#554 – PLEASE be a die shrink

Chris Gammell: This is The Amp Hour Podcast. Released August 15th, 2021. Episode 554. Please be and die shrinking.

Dave Jones: Welcome to the Amp Hour. I'm Dave Jones from the EEV blog. Howdy, y'all. I'm in North Carolina now.

Chris Gammell: That was terrible.

Dave Jones: At your sprawling mansion in North Carolina. The podcasting business is good.

Chris Gammell: If you don't know, I'm Chris Gammell. I've been here a while. Yeah, right. I recently moved to Durham, North Carolina, and that was my terrible, terrible accent.

Dave Jones: Right.

Chris Gammell: I will not do that again. Right.

Dave Jones: Has any one mentioned? Because what accent do you have, technically? Because I don't know the various. I don't really have. You don't really have one? Just a generic Yankee?

Chris Gammell: Western New York is where I grew up.

Dave Jones: Right. Okay.

Chris Gammell: Yeah. I don't know. You know, just generic nerd.

Dave Jones: So is there a specific, there's a specific accent there, is there, in North Carolina, or is it?

Chris Gammell: Yeah. I mean, you start getting, like, if you go, like, yeah. A bit more twangy, a bit more banjo-y. Yeah, a little more twangy.

Dave Jones: Right. You know, it's very charming.

Chris Gammell: I like it.

Dave Jones: Excellent. Well, you have to, because you bought a sprawling mansion there. So, you know, you've kind of invested big time there. So, yep. Yeah. Ah, boy.

Chris Gammell: But I'm here.

Dave Jones: Excellent.

Chris Gammell: I had my last meetup in Chicago, a nice little outdoor thing. You know, meetups are probably not looking too good in the US for a little while, because of the Delta variant.

Dave Jones: Oh, okay. Same thing happening there again, is it? Yeah, unfortunately. Because you guys were starting to open, and then we were, of course, open for the last year, and then all of a sudden we've... Kamigatza? Wah, wah, wah.

Chris Gammell: Yep. Yeah. So, I don't know when it'll happen, but I would like to start a meetup here at some point. There are some. There's actually one here called Raleigh IoT, I think. It's Riot. So, they have, like, a lot of virtual events. So, anyone can actually do those. But I'd like to do some in person. You know, just show up. Yeah, yeah, of course.

Speaker ?: Bring your hardware.

Chris Gammell: Usual. Yep. Usual shtick. Yep. Well, you know, if there's people in the area, let me know. I'd love to hear from you.

Dave Jones: Maybe you could have, like, a hardware camp or something.

Chris Gammell: Sure. Yeah. I mean, I think the real thing, right now, it's like, who knows? But at some point, yeah. You said hardware camp? What would that be? Camp?

Dave Jones: I don't know. It's North Carolina. Don't you? Aren't there lots of woods there and stuff? Oh, yeah, yeah, yeah. I imagine.

Chris Gammell: That's true. I mean, now where I am. Right.

Dave Jones: Okay. Well, yeah, I'm sure if you head out of town for half an hour.

Chris Gammell: Yeah, yeah, yeah, yeah.

Dave Jones: Jeez, you wouldn't even be half an hour out of there, would it?

Chris Gammell: Yeah, probably not. Yeah. I went to that tour camp, which is the one two years ago. Yep. And that was a lot of fun. Two or three. No, three years ago now. That was a lot of fun. And, yeah, you know, like, just hanging out with nerds in the woods. It's good times. Cool. So maybe we'll do something like that here. We'll see.

Dave Jones: Although I suspect most nerds have never camped in their lives, so, you know, it's like.

Chris Gammell: Well, you know, you can go glamping where you, like, bring a trailer. Yeah, right. You know, like a. Yeah, right. Yeah. Yep. What do you call those, caravans? Is that what. Caravans? No. Isn't there like a camper, like a dragging, like a tow?

Dave Jones: There's a camper van. Like the powered. Basically, the camper van is one that powers itself. It's its own, you know. I see. Got it. It's a Winnebago in the US speak, right?

Chris Gammell: RV. Yeah, that sort of thing.

Dave Jones: It's an RV. Yeah. Yeah. Right. Yeah.

Chris Gammell: Okay.

Dave Jones: So, whereas a caravan, a caravan here is one that you tow.

Chris Gammell: Yeah, that's what I was thinking. Yeah. Okay. Yeah. With a ute. You tow a caravan with a ute, right? Right. That's right.

Dave Jones: Or you're holding Kingswood. Yep.

Chris Gammell: Yeah. Sure. Yeah.

Dave Jones: You have no idea what I'm talking about. I know that one. Yeah. That's all right. That's all right. No worries. Yeah, yeah.

Chris Gammell: Anyway.

Dave Jones: Yep. So, you couldn't even find a camera. So, you're literally surrounded by boxes, are you?

Chris Gammell: Oh, just boxes. Yeah. Just, it's... Right. You know, it actually... I had gotten to the point where I had all the boxes kind of like lined up on the walls and all pretty well. You know, everything's in the right room, at least. Yeah. You know, I did like a manifest. I tracked all like that.

Dave Jones: Manifest? Wow. Oh, yeah.

Chris Gammell: I used my label printer. Everything had a number.

Dave Jones: Oh, wow.

Chris Gammell: Tagged with three... Noodle it. I used three numbers. I had three labels for each one. It had the box number. It had my phone number on it. And then you could... So, you could put it on three different sides of the box. So, it was almost guaranteed to be seen. I printed custom, like heavy labels and fragile labels. I love label printers. If I haven't gotten that across, I love my label printer. I use it way too much. And then the manifest allows you to have a lot of detail.

Dave Jones: I'm sure there's professionals you can speak to about this, Chris.

Chris Gammell: Oh, I'm sure.

Dave Jones: About this problem. Like go to like therapy or something. Yeah, yeah. Yeah.

Chris Gammell: Yeah. Yeah. It's a codependent relationship. Right. I click print and the label prints and all is right in the world.

Dave Jones: Sounds a bit master slave-y.

Chris Gammell: Yeah, maybe. Yeah. Yeah. You know, there's definitely like a... Right. There's some weirdness there.

Dave Jones: Yep.

Chris Gammell: Yeah. The only thing I wish is that there was better like Linux support for this. I have like a really old brother, like a QL 1050. I got it for like 150 bucks on eBay. But let me tell you, label printers, really worth it. Get the kind that don't have like fixed labels. Get the kind that are like continuous labels and then they cut. Yep. Alvaro from the Unnamed Reverse Engineering podcast. He turned me on to these things because he outfitted his whole lab like with... He labeled everything in his lab.

Dave Jones: Yep. Oh, I love my... Zebra printer there. Yeah. Like the de facto industry. Oh, I've got several. Yeah. They're great. Yep. Yep. Terrific. And I've got the little... The smaller brother... Not brother. Dymo one as well. Oh, yeah. Like the Dymo 450. That's all you need. You can get it for like 50 bucks or something. And you can get all the different label types in there. And it's absolutely fantastic.

Chris Gammell: The Dymo is like the thin strip label? Is that like the... No, no, no.

Dave Jones: Well, no. They're the Dymo tape ones. They're the ones with the tape machine, you know, the handheld tape machines. But no, Dymo make a desktop version. It's a Dymo 450 turbo. It's been around for... I've been using it for 12, 13 years now or something. It's been around for a long time. It's like... I don't think they've ever upgraded the thing apart to go from regular to turbo. And yeah, it's the same thing. You put in the... But it isn't as wide as a Zebra or that brother one you've got. So it isn't as wide. It's a bit more narrow. But it has the reels in there. And you can get all the different labels. You know, you can get individual labels. And I use that for all my mailing. I've used it for...

Chris Gammell: Yeah, exactly.

Dave Jones: ...wedding invitations.

