and of course it worked. And then I kept on checking and I went, I don't think I've got this wrong. Yeah. And then I looked at, downloaded the latest data sheet that had just come out. Sure enough, in the errata section was the, you know, yep, was the, by the way, yeah, we've swapped the, and now the pinouts changed. Yeah. Yeah. The main pinouts changed between rev C and rev D of the data sheet. Like, oh God.

Chris Gammell: Yep. Yep. Yep.

Dave Jones: You know, and, but that would have saved me. If that chip had the reroutability function, I wouldn't have had to do a physical, you know, route or something like that.

Chris Gammell: So a physical mod. Yeah. I mean, and then just even the layout is easier too, because you just pop it right out of the pins and stuff like that. Yep. Yeah.

Dave Jones: And yeah, like speaking of which is like, as we were talking about before, the, um, Hikro supply with that flex PCB, that new processor on there. Um, oh, by the way, an update on that.

Chris Gammell: What's that?

Dave Jones: Uh, that just came through last night is that David did manage to get everything into the 32 K.

Chris Gammell: Oh, great. Okay.

Dave Jones: With, with the help of, of the rich tech, uh, people who are the ones that make the USB PD, uh, library, they've been working with him to really get the, uh, code down. Um, but in the end it still wouldn't fit. He was very close to, I, I, no, I think he had just squeezed in the USB PD with their help, um, but couldn't fit anything else. But then he went back to his own code and he was able to shave five or six K off his own code, um, by optimizing some library, you know, somewhere. And, um, yeah, he's now got everything to fit in the 32 K, but for prudence, we're still going to design in the larger chip. So that means we have to relay out the board and put in the quad flat back because we're not going to use this flex PCB as a production solution. This is only for the prototype boards because we've already like, cause they're quite complicated boards, um, which we had a factory assembled by the way. Um, cause we didn't want to do them by hand and we got half a dozen of them made. And, um, yeah, so we didn't want to scrap those and rebuild them. So that's why I did the flex PCB, but anyway, it fits and we're picking and now we're saying, oh, well, do we use the 64 K or do we use the 128 K to future proof this thing?

Chris Gammell: Are they, uh, and it's like, are they pin compatible or no?

Dave Jones: Yes, they're pin compatible. So it doesn't probably, but, but the problem is the problem is, is that once you ship them, you can't update them. Yeah. So do you put in the bigger chip from the get go so that you don't like, you don't alienate your early customers who will buy the product? Oh, it's only got a 64 K part, but damn, somebody has written this new great new firm, which has all these features, but it needs 128 K part.

Chris Gammell: So yeah, I've been, I've been dealing with something like that recently too, where I've been like stressing about something and I'm like, you know what, this is not the time to cost down. Like cost down is a whole separate design thing, you know? Yes. It's gotta be working first. It's gotta be shipped. It's gotta be, and like, you give up a little bit on margin, you give up a little bit on everything else, but like that is.

Dave Jones: But you gain flexibility.

Chris Gammell: Unless you're like, unless you're a toy designer or something where like, you know, upfront that cost is the most important thing. Then yeah, I think that that's, I'd ship, if it was me, I'd ship the 128 unless it's like such a glaring cost adder that it's like, what's it like a dollar difference?

Dave Jones: Absolutely. If it was double the price or something, yeah, you'd, you know.

Chris Gammell: Even then maybe it's worth it. You may not do it. Maybe it's worth it.

Dave Jones: Maybe.

Chris Gammell: And then the people that have the early one, you know, they, they have something special, you know?

Dave Jones: Right. Yeah. Something special.

Chris Gammell: But yeah, that's, that's something. Yeah. You can't really, you can't go back on that.

Dave Jones: So Mike, Mike's electric staff did a video on leads. This is actually, we've gone over that before. Isn't that an old one?

Chris Gammell: No, this is his presentation from Supercon.

Dave Jones: But I thought he's given this lead presentation before everything I've learned about leads.