Chris Gammell: You've got to be prepared to get taken to the cleaner on some of the printing supplies and stuff like that. But yeah. The thing I really want to do is the custom... Like the acid wash safe label printers for circuit boards. You can get those kind of things. Ah, right, yes. The polymide style ones. Assemblers use them. And there's QR codes. So I used Screaming Circuits for a build a couple of months ago. And they actually apply them. So I just adopt their unique ID system, that sort of thing. Yep. But I'd love to actually have my own printing system. And then the Zebra ones, like you mentioned, those are Python scriptable and whatever. So there's a bunch of articles out there about that stuff. I just haven't had the tooling necessary for that sort of thing. But at some point, I'd like to.

Dave Jones: I totally agree. Label makers are the king. I love them. Yeah. I'm not an organized person. But it makes me feel like an organized person. That's really what's important here.

Chris Gammell: It's like I'm cosplaying as an organized person.

Speaker ?: Right.

Chris Gammell: Yeah. So anyways. Yeah. Everything was in boxes or still is in boxes pretty much. And the problem was that I started opening the boxes. And then it's just chaos.

Dave Jones: Right.

Chris Gammell: You know, it's just like, oh, look, there's a lot of stuff everywhere.

Dave Jones: No, that's why most people leave me boxes. Yeah. It's just easier. You know. Yeah. Totally.

Chris Gammell: Yeah. Yeah. And then they'll be packed for the next time I move. Exactly. You know, I was digging through. Thank goodness I knew where my old ass J-Link was because my old ass microphone also uses USB-A. I couldn't find my USB-A cable otherwise. And, you know, sometimes you get a pill for cables from one another. But also, people still use USB-A apparently.

Dave Jones: Including me. Label printers. Because all my label printers use USB-A. Oh, true. That's a good point. Actually, yeah. Everything. Yeah. That would have been another source of one. Yeah.

Chris Gammell: That's a good point. Yeah. Yeah. I don't know. Yeah.

Dave Jones: Well, it's still a thing. Why not? Right? It's... I don't think I've ever broken a USB-A. I guess. Yeah.

Chris Gammell: I mean, it's just D plus, D minus and a five volt and a ground. Yeah, I know.

Dave Jones: I don't think I've ever broken a USB-A. Has anyone ever broken a USB-A?

Chris Gammell: No.

Dave Jones: I mean, they're just so big. You know?

Chris Gammell: Yeah.

Dave Jones: It's just...

Chris Gammell: Yeah. Yeah.

Dave Jones: I don't think I've ever had a dodgy one.

Chris Gammell: I don't think so. Yeah. I think you would have to really, really stomp on that thing to get it to like... You know, I've definitely broken the other side. Oh, yeah.

Dave Jones: The micros and Bs and... Well, no, no. Sorry.

Speaker ?: Oh, no.

Dave Jones: The other USB side. The side that plugs into the computer. Yes. Yeah. Yeah. Exactly.

Chris Gammell: I forget how the terminology goes. But... Yeah. Yeah. I've definitely broken that because, you know, sometimes you have the lever force. Yeah. Yeah. You catch your computer from falling off the desk with the USB cable and it just kind of cranks open. Yeah. Yeah. I don't know about on the USB-A side or the device side or whatever it is that plugs into the label printer or the microphone. Yeah. So... Yeah. Beefy. Beefy.

Dave Jones: That happened to Barnacles on Twitter. He's another YouTuber. And he posted a photo of this and he goes, oh, I don't know. I'm not going to be able to solder these four pins. And everyone's going, you can do it. You can do it. Go. Four pins. It's like, yep. Yep.

Chris Gammell: Yeah.

Dave Jones: Easy peasy.

Chris Gammell: Well, speaking of USB and D plus, D minus, I have been checking out the M.2 spec. It's been three weeks since we recorded. Last week, it didn't work because I was moving. And love that show you did with Shario. That was fantastic. Always love when Shario's on the show.

Dave Jones: Yeah. He's great. Also available as a video version, by the way, for those who want to see it. Oh, that's right. Yep. Yep.

Chris Gammell: I've been doing M.2 spec. Have you ever done M.2? Have you looked at that specification?

Dave Jones: I have not even looked at the spec. I'm not into that. Yeah. No. It's too interesting. It's too new. Well. Yeah.

Chris Gammell: Yeah. I mean, it has. So like, it has like a bunch of, it's basically like a mini version of PCIe. It has PCI multi-lane and whatever, but the confusing thing to me, and if you, so there is a wiki page I was tweeting about as I was kind of learning it and stuff. So I, people can check out that chain or whatever you want to do, but there's a wiki page. And at first it didn't make any sense to me, but the thing that kind of brought it all home is that basically the M.2, whoever came up with this, the, whatever bot governing bodies. And it's basically, they, they took this connector, which is like a, I think 67, 88 pin. I don't remember what it is. It's like a, it's a very fine pitch 50.

Dave Jones: It depends on the width because you can get different widths, can't you? Or are they?

Chris Gammell: Well, that's, that's the thing. No, it's actually, it's where it's keyed. And so that's the idea is it moves across, the key moves across. And basically it's like where they punched out certain pins. And so there's like a, a through, I should probably just look this up, but basically it goes like A, B, C, D, E, F, G all the way across.

Dave Jones: Yep.

Chris Gammell: And it's about how many keys they put or how many pins they punt, or this is the same number of pins, but it's where they're punching them out. Oh, okay. So the common ones are like A, B, E, and then M. I think those, and then there's also a combo B, M. Oh God. And B, M basically you see that for, I think that's the NVMe. So that's one part of it. And then the other part of it is there's a numeric size piece of it.

Dave Jones: But what's the advantage to that? Now, what's the advantage if they're the same number of pins and the same width? What is the difference with just notching out? I don't get it.

Chris Gammell: So the notching determines what is actually assigned to those pins. There's actual standardized pin assignments.

Dave Jones: Oh, okay. So it could, right. So the actual pinout could change or the functionality of the pinout changes based on the notch, even though it's physically the same connector and the same width. Yes. Okay.

Chris Gammell: I thought they were available in physically different widths. There may be bigger ones, but the ones I've seen have all been the same size connector. And then you have to, and I was like staring at these connectors and digi-key and just like looking at it, looking at it. I'm like, what is the difference? What is the difference? And then I went and did all this other research and I went back and like in like big block letters on the top of the connector, it actually has the letter on there. And I was like, oh, A, B, M, whatever. You know what I mean? So it's like, and so they're basically on the digi-key or sorry, the actual part number, it's like a dash. So that's where it gets confusing. It's like a dash one, dash two, dash three, dash eight. And you need to be careful buying that stuff. And then also they're not super well labeled on distributor sites about what keying it is, but the keying is very, very important.

Dave Jones: Yeah.

Chris Gammell: So that's one piece. And so if you buy an A-keyed thing and like A is, I was doing it for cellular, I was looking at cellular modems and so they're mostly B-keyed and it's got things like, I think like a couple lanes of PCI-E. It's got USB D plus D minus. It's got USB 3 TXRX. It's got a bunch, it's got SIM card connectors. And I should also say this is Timon Scrooge or Timon Scu on Twitter, who was helping me out with a bunch of this stuff. And he did this on the Pi Unora board. So if people haven't seen that, they should check it out. And he basically had hooked all this stuff up to a CM4. And so thanks to Timon for walking me through some of this stuff. But basically you can, so then you can put a SIM card connector on a board and then there's a standard pinout for the B-type connector. And so when you're trying to look at like SIM I.O., SIM power, you know, you apply usually the device that's talking to the SIM card will send it power. So it can also power it down, that sort of thing. So that's one thing. And so I was using the B-type and B-type is used, it can be used for many things, but it is almost always used for the cellular modems. Right. Other ones are used for things like NVMe, not NVMe, sorry, the solid state drives. So I think that's the combo BM. I should just have these. I should have had these in front of me. Sorry about that. But then the other piece that's confusing about this is there's also a length spec. And so basically then there's also like how long it is. Right. And that's where the, then there's like usually a.

Dave Jones: And there are standard mounting points. There's usually a mounting, a single mounting point. Anyone who's used a latest PC motherboard knows about, you know.

Chris Gammell: Yes. And that's where it gets confusing too, because it's like, so people will buy what they think is the right M.2 connector, but it's maybe keyed for an A or they put the mounting point at the wrong point, that sort of thing. So you need to kind of be careful about that sort of thing. You also need to be careful.

Dave Jones: The M key would be for memory, would it?

Chris Gammell: I'm not sure.

Dave Jones: Would that stand for memory? M key?

Chris Gammell: That would maybe make sense. It kind of seemed like they were just kind of going through and like assigning them over time. Right. Okay. You know? Yeah. And so let's see. So if you go to the wiki page, and I can link this to you as well. If you go to the wiki page there, are you already there?

Dave Jones: Yep.

Chris Gammell: Yeah. The A is assigned two PCI-Es, a USB 2.0, I2C, and DP, or DisplayPort. PCI, the B type is two PCI-E, SATA, USB 2 and 3, audio UIM, which I believe is, that's actually the SIM card. Instead of UIM, it's SIM. SS-IC, I'm not sure what that is. And then basically some like some low power IO. I think that's basically what that stuff is. And then M is, yeah, it's a four lane PCI-E and SATA. So that would make sense for if you're doing a hard drive, that sort of thing.

Dave Jones: Which is the most common use for these in PCs is a solid state drive. So that'd be used in the M.

Chris Gammell: Yeah. I think a lot of people, when you've opened up a, if you've opened up a laptop and replaced a hard drive, you've almost certainly seen this sort of thing or dealt with it. What's confusing about it is then you start seeing people using this other places, but they're actually just using the connector as well. So it's not necessarily conforming to the spec. So Blue's Wireless, which is a new cellular offering that's out there, is pretty cool. They use M.2. The new MicroMods from SparkFun also use M.2, but I don't believe they're actually using, they're just using the connectors because they're cheap.

Dave Jones: They're not any standard interface. Yes. They're just using the connector. Yep. Yeah. Which is common. There's a huge history of this going back to day dot, you know?

Chris Gammell: Yeah. Yep. To what? Sorry.

Dave Jones: To, well, basically going through the annals of time, everyone's used, everyone's used commercial computer connectors as their own thing, as their own module I.O. thing. I can remember when-

Chris Gammell: I thought you said a name though. No.

Dave Jones: I was asking about it.

Chris Gammell: No. Okay.

Dave Jones: I can remember back when, before the Arduino craze, you know, took off. You could get modular computers. You probably still can, if you look around, modular computers in a SIM. Oh, yeah. Hemery format, right? They would actually have the computer on a SIM module. And that, so they were using the, they were using the memory interface connector as the actual computer interface footprint. Interesting. So, you know, yeah. Yeah.

Chris Gammell: I remember the first time I saw this and I was very confused by it was with the Wi-Fi module. Oh, I can't think of it now. It wasn't Wicked Devices. They had like a devil logo though. And I remember it flashed with a IR LED. Anyways, we talked about it a long time ago on the show, but it was like an SD card form factor. And I was very confused about it because it like, if you plugged in an SD card, it did not work. Right. Yep. Yeah. Yeah.

Dave Jones: So anyway, yeah, this, you know, has like, and the PC 104 standard that was used for all sorts of industrial stuff as well. And, you know, so.

Chris Gammell: Yeah.

Dave Jones: Yep.

Chris Gammell: Yep. Yep.

Dave Jones: Cool bananas. Yeah.

Chris Gammell: So anyways, it's, it's interesting to dig into. I've, I've never, I had not played with it before this. It was definitely a little confusing.

Dave Jones: I've never played with it. Yeah. No, that, that, that would be, that'd be, I thought that it was just the one, I didn't know it was notched. I thought it came in different widths or something, but I thought it was, I didn't know it was this complicated. I didn't know there were this many variants. So yeah, that's. Wow. Yeah.

Chris Gammell: And there's a bunch that are like, so if you, again, if you look at the wiki pages, that CD, HJKL are all reserved. So like basically they just haven't been, they haven't been used before. So, oh, and actually one of them, it says G is reserved for custom use unused in the M.2 specification. So maybe that's a, maybe that's what the, the micro mods of the world are using that sort of thing.

Dave Jones: Right. That'd be interesting. Yeah. Yeah. Surely you would pick one. You wouldn't just like use the memory interface one or something, although that might be the easiest to get connector, right? That might be the cheapest to get.

Chris Gammell: That's kind of the conundrum is like, would any manufacturer make low cost connectors if you weren't going to be doing super high volume? So it's like, yeah, okay. That's right. That doesn't really make as much sense.

Dave Jones: When you're searching for these, did you notice any availability difference between the different types?

Chris Gammell: Cost differences for sure. Yeah.

Dave Jones: Oh, okay. Right.

Chris Gammell: Yeah. A, B and M were the easiest to get.

Dave Jones: Okay.

Chris Gammell: But again, it was hard to tell, at least on DigiCute, it was hard to tell because there's not like a column for it or maybe I missed it, but it was kind of dashed out. And then you can, I think they maybe were just showing where they were blanking out the connectors. Anyways, be very careful when you're ordering these things and double check all your pins. It's easier to tell. Basically the way to tell is kind of the number of teeth on the connector on the, uh, on the card side connector. That's, that's the easiest thing to see. If it's only got three pins on the top, it's probably an A connector.

Dave Jones: Right.

Chris Gammell: And if it's got, I think five or five, it's a B. So if it's got five on each side, that's a B slash M. Right. Cause that's got both notched. If it's not for both.

Dave Jones: Okay. Yeah.

Chris Gammell: So, yeah, I mean, it's just like a thing that you'll start to see over time, but I was only looking at the cellular ones and then kind of just digging this out.

Dave Jones: And these are about one buck Aussie on TG key. So they're fairly cheap.

Chris Gammell: Yeah. Yeah. I think they're good. What's the pin pitch on those? That's 50 mil. Sorry. 50 mil. 50 thou. Yeah.

Chris Gammell: Yeah.

Dave Jones: You know, 50 thou. Okay. So it's Imperial. None of this metric rubbish.

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

Dave Jones: Nice.

Chris Gammell: Old school. That's like the standard. I think that's the same. That's on a lot of fine pitch connectors. I think so.

Dave Jones: There's a lot of metric connectors out there.

Chris Gammell: Hmm. Yeah. Maybe I'm actually getting this wrong then. I think you still cut my design and tell you. Let's see. Yeah. It might be. Yeah. Maybe it's a 0.05 millimeters. I don't know. Let's see the connector. I am using here.

Dave Jones: 0.5 millimeter pitch. There you go.

Chris Gammell: 0.5 millimeters. You see? 0.5 millimeters. Yep. Never trust an American with the. No. No. With the metric.

Dave Jones: Because then you crash probes on Mars and stuff.

Chris Gammell: Actually, I was just. I just finished reading Hail Mary. The second Andy Weir book. It's about. It's about a space mission. And another space mission. He wrote The Martian. If you don't remember. Yeah.

Dave Jones: I'm not aware of another book he wrote.

Chris Gammell: Oh, yeah. It's pretty good. Actually. It's. It's. It's very. It's. It's much more out there. It's much more like. Cool.

Dave Jones: I got to check it out.

Chris Gammell: It's much more out there. But I wasn't a huge fan of the ending. But. Right. The interplay stuff was pretty good. Got it. Every time he brings up metric. He's like. You know. He just keeps making. Yeah. Jokes about it. He's like. Sorry. It was like. When the. When the astronaut. Who's American. Gets stressed. He starts thinking in Imperial. Yeah. It's like. Like. Oh, yeah. I feel that. Yeah.

Dave Jones: It reminds me of. Firefox. Have you seen. The Clint Eastwood movie. Firefox. I've never seen it. No. From the 1980s. Where he's a. He's a pilot. And he has to go into Russia. To steal Russia's new. Super secret plane. And he's going. Right. I've got to. Because it's. It's actually controlled by your brain. Like. There's no joystick. Or whatever. Like. So. And he goes. But. But it only understands Russian. So I've got to think. In Russian. So I've got to like. You know. So I can't translate in my head. I've got to. Actually think Russian. You know. Yeah. So. Anyway. Yep. Great. Classic. Yeah. Cool. There you go. All right. M.2.

Chris Gammell: People should check out M.2. It's. It's pretty interesting.