Chris Gammell: I'm sure I've. Possible. Yeah.

Dave Jones: Watched it before. I think he's given it before at some other conference.

Chris Gammell: Well, I mean. Hasn't he? Am I wrong? Can he give a different talk about everything he's learned about leads? Oh, maybe.

Dave Jones: I, you know. Okay. So, so it is a new one. All right.

Chris Gammell: No, I mean, I don't know if it's anyone, honestly. I mean, I just, I just, you know, link things indiscriminately, Dave. Right. Okay. No worries. One, a new video that is out is the Great Scotland actually, that's kind of related. I saw this one. It was about the USB PD, which is why I was interested because.

Dave Jones: Yeah. I skipped through it, but it was kind of, oh, he's just using a demo board. And my comment. He is, yeah.

Chris Gammell: But it was on, it was on the same one as the, I thought the 72 was the same chip you guys were using.

Dave Jones: Uh, yes. But I, does it use the Rich Tech chip? I'm not sure for the USB PD.

Chris Gammell: I don't know if it's that, but I just, I was just looking at the STM, all I noticed was the STM 32 072. Yep. So. Yep. That looked familiar. Let's see. It's the F4 P6.

Dave Jones: Yes, it is, but it might use a different, anyway, I haven't, yeah, I don't remember details, but.

Chris Gammell: Yeah.

Dave Jones: Yeah. My, my, someone posted this on the EV blog forum and my comment was, I'd love to see him do a from scratch. Oh, sure, sure, sure. USB PD design. Yeah. Because he makes it look easy. Yeah, that's not really what he does. And their demos make it look easy, but when you get down to the details to design a fully compliant USB PD.

Chris Gammell: Someone's being defensive here, folks. This is what this.

Dave Jones: Yeah. Because it's hard. Of course. Of course. Custom is always harder. This is why products do not have USB PD in them. This is why they sold in irons. And they have USB quick charge instead of USB PD. Because to do it properly, fully compliant is actually really hard. Sure.

Chris Gammell: Hey, nobody's, nobody's debating that, Dave. You guys are.

Dave Jones: So I, so I don't want people to like, because I've like spent a lot of time and money on this personally. Sure. Sure. Right. So it's like, so just be wary. That's what I'm saying. Great video, but just be wary of thinking that's trivial to put USB PD into your product.

Chris Gammell: Right. Right. I agree. I'm saying it's not.

Dave Jones: Although he's saying USB type C. He's just saying USB type C, which, but it's mostly on USB PD, isn't it? I think.

Chris Gammell: Like he goes over, he goes over the differences and that's why it's a good video. I mean, like, so.

Dave Jones: Yeah, yeah, yeah. No, of course. No, it's fine. I just, yeah. Want people to be aware that it's, yeah, can be a deep trap. Right. So, yeah. Anyway. Yeah. And, and everyone's coming out and saying the same thing. It's like, you know, I've, I've learned the lesson the hard way. David and I have learned the hard way. And everyone else comes out and goes, oh yeah, USB PD's a pain in the ass. And no, you know, some people say no manufacturer does it right. No manufacturer has a fully compliant library or something like that. They're all bogging it and they're doing it. Cause you go like it's 300 pages, the USB PD specification. It's.

Chris Gammell: That's what has amazed me too about like Bluetooth and stuff. Like, you know, they talked about like, oh, well, Apple Bluetooth is different than the other people do Bluetooth. And it's like, wait, that's when I started to realize that standards are like, they're just suggestions, you know, like, and it's about interoperability, but, but yeah, if you, if you say, Hey, well, I mean, if you have to be the most valuable hardware company in the world to, uh, you also get to do other things, but like, um, but yeah, it's not like, it's not like it has to do A, B, C, and D. It's just that if you want to interact with the largest number of things, then yeah, you better, you better, you know, follow the spec as close as possible.

Dave Jones: So, and apparently USB PD is the mother of load of, you know, issues like that. So, yep.