Dave Jones: Is that the connector of the week?

Chris Gammell: I guess so.

Dave Jones: I guess it is.

Chris Gammell: So this is what the article. I shared on the subreddit. About routing PCIe. I was actually. I was surprised. It's 85 ohms characteristic impedance. Instead of 90 or 100. And you know. It's just always interesting.

Dave Jones: Right.

Chris Gammell: But it had a nice trace on that. That article I shared too. It was just one of the time based traces. I don't have it open here. It's 85. PCI express is 85 ohms really needed. It's a Samtech article. And they show. A TDR. And so like showing. Showing a TDR is pretty cool. Yeah. Kind of like showing where the discontinuities are. And first time I saw this. I was like. What am I even looking at here? But it's always cool to see like. How you can just. See. Like in terms of distance. Yeah. Where a discontinuity is. Just because it's like. Oh it's. However many. Nanometers further away. It's like. That's based on the time. It's basically like. It's almost like a sonar ping. You know. Like to see how that. That operates.

Dave Jones: Of course.

Chris Gammell: Yeah.

Dave Jones: It's very cool. Great. We'll link that one in.

Chris Gammell: Yeah. Definitely. Yeah.

Dave Jones: Controlled impedance routing. It's important. Folks. Trust us. When you get to like the bleeding edge of stuff. You know. It really matters.

Chris Gammell: Yeah. I'm not there. Personally.

Dave Jones: It's. It's mattered. Like I've done like. Boards.

Dave Jones: Boards.

Dave Jones: Boards. Boards.

Dave Jones: Boards.

Dave Jones: Boards.

Dave Jones: administered administered administered administered !

Dave Jones: meet the impedance target or whatever, yeah, you would see that. You would actually see it come out. It'd work in quote marks, right? But you would get a larger rate of errors and stuff like that. So, yeah.

Chris Gammell: Right. And that's going to be like a real ghosty kind of error. It's going to be like, oh, you know, we don't make our packets every time or whatever.

Dave Jones: Right. So it halves your bandwidth or something like that. Yeah, exactly. And you're wondering why, you know, it's because, yeah, it's got to resend or whatever. So, yep.

Chris Gammell: Crazy. Crazy.

Dave Jones: Nuts. Ah, boy. You want to know what else is crazy? What else is crazy, Dave? A video I did yesterday. Completely innocent. But apparently there are people out there who didn't like the fact that I showed a product, namely a multimeter of mine that could shock horror. This is going to stun you that it can be actually the microcontroller and it can be factory reprogrammed. It's a shock and it can be done by dealers too. It's a real shock. I'm not sure the, what's the controversy here? You're not sure of the context. Okay. Yeah. Yeah. Because, right, the, my new BM786 meter, right, from Bryman, it's got a programming port on it. Right. If you open up the battery compartment, there's a little pin header in there. There's a tiny little, you know, tiny pitch five-way pin header in there. And obviously it's a programming port, right? It's a programming port for the internal micro. Okay. It's like obvious. Everyone knows this, right? Right. It's not rocket science, but Bryman do not let you, but they do not give details of it. Right. It's like, no, sorry, this is only for factory programming.

Chris Gammell: Right.

Dave Jones: It's not for user programming. Sorry. We will not provide any details on it.

Chris Gammell: Oh, that's, come on guys. That's just a, that's just a challenge. That's just, yeah, exactly.

Dave Jones: No, but anyway, there were, there were a lot of people that were quite upset that I actually showed that it was possible to actually program because I, I have the programmer for this because I'm a dealer, right? I'm a dealer of this meter. I need to have this programmer just in case. Dave's the man.

Chris Gammell: That's what he's saying here, folks.

Dave Jones: People didn't like it.

Chris Gammell: You gotta, you gotta just work around these constraints. I mean, yeah.

Dave Jones: People seem shocked that a program, like they, they don't, but, but any multimeter that has firmware updating should be make publicly available information. It's like going, no, I guarantee that the meter that you bought, the meter that you bought last year has, and if you buy the same meter again, now it probably has different firmware in it. Yeah. I would agree with that. It's just programmed at the factory. Right. And you don't know about it. And it's got some internal, you know, pad header or something like that. You know, there's some like bed of nails. Yeah. Like bed of nails. Programming, you know, things like that. Yeah. Right. It's, it's absolutely no different, but because, because the connector.

Chris Gammell: So is the problem here just that you showed it, like it's like, uh, now that it's out there.

Dave Jones: The problem is, is that I dared to show it. Yes. And I dared to show it. And then I dared to tell people, sorry, don't ask for the information. I can't give it to you. Right.

Chris Gammell: So just put a redacted sticker over it. Yeah. I mean, yeah, I think, well, I think one of the things is people expect you, I mean, you've talked about open source. You've talked about that.

Chris Gammell: Of course.

Dave Jones: That was another thing. Something people want. Yep.

Chris Gammell: I think the real question is to people actually, would they use it? Right. I mean, I feel like that's a, that's the real thing.

Dave Jones: Oh, they probably wouldn't. No, probably not. Yeah. No. I mean, majority won't. No. They, and a lot of people said, oh, but you don't, you don't support right to repair. It's like, really? Really? I don't support right to repair. And it's like, but this is a right to repair issue. It's not really. No. A, a schematic is vastly, vastly more valuable than being able to program the chip on the board. Right. In terms of a repair scenario, in terms of right to repair, a schematic is vastly more valuable. Right. Yeah.

Chris Gammell: I think that is, that is an interesting kind of like, where is the line delineated? I don't know, like if Lewis or the crowd has really talked about that before, like I would think that what, what would happen maybe as a hypothetical. So say someone blew up the micro on board.

Dave Jones: Yep.

Chris Gammell: Probably shouldn't happen. Right. Cause of the section of DMM, whatever.

Dave Jones: But in theory, the micro could fail. Right. Yeah.

Chris Gammell: Yeah.

Dave Jones: Yeah.

Chris Gammell: So what happens then?

Dave Jones: Well, A, you've, A, you've got to know what micro it is to begin with and B, which, which you can't in this case, cause it's a Breiman branded.

Chris Gammell: Right. Okay. So that would actually, I think then that actually is a fair argument from them to say that it is not repairable. If you can't source the chips, it's not repairable.

Dave Jones: That is correct. Right. But this is nothing new. Breiman have done this for decades. Right. This is, they've done it from day dot.

Chris Gammell: But I'm just saying, yeah, no, I'm not saying it's not like right or wrong. I'm not giving any judgment. I'm just saying that like that argument is somewhat valid. I would think you could maybe, you could offer. So like, again, to like, usually it's going down to like Apple, right? It's like, Oh, I want to be able to repair a phone or whatever. It needs to be out there, whatever. But I think that if a chip is on offer with the latest firmware pre-programmed under the chip itself, then maybe that is still in the repairable realm. Maybe it's like.

Dave Jones: Oh, of course. I don't think anyone would. You don't need access to the bootloader.

Chris Gammell: You don't need access to the firmware or whatever. But like, if you have.

Dave Jones: As long as it was fairly cheap and it was available. Yeah. Yeah. That'd be. Yeah. Of course.

Chris Gammell: I would say that's repairable. That's within the realm of repairable. Don't you? I mean, like that's. Oh, yeah. So.

Dave Jones: Absolutely.

Chris Gammell: That would obviate the argument.

Dave Jones: Absolutely. But people made the argument that, oh, I shouldn't be supporting a company like this. Well, nothing's changed. Nobody like nobody complained at all when about the original Bryman that I've been selling for half a decade, which is exactly the same thing. Nobody complained. Yeah. I don't.

Chris Gammell: I don't know. I don't know if that's always a good argument. Like the way it's always been done is not always a good argument, but.

Dave Jones: No, it's not. But in theory, this is a step forward because it's reprogrammable. Right. Whereas previous ones have been one time programmable. So at least it's a little step forward. Yes, I agree. In an ideal world. Yes. That they should let us have the schematic. Right. And they should let us know what chip it is, of course, in an ideal world.

Chris Gammell: I mean, I would think that. OK. Again, take this to a conclusion. It would be like they shouldn't really support it then. Right. If they're not. If they're looking to buy. You know, I think there's always going to be a cost trade off.

Dave Jones: No, that's right. I totally support their right not to buy a Bryman in that case. Go for it. Go and buy a Fluke. Oops. They're exactly the same. Go and buy some other brand. Oops. They're exactly the same. Once again, it comes down to that argument just because it's always been done like that. Is no excuse or reason to do it. But I'm just saying the practical reality of it is that basically the most open and ironically, the most open meter on the market is my one is is the 121 GW. There's a schematic. Right. There's all the programming and calibration information, which almost no manufacturer gives you. There's the protocol interface, which is open. And all the application is all completely open source. All the wireless application software is all open source. So it's like, well, OK, buy that one instead.

Chris Gammell: Maybe you should start start promoting it as the most you could say 121 GW, the colon, the most open DMM question mark.

Dave Jones: The most open. But there's a lot of people. This is once again, it's funny. People complained that, oh, I claimed it would be completely open source. Where are the PCB files? Where are the, you know, all this?

Chris Gammell: I mean, there's always going to be this always going to be this argument. Right. Yeah. I think that, you know, and we, I think we like when there's open stuff, but it's.

Dave Jones: Yep. But that doesn't mean I'm not going to support a company where, which nobody, everyone loves Brian Mometers. They think they're fantastic value for money and feature set and quality and everything else. Nobody complained until I simply showed that, oh, goodness, a dealer can update the firmware. Like, wow. Like, you know, all hell broke. Well, not all hell broke loose. There's half a dozen. Yeah. I think there's people that were upset. Right. Yeah. Yeah. There were people that are upset that I showed this, but, but this is a thing, right? It's they, they wouldn't have known. This actually exists. It's like, I'm okay. You're I'm, I'm, I'm educating you. You're, you're now aware that this sort of thing happens. Dealers often have access to updating products that, that the user doesn't. And that's for your benefit. Yes. It's not an ideal world. I would, I've asked for the schematic for the Brian Meters. It's not like I haven't, it's not like I haven't asked them to do it.

Chris Gammell: In a best case scenario, someone would come out with a fully open meter. Right. And it's got everything, everything. And if that's the best solution, then that would, you know, in a market system, that would be the, the winner. Right. But I, I think, I think the truth of the matter for a lot of this stuff is people aren't willing to pay that cost or, you know, support a small run that might be more likely to be completely open. They just want to say.

Dave Jones: Or it doesn't support their, what features they're after or whatever. Right. Insert problem here. Right. There is a reason that there's literally a thousand different types of multimeters because there's one for every need. All right. You can't have one perfect meter or one perfect scope.

Chris Gammell: I'm starting to think that people might not be, might not be satisfied all the time, Dave. It's.

Dave Jones: Oh, I, it's a shock. Yeah. Yeah. Yeah. So anyway, I thought that was interesting that I simply showed something that's been around forever and is currently in the current meter that they own. I guarantee it because I showed it is the problem.

Chris Gammell: Yes. So now here's, here's the real test is show like a, I think what you need to do maybe on April fools or otherwise is show like another programming port that doesn't, that goes to nowhere and then see if we can get mad about it or some other feature where like, do people just get mad when you show a feature, right? You need like a control test in this, in this scenario.

Dave Jones: Right.

Chris Gammell: So it's like, do you show a non-replaceable fuse or something like that? And it just goes nowhere, right? You've just glued it onto the board. Right, right. Here's that, here's the non-replaceable fuse. And it's just like a tic-tac that's glued to the board. Right. Okay. That would be a good test. I think. Yep. Yeah. You know, people, people.

Dave Jones: Yeah. Yeah. There's people. It's, you know, people just got angry because I showed something that was there.

Chris Gammell: Yeah. Like humans are the problem, Dave. Yeah. I know.

Dave Jones: Yeah. Humans are the problem. Yeah. I agree. But no, I, I, I totally understand where they're coming from that. Why show it? Well, I'd rather live in ignorance that, that this thing's not possible. You know, like, like people called me a tease, right? Oh, I'm just tea. I'm going, ha ha. I'm making fun of people that I can do this and you can't. No, no, no, no, no. Right. It's like, well, that wasn't the intention of the video. Some, some people asked to see it because everyone, you know, everyone on the forum is talking about this meter knew that this thing, and they knew I had the programmer and they wanted to see their work.

Chris Gammell: You know, Dave, you just got to be your own person, you know, just. Yeah.

Dave Jones: No kidding.

Chris Gammell: Yeah.

Dave Jones: What do you think I've been doing for 12 years? Anyway, I, I think it's funny when it, you know.

Chris Gammell: Don't let them grind you down, man. Don't let them grind you down.

Dave Jones: Dude. Once again, I've been doing this for 12 years. You think I would have been grounded to dust by now if I didn't. Yeah. You'd think so. I just think it's an interesting discussion, you know. Yeah. No, I think it is. You're right. I think you're right. I think you're right. Why are people upset about something that's already, you know, that's just there? It's not, it's not like this is some new thing designed to, you know, it's, it's just done in virtually every meter on the market has this. It's like, what, what is different? It's because I showed it and they're upset. So it's the internet. Anyway, I thought it was.

Chris Gammell: Speaking of test equipment and the internet, a past guest, Nash Riley, wrote about whether or not to make a V2 of the B-Funk, which was a function generator that he put out there. We thought this was an interesting article. Basically, he went through some of like the, the economics of it and the, the logistics of doing low run and like whether or not to, you know, we would fit in a marketplace. You have a video that was about this in the past though, as well.

Dave Jones: I do. This video, very popular, I believe. The economics of selling hardware, EEV blog 887. And I go through the, well, exactly as the title says the, the economics of selling hardware, the pros, the cons, how much gross margin to make the cost of products, the costs of goods and you know, all this aware, should you sell them direct through resellers and all that sort of thing. And that was very well received video. Yeah. This is almost like a, a, a follow-up. Like it's, this is almost a, like, should you do it in the first place kind of thing, rather than the economics of making it work. Like, because the too long didn't read of this article is that basically he made a loss on it. He made a loss of 600.

Chris Gammell: He, yeah, he lost about 700 bucks.

Dave Jones: It cost him 1300 bucks to make them and he got 600 bucks back. So it cost him 700 bucks.

Chris Gammell: Well, I think, I don't know if that's the TL, that's the TRD, TL, TLDR of the V1, I think. But I think the real question is then like, okay, looking forward, right.

Dave Jones: Could, could he do V2?

Chris Gammell: Yeah. And I think that actually kind of rolls into the, what you were talking about earlier too, of like test equipment's kind of its own realm, you know, and it's like, there's a lot of expectations. There's a lot of players that are well-funded. Yep. The low end is tough because you're usually playing with a constrained feature set. And so these are all things that Nash goes through in his article. And, you know, especially with function generators where it's like, okay, so what is the minimum people are going to need? What are they willing to pay for? You know, if you're then working against like some of the Chinese manufacturers that have already dialed this in with like ASICs and, and similar things, it's like, okay, so like, where, where do you fit? You know, like, do you go up market? Do you go down market? He brings up a good point about Celié as well. Like, it's like a, you know, there's a ton of logic analyzers, but Celié differentiates because of UI. I think that's a, that's a great point. Celié is still going strong because they went more up market. If people haven't noticed, if you go to buy a Logic Pro, it's a lot more than it used to be. And it should be. It's like, it's a high end that what you're paying for with the Celié is not like just the chips inside, you know, the FPGA that's in there, whatever you're paying for, like, You're paying for the software development. Really cool software. Yeah. You're paying for the software development of it.

Dave Jones: But unfortunately that model allows you to buy a clone of it for 10 bucks on eBay or AliExpress or anything. No, not anymore. Oh, really?

Chris Gammell: Oh, they got around that, did they? Yeah. That used to be with the Cypress C2, whatever the chip they used to use that was in there. Right. Yeah. So it's all like FPGA based now. And it's, I mean, it's better too. Like, so there was that Cypress chip that kind of every low end logic analyzer, that was the Logic. The Logic 8 maybe. Yeah, something like that. But the Logic Pro is what they've worked on. And that's when Joe and Mark were on the show five plus years ago as well. They also were talking about kind of like what's next. And they're well into that.