Chris Gammell: Yeah. Very complicated standards. USB in general is, is a pretty tough on its own. And then, and then you add a lot of other. Yeah.

Dave Jones: And then you add USB. Yeah.

Chris Gammell: Yeah.

Dave Jones: Which you wouldn't think it'd be that hard because, oh, it's just power delivery. All it is, is just sending a few commands to go, oh, please give me this power. And that's it. Job's done. But no, apparently it's, it's, you know, well, the most 300 pages complicated. So, yeah, it's, it's nuts. Anyway. Yep. Cool. Any last items? Because our time is off.

Chris Gammell: Yeah, it is. Another good video from the Flex LED 2.0. This is a good one. This is from.

Dave Jones: All we're doing is pointing out videos and here we are doing a radio show.

Chris Gammell: I know. This is from Carl. That was a good one. I liked it from, it was like, he did a lot of Altium stuff. So I, you know, just kind of seeing the walkthrough on the Altium stuff is interesting. Well, I guess the sped up ones. But he's doing like a magnet based POV display. So that's kind of cool. Okay. Do we need to talk about Yagio buying Kemet? I don't even know if that's how you say it. Yagio? Yagio?

Dave Jones: Yaygo.

Chris Gammell: Yaygo?

Dave Jones: Yeah, sure. That's how I say it.

Chris Gammell: Yeah. Okay. I don't know if it matters at all.

Dave Jones: They're buying Kemet. Yeah. You know.

Chris Gammell: It's interesting given the, you know, like how hard it was to get parts for a long time. Yeah. Caps. Yeah. We're talking about. Yeah. Of course. We talked about that. And, you know.

Dave Jones: Do Yagio do anything else? Do Kemet do anything? Are they just one capacitor manufacturer buying another?

Chris Gammell: Yagio does a lot of resistors and other passives. Yes. I thought so. Yeah, yeah, yeah.

Dave Jones: Yeah. Right. Yeah. But is that like all like passives? That's pretty much all they do?

Chris Gammell: I think they might do like antennas and stuff as well.

Dave Jones: Okay.

Chris Gammell: But it's definitely, they don't make any silicon as far as I know. Or any smart silicon. Maybe some active devices like diodes, but.

Dave Jones: Right.

Chris Gammell: They're like a clearinghouse of, you know, like a Vachey or anyone else too. You know, there's like another one of them.

Dave Jones: Well, I guess all I can say is that, well, it's a Taiwanese company instead of Chinese. Sure. So, you know.

Speaker ?: Sure.

Dave Jones: I guess. That gets around the current political dilemma to do with China, I guess. So, yeah. Yeah. Anyway. I don't know. Like, do any of us care? I don't, you know. Yeah.

Chris Gammell: I don't know if it's a direct impact kind of thing. I honestly didn't hear about it until like a week and a half after it happened.

Dave Jones: Oh, Kemet also do AC line filters, EMI cores, flex suppressors, electrometer, relays. Kemet do relays?

Chris Gammell: I did not. Yeah. I mean, like, that's the thing. Like, at a certain point, like, it's going to go on your bomb. It's going to be like, you know, be in your parts or be in your products rather. But, like, I don't have much loyalty to any particular brand other than, you know, very, very specific ones.

Dave Jones: Not really. Yeah. Meh. That's all I got to say to that.

Chris Gammell: Well, then, yeah. I guess we're good then.

Dave Jones: We're good. Yeah. Everybody happy.

Chris Gammell: Yeah. I'm going to go try and learn some FPGA stuff. We'll see how that goes. That and micros and RF and everything else, you know. Learning. Right. Got to keep learning. Yep. All right, Dave. Talk to you next time.

Dave Jones: Catch you next time.

Chris Gammell: No way. I say I'll talk to you soon, I guess.

Dave Jones: Yeah. And I say catch you next time.

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

Dave Jones: My catchphrase. Catch you next time.