Dave Jones: Right. Okay. So they've done that. So the Pro is, so there's no clones of the Pro out there?

Chris Gammell: Oh, I don't know. Maybe. I haven't seen any, but I think the other thing too, is that like at this point, I think it's beneficial to just buy into that ecosystem anyway. And so I don't think that the software would work with any clones. Okay. Because there's more validation than it used to be. Like SigRock used to work with the old Logic 8s, the old lower end Cypress based ones as well. You know, the open source logic analyzer stuff. But the new stuff is, you know, it's got, it's got feature, it's got stuff built on top of it as well. And I think what Nash was bringing up is like, yeah, that would, that's kind of what you also need to do. It's like moving up market. And so is that even possible with a logic, with a function generator? I don't know. Like what, what do we really need out of function generators? Right. It's like, you know, if it's just sine waves, square waves, triangle waves, like, oh yeah, well that's, so it's basically then like, how clean is your signal? How, how fast does it go? You know, how much can it drive current wise? Like there's not a huge software component to that normally.

Dave Jones: I've been there. I've done that. I sold a function generator kit way back in there. Oh yeah. Oh yeah. This was 20 plus years ago. Yeah. Yeah. It was published in a Silicon chip magazine. Okay. In fact, I probably use the same analog devices, AD9837 DDS chip. So it's all near identical to this. I think. Anyway, mine was a sine triangle, sine square triangle up to 20 megahertz. Okay. Yeah. Which was the limit of the DDS chip at the time. This one, this one might be a slight variant, but it's all the same family of DDS chips. You basically have a microcontroller, which sends in the spy commands into the analog devices, DDS chip. And then the DDS chip does it all its magic and outputs any frequency from basically DC to the upper frequency limit of the DDS.

Chris Gammell: Yeah. Sure. Sure. And I think that, you know, that's, so that's a great point too, right? It's like, if you're using the chip to its specs, then the only thing that's stopping a competitor or, you know, a potential competitor, maybe someone hasn't even done it. It's like, well, can they also re-implement what I've done? Like, what have I, what is the value add that's on top of that?

Dave Jones: It's the user interface. Yeah. Yeah, exactly. Mine was a push, mine was a self-contained jiffy box with push buttons and a seven segment display to actually read the frequency and stuff.

Chris Gammell: Probably like Nixie tubes back in the 50s. Yeah. Right. Yeah.

Dave Jones: That's it.

Chris Gammell: Yeah. Don't knock the Nixies. Yeah. Right. Sorry. I was imagining like the Doc Brown, like making the, you know, the tube based version of the Deluxe. Yeah. Anyway, so yeah. Oh boy. Yeah. I mean, yeah. So I think that this is a, I mean, you know, Nash wrote a great write up of this and it's yeah, I think in any kind of thing, and maybe this is a good follow on video to your sort of thing too. It's like, what, what is that value add on top of just the heart? Like the Harbor keeps getting easier for people, right? I mean, chips are out there. Well, maybe the hard part is to source chips these days, but we'll move past that. You know, putting stuff onto boards is not what it used to be. Sourcing stuff, making maybe even, you know, small hardware businesses out of this. It's not as hard as it used to be. So then it does come down to the firmware, the heart, the software, the network stuff that might be in there, the web UI UX type stuff. And it's like, oh, okay. That's like moving up market. And it might be a lot more stuff than hardware people want to do. And it's like, all right, well then that means you might have to move into different space than you thought you did. And I don't know what the ultimate outcome of that is from an economic perspective, but I think from a, you know, the products in the marketplace, it's going to be, there's going to be more demand for higher level features if you're not trying to go low end.

Dave Jones: Well, this one doesn't have any user interface. It's just a board. It's just a board, isn't it? It's just a USB connector and a board. And presumably it hooks up to a piece. Maybe there's there a, is it like a command line, like a serial interface, serial command interface, like a, you know, a UART kind of, I don't know how you actually control it. Yeah. I think it's scriptable. I think so. It's scriptable. Is it? Yeah. Yeah. I'm not sure of the exact way it works, but yep. That's, that's the problem. You have to differentiate with a, you know, like you've got a value add somewhere else because a board can be ripped off, right? This, this board could be ripped off and sold on eBay for 10 bucks, right? From someone on, you know, someone in China or somewhere else. Anyone could take this and just, yeah. Well, I presume is it open source?

Chris Gammell: I don't know.

Dave Jones: If it's open source, then anyone can take it and expand on it.

Chris Gammell: And once again, it is, sorry, it is open source.

Dave Jones: And that is another risk, which I talk about in my video. I think from memory is that, yeah, that that's one of the risks is that once you open it up, you've got to stay one step ahead of your competitors, so to speak. If you want to make this, you know, people need to put food on the table, right? If you want to get into this as a business, you've got to make money, right? So, and if it's open source, you can't complain that somebody else has taken your thing and they're selling it and they're doing a better job, right? You can't complain. That's just the, you have to be one step ahead.

Chris Gammell: Yeah. Yeah. I think Nash also brings up the unit costs of like, if, you know, if he did some math on like the target costs, I think he said 12 to 15 bucks. So then like, that would be like the bomb cost rather and what that would end up being, you know? So the thing I think about is at the end of the day, like not the end of the day, in my old days, rather like, you know, seeing my stuff at Keithley and like, what were the value adds at Keithley? You know, there was definitely circuit design, interesting novel circuits there. But then really a lot of the magic came in at the sourcing level, right? Sourcing. And then also some manufacturing level stuff too, of like further refinement of chips that are off the shelf. You've done a lot of teardowns. You've seen some of the parts that are in there, but then there was further refinement, maybe further sourcing of higher specced parts than you could buy on the open market.

Dave Jones: Or they would bin them. You would bin the parts and you're testing out. If you're making a really high end ultra end meter, for example, you might be, or a voltage reference, for example. Right. Exactly. It's just like, who, who, who makes that? What's the de facto standard voltage, hobbyist voltage reference out there? I can't remember.

Chris Gammell: The LTZ 1000? Yeah, yeah.

Dave Jones: No, but that's, that's the chip. But who use, who, who actually makes it? Somebody made them in their garage and they sold, you know, thousands of these things, but they would individually hand calibrate each one and burn them in under. And that's what you were paying for, right? You could buy the exact same board on eBay. Yeah. You could buy the exact same board on eBay for 10, 15 bucks or something. And his one might, might cost a hundred or two. It's because it was burned in and calibrated against his traceably calibrated eight and a half digit meter. Right. And it was all, it came guaranteed. It came with a little, you know, each one was hand done. So it's, you know, it's 5.0001 volts, you know, and it's like that kind of thing. That's what you're paying for. You weren't paying for the actual.

Chris Gammell: Yeah.

Dave Jones: Yeah. It's just the parts.

Chris Gammell: Yeah. Metrology is another value add.

Dave Jones: Metrology is a value add. Yes. Having traceability. Yeah.

Chris Gammell: And like having a, you know, guarantee. Having it pass all the way back through to a NIST standard is like, yeah, there's a lot of, there's a lot of costs there because they're also paying to have their 84, whatever the hell, the 83, 57 or whatever.

Dave Jones: I do that with my microcurrent, right? I've been doing that for more than a decade. It's because it's a precision. It uses precision parts, right? Heck, one single resistor on there is four bucks, right? Just for one resistor, right? And there's a reason. And there's a reason it has the specs it does because they're built into the parts. And there are people, even because it's open source, you can go make it and sell it and everything. Just don't use my name. Yep. And they do, and they put lower precision parts on it and it's not the same spec. And then I've, and then I've got custom built test gear to actually test each one, you know, and guarantee that's within specification and things like that. So yeah.

Chris Gammell: Right. Yeah. So then I think like the, so then the, so the value add there is one thing is the calibration, the testing, all that other stuff. I think the other thing is that sourcing and really unit economics then of like, can you get enough that you're not paying for, you're paying four bucks a resistor instead of maybe the single piece price for one would be like 12 bucks a resistor. And then it's like, okay, that eats eight bucks your margin. Are you willing to take that hit if you're doing low volume? So it's like, yeah. And Nash talks about this as well. Oh, okay. And it's just like, so you need to, I mean, you just need to make these balance points and, and make the decision of whether this is a, I think one thing it also points to is that like, it's a bit of a grind, right? I mean, if you wanted to do all this stuff. Oh yeah. Oh, totally. You need to sell a thousand DMMs versus one. And it's like, okay, that's tough, right? That becomes a bigger business that you have to kind of get into. And it's like, yeah, that's. Yeah. So good discussion from Nash. You can hear Nash on the show when he was at Sonos and he's no longer there, but yeah, it's, I love following Nash's blog too.

Dave Jones: As someone who's been in the kit business for 30 years before it was even called open source software, I can assure you that. Yeah. It's a, yes, it's a grind, but you either. Enjoy it or you don't, you know, it's like, you know, yep. I actually enjoyed it. I actually liked actually selling kits. I liked actually kitting them up back when I would, you know, individually kit all the parts. I'd have all the bins there and each part and they'd each go into a bag and, you know, I'd actually sell them as a kit, not just build up and actually sell kits as well. And then you kit the things up and it's, you know, it's kind of a fun thing, but other people think it's a grind and sometimes it's a grind, but you know, yeah, it's always, well, it's always a grind. There's a lot of work in getting a product actually produced and the logistics of getting it produced and the logistics of handling and shipping and, you know, sales and support and everything else. It's like, yep.

Chris Gammell: Right.

Dave Jones: Yep. Totally. And how one person can ruin your day. You know, if you get some dumb ass who comes back and say, I built my kit and it doesn't work. What's wrong with it?

Chris Gammell: I mean, it might not just be a person. It might also be Vichay removing markings and messing you up. So what's the story here, Dave?

Dave Jones: What's the story here?

Chris Gammell: Well, Vichay is removing high marks.

Dave Jones: We can get to that, but that reminds me is that I've done a video on this in that it involved my microcurrent and I'm not sure if it was Vichay or not. It was one of the resistor manufacturers and the resistors actually came in the wrong spec. I ordered them.

Chris Gammell: Yeah. I remember. I think you've talked about that on the show at one point.

Dave Jones: I was ordering like 0.05% parts and they ended up being like 0.1% and that was throwing me out. That was, you know, I was getting like 30% reject rate.

Chris Gammell: Right. Because when you do that, that value add metrology stuff, you're like, my yield just went to 50% or less. It's like, holy shit.

Dave Jones: And it took me ages to figure it out. And then I actually went to the original reel I got and measured it and I go, holy shit. Like, you know, X percent of these are out. Like they've, they've actually reeled them wrong. And they, you know, I contacted them and they apologized, I think, and said, yep, oops. And they sent me a new reel or something, but you know, it's like, holy crap. You know, how much time and money did that cost me? Right.

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

Dave Jones: Anyway. Yes. Who was, I can't, where's the link for the, oh yes, there it is. Who's this on Twitter? Who pointed out. No. So yeah, somebody sent it to me. Sorry. I forget your name. They actually emailed me and said, Hey, they got this.

Chris Gammell: A PCN product change notification.

Dave Jones: A product change notification. Yes. From, if, if, if you don't know, if you're buying these parts from a good supplier like Mouser or DigiKey, they actually will, if you bought these parts before, I think they will forward you. I've had it several times. I'll forward you product notifications or this product's been last buy, you know, they'll send you a last buy notice, you know, this product's been, been discontinued.

Chris Gammell: It's basically the distributor kind of acting as your, uh, as your obsolescence. Yeah, exactly. So anyway, which is a lovely service. We should say this is a very nice service.

Dave Jones: No, no, it's great. It's also horror when you get that email pops up in your inbox and it says from DigiKey or Mouser and it's like, there's a product change. It's like, Oh, holy crap. You know? Like, yeah, yeah, yeah. That can ruin your day.

Chris Gammell: Yeah. Please, please be a DiceRink. Please be a DiceRink. Please be a DiceRink. Yeah.

Dave Jones: That could be the name of this episode. Please be a DiceRink. Within all our caps, you know, please, please in all caps, please be a DiceRink. Yeah. Okay. Okay. All right. And anyway, the product change notification is the discontinuation, not of the resistor, but of the resistor value markings for their TNPW E3 series for those playing long at home, which is their highest, which is their highest to build. I'm sure there's lots of E3 series. Oh, yeah, yeah, yeah. Yeah. Yeah. Come on. High stability.

Chris Gammell: People sitting there in their Vachet gear. They're just like, oh, yeah. Actually, I'm more of a Draloric, you know, originalist. Oh, boy.

Dave Jones: Look, come on. There's people out there. Yeah, yeah, yeah.

Chris Gammell: You're right. There probably is someone. I know. I know.

Dave Jones: Anyway, their high stability thin film, 0603 to 2512 film resistors, right? So it's just a boring resistor. So what they're doing is they're removing the markings from, you know, they would have had the, you know, the value marked on them.

Chris Gammell: I bet this is because other things are going, you know, I can imagine this is pure conjecture, of course, you know, things keep getting smaller and smaller. You can't mark stuff on 0402 and smaller. So they're not doing it there. No, that's right. And then if your volume goes down enough and then your machine breaks, you're like, oh, I guess we just won't mark the 0603 to 1212, 2512s anymore. It's a process step. We'll just send out a PCN. It's just Dave.

Dave Jones: Well, and they're saving X cost. Come on. It's a product saving. You know, the reason is product saving, right? Because it saves a whole step of having to move those components to another line, potentially. I don't know how they do it, whether or not it does when it comes off the line or whether or not they reel them and then shoot them through. And I don't know if anyone works at one of these resistor factories, please let us know. But anyway, yeah, they're removing the markings. And what's the big deal? It's well, you know, there could be potential impacts. One of the big ones might be optical inspection. For example, if you've at your factory, if you've got optical on these automated optical inspection machines, you actually program for those who don't know, this is a service you can get at your board, at your board assembler, right? Usually, yeah. Any good board assembler will have these optical inspection machines. And what you can do is get a known golden reference, you'll call it, right? A golden reference board that you'll put into this machine. And it's been, you know, you've checked every component that it's around the right orientation and everything. And then they take a high res photo of it, although optically, like the camera can move around. It's on like an XY platform, you know, can zoom around. Or you can do it as a photograph of the whole board. If you have the resolution or whatnot. Anyway, regardless of how they do it, they optically take a reference of that board. And then every board that comes off the production line, when it comes out of the conveyor, the oven conveyor belt, then it'll continue on the conveyor belt and it'll go immediately into an optical inspection machine. This then takes a photo or puts the scans, the camera over the board and checks for faults on individual components, whether or not they're the wrong value, because it can actually see the number on top of the chip, right? You can see the part number or you've put the wrong, you know, somebody, some idiots loaded the wrong value reel on the line, right? They put it into the machine. And that's a common fault. Like the, the machine, the pick and place machine assumes that you've put the correct part in the correct slot. It doesn't know, right? It doesn't know that it's doing that. So you only pick that up later after you've assembled it at the optical inspection station. And it, that can be one of the things that it picks up. And if you suddenly change your resistor, then whoop, you have to reprogram your optical machine. Or if they did this and you, you know, then your parts are all flagged and the production line stops and, you know, you could be days or weeks, you know, out of, you know, just because the manufacturer decided to not put the value on the resistor anymore. That's just one of the issues, right? So yeah. Anyway.

Chris Gammell: Yeah, totally. Yeah.

Dave Jones: It's yeah. Bastards.

Chris Gammell: Yeah. Yeah. I mean, it'll, it'll, it'll, it'll ruin your day. You know, PCNs in general.

Dave Jones: Anyway. Yes. Why they've done it. I don't know. Well, it's cost. Right. Come on.

Chris Gammell: Yeah. It's probably, probably right. Yeah.

Dave Jones: But it's, it's, it's only on that series, which is interesting. It's not like they've got, well, maybe there's a product change notification for all their 06, 03 and bigger resistors as well. I don't know.

Chris Gammell: Yeah. I mean, yeah, unless you're, you're not going to get the markings unless you are not going to get the notices unless you're buying that individual part, but that's why you buy one of each, you know, one of each.

Dave Jones: And then just from a hand assembly point of view as well, it's a pain in the ass. If you drop the thing or they're sitting on your desk and you get them mixed up, you don't know which one's what. And ah, that'll ruin your day too. Yeah. Yeah. Yeah. Anyway, sneaky bastards.

Chris Gammell: Well, speaking of a die shrink, well, this isn't a die shrink. This is a die switch, but ESP or Espressif, I have a new chip coming out. I'm pretty excited about it. It's Zigbee ThreadX and, or sorry, ThreadX thread and 802.15.4. So basically a new, instead of wifi, it'll basically be a different way of getting your signal out there. So it'll basically have different fies on there instead of just wifi. It'll have the way to do 802.15.4 and stuff like that. It'll have other support on there. So you can basically make a different type of ESP widget. That's one thing that is exciting. Another thing is that I have been seeing out in the wild, the C3, the ESP32-C3. And you know what that is?

Dave Jones: I have no idea, Chris, because I'm not in this world.

Chris Gammell: It is the RISC-V version of Espressif. And so instead of using the, I forget the name of their old core.

Dave Jones: Yeah, but that changes everything, right? It's not compatible anymore, right?

Chris Gammell: It will be a different compiler. But no, I mean, basically it should be a switch in terms of how you're targeting it. And then, yeah, it should, it really shouldn't actually impact too much. It's, you know, so there's different instructions. You would have to compile it. You have to compile different firmware.

Dave Jones: Okay. All right. Jeez, I'd be.

Chris Gammell: But what I'm excited about, this is very likely the first RISC-V part that I will use in a real capacity. Right. Okay. Got it. You know, RISC-V, we've been talking about on here. It's out there. You can buy it. There's, you know, there's a lot of stuff that's being commercially made. There's some stuff in the hobbyist market you could get. You can get PSY-V parts for a while. Personally, they, you know, they're interesting. There was some high-end stuff. There's some low-end. There was like a low-end one that went into Arduino kind of application. There's like a super high-end Linux-y one. And I think BeagleBoard now has a RISC-V based Linux computer as well. But they're not really, you know, like they're not, it's just becoming so commonplace that it's even in Expressive now. And I guess they had a RISC-V based like subprocessor, but it's going to be the main processor now. And yeah, I think this is going to be the first chip that a lot of people, especially in the hobbyist community, are going to be kind of have hands on with. Or I'll be told that it's already been used otherwise.

Dave Jones: Have we actually reached the point where the architecture doesn't matter? Because like they're, like I've not known another chip ever in the history of microcontrollers and microprocessors ever to use the same name. It's got the ESP32, right? But it uses, but this H2 version or whatever would dash, it just has a dash on the end. And it's a completely different architecture processor.

Chris Gammell: Yeah, I mean, it's really the peripherals that are going to be very similar.

Dave Jones: Yeah, it's like, this is like a first. This is like a new paradigm. Yeah, maybe. I don't know. Name another one.

Chris Gammell: Yeah, I think a lot of other companies would have, instead of dashing it out, right? Yeah, no, it wouldn't be. Instead of saying ESP32 dash something, it would have been, yeah, completely different product line.

Dave Jones: Like a PIC, for example, like a PIC32 is a different architecture to a PIC24, which is a different architecture to a PIC8, right?

Chris Gammell: Yeah, I'd say that's pretty close still. I mean, like that's a, yeah, PIC32 versus PIC24.

Dave Jones: No, one's a PIC architecture and one's a MIPS architecture. They're very different.

Chris Gammell: No, I know. I'm just saying that it's like PIC and the number is like not that different, right? I guess it would be like if there was a PIC32 dash A and a PIC32 dash B and one was MIPS and the other was ARM-based, then yeah, that would be.

Dave Jones: Yeah, but this one's exactly the same. It's ESP32 and you might think, oh, it's an expressive architecture or whatever, you know? And it's like. Yeah, I'm trying to find the name of that. I keep forgetting. It's actually Risk. So anyways.

Chris Gammell: There's a lot of other parts coming out. They're definitely growing like crazy.

Dave Jones: Strange times we live in.

Chris Gammell: Yeah, definitely.

Dave Jones: Resistors with no bloody labels on them. Yeah, pretty much the same thing.

Chris Gammell: No resistors without markings and Risk-V cores.

Dave Jones: And ESP32s with different architectures. What's the world come into? I don't know. When I was a boy, it's much simpler. What do we got last up? Come on. We've done our AMPR. We've done our dash.

Chris Gammell: I think that's it. Oh, former guest Bill Hurd.

Dave Jones: Oh, yes. Bill Hurd's book. Yes. Yeah, he's got a book. Yeah, he's got a new book. Yeah, former guest. But yep, that's on my get list. It's available in a Kindle edition. So I'll be getting the Kindle edition.

Chris Gammell: Critical. Yeah. Yeah. Yeah. And people can hear Bill talking about the C128 on the Amp Hour many years ago. Yep. Bill also does a podcast with Ben Jordan, formerly of Altium now. Yes, that's right. Yep. Autodesk. And they do some live streams and stuff like that. It's some fun stuff. And I guess in terms of other former guests doing podcasts, actually, this was a really great episode and I highly recommend it. It was Chris Denny from Worthington. Worthington and Circuit Hub have a podcast together. And Chris was talking about the small PCB layout decisions that have significant impact on assembly. It's called Pick Place Podcast. But it was just a really good rundown on kind of DFM type stuff. And Chris works at Worthington and so they see a lot of that stuff. And I really liked it. They've been doing a lot of episodes now. They have super, super niche stuff. They had two consecutive episodes with a guy that just does board washing. Right. It's great. So if you really want to get deep on manufacturing.

Dave Jones: Yeah, it's important. Board washing.

Chris Gammell: Yep. Yeah. I mean, yeah, totally. It is. And when else? I think the only time you would otherwise have those conversations is if you're waiting around as a machine got installed and you're talking to the applications engineer from the board wash thing. No, no.

Dave Jones: That's rare as hence teeth, right? Yeah. That's something you'll never hear.

Chris Gammell: Yeah. That's great. And Chris has been on the show. He's great. So highly recommend that podcast. Excellent. So yeah. Congrats to Bill on the book and Chris. Nice job on the podcast here.

Dave Jones: So the book is like the history of his time at Commodore or is it a complete history? I haven't got the read the full synopsis yet.

Chris Gammell: I think it's just Bill's story.

Dave Jones: I don't think it's not a complete history of actual Commodore from go to work.

Chris Gammell: I think there's going to be broader history, but I think it's mostly Bill's story. Awesome. Yeah.

Dave Jones: Yep.

Chris Gammell: Yep.

Dave Jones: Is there an audio version? Oh, I'm going to have to hit him up for an audio version.

Chris Gammell: Maybe you could record the audio version.

Dave Jones: Who wouldn't want to hear Dave read an audio book for eight hours? We would have to pay people to actually listen to it. It'd be a reverse business model. Yeah. Yeah. Uh-huh. Yep. Yep. Yep.

Chris Gammell: Brewster's million scenario where you're trying to get rid of money, $30 million. Yep.

Dave Jones: Oh, boy. Yep. That's great. Is he on Twitter?

Chris Gammell: Bill is on Twitter, I think. Right. Not super active, but Bill and Ben have a... Their podcast also has a Discord channel you can join.

Dave Jones: Right. There you go. I'm not really into the Discord thing. Yep. No. I want bloody... I want BBS forums. I want labels on my resistors, and I want my processors to be the same, please. I don't think it's too much to ask in the modern world.

Chris Gammell: Dave Jones, the progressive engineer. The engineer of the future, folks. Right here. Right here.

Dave Jones: Catch you next time. See ya. See ya. See ya.

Speaker ?: See ya. See ya. We'll be right back